Matrimony.Com Limited vs Google Llc & Others (30/2012) … on 31 January, 2018
Competition Commission of India
(07/2012) Matrimony.Com Limited vs Google Llc & Others (30/2012) … on 31 January, 2018
PUBLIC VERSION
COMPETITION COMMISSION OF INDIA
Case Nos. 07 and 30 of 2012
Case No. 07 of 2012
In Re:
Matrimony.com Limited Informant
And
1. Google LLC Opposite Party No. 1
2. Google India Private Limited Opposite Party No. 2
3. Google Ireland Limited Opposite Party No. 3
WITH
Case No. 30 of 2012
In re:
Consumer Unity & Trust Society (CUTS) Informant
And
1. Google LLC Opposite Party No. 1
2. Google India Private Limited Opposite Party No. 2
3. Google Ireland Limited Opposite Party No. 3
C. Nos. 07 & 30 of 2012 1
PUBLIC VERSION
CORAM
Mr. Devender Kumar Sikri
Chairperson
Mr. S. L. Bunker
Member
Mr. Sudhir Mital
Member
Mr. Augustine Peter
Member
Mr. U.C. Nahta
Member
Mr. Justice G. P. Mittal
Member
Appearances: Shri Rajshekhar Rao, Shri Naval Satarawala Chopra, Shri
Yaman Verma, Shri Aman Singh Sethi, Shri Sapan Parekh
and Shri Varun Mishra, Advocates alongwith Shri
Murugavel J. Ravichandran, Founder & CEO and Shri S.
Ravichandran, General Manager (Legal & Regulatory) for
Matrimony.Com Private Limited.
Shri Vijay Singh, Centre Co-ordinator for Consumer Unity
& Trust Society.
Dr. Abhishek Manu Singhvi, Shri Arvind Nigam and Shri
Arun Kathpalia, Senior Advocates with Shri Samir R.
Gandhi, Shri Suhail Nathani, Ms. Kamya Rajagopal, Shri
Ravishekhar Nair, Shri Arjun Khera, Ms. Aditi
Gopalakrishnan, Ms. Kadambari Chinoy, Ms. Krithika
Ramesh, Ms. Anuja Agrawal and Shri Aakarsh Narula,
Advocates alongwith Shri Munesh Mahtani, Senior
Competition Counsel, Ms. Gitanjali Duggal, Legal
C. Nos. 07 & 30 of 2012 2
PUBLIC VERSION
Director and Shri Shekhar Sharad, Product Manager for
Google.
Order under Section 27 of the Competition Act, 2002
1. The information in Case No. 07 of 2012 was filed under Section 19(1)(a)
of the Competition Act, 2002 (‘the Act’) by Consim Info Private Limited
(‘the Informant’/ ‘Consim’, now known as Matrimony.com Limited)
which is stated to provide internet as a vehicle/ platform for prospective
marriage alliances, against Google Inc. and Google India Private Limited
alleging contravention of the provisions of Section 4 of the Act.
2. The information in Case No. 30 of 2012 was filed under Section 19(1)(a)
of the Act by Consumer Unity & Trust Society (CUTS) (‘the Informant’)
which is stated to be a non-profit, non-governmental organisation working
on public interest issues including those related to consumer protection
and competition – against Google Inc. (now Google LLC) and Google
India Private Limited alleging inter alia contravention of the provisions
of Section 4 of the Act.
Facts
3. In Case No. 07 of 2012, it was stated by the Informant that Google runs
its core business of search and advertising in a discriminatory manner,
causing harm to advertisers and indirectly to the consumers. It was alleged
that Google is creating an uneven playing field by favouring Google’s
own services and partners, through manually manipulating its search
results to the advantage of its vertical partners.
4. It was pointed out that in addition to running the world’s most popular
search service, Google also provides a large number of vertical search
C. Nos. 07 & 30 of 2012 3
PUBLIC VERSION
services, including video (YouTube), news (Google News), maps (Google
Maps), etc. It has been averred that in order to promote Google’s own
vertical search sites, Google started mixing many of its vertical results into
its organic search results. Therefore, when a user searches, for example,
the name of a song on Google, he receives links to videos of that song
from Google Video or YouTube, both of which are properties owned by
Google.
5. It was further pointed out that Google’s own sites would appear
prominently on the search results page irrespective of whether they are the
most popular or relevant sites to the search and Google will not place
results from any other vertical search sites as prominently as Google’s
own vertical search sites in its list of results.
6. It was further averred by the Informant that acquisition of various software
products by Google to complete its vertical integration further fortifies its
monopolistic position and tendency to eliminate competition. Google’s
dominance in algorithmic search market leads to its status as an
unavoidable trading partner in search advertisement market.
7. It was stated that the search algorithm used by Google to determine where
a website appears on a search results page the algorithm used to calculate
an advertiser’s quality score which forms the basis for determining where
an advertisement appears, the history of changes to Google’s search and
quality score algorithms, the basis for placing Google’s vertical properties
towards the top of the search results pages and what caused Consim’s
quality scores to fall, the ad hoc behaviour of Google etc. will only come
to light if and when Google is directed to provide this information by the
Commission/ Director General during the course of an investigation.
C. Nos. 07 & 30 of 2012 4
PUBLIC VERSION
8. In view of the above, it was averred that Google has abused its dominant
position in the relevant market in India thereby contravening the
provisions of Section 4 of the Act.
9. In Case No. 30 of 2012, the Informant averred that based on market
structure for internet search and internet search advertising market, the
relevant markets for the purposes of the present information are online
search market and online search advertising market in India.
10. It was further averred that Google, because of its market share, size,
resources, reputation etc., is widely recognized as enjoying a dominant
position in the market for online search and online search advertising
world-wide including in India. As such, Google enjoys a position of
strength in the relevant market which enables it to operate independently
of competitive forces and to affect its competitors, consumers and the
market in its favour.
11. The Informant alleged that Google is indulging in abuse of its dominant
position in the market for online search through practices leading to search
bias, search manipulation, denial of access to competing search engines,
refusal to license content to competing search engines and creation of
entry barriers etc. Allegations were also made about the abusive conduct
of Google in the market for online search advertising through imposition
of unfair and discriminatory conditions on its customers etc.
Directions to the DG
12. The Commission, after considering the entire material available on record,
vide its order dated 03.04.2012 passed under Section 26(1) of the Act in
Case No. 07 of 2012 directed the Director General (DG) to cause an
investigation to be made into the matter.
C. Nos. 07 & 30 of 2012 5
PUBLIC VERSION
13. Subsequently, the Commission passed another order under Section 26(1)
of the Act in Case No. 30 of 2012 on 20.06.2012 directing an investigation
by the DG. Further, the Commission directed this information to be
clubbed with Case No. 07 of 2012 in terms of the provisions contained in
the proviso to Section 26(1) of the Act as similar issues were pending
investigation before the DG in that case.
14. During the course of investigation, it was observed by the DG that Google
Ireland Ltd., Ireland (GIL), a subsidiary of Google Inc., was playing an
important role in the operations of Google in India and as such, on a
reference made by the DG, the Commission vide order dated 03.09.2014
directed Google Ireland Limited to be included as an Opposite Party in the
cases. Accordingly, investigation was conducted against Google Inc.,
Google India Pvt. Ltd. and Google Ireland Ltd. (collectively, “Google”).
15. The DG, after receiving directions from the Commission, investigated the
matters and filed confidential version of a common Investigation Report
dated 27.03.2015 in both the cases. Subsequently, a non-confidential
version of the Investigation Report was filed by the DG on 14.07.2015.
16. It may also be noted that the Commission, vide its order 31.07.2013 took
on record the letter dated 12.07.2013 of the Informant in Case No. 07 of
2012 whereby it was informed that the name of the Informant company
has been changed to Matrimony.com Private Limited. Subsequently, it
appears from the filings of this Informant that the name has now been
changed to Matrimony.com Limited. Accordingly, the cause title
mentions this name in the order.
C. Nos. 07 & 30 of 2012 6
PUBLIC VERSION
Investigation by the DG
17. To investigate the allegations, the DG first determined the relevant
market. Based on the analysis of characteristics, intended use and price, it
was found that there is no substitution between Online General Web
Search Services and Specialised/ Vertical Search Services or Site-Specific
Search Services. Further, analysis of characteristics, intended use and
price revealed that first of all, online advertising is distinct from offline
advertising and secondly, online search advertising is distinct from other
forms of advertising like display advertising consisting of text, images,
graphics and videos, social network advertising, email-based marketing
and advertisements on mobile applications. Online search advertising is
linked to user-initiated search query that reveals user’s immediate interest
in the subject matter of the query and thus allows very effective means for
targeting potential customers. On the other hand, online non-search
advertising is more appropriate to improve brand awareness and brand
building.
18. It was further noted that the fact that these distinct categories exist and
many advertisers opt for several types of these advertising forms
simultaneously clearly show that they provide different kinds of
opportunities for advertising to target eyeballs. These are, in fact,
complementary in nature. It was also found that Online Search and Search
Advertising are complementary and do not form a part of the same
relevant product market.
19. Therefore, in terms of the provisions of Section 2(r), Section 2 (s) and
Section 2(t) read with Section 19(6) and Section 19(7) of the Act,
following relevant markets were delineated by the DG:
C. Nos. 07 & 30 of 2012 7
PUBLIC VERSION
a. Relevant market of Online General Web Search Service in
India.
b. Relevant market of Online Search Advertising in India.
20. On the issue of dominance, Google was found to be a dominant enterprise
in both the relevant markets of Online General Web Search Services and
Online Search Advertising in India. In the relevant market of Online
General Web Search, Google’s market share, estimated using various
parameters such as number of search queries and page views and across
device categories, was found to be consistently more than during
2009 to 2014. This was the case despite the long standing presence of
other competitors like Yahoo! and Microsoft Bing and the entry of new
players in the market. In the relevant market of Online Search Advertising
as well, Google was found to consistently maintain an estimated market
share of more than in India during the period 2009 to 2013.
21. Other factors like Google’s size and resources, economic power and
commercial advantages, entry barriers, etc. further reinforced Google’s
position of dominance. As per the DG, there exist significant entry barriers
in the nature of high cost, technology, network effects, minimum scale
requirements, and contractual restrictions etc. that bestow substantial
economic power on Google and place it at a major advantage. For these
reasons, Google has been able to establish itself as a critical platform for
all the stakeholders. The competitors have not been able to dent its market
position despite long presence and do not pose any significant competitive
constraint upon it. The DG thus, concluded that Google enjoys a position
of strength in these markets which enables it to operate independently of
competitive forces and to affect its competitors/ consumers as well as the
relevant markets in its favour.
C. Nos. 07 & 30 of 2012 8
PUBLIC VERSION
22. Adverting to the issue of abuse of dominant position, it was noted by the
DG that apart from Online General web search services, Google also
offers specialised search services like Google News, Google Maps,
Google Flights etc. It also provides other products and services such as
Google Reviews, Google+, YouTube etc. Google was found to be
indulging in practices of search bias and by doing so, it causes harm to its
competitors as well as to users. Investigation has revealed that Google
integrates/ blends its own specialised/ vertical search
services/options/features in its online general web search services in
Universal Results and Commercial units using mechanisms that do not
apply in an equivalent manner to non-Google websites/ web content.
Moreover, it offers its own specialised search features (Commercial Units,
Universal Results etc.) at prominent ranks or positions on the Search
Engine Results Page (SERP). Top results receive higher user attention and
are critical for online visibility. Through this practice, Google steers users
to its own products and services, and produces biased results. This
structure offers abundant opportunities for leveraging and has also raised
issues of conflict of interest. Thus, the users may not receive the most
relevant results. This also adversely affects the competitive landscape in
the markets for online general web search and search advertising as well
as adjacent markets such as travel, maps, social networking, e-commerce
etc. Such actions deter innovation in a wide-range of these online ancillary
markets. Consequently, users may be devoid of additional choices of
results. This shows that competition is hampered in the market-impeding
innovation, and thus, harming consumers. Therefore, Google’s conduct
was found to be anti-competitive in terms of Section 4(2)(a)(i), Section
4(2)(b)(ii), Section 4(2)(c) and Section 4(2)(e) of the Act.
23. Further, it was observed by the DG that Google does not disclose to the
advertisers the details of their quality score or quality scores and bids
received from various advertisers for a particular keyword in an auction,
C. Nos. 07 & 30 of 2012 9
PUBLIC VERSION
even on historical basis. While it would not be appropriate to disclose the
bid amount on real time basis to others to prevent collusive behaviour, in
the absence of disclosure of historical data, any aberrations in the process
on account of errors of bonafide nature or manipulations can remain
undetected by the concerned advertisers over long period of time. The
advertisers may be oblivious to the fact that demotion of their ads is due
to system faults/ possible quality score manipulations, thereby subjecting
them to unfair and discriminatory conditions. Investigation revealed that
there exists technical feasibility of sharing greater details of quality scores
of individual campaigns as well as of historical data. Complex nature of
determination of Quality Score coupled with non-disclosure of adequate
information to advertisers renders the entire process opaque and non-
transparent and susceptible to manipulation which amounts to imposition
of unfair conditions on advertisers in violation of Section 4(2)(a)(i) of the
Act. Google was, thus, found to be abusing its dominant position by not
making such details available to its advertisers and following non-
transparent procedure.
24. It was also noted by the DG that Google’s policy regarding compensation
is entirely discretionary and does not place any obligation on it to
compensate the advertisers for losses that can be attributed to Google’s
system error. While no site may be guaranteed to win a particular position
under the AdWords mechanism, nevertheless it should be treated in a fair
and non-discriminatory manner and Google needs to take up
responsibility for any aberrations in the system which subjects them to
unfair treatment.
25. The DG also recorded that Google is not required to pay any monetary
consideration for its House Ads which gives it an additional competitive
edge. Further, Google has access to information on quality score that the
system assigns to other websites including those of its competitors.
C. Nos. 07 & 30 of 2012 10
PUBLIC VERSION
Google being aware of these is in a position to ensure that its House Ads
are assigned higher quality score than its competitors and ensure that
invariably its House Ads appear in the top slots and above third party ads
particularly of its competitors. The checks referred to by Google were
found to be inadequate to address and prevent preferential treatment of its
own properties in paid results. It was thus, found that there does not exist
a level playing field for third parties competing with Google’s House ads
to appear in paid results in response to user searches. Such conduct also
amounts to imposition of unfair and discriminatory conditions on third
party advertisers using AdWords and was found by the DG to be in
violation of Section 4(2)(a)(i) of the Act.
26. Google was also found to be abusing its dominance in online web search
and online search advertising markets by imposing unfair conditions upon
trademark owners (particularly those who have notified their trademarks
to Google) whose trademarks are being allowed to be bid as keywords by
third parties in online search advertising, in violation of Section 4(2)(a)(i)
of the Act. As per the AdWords mechanism, ads that appear first may not
be the most relevant for users and may appear at that position due to higher
bid placed by an entity. The competitors get opportunity to free ride on
the goodwill and brand value of the trademark owner thereby hampering
fair competition. This practice also creates a significant risk of causing
confusion and deception in the minds of the users thereby causing
consumer harm. Unsuspecting consumers may be misled to believe that
there exists an association between the owner of the trademark and its
competitors (whose ads appear in response to searches on their brands)
and this may divert traffic. In such a scenario, the owner of the trademarks
are compelled to participate and outbid competitors (for their ads to appear
before them) thereby augmenting their advertising budgets.
C. Nos. 07 & 30 of 2012 11
PUBLIC VERSION
27. Further, it was observed that there is scope for Google to use the system
in a discriminatory manner and ensure that its own trademarks are not
subject to the same unfair conditions as those imposed upon third parties,
where bidding under AdWords. Such conduct of Google was therefore
found to be unfair and discriminatory in terms of the provisions of Section
4(2)(a)(i) of the Act.
28. It was further noted that though Google’s AdWords policy restricts usage
of notified trademarks in Ad text of competitors, investigation brought out
that on various occasions, ads of competitors using Consim’s trademarks
in Ad text in response to searches on these trademark terms despite
notification. Further, Google allowed usage of minor variations of
Consim’s notified trademarks in Ad text of competitors under its
AdWords program despite repeated complaints of Consim and this led to
unfair bidding between trademark owner and other advertisers. This was
in complete disregard of anti-competitive effect and appeared to be driven
by Google’s commercial interests. Consim’s own ads were blocked for
searches on its trademark terms even after it complied with Google’s
requisite procedures. Google thus, was concluded to have abused its
dominant position and imposed unfair conditions on Consim in violation
of Section 4(2)(a)(i) of the Act. Such conduct also resulted in unfair
business gains to Consim’s competitors as well as to Google itself.
29. The DG also found that apart from online web search services, Google
also offers online search and advertising on other websites through
Syndication/ Intermediation services. With regard to advertising,
intermediation can take place for both search and non-search advertising.
Google offers Syndication services under its AdSense program. The
Online Search and Advertising Syndicate Services constitute distinct
relevant markets. By virtue of its position of strength in the relevant
markets of Online General Web Search Services and Online Search
C. Nos. 07 & 30 of 2012 12
PUBLIC VERSION
Advertising Services, Google is also a preferred Syndicate service
provider for publishers wanting to offer search and advertising services on
their websites. Google was found to be using its dominant position in
Online General Web Search Service and Online Search Advertising
Service to impose certain restrictive conditions in its agreements for
syndicate search and advertising services in violation of Section 4(2)(e)
and Section 4(2)(c) of the Act. The nature of restrictions varied across
types of agreements. While in some cases, partners are prohibited from
using competing services, in other there are restrictions relating to the
manner of placement of ads of competitors. These restrictions prevented
competing service providers from achieving necessary scale which results
in creation of entry barriers for them.
30. The policy and conduct of Google, prior to May 2010, for not disclosing
AdSense Revenue to online AdSense partners was concluded by the DG
as amounting to imposition of unfair conditions on them resulting in
infringement of Section 4(2)(a)(i) of the Act.
31. Such agreement with AdSense partners was also held to be one-sided and
providing enough scope for arbitrary conduct without fair opportunity to
the other party. This amounts to imposition of unfair conditions by Google
within the meaning of Section 4(2)(a)(i) of the Act.
32. Google also entered into agreements with customers licensing AdWords
API from it. Google’s AdWords API agreements with third party tool
developer entities contain certain restrictive clauses that have anti-
competitive effects which the DG found to be in violation of the
provisions of Section 4(2)(a)(i), 4(2)(b)(ii) and Section 4(2)(c) of the Act.
These restrictions have the potential to be used as a tool for discouraging
advertisers from multi-homing thereby being instrumental in denial of
market access to competitors and causing other anti-competitive effects.
C. Nos. 07 & 30 of 2012 13
PUBLIC VERSION
Further, inclusion of a provision on termination without reason in the
AdWords API terms amounted to the imposition of an unfair condition on
AdWords API users in violation of Section 4(2)(a)(i) of the Act.
33. Resultantly, it was concluded by the DG that Google abused its dominant
position in the relevant markets of “Online General Web Search Service
in India” and “Online Search Advertising in India”, in violation of Section
4(2)(a)(i), 4(2)(b)(ii), 4(2)(c) and Section 4(2)(e) of the Act.
Consideration of the Investigation Report by the Commission
34. The Commission considered the Investigation Report submitted by the
DG in its ordinary meetings and decided to forward copies thereof to the
parties for filing their respective objections/ suggestions thereto. After
disposal of various procedural applications, the matter was finally heard
by the Commission on January 12, 18, 19, 20, 23 and 24, 2017 whereupon
the Commission decided to pass an appropriate order in due course.
Replies/ Objections/ Submissions of the parties
35. On being noticed, the parties filed their respective replies/ objections/
submissions to the Investigation Report of the DG besides making oral
submissions.
Replies/ objections/ submissions of the Opposite Party
36. Google filed its response to the Investigation Report and the same shall be
referred to and dealt with while analysing the matters on merits.
C. Nos. 07 & 30 of 2012 14
PUBLIC VERSION
Replies/ objections/ submissions of the Informants
37. The Informant in Case No. 07 of 2012 filed its objections and broadly
supported the findings of the DG. The objections filed by this Informant
shall be appropriately referred to in the order. The Informant in Case No.
30 of 2012 filed its brief preliminary comments/ submissions and
supported the findings of the DG.
ANALYSIS
38. Briefly stated, it was inter alia alleged in the informations that Google
runs its core businesses of search and search advertising in an unfair and
discriminatory manner, causing harm to the publishers and advertisers,
and to the consumers. Further, it was alleged that Google was creating an
uneven playing field by unduly favouring its own services. Google is
leveraging its strong position in various online search market to enter into
and enhance its position in ancillary markets. Not only does that cause
direct harm to competitors in vertical markets, it also causes direct harm
to other website owners, since their websites are moved down on SERP
and hence, they receive less clicks as a result of lessened traffic. Further,
this also harms consumers as they no longer receive the most relevant
results at the top of SERP.
39. To examine the allegations, it would be appropriate to highlight the
working of search engine platforms. General web search engines like
Google, also known as “Horizontal Search Engines”, crawl and index
websites/ web content and other online content and display results
algorithmically on SERP on the basis of relevance in response to a user
query entered in the search box/ input bar. They deliver information on a
wide variety of topics/ content categories for a given user query.
C. Nos. 07 & 30 of 2012 15
PUBLIC VERSION
40. A search engine service provider brings together the following categories
of users: internet users who key in search terms to find relevant results on
the web; websites and information/ content providers on internet whose
links and content appear as results on the SERP; and Advertisers whose
ads appear on SERP in response to search queries.
41. Service providers monetise their online search business through online
search advertising. Internet users who utilise search services form
consideration by providing their attention or “eyeballs” to the search result
pages containing links to web content. In addition, they allow the search
engines to collect their information for use. By doing so, users facilitate
the generation of revenues through sponsored advertisements. On the
other hand, by allowing links to and content from their webpages to appear
in search results, websites enable search engines to maintain and enlarge
their internet user base. This inter-relationship amongst internet users,
websites and advertisers is reflective of the widely accepted multi-sided
character of the search engine business. In a multi-sided market, there
exist multiple distinct customer groups that have inter-related demand and
so one or more groups impose a positive externality on the other group/s.
In such a scenario, there exists possibility of subsidizing provision of
services to a certain set of consumers by charging another set of
consumers.
42. Against the aforesaid backdrop and on perusal of the information, the
Investigation Report of the DG, replies/ objections filed thereto, the
submissions made by the parties and other materials available on record,
the following points arise for consideration and determination in the
matter:
C. Nos. 07 & 30 of 2012 16
PUBLIC VERSION
(i) What is the relevant market(s) in the present case?
(ii) Whether Google is dominant in the said relevant market(s)?
(iii) If finding on Issue No. (ii) is in the affirmative, whether Google has
abused its dominant position in the relevant market(s)?
43. Prior to examining the aforesaid issues, it would be appropriate to mention
that throughout the analysis the term “Google” shall be used to denote
collectively the three Opposite Parties viz. Google LLC, Google India
Private Limited and Google Ireland Limited. As has been brought out by
the DG in the Investigation Report, these three entities constitute a
“Group” as defined in the Explanation (b) to Section 5 of the Act and no
objection has been raised by Google against this finding of the DG.
Relevant Market
44. The DG has identified the relevant market after considering both the
relevant product and geographic markets in accordance with Section 2(r),
Section 2(s) and Section 2(t) read with Section19(6) and Section 19(7) of
the Act:
Relevant Product Market
Online General Web Search Services
45. While delineating the relevant product market, the DG has conducted a
detailed analysis of the factors set out in Section 19(7) of the Act.
46. The DG considered the question of substitutability of online general web
search services and search advertising services and concluded that online
general web search services and search advertising did not constitute the
same relevant product market on account of wide variations in the
C. Nos. 07 & 30 of 2012 17
PUBLIC VERSION
mechanism for generation and display of results and also clicking
behaviour. Further, the DG noted that these services served distinct goals
and are perceived differently by various categories of users, namely,
publishers (websites) and internet users entering search queries. The DG
also observed that these services constituted complementary services from
the point of view of websites striving for eyeballs.
47. The DG also considered the issue concerning substitutability of direct
search option by typing the Uniform Resource Locator (URL) of a
particular website in the internet browser with online general web search
services. The DG acknowledged the limitations associated with the usage
of the direct search option as it required the users to not only be aware of
the websites that offer the relevant information but also required the users
to remember URL. The DG also noted that given the large number of
websites that are in existence coupled with the voluminous nature of
information available on the webpages, it is virtually impossible for the
users to be aware of all the websites that might be able to provide them
with the desired information and their respective URLs. In fact, most of
the users might not remember more than a handful of URLs of the
websites. Additionally, internet users may prefer a general web search to
explore the vast alternative online sources of information. On this basis,
the direct search option involving URL has not been found by the DG to
be substitutable with a general purpose web search option.
48. The DG also considered if online general web search services were
interchangeable or substitutable with online specialized search services.
While general purpose search engines allowed internet users to search
information on a wide range of topics, specialised search services
permitted online searches for information limited to particular topics or
areas such as news, shopping, travel, entertainment, etc. Further, in
response to a search query, general purpose web searches display
C. Nos. 07 & 30 of 2012 18
PUBLIC VERSION
information from across the web while specialised search results return
with information from a limited source, i.e., either its own contents or
from the contents of certain specified websites. Additionally, pricing and
registration requirements stipulated by the general purpose online
searches and specialised searches are different. Accordingly, the DG
concluded that online general web search services were not substitutable
with site-specific search and specialised search services as there were
variations in terms of their characteristics, intended use, price etc.
49. In view of the above, the DG identified online general web search services
as a distinct relevant product market in accordance with the provisions of
Section 2(s) read with Section 19(7) of the Act.
Online Search Advertising Services
50. Further, the DG examined if consumers (who use advertising services)
regarded online and offline advertising as substitutable. Online
advertising is undertaken using internet as a medium. Therefore, its
coverage is largely dependent on reach of the internet. Given that a very
large number of people in our country still do not have access to the
internet, online advertising is not substitutable with newspapers, radio or
television for advertisers who seek to target areas or user groups with
limited internet access. In addition, advertising rates are significantly
lower for online advertising in comparison to traditional media.
Furthermore, online advertising allows advertisers to accurately monitor
the effectiveness of their advertisements on the basis of actual number of
users that it reaches whereas for offline advertisements, the advertisers
rely on estimated number of views and not actual views. On account of
vast differences in the characteristics of online and offline advertising, the
DG noted that these are not substitutable and therefore not a part of the
same relevant market.
C. Nos. 07 & 30 of 2012 19
PUBLIC VERSION
51. Further, the DG considered the substitutability of online search
advertising and online non-search advertising from consumer’s (i.e., user
of advertising services) perspective. Search advertising is unique from
advertiser’s perspective as it allows highly targeted advertisements by
providing advertisers with clear insights into user’s intents and desires.
Search and non-search advertising fulfil different marketing functions.
Typically, search advertisements are used for demand fulfilment while
non-search advertisements are for brand awareness or recognition.
Further, the two kinds of advertisements are priced using different pricing
mechanisms. For example, search advertisements are generally paid on a
cost-per-click basis, while non-search advertisements are usually paid on
cost-per-thousand-impressions basis. Therefore, the characteristics,
intended use and price of search and non-search advertising were found to
be different from one another. Further, the advertisers simultaneously
used many different forms of advertising on the basis of their specific
advertising objectives. Accordingly, the DG held that one form of online
advertising did not serve as a replacement for the other and the two are,
therefore, complementary in nature.
52. In view of the aforesaid, the DG delineated online search advertising
service as a distinct Relevant Product Market in accordance with the
provisions of Section 2(s) read with Section 19(7) of the Act.
Relevant Geographic Market
Online General Web Search Services
53. The DG noted that local specification requirements, language differences
and consumer preferences of users were relevant factors for the supply of
online web search services to them. Supply was also dependent on the
legislative framework of the country. Therefore, the conditions for supply
and demand of online web search services in India are distinct from the
C. Nos. 07 & 30 of 2012 20
PUBLIC VERSION
conditions prevailing in other areas. Accordingly, India has been
considered as the relevant geographic market for online general web
search services in accordance with the provisions of Section 2(t) read with
Section 19(6) of the Act. It may, however, be mentioned that the relevant
geographic market in any event could not have been taken as global since
from a plain reading of the Explanation to Section 4 of the Act, ‘dominant
position’ means a position of strength, enjoyed by an enterprise, in the
relevant market, in India, which enables it to operate independently of the
competitive forces prevailing in the relevant market or affect its
competitors or consumers or the relevant market in its favour. Similar
view has been taken by the Commission in previous cases including the
Coal cases.
Online Search Advertising Services
54. The DG considered the conditions for demand of online search advertising
services and those for supply of online search advertising (in terms of
legislative framework, presence of local distribution entities and
variations in applicable terms and conditions etc.) and held India to be the
relevant geographic market for online search advertising services in
accordance with the terms of Section 2(t) read with Section 19(6) of the
Act.
Relevant Market
55. In sum, it may be noted that the DG has identified the following two
relevant markets in the present case:
(a) Market for Online General Web Search Services in India
(b) Market for Online Search Advertising Services in India
C. Nos. 07 & 30 of 2012 21
PUBLIC VERSION
Google’s Response
56. Challenging the delineation of the relevant market by the DG, it was
submitted on behalf of Google that the DG’s market definition for search
is flawed because there is no relevant market for general search. It was
submitted that users do not conduct “general searches”. They search for
specific things, such as people, places, recipes, products, or local
businesses. Within each of these query categories, Google competes with
all types of services that are able to answer that query.
57. Elaborating, it was argued that for product queries, Google competes with
all services that allow users to search for such products. For local queries,
Google competes with all services that allow users to search for such local
information. And for travel queries, Google competes with all services
that allow users to search for travel information. The Investigation Report
is therefore correct when it acknowledges that “there exist websites that
provide search services for specific content categories that compete with
Google in those areas”. Yet, it was pointed out that the Investigation
Report claims that vertical search services do not compete with general
search services. The DG bases this conclusion on technical differences
(such as data inputs and traffic sources) without analysing how or why
these differences are relevant for users searching for information. This is
wrong. Under Section 2(t) of the Act, relevant product market comprises
of services “regarded as interchangeable or substitutable by the
consumer”. The Investigation Report should have analysed whether users
consider vertical search services and general search services as
substitutable for individual queries; however it has failed to do so.
58. It was further submitted that the Investigation Report also ignores the ad-
funded nature of Google’s business, which means that – as confirmed by
international precedent – the relevant market can only be advertising, not
free search. In fact, the Act, consistent with international precedent,
C. Nos. 07 & 30 of 2012 22
PUBLIC VERSION
requires the existence of trading relationship between a company and its
customers as a pre-condition for defining a relevant market and
establishing dominance. Because search is free, Google has no trading
relationship with the users of its search service, and so the basis for
establishing dominance is absent.
59. Also the DG’s market definition for advertising is flawed because the
relevant market includes all forms of advertising. The DG’s claim that
relevant market constitutes only online search advertising is incorrect. An
advertiser who wants to run an ad campaign can take advantage of a
multitude of different advertising opportunities, both online and offline.
These include TV, radio, newspapers, billboards etc. All these ad
opportunities serve the same purpose of attracting awareness to the
advertiser’s product or service. Based on a relative assessment of their
costs and Return-On-Investment (“RoI”), these forms of advertising are
interchangeable from an advertisers’ perspective. They, therefore, form
part of the same relevant market.
60. The reasons given by the DG for excluding offline advertising from the
relevant market suffer from the following methodological flaws:
(i) The Investigation Report applies the hypothetical monopolist test
(SSNIP test) in the wrong direction
The Investigation Report erroneously dismisses offline advertising
as a constraint based on the low level of Internet coverage
throughout India, applying the hypothetical monopolist test in the
wrong direction and therefore commits a major economic error: the
test should start with the narrowest candidate frame of reference
and then examine the extent to which offline advertising would
constrain a hypothetical monopolist in that candidate market.
C. Nos. 07 & 30 of 2012 23
PUBLIC VERSION
(ii) Alleged differences in monitoring and targeting ignore recent
developments
The alleged differences in monitoring and targeting fail to take into
account the recent developments in digital analytics and targeting
technology that are available for offline ads. The Investigation
Report points to differences in price, but erroneously ignores that
products with price differences can and do compete through
adjustments in quality to reflect price differences. In fact, the DG
in its Investigation Report even recognises that offline ads may be
more engaging than online ads. This is precisely the sort of
difference that can explain online ads’ lower prices.
(iii) Given recent developments, older decisions from other
jurisdictions are not probative
The DG in its Investigation Report relies on older decisions from
other jurisdictions that identify separate markets for online and
offline advertising. But given recent developments in media, mix
optimisation modelling and ad technology, these older decisions
cannot be relied upon to delineate such separate markets in India.
61. It was argued that a full assessment of up-to-date evidence shows that the
relevant market encompasses all forms of advertising. In this properly
defined market, Google’s share in India is just .
62. It was also argued that the Investigation Report claims that search and
non-search advertising should be differentiated based on differences in
characteristics, without explaining how or why these affect advertiser
demand. The Investigation Report ignores the substantial recent
convergence between the two mediums. Non-search ads are increasingly
used to elicit a direct response, and non-search ads exhibit sophisticated
targeting capabilities.
C. Nos. 07 & 30 of 2012 24
PUBLIC VERSION
The Informant’s Response
63. The Informant, on the other hand, supported the determination and
delineation of relevant product markets by the DG. Controverting
Google’s submission that it has no trading relationship with search users
or websites because the search service is free, the Informant submitted that
free services are covered within the ambit of the Act. There is no
requirement for a monetary consideration for services anywhere under the
Act and, therefore, where users of search engine are providing data as well
as “eyeballs” to the search engine as a consideration, a commercial
activity is clearly taking place and therefore the Act applies.
64. On the applicability of SSNIP test, the Informant submitted that a relevant
market may be defined after taking into consideration any or all of the
various factors listed in Section 19(6) and Section 19(7) of the Act. It is
not necessary to use the SSNIP test, which is merely one tool used to
determine what the relevant product market may be. While looking at
online general web search, it may not be appropriate to use SSNIP test.
Further, all other evidence suggests that online general web search is the
relevant market for consideration.
65. Further supporting the delineation of the relevant product markets by the
DG, the Informant submitted that there are several key differences
between online search and search advertising, which make them separate
products for advertisers such as the Informant. In fact, the Informant
considers online search and search advertising to be complementary
products, and not substitutes. It strives to appear at the top of the SERP
for search terms that are relevant to its business, both through the online
general web search results as well as the search advertising results. Google
itself earmarks separate space for advertisements and online general web
search results, which suggests that they are distinct. Further, users also
have a greater trust in the results that are “organic” and not paid for and
C. Nos. 07 & 30 of 2012 25
PUBLIC VERSION
therefore, general web search results are likely to be viewed as distinct
from advertising results by the users as well. Therefore, the Informant
submitted that online general web search and search advertising services
are separate relevant product markets.
66. Further, the Informant submitted that using a URL bar would also be
entirely meaningless for the purpose of discovering new information or
websites online. A user can only type in the URL when he/she is aware of
the URL or has visited the website before. A search engine provides access
to websites that may have the specific information that a user wants, even
though the user may never have been to, or heard of such a website. The
URL bar cannot substitute this purpose, which is also why URL bars are
now linked to a search engine such as Google. Therefore, if any phrase or
word that is not a URL is typed in the URL bar, the search engine linked
to the browser will carry out a search and the user will be shown the SERP
for that query. This establishes that URLs do not serve the same purpose
as a search engine, which is why URL bars need to link back to a search
engine so that the user can view the SERP when a search is carried out.
67. It was pointed out by the Informant that while Google provides a general
search service, the Informant’s website provides a specialised search
service. Google is one of the major sources of new users for the Informant,
and the Informant is not a substitute for Google’s online general web
search business in any way, as the Informant only provides a search
service for matrimonial needs. Further, site-wide search services offered
on the Informant’s website also cannot be compared with Google’s online
general web search, as the Informant is limited to searches on its site,
while a search on Google’s website provides results from all over the
internet.
C. Nos. 07 & 30 of 2012 26
PUBLIC VERSION
68. Therefore, the Informant supported the DG’s conclusion that the relevant
product market is the market for online general web search, and this is
distinct from search advertising, other sources of information online, as
well as vertical search services.
69. Commenting upon offline and online mode of advertising, it was
submitted by the Informant that from its perspective, while both online
and offline advertising are constrained by a common overall company
budget, the objectives of search advertising and offline advertising (such
as television campaigns) are entirely different. The Informant uses offline
advertising for creating brand awareness, while the objective of search
advertising is to convert a real-time query from a user into a click through
to the Informant’s website. In fact, the Informant expects an increase in
searches on Google’s website for keywords that relate to it, right after it
launches a television campaign. Therefore, these forms of advertising are
also seen as complementary, and not substitutes.
70. Similarly, distinguishing online Search and non-search advertising, it was
submitted on behalf of the Informant that search advertising provides
unique opportunity to convert a user query into a click on website. It is the
only form of online advertising with specifically timed targeted
advertising, which makes it a unique value proposition for the Informant
(and other advertisers). Although display advertising can, in some cases,
be targeted to the general preferences of the recipient of the advertising, it
is not targeted in real time.
71. Therefore, the Informant submitted, in line with the DG’s conclusion, that
the relevant product market is the market for online search advertising,
and this is distinct from offline advertising, as well as online display/ non-
search advertising.
C. Nos. 07 & 30 of 2012 27
PUBLIC VERSION
72. On relevant geographic market, it was submitted by the Informant that
Google has a separate entity to run India operations and its search results
on the India specific URL (www.google.co.in) are customized to show
pages mostly from India. Further, Google is yet to launch several services
in India which are currently available elsewhere in the world like Google
shopping. Accordingly, India is a separate and distinct geographical
market.
73. In conclusion, the Informant submitted that the DG correctly delineated
the relevant markets, as noted earlier.
The Commission’s Analysis
74. The Commission has carefully examined the rival submissions on the
issue of determination of the relevant market.
75. As per Section 2(r) of the Act, ‘relevant market’ means the market which
may be determined by the Commission with reference to the relevant
product market or the relevant geographic market or both. Further, the
term ‘relevant product market’ has been defined in Section 2(t) of the Act
as a market comprising all those products or services which are regarded
as interchangeable or substitutable by the consumer, by reason of their
characteristics, prices or intended use. The term ‘relevant geographic
market’ has been defined in Section 2(s) of the Act to mean a market
comprising the area in which the conditions of competition for supply of
goods or provision of services or demand of goods or services are
distinctly homogenous and can be distinguished from the conditions
prevailing in the neighbouring areas.
76. For determining whether a market constitutes a ‘relevant market’ for the
purposes of the Act, the Commission is required to have due regard to the
‘relevant geographic market’ and the ‘relevant product market’ by virtue
C. Nos. 07 & 30 of 2012 28
PUBLIC VERSION
of the provisions contained on Section 19(5) of the Act.
77. To determine the ‘relevant geographic market’, the Commission, in terms
of the factors contained in Section 19(6) of the Act, is to have due regard
to all or any of the following factors viz., regulatory trade barriers, local
specification requirements, national procurement policies, adequate
distribution facilities, transport costs, language, consumer preferences and
need for secure or regular supplies or rapid after-sales services.
78. Further, to determine the ‘relevant product market’, the Commission, in
terms of the factors contained in Section 19(7) of the Act, is to have due
regard to all or any of the following factors viz., physical characteristics
or end-use of goods, price of goods or service, consumer preferences,
exclusion of in-house production, existence of specialized producers and
classification of industrial products.
79. In light of the aforesaid statutory landscape, the Commission proceeds to
determine the relevant market in the instant case.
80. Before delving deep into this aspect, it is apposite to first deal with the
contention raised by the learned senior counsel appearing on behalf of
Google that to attract Section 4(a) of the Act, there has to be purchase or
sale of goods or services. It was argued that in case of online search, there
is no purchase or sale of goods or services as Google provides search
services to users for free and hence, the question of applicability of Section
4 does not arise.
81. Be that as it may, for the present purposes, the Commission is of
considered opinion that the plea of Google that there is no sale or purchase
is not only flawed but altogether ignores the role of big data in the digital
economy.
C. Nos. 07 & 30 of 2012 29
PUBLIC VERSION
82. Coming to the facts of the present case, the Commission has no hesitation
in holding that users offer indirect consideration to Google by: (a)
providing their attention or “eyeballs” to SERP; and (b) allowing Google
to collect and use their information, both of which facilitates generation
of revenues by Google as it attracts more advertisers.
83. The matter can be looked at from another angle also. The Commission
observes that online platforms such as Google, Bing, Yahoo etc. which
provide search services are intermediaries that act as an interface between
search users and advertisers. On the one side of the market are users who
are looking/ searching for information from the World Wide Web and on
the other side of the market are advertisers/ business firms/ merchants who
seek to advertise their goods or services to consumers. Examples of such
market are newspapers, credit card network, video game consoles etc.
Such markets can have two or more sides which are linked by the
intermediary or platform. In the instant case, when a user seeks for some
information that is available on the World Wide Web by accessing a
search engine such as Google and places a requisition for a particular
keyword or phrase, a search request gets initiated on the search side.
Correspondingly, the search engine scans millions of websites or billions
of web pages through its bots or spiders or crawlers and harvests relevant
information and then sorts them based on some criteria which would then
be displayed by the search engine on a dynamically generated SERP.
These results displayed are known as organic results. On the other side of
the market, the platform solicits bids for an auction from advertisers/
business firms/ merchants for display of relevant advertisements that
match with the information sought on the search side and the top three or
four bids received from advertisers are then prominently displayed along
with the organic results in the dynamically generated SERP. The ordering
of advertisements that are displayed against the top three or four positions
would depend on certain criteria based on which the adspace auction is
C. Nos. 07 & 30 of 2012 30
PUBLIC VERSION
designed. The two sides of the market described above complement each
other and they are interdependent. In the absence of one side or if the size
of one side is too small the other side would not exist. A platform that is
able to deliver relevant search results and that is able to attract a large
number of search users would only be able to attract advertisers, apart
from that the platform’s pricing mechanism for its advertisement service
is a crucial one.
84. The Commission observes that it has been contended by Google that the
search services offered by it is free and hence there is no purchase or sale
of goods or services. In markets that are characterised with more than one
side, any market assessment that relies only on the side where the service
offered is free to consumers distorts the true picture and leads to a biased
assessment of the nature of competition in such markets. Whenever any
users places a search requisition for a particular keyword or phrase
through a search engine, the search platform seeks certain information
from such users such as IP address, device information, location,
information regarding Operating System etc. apart from the information
with respect to date and time of search and the keyword or phrase searched
for. The huge volume of such information generated from each and every
search conducted on such platforms constitutes what is known as ‘big
data’ by aid of which search platforms are able to attract advertisers, target
relevant ads and conduct their search business. The submission made by
the learned senior counsel completely misses the role and nature of ‘big
data’ i.e. an aggregate of eyeballs/ choices which is being provided by
users while availing the search services offered by a search engine.
85. It may be noted that rise of new business models based on collection and
processing of Big Data is currently shaping the world. With the
development of data mining and machine learning, businesses are able to
offer innovative, high-quality and customised products and services at low
C. Nos. 07 & 30 of 2012 31
PUBLIC VERSION
or even zero prices, with great gains for consumers. At the same time, it
cannot be denied that the benefits of providing Big Data do not come
without a cost. Consumers may be increasingly facing a loss of control
over their data and are exposed to intrusive advertising and behavioural
discrimination.
86. In fact, it would not be out of place to equate data in this century to what
oil was to the last one. The Commission is not oblivious of the increasing
value of data for firms which can be used to target advertising better.
Moreover, the data can be turned into any number of revenue generating
artificial-intelligence (AI) based innovations.
87. The Commission notes that while the aforesaid data is collected from the
search side of the market, on the other side also some crucial events
happen once an SERP is displayed. The search user by clicking on the
results displayed reveals her choice and preferences and the most
important aspect to be noted is the link on which the search user clicks. If
the search user clicks on an organic result then the search engine lands the
user in the appropriate webpage and the platform correspondingly does
not receive any monetary payment for its services from the advertisement
side. But on the contrary, if the search user clicks on the advertisements
displayed then the search engine lands the user on the advertisers/ business
firms/ merchants website and correspondingly the search engine receives
a payment according to the bid price consummated in the adspace auction
for its advertisement services which depends on the pricing mechanism
that is pay per click. The revenue earned by search platforms’ through
provision of search based adservices bears testimony to not only to the
potential of adservices offered by them but also negates the view that
search services offered by such platforms are free. In view of the above,
the Commission disagrees with the contention raised by Google that in
case of online search there is no purchase or sale of goods or services and
consequently holds that online search falls within the ambit of Section 4
C. Nos. 07 & 30 of 2012 32
PUBLIC VERSION
of the Act.
88. The Commission notes that it is not unusual for one-side in a multi-sided
market to receive services subsidised by customers on the other side of the
market. Several mobile applications and websites work through an
advertiser funded model and free-to-air television channels are also based
solely on advertising revenue. This, however, is not suggestive of the fact
that users are not providing any consideration for availing these products
and services. In such cases also, a commercial relationship exists and the
conduct of the participants in such commercial relationships can be
examined within the four corners of the Act.
89. Having, thus, disposed of the preliminary objection raised by Google, the
Commission now proceeds to examine the relevant market in the present
case.
90. First of all, the Commission notes that on the question of substitutability
of online general web search services and search advertising services, the
DG concluded that online general web search services and search
advertising did not constitute the same relevant product market on account
of wide variations in the mechanism for generation and display of results
and also the clicking behaviour. Further, the DG noted that these services
serve distinct goals and are perceived differently by the various categories
of users, namely, publishers (websites) and internet users entering search
queries. The DG also noted that these services constitute complementary
services from the point of view of websites striving for eyeballs.
91. Further, the DG also considered the issue concerning substitutability of
direct search option by typing the URL of a particular website in the
internet browser with online general web search services. The DG
acknowledged the limitations associated with the usage of the direct
search option as it requires the users to not only be aware of the websites
C. Nos. 07 & 30 of 2012 33
PUBLIC VERSION
that offer the relevant information but also requires the users to remember
the specific URLs. The DG also noted that given the large number of
websites that are in existence coupled with the voluminous nature of
information available on the webpages, it is virtually impossible for users
to be aware of all websites that might be able to provide them with the
desired information and their respective URLs. In fact, most of the users
might not remember more than a handful of URLs of the websites.
Additionally, internet users may prefer a general web search to explore
the vast alternative online sources of information. On this basis, the direct
search option involving the URL has not been found to be substitutable
with a general purpose web search option.
92. The DG also considered if online general web search services are
interchangeable or substitutable with online specialised search services.
While general purpose search engines allow internet users to search
information on a wide range of topics, specialised search services permit
online searches for information limited to particular topics or areas such
as news, shopping, travel, entertainment, etc. Further, in response to a
search query, general purpose web searches show information from across
the web while specialised search results yield information from a limited
source, i.e., either its own contents or from the contents of certain
specified websites. Additionally, pricing and registration requirements
stipulated by general purpose online searches and specialised searches are
also different. Accordingly, the DG has concluded that online general web
search services are not substitutable with site-specific search and
specialised search services as there are variations in terms of their
characteristics, intended use, price etc.
93. The Commission finds no reason to differ with the analysis of the DG and
agrees that online general web search services cannot be substituted with
direct search option by typing URL of websites in the internet browsers.
C. Nos. 07 & 30 of 2012 34
PUBLIC VERSION
Users may not be aware of URLs of all websites that offer the information
they are searching or looking for. In these circumstances, search engines
become the first port of call for a user looking for information online. Any
comparison of general web search service with direct search option would
be thoroughly misplaced. Further, the Commission notes that online
general web search services cannot be equated with specialised search
services.
94. Resultantly, the Commission holds online general web search services to
be a distinct relevant product market in accordance with the provisions of
Section 2(s) read with Section 19(7) of the Act.
95. Next, the DG examined if the consumers (who use advertising services)
regard online and offline advertising as substitutable. On a detailed
analysis, the DG noted that these are neither substitutable nor part of the
same relevant market.
96. The DG has considered the substitutability of online search advertising
and online non-search advertising as well from consumer’s (i.e., user of
advertising services) perspective and noted that one form of online
advertising does not serve as a replacement for the other and the two are,
therefore, complementary in nature.
97. The Commission is of the considered opinion that online and offline
advertising services are not comparable. Online advertising is undertaken
using internet as a medium and, hence, its coverage is largely dependent
on reach of the internet. Similarly, online advertising is not substitutable
for newspapers, radio or television for advertisers who seek to target areas
or user groups with limited internet access. As pointed out by the DG,
advertising rates are significantly lower for online advertising in
comparison to traditional media. Not only that, online advertising allows
advertisers to accurately monitor the effectiveness of the advertisement on
C. Nos. 07 & 30 of 2012 35
PUBLIC VERSION
the basis of actual number of users that it reaches whereas for offline
advertisements, advertisers rely on estimated number of views and not the
actual views.
98. Adverting to the issue of search and non-search advertising, the
Commission notes that search advertising helps advertisers in targeting
specific users. Typically, search advertisements are used for demand
fulfilment while non-search advertisements are for brand awareness or
recognition. For that reason, both the advertisements are priced using
different pricing mechanisms. For example, search advertisements are
generally paid on a cost-per-click basis, while non-search advertisements
are usually paid on a cost-per-thousand-impressions basis. Thus, the
characteristics, intended use and price of search and non-search
advertising are different from one another.
99. In view of the aforesaid, the Commission holds that online search
advertising services to be a distinct relevant product market in accordance
with the provisions of Section 2(s) read with Section 19(7) of the Act.
100. On relevant geographic market in the context of online general web search
services, the DG noted that local specification requirements, language
differences and consumer preferences of users are relevant factors for the
supply of online web search services to them. Supply is also function of
the legislative framework of a country. Therefore, conditions for supply
and demand of online web search services in India are distinct from those
prevailing in other areas. Accordingly, India has been considered as the
relevant geographic market for online general web search services in
accordance with the provisions of Section 2(t) read with Section 19(6) of
the Act. The relevant geographic market in any event could not have been
taken as global since from a plain reading of the Explanation to Section 4
of the Act, ‘dominant position’ means a position of strength, enjoyed by
C. Nos. 07 & 30 of 2012 36
PUBLIC VERSION
an enterprise, in the relevant market, in India, which enables it to operate
independently of competitive forces prevailing in the relevant market or
affect its competitors or consumers or the relevant market in its favour.
Since the Commission is concerned with the market in India, it holds the
relevant geographical market in case of online general web search services
to be limited to India.
101. Similarly, in respect of online search advertising services also, the DG
considered the conditions for demand for online search advertising
services and those for supply of online search advertising in terms of
legislative framework, presence of local distribution entities and
variations in applicable terms and conditions etc. and held India to be the
relevant geographic market for online search advertising services in
accordance with the terms of Section 2(t) read with Section 19(6) of the
Act. For the reasons noted earlier, the Commission agrees with the DG’s
determination.
102. In the result, the Commission determines the following two relevant
markets in the present case for examining the alleged abusive conduct of
Google:
(a) Market for Online General Web Search Services in India;
(b) Market for Online Search Advertising Services in India.
Assessment of Dominance
103. After having delineated the relevant market(s), the Commission proceeds
to assess Google’s dominance in the same. It may be noted that by virtue
of explanation (a) to Section 4 of the Act, ‘dominant position’ means a
position of strength enjoyed by an enterprise in the relevant market in
India which enables it to operate independently of competitive forces
C. Nos. 07 & 30 of 2012 37
PUBLIC VERSION
prevailing in the relevant market; or to affect its competitors or consumers
or the relevant market in its favour.
104. Further, to analyse dominance, the factors enumerated in Section 19(4) of
the Act are to be considered, namely, market share of the enterprise; size
and resources of the enterprise; size and importance of the competitors;
economic power of the enterprise including commercial advantages over
competitors; vertical integration of the enterprises or sale or service
network of such enterprises; dependence of consumers on the enterprise;
monopoly or dominant position whether acquired as a result of any statute
or by virtue of being a Government company or a public sector
undertaking or otherwise; entry barriers including barriers such as
regulatory barriers, financial risk, high capital cost of entry, marketing
entry barriers, technical entry barriers, economies of scale, high cost of
substitutable goods or service for consumers; countervailing buying
power; market structure and size of market; social obligations and social
costs; relative advantage, by way of the contribution to the economic
development, by the enterprise enjoying a dominant position having or
likely to have an appreciable adverse effect on competition; or any other
factor which the Commission may consider relevant for the inquiry.
105. The DG has assessed the question of Google’s dominant position in both
the relevant markets identified in terms of Explanation (a) to Section 4 of
the Act after a detailed analysis of the above-stated factors enumerated in
Section 19(4) of the Act.
106. The DG has found Google to be a dominant enterprise in both the relevant
markets for online general web search services and online search
advertising services in India.
C. Nos. 07 & 30 of 2012 38
PUBLIC VERSION
Google’s Response
107. Google denied its dominance in the relevant market and submitted that
regardless of market definition, Google is not dominant in online search
services for the following reasons:
(i) Google faces substantial competitive constraints in each query
category
An empirical analysis of the search alternatives available to Indian
users reveals that Google is just one search option among many
popular providers. In fact, in product search, local search, and
travel search, Google has usage shares of only , , and
respectively. These low shares preclude dominance in online
search market.
(ii) Usage shares are not, in any event, a proxy for market power over
quality and innovation
Further, usage shares of a free service, in any event, are not a proxy
for market power over quality and innovation. This is because
usage shares – unlike shares of sales – do not reflect investment or
commitment by users. It is very easy for users to shift to another
service provider. Usage shares, therefore, do not indicate whether
an enterprise can “behave independently” within the meaning of
Section 4 of the Act.
(iii) Google’s high innovation rates exclude dominance
Because search is free, the Investigation Report correctly
recognises that the indicator for dominance is “whether Google
would lose [usage] share materially if it were to reduce
innovation”. Google cannot plausibly reduce its innovation rates
without users at the margin switching to rival services. This is
borne out by Google’s consistently high innovation rates. The
C. Nos. 07 & 30 of 2012 39
PUBLIC VERSION
Investigation Report acknowledges that Google’s “significant
innovations […] have helped users all over the world”, and CUTS
also recognises Google’s “tremendous record in innovation”.
(iv) The Investigation Report ignores the constraint from user
switching
The fact that users can switch to rivals with a single click provides
an additional reason why Google cannot act independently of
market forces. The Investigation Report dismisses such constraint
from switching of users by pointing to Google’s high usage share
in general search. However, this conflates users’ ability to switch
with their incentive to do so. Multihoming evidence shows that
users in India often try search services other than Google as well –
with of Google’s users visiting another general search
service in a single month. If users try other general search services,
yet chose to return to Google, this can only demonstrate that
Google is successfully competing by offering innovative and high-
quality service showing the most relevant results.
(v) None of the “other factors” under Section 19(4) of the Act give
Google market power
None of the other factors the Investigation Report relies on can
establish Google’s dominance in search. First, Google’s size and
resources do not distinguish it from its rivals, such as Microsoft,
Yahoo!, and Amazon. Second, search is not subject to direct or
indirect network effects. Third, returning high-quality search
results is largely dependent on technological capabilities unrelated
to query scale; the benefits of scale are subject to significant
diminishing returns. Fourth, search is not characterised by
substantial barriers to entry and expansion – particularly given
recent developments in cloud computing and open-source
C. Nos. 07 & 30 of 2012 40
PUBLIC VERSION
software. Fifth, Google is not vertically integrated in a way that
conveys market power in search. Sixth, users and websites do not
“depend” on Google – as shown by traffic source data contained
in the Investigation Report itself.
108. Similarly, it was submitted on behalf of Google that regardless of market
definition, Google is not dominant in advertising as well. It was argued
that the Investigation Report ignores the competitive pressure Google
faces when seeking to attract advertising revenue, erroneously dismisses
the constraint from user multihoming, and overstates barriers to entry:
(i) The Investigation Report ignores extra-market constraints
The Commission recently found in Cloudwalker that online
advertising may compete with offline advertising, even if they
might not be in the same market. Similarly, the EU Commission
held in Google/DoubleClick that search and non-search advertising
constrain each other. Even if Facebook is not considered to be a
part of the same online advertising market as Google, it cannot
reasonably be ignored in a competitive analysis of online
advertising in India. Given Facebook’s growth and strength in
online advertising, it is not plausible that Google can operate
independently of competitive forces.
(ii) The Investigation Report ignores constraint from advertiser
switching
Advertisers frequently multihome between different advertising
platforms. In September, 2015, advertisers accounting for of
paid clicks on Google also advertised on Bing/Yahoo!. This
multihoming precludes dominance because if Google tries to raise
prices or reduce quality of its ads, Google would lose a substantial
proportion of its advertisers (who have easy access to rival ad
C. Nos. 07 & 30 of 2012 41
PUBLIC VERSION
platforms).
(iii) Barriers to entry are not material
The Investigation Report overstates the relevance of capital costs
and scale as evidence of barriers to entry, which are no greater than
in other technological industries. The Investigation Report’s
references to scale and to the need for a “critical mass” of
advertisers are erroneous. Advertisers can run multiple campaigns
through publishing platforms, placing smaller proportion of their
advertising budget with a new entrant than with established players.
If a new entrant has an innovative product and succeeds in
generating a good ROI, marketers will be driven to place greater
proportions of their budget with the new entrant, and over time it
will grow.
The Informant’s Response
109. The Informants, however, supported the DG’s findings on dominance.
The Commission’s Analysis
110. The Commission has carefully examined the issue of dominance and
considered the rival submissions on this point.
Market Share of the Enterprise
111. In coming to its conclusion on the market share of Google in the relevant
markets of online general web search and online search advertising in
India, the DG has taken into account: (a) volume of search business; and
(b) total revenues generated in India, as the basis of estimation.
Google’s market share in online general web search services in India
112. The DG has considered the market share data (at pan-India level) on the
basis of: number of searches; and number of search result page views; to
C. Nos. 07 & 30 of 2012 42
PUBLIC VERSION
conclude that Google has consistently maintained a significant market
share of more than during 2009 to 2014 in the relevant market of
Online General Web Search in India. This has been the case despite the
long standing presence of other competitors like Yahoo! and Microsoft
Bing and market entry of new players such as Ask.
113. The DG has also noted that even on a worldwide basis, Google has
consistently held around market share during 2010 to 2014 among
all search engines offering online web search services despite the presence
of several other online general search service providers.
114. The Commission notes that though the Act does not contemplate a market
share threshold beyond which dominance is presumed, it is critical to
observe that irrespective of the metric applied to assess market share,
Google has the highest share, and this share is exponentially greater than
its nearest competitor. Market share of Google in excess of on
various metrics coupled with very low market shares of competitors,
clearly evidences that Google enjoys indisputable dominance in the
relevant market.
Google’s market share in Online Search Advertising Services in India
115. After considering the revenues earned by Google from search advertising
services under both the AdWords and the AdSense programmes, the DG
concluded that Google’s market share has been more than during
the period of investigation in the relevant market for online search
advertising services in India.
116. Additionally, the DG dealt with Google’s submissions on market share in
the context of different relevant markets proposed by it. It may be noted
that the DG considered Google’s claims relating to its market shares in:
(a) online web search services, including, vertical/site searches and social
networking site searches; and (b) online advertising services including
C. Nos. 07 & 30 of 2012 43
PUBLIC VERSION
advertisement services on mobile applications, and has recorded sound
and cogent reasons for rejecting such submissions of Google.
117. Further, the DG also considered the case laws referred to by Google in its
submissions and has stated that there is no denial of the fact that high
technology services demand continuous innovation. In the present
context, the question is whether Google would lose market share
materially if it were to reduce innovation. Given high barriers to entry and
Google’s insurmountable scale advantage, the DG held that it is unlikely
that a large number of users would switch to a competing search engine in
a short or medium term if Google reduces the quality and innovation of its
services. Further, in a multi-sided market, effect of marginal switch of
users on one side of the market (like search service) may not lead to a
switch on other sides (like search advertising).
118. The Commission observes that while high technology services demand
continuous innovation, given barriers to entry and Google’s scale
advantage, it is unlikely that a large number of users would switch to a
competing search engine in the short or medium term.
119. Thus, it is clear that the market conditions, including barriers to entry and
Google’s scale, reinforce the effect of Google’s market shares in
establishing Google’s dominance in the online general web search and
search advertising markets. These factors have been analysed by the DG
in its Report and the same are noted in the succeeding paras.
Analysis of other factors for assessing Google’s dominance
120. The DG noted that the market shares of Google and its competitors show
that none of Google’s competitors has so far been able to match Google’s
market strength either in general purpose online web search services or
online search advertising services in India. Further, the DG rejected
Google’s claim that Facebook is one of its main competitors in the online
C. Nos. 07 & 30 of 2012 44
PUBLIC VERSION
advertising market owing to its strong presence in the display
advertisements segment.
121. As the relevant market has been defined by the DG to include online
search advertising services only, and not online advertising services
generally, Facebook and other entities that are engaged in non-search
advertising are not Google’s competitors in the search advertising market.
Further, Facebook has also explicitly stated that it does not engage in
online search advertising business. Not only this, the overall revenues
earned by Facebook from online advertising in India are substantially
lower than that of Google.
122. The Commission observes that for a search engine, it is extremely
important to be able to crawl the web and index the data. Google has a
significant head start in this regard, and the cost of crawling the entire
internet, in terms of servers and technology, is prohibitive for a new
entrant. As Google has an insurmountable scale advantage and given that
only market participants in the online general web search market can
compete in the search advertising market, the barriers in the online general
web search market also effectively restrict entry into the search
advertising market.
123. Additionally, the DG has acknowledged the competitive advantage
enjoyed by Google on account of vertical integration and takes note of the
critical position of Google as a platform for users, publishers and
advertisers. The DG has also noted that the technology based market
structure and size of technology industry confer upon Google a position
of competitive strength.
124. The DG took note of the absence of countervailing buyer power in the
market for search advertising services. In this regard, the Commission
notes that due to dependence of its customers, as well as the disparity in
C. Nos. 07 & 30 of 2012 45
PUBLIC VERSION
size and resources between the average user and advertiser in comparison
to Google, none of them has countervailing buyer power. This is true for
the largest companies that advertise on Google. They might be of a similar
economic stature generally, but they are highly dependent on Google for
search advertising and, therefore, do not exercise any buying power over
Google.
125. The Commission acknowledges the fact that Google has revolutionised
the manner in which users access information on the Internet and agrees
with the DG that the market strength has been acquired by Google over a
period of time. In the high technology markets, innovation is key and in
multi-sided markets, market shares should be transient. However,
Google’s market shares have been consistently high, which suggests that
it has got other advantages, besides technical advantages, which insulate
its market position. The structure of the market is both indicative of and
conducive to Google’s dominance. In view of this, the Commission has
no hesitation in holding that Google is dominant in both the relevant
markets i.e. market for online general web search services and market for
online search advertising services in India.
Abuse of Dominant Position
Procedural Deficiencies
126. Before examining the Investigation Report and the findings contained
therein in respect of the alleged abusive conduct of Google, it would be
appropriate to deal with the procedural deficiencies highlighted by Google
in its response.
127. Google has argued that the DG’s investigation was deficient on procedural
grounds and thereby provides an additional reason in itself as to why the
Investigation Report cannot form the basis of an infringement finding
C. Nos. 07 & 30 of 2012 46
PUBLIC VERSION
against Google.
128. It was submitted that the DG’s market consultation was flawed. It failed
to consult a representative cross-section of affected participants, and it
asked leading questions to the witnesses “formulated with the sole object
of finding fault” with Google contrary to the principles set out in Tamil
Nadu Film Exhibitors Association v. Competition Commission of India,
Appeal No. 19 of 2014, decided on 28.04.2015 by the Hon’ble
Competition Appellate Tribunal (COMPAT) whereby the order of the
Commission against the appellant was set aside inter alia on the ground
that the DG asked questions which were “clearly loaded against the
appellant” that prompted the respondent to support the Informant’s legal
theories.
129. Further, it was pointed out that the Investigation Report “cherry picks” the
evidence. It does not properly assess the evidence in this case. It relies
almost exclusively on submissions of parties with an obvious commercial
motive to obstruct Google’s business. It ignores the evidence that the DG
collected from third parties and Google that is exculpatory of Google. The
DG failed to engage with Google on the complainants’ allegations, which
was especially problematic because these allegations concerned Google’s
technologies. The Investigation Report exacerbates these deficiencies by
citing the draft EU commitments as if they were evidence of infringement
in India, while at the same time disregarding decisions by courts and
authorities in the UK, Germany, Brazil, Taiwan, Canada, Egypt, and the
USA, finding that Google’s conduct is procompetitive.
130. The Investigation Report relies on mere hypothetical speculation. It was
pointed out that the DG is tasked with a fact-finding exercise to determine
whether Google in fact abused a dominant position. However, the DG has
failed to fulfil this duty. Instead, it identified concerns based on “mere
conjecture and imagination” and “possibilities”.
C. Nos. 07 & 30 of 2012 47
PUBLIC VERSION
131. The Investigation Report’s conclusions are based on information to which
Google has not been granted access. Google has not been granted access
to critical aspects of the Investigation Report. Google cannot understand
the precise information relied upon against it, understand the Investigation
Report’s reasoning, or verify the conclusions that the DG seeks to draw
from this information. Without access to such information, any
infringement decision against Google would be contrary to the principles
of natural justice and the jurisprudence of the Hon’ble COMPAT.
132. Lastly, it was argued that the DG has exceeded its statutory authority. It
was argued that the Investigation Report may not be used to support an
infringement finding against Google in relation to conduct or issues that
the DG had no jurisdiction to investigate or determine. Specifically, the
Investigation Report cannot be used as the basis of an infringement
finding: (i) in relation to investigated issues that were not included in the
Informations or Section 26(1) Order, (ii) against Google Ireland Limited,
and (iii) in relation to matters of Intellectual Property law.
133. It was submitted that an unwarranted expansion of the scope of
investigation is ultra vires and illegal. This principle applies to both the
DG and the Commission. Referring to a decision of the Hon’ble
COMPAT in Andhra Pradesh Film Chamber of Commerce v.
Competition Commission of India, COMPAT Order dated 14.10.2015 in
Appeal No. 15 of 2013, it was argued that the Hon’ble COMPAT ruled
that the DG has no jurisdiction to deal with issues that are not raised in the
Information originally filed before the Commission or in the
Commission’s referral order. This principle was confirmed in Grasim
Industries Ltd. v. Competition Commission of India, W.P (C) No. 4159 of
2013.
134. It was contended that Section 26(1) of the Act provides that the DG’s
authority to investigate is exclusively conferred by the Commission.
C. Nos. 07 & 30 of 2012 48
PUBLIC VERSION
Regulation 20(4) of the Competition Commission of India the (General)
Regulations, 2009 (General Regulations) also does not permit the DG to
add issues of its own accord and thereby expand the scope of the
investigation. Therefore, it was argued by Google that any allegations that
form part of the unwarranted expansion of scope are ultra vires and
findings on such matters are not sustainable.
135. Specifically, it was pointed out by Google that the DG expanded the scope
of investigation by investigating multiple issues that did not form a part
of these prima facie orders (or the Informations). These issues include:
(i) Google’s use of third-party content in Google search results
(content use or scraping).
(ii) Google’s treatment of one particular AdSense partner
(Octathorpe).
(iii) Google’s enforcement of its AdWords policies in relation to a
specific category of advertisers (technology support
advertisers).
136. It was submitted that any findings on these issues are without jurisdiction
and may not form part of any infringement finding against Google.
137. It was next contended that the Investigation Report cannot be used to
support an infringement finding against Google Ireland Limited. The
Commission’s prima facie orders dated April 03, 2012 (Consim) and June
20, 2012 (CUTS) were directed solely against Google Inc. and Google
India Private Limited. There is no provision for the Commission to have
ordered an investigation against Google Ireland Limited (“Google
Ireland”) without first making a prima facie determination that Google
Ireland has contravened Section 4 of the Act (i.e., without first issuing a
Section 26(1) order).
C. Nos. 07 & 30 of 2012 49
PUBLIC VERSION
138. In contravention of the provisions of the Act, the DG requested the
Commission by a letter dated September 1, 2014, that Google Ireland be
added as a party to the investigation. Google has not, at any stage during
the investigation, had access to a copy of this letter. The Commission
agreed with the DG’s request in a meeting held on September 3, 2014.
139. Neither the Commission nor the DG has communicated reasons for the
decision to join Google Ireland which are sufficient to establish that
Google Ireland had prima facie infringed Section 4 of the Act. Without
answering that question, there can be no infringement finding against
Google Ireland and any penalty imposed upon Google must not include
Google Ireland’s revenues for computation.
140. Further, it was canvassed that the Investigation Report cannot be used to
support a finding on trademark law. The Commission and the DG are
charged with the application and enforcement of competition law only.
They have no statutory or common law jurisdiction to assess and apply
intellectual property laws. Yet the Investigation Report makes findings in
relation to trademark law. The Investigation Report and the DG’s
assessment of trademarks cannot form part of an infringement decision
against Google. This is particularly so because courts of competent
jurisdiction (the Hon’ble Madras High Court) is currently seized of
exactly the same legal question and any finding on trademark law by the
Commission facilitates forum shopping and creates the risk of
irreconcilable judgments and conflict of laws.
141. Lastly, an objection was taken that the Investigation Report has been
issued in contravention of the requirements of the Act and is therefore
ultra vires.
142. Elaborating, it was contended that the DG Investigation Report is ultra
vires the Act. Section 26(3) of the Act statutorily obliges a specific
C. Nos. 07 & 30 of 2012 50
PUBLIC VERSION
persona designata, the DG to whom the Commission may refer a
complaint under Section 26(1) and who is entrusted under Section 26(3)
to submit an Investigation Report containing his findings. No other
persons are conferred with these duties. Section 16 of the Act provides the
manner, mode, and qualifications for appointment of DG and other
officers in the DG’s department. Section 16(2) makes clear that the DG’s
department may discharge its statutory functions through its staff
(including Additional Director General, Joint Director General, Deputy
Director General) provided that those staff are under the “general control,
supervision and direction” of the DG. It is well established that when the
statute requires the Investigation Report to be submitted by a persona
designata, only he/ she can discharge that function. It was alleged that in
this case, Ms. Jyoti Jindgar, Additional DG, signed the Investigation
Report Google received on 24 August, 2015. There is no evidence that the
DG was involved in the investigation or applied his mind. There has thus
been an illegal delegation of power under Section 26(3) of the Act to a
person not entrusted with such power under law and not recognised by the
Act as being able to exercise and execute that power. It was submitted that
this assumed added significance since the DG’s Investigation Report,
though not binding on the Commission, has caused serious prejudice to
Google, including publicly, and can result in Google being visited with
grave and irreparable consequences.
143. Based on these procedural defects, it was submitted that the Commission
may not safely rely on the Investigation Report and must close the inquiry
against Google.
The Commission’s Analysis
144. The Commission has examined carefully each of the procedural
objections taken by Google.
C. Nos. 07 & 30 of 2012 51
PUBLIC VERSION
145. At the outset, it may be noted that the DG conducted a wide-ranging
investigation against Google examining its various business aspects.
Some of these aspects were not even subject matter of information filed
by the Informants. It appears that during the course of investigation,
several third parties provided information to the DG pursuant to the probe
letters issued by the Office of the DG. In their submissions, the parties
apart from furnishing the information as sought for by the Office of the
DG, have made other allegations against Google. The DG has investigated
such allegations and returned findings on them. This is manifested in the
Investigation Report on the following counts:
(i) Google’s use of third-party content in Google search results (content
use or scraping).
(ii) Google’s treatment of one particular AdSense partner (Octathorpe).
(iii) Google’s enforcement of its AdWords policies in relation to a
specific category of advertisers (technology support advertisers).
146. Google has objected to the DG enlarging the scope of investigation,
however, the Commission is of the opinion that scope of the investigation
cannot be unduly restricted and the DG may be justified in looking and
examining the aspects which have not been specifically directed to be
investigated by the Commission. It needs no reiteration that the purpose
of an investigation is to cover all necessary facts and evidence in order to
see as to whether there are any anti-competitive practices adopted by the
persons complained against. For this purpose, as held by the Hon’ble
Supreme Court of India in Excel Crop Care Limited v. Competition
Commission of India & Anr., Civil Appeal No. 2480 of 2014 decided on
08.05.2017 that “…the starting point of inquiry would be the allegations
contained in the complaint. However, while carrying out this
investigation, if other facts also get revealed and are brought to light,
C. Nos. 07 & 30 of 2012 52
PUBLIC VERSION
revealing that the ‘persons’ or ‘enterprises’ had entered into an
agreement that is prohibited by Section 3 which had appreciable adverse
effect on the competition, the DG would be well within his powers to
include those as well in his Investigation Report. Even when the CCI forms
prima facie opinion on receipt of a complaint which is recorded in the
order passed under Section 26(1) of the Act and directs the DG to conduct
the investigation, at the said initial stage, it cannot foresee and predict
whether any violation of the Act would be found upon investigation and
what would be the nature of the violation revealed through investigation.
If the investigation process is to be restricted in the manner projected by
the appellants, it would defeat the very purpose of the Act which is to
prevent practices having appreciable adverse effect on the competition.
We, therefore, reject this argument of the appellants as well touching upon
the jurisdiction of the DG”.
147. Having said that, the Commission is not giving any opinion on them. They
were not specifically alluded to in the information and the parties did not
address them during final hearing.
148. The Commission also notes that Google has taken objection to joining of
Google Ireland as party to the case. Yet nothing turns upon such a plea as
such a course was clearly sanctioned by the order of the Commission. The
further plea of Google that the Commission did not provide any reasons
while allowing the DG’s request to join Google Ireland as a party, is also
misplaced. It may be noted that the DG clearly provided reasons in its
request to join Google Ireland as a party and the same were duly
considered by the Commission. There is no requirement under the law to
pass separate detailed orders at such intermediate stage when the
Commission has already issued prima facie orders detailing the abusive
conduct for investigation. Joining of such a party actually relates back to
the stage of prima facie orders and it would be futile to record detailed
C. Nos. 07 & 30 of 2012 53
PUBLIC VERSION
reasons therefor. In fact, the Hon’ble Supreme Court of India in the
Competition Commission of India v. Steel Authority of India Limited &
Ors., Civil Appeal No. 7779 of 2010 decided on 09.09.2010 has clearly
laid down that neither any statutory duty is cast on the Commission to
issue notice or grant hearing, nor any party can claim, as a matter of right,
notice and/or hearing at the stage of formation of opinion by the
Commission, in terms of Section 26(1) of the Act that a prima facie case
exists for issuance of a direction to the DG to cause an investigation to be
made into the matter.
149. Thus, when the Commission is not obligated to issue notice or grant
hearing at the stage of formation of prima facie opinion, it can be said that
the Commission is much less required to record detailed reasons while
joining a party to an ongoing investigation. In the result, the Commission
finds no merit in the contention of Google that the DG had no jurisdiction
to investigate Google Ireland Limited in the instant case.
150. As for the other alleged procedural deficiencies highlighted by Google i.e.
the market consultation by the DG was flawed, the Investigation Report
cherry picks the evidence and the Investigation Report is based upon
hypothetical speculations, the Commission is afraid that the alleged
deficiencies touch upon the merits of the case and cannot be termed as
mere procedural deficiencies in any way. These would be taken into
consideration while examining the matter on merits.
151. Further, the Commission finds objection of Google that it has not been
granted access to certain critical aspects of the Investigation Report, quite
egregious. In this regard, the Commission notes that Google was granted
access to the confidential information of the parties after following an
elaborate procedure.
C. Nos. 07 & 30 of 2012 54
PUBLIC VERSION
152. Lastly, the Commission notes that the plea raised by Google that the DG’s
Report is ultra vires the Act as Section 26(3) of the Act statutorily obliges
a specific persona designata, namely, the DG, to whom the Commission
may refer a complaint under Section 26(1) and who is entrusted under
Section 26(3) to submit an Investigation Report containing its findings, is
of no consequence. In the present case, the investigation was done and the
Investigation Report was submitted and signed by Additional DG. It was
therefore alleged that there has thus been an illegal delegation of power
under Section 26(3) of the Act to a person not entrusted with such power
in law and not recognised by the Act as being able to exercise and execute
that power.
153. The plea is ex facie frivolous. In terms of the provisions contained in
Section 26(1) of the Act, the Commission may direct the DG to cause an
investigation to be made into the matter. The term “Director General” has
been defined in Section 2(g) of the Act as meaning the Director General
appointed under sub-section (1) of Section 16 and including any
Additional, Joint, Deputy or Assistant Directors General appointed under
that section. Thus, it is apparent that while the Commission may issue
direction to the DG to conduct an investigation, the DG can, in turn, mark
the case to any of the competent officers in his office such as Additional
DG, Joint DG, Deputy DG or Assistant DG subject to his general control,
supervision and direction. This is borne out from the scheme of Section
16 of the Act.
154. In view of the above discussion, the Commission finds no merit in the
procedural deficiencies alleged by Google.
155. At this stage, the Commission would like to address one more plea which
was urged by the learned senior counsel appearing on behalf of Google
and which has a bearing upon the findings recorded by the DG on multiple
C. Nos. 07 & 30 of 2012 55
PUBLIC VERSION
counts. It was submitted that on many occasions, the DG while recording
its findings was greatly swayed by the draft voluntary commitments made
by Google before various authorities globally. It was submitted that such
a course was impermissible in law for the DG to have adopted. The
learned senior counsel cited various statutory provisions and the decisions
in support of the plea.
156. The Commission has examined the plea and is of the considered opinion
that reliance by the DG upon voluntary commitments given by Google
elsewhere was against law and public policy. It may be observed that if
voluntary concessions offered by the parties in separate proceedings and
different jurisdictions in different settings are referred to and relied upon
by the investigators in another jurisdictions to base findings of
contraventions, the consequences would be enormous. On the one hand,
it may unduly prejudice the cause of such party and thereby may have the
potential to vitiate the findings. And on the other hand, such an approach
would run contrary to public policy as the parties would be deterred from
making even reasonable concessions to avoid a costly and protracted
litigation.
157. A brief survey of case law on the issue would also be appropriate.
158. In Wilson v. Howard, (1878) ILR 4 CAL 231, Hon’ble High Court of
Calcutta observed:
Then, as regards the letter itself, upon which the Learned Judge in
the court below has laid so much stress, it is perfectly true that it
was a very improper thing for the defendants’ attorneys to use a
letter in evidence which was written without prejudice, and
obviously in the course of negotiation between the attorneys on
both sides for an amicable adjustment of the plaintiff’s claim
[Para 11].
C. Nos. 07 & 30 of 2012 56
PUBLIC VERSION
Communication such as these are clearly inadmissible in
evidence. They are excluded on grounds of public policy and
convenience; and the rule of law which excludes them is as binding
upon the arbitrators as upon Courts of Justice, notwithstanding
Section 1 of the Evidence Act (see Taylor on Evidence, 7th edition,
Section 795, and the authorities cited therein. One is only
surprised that a rule, so well-known amongst professional men,
should have transgressed in this instance by the defendants’
attorney [Para 12].
159. Similarly, the Hon’ble Bombay High Court in Sanjay Kumar Agarwal v.
Central Bank of India, 2016 SCC Online BOM 10368 held:
“… The correspondence written by the plaintiff are all on without
prejudice basis. Therefore, in my view, we cannot even place
reliance on such documents to conclude that there was a
concluded understanding.” [Para 15]
160. Lastly, the decision of the United Kingdom House of Lords in Rush &
Tompkins Limited. v. Greater London Council and Others, [1988] 3 All
ER 737 may be noted. In this case, the House of Lords was considering
the settlement negotiations between two parties wherein one of the parties
sought discovery of documents that came into existence for the purpose
of settling the dispute between the parties. It held:
…the “without prejudice rule” is a rule governing the
admissibility of evidence and is founded upon the public policy of
encouraging litigants to settle their differences rather than litigate
them to a finish… the rule applies to exclude all negotiations
genuinely aimed at settlements whether oral or in writing from
being given in evidence.. Nearly all the cases in which the scope
of without prejudice rule has been considered concern the
admissibility of evidence at trial after negotiations have failed. In
C. Nos. 07 & 30 of 2012 57
PUBLIC VERSION
such circumstances no question of discovery arises because the
parties are well aware of what passed between them in the
negotiations.
161. In view of the above, the Commission is of the considered opinion that no
reference can be made or reliance placed upon the voluntary draft
commitments given by the parties in different jurisdictions and in different
statutory architecture to base conclusions of contraventions in the instant
matter lest the same prejudice or vitiate the final determination.
Accordingly, any reference or reliance made by the DG to such draft
voluntary commitments given by Google before different Authorities or
Agencies or Courts in the Investigation Report shall stand deleted from
records. Each of the findings of the DG would be independently assessed
and examined by the Commission based on other material available on
record.
Description of Key Terms
162. Before framing the issues which fall for consideration in the present
inquiry, it would be apposite to describe a few key words/ terms which
would be used throughout the decision. These key terms have been
explained by Google in its response and the same are noted below for easy
reference. It is made clear that any qualitative comment provided by
Google while explaining and describing such words shall have no bearing
upon the merit of the case and its alleged abusive conduct shall be assessed
on merits separately and independently.
Online search services
163. An online search service (or search engine) is a service that searches for
information on the World Wide Web and responds to user’s search
queries. General search services and vertical search services are types of
online search service:
C. Nos. 07 & 30 of 2012 58
PUBLIC VERSION
General search service
General search services, like Google, seek to answer queries for
all types of information, such as products, local businesses,
recipes, historical figures etc.
Vertical search service
Vertical search services focus on one (or more) categories of
information, such as products, local businesses, or travel.
Search results
164. Search services return search results in response to users’ queries. They
typically use complex algorithms to find relevant search results from
among the trillions of different webpages. There are two main types of
search results:
Free search results
Free search results (also known as organic results) are search
results whose ranking does not include a paid element.
Paid search results or ads
Search result formats
165. There is no single way to present search results. Different types of
information benefit from different treatment. Google therefore presents
search results in different formats for different types of information:
Generic results or blue links
Generic results are search results that appear as blue links and
can cover any category of information.
C. Nos. 07 & 30 of 2012 59
PUBLIC VERSION
Universal Results
Universal Results are groups of search results for a specific
category of information, such as news, images, or local
businesses.
OneBoxes
OneBoxes provide factual answers to users’ queries. OneBoxes
return direct answers to, for example, queries about
mathematics, stock quotes, local time, currency conversion, and
the weather.
Commercial Units
Commercial Units are result types that Google sets apart in ad
space and distinguishes from search results with a “Sponsored”
label.
Specialised result designs
Specialised result designs refer collectively to Universal Results,
OneBoxes, and Commercial Units.
Snippets and image thumbnails
166. Like other search services, Google shows links to relevant search results
along with text snippets and, in some cases, image thumbnails. Showing
snippets and image thumbnails helps users to evaluate different search
results and choose the ones that are most useful to them.
AdWords
167. AdWords is Google’s ad platform. AdWords enables advertisers to bid for
ads to appear on Google’s pages.
C. Nos. 07 & 30 of 2012 60
PUBLIC VERSION
Ad Rank
The position that an ad appears in Google’s ad space is
determined by its Ad Rank. Ad Rank is calculated as a function
of an advertiser’s bid and auction-time predictions of the ad’s
relevance and quality.
1-10 Quality Score
Google provides advertisers with an illustrative 1-10 Quality
Score as a simple and easy-to-understand metric to get an
indication for the quality of their ads.
House Ads
House Ads refers to ads placed by Google for its own products
in AdWords.
Keyword Bidding Policy
AdWords advertisers bid for search terms (“keywords”) for
which an ad will be shown. Google’s Keyword Bidding Policy
does not prohibit advertisers from bidding on keywords that
may be trademarked.
Ad Text Policy
Google’s Ad Text Policy defines the rules that apply to the text
of ads that Google shows. Under Google’s Ad Text Policy,
Google investigates properly notified complaints concerning
unauthorised use of trademarks in the text of ads.
AdWords API
168. Google licenses an Application Programming Interface (“API”) for
AdWords. The AdWords API allows advertisers and third-party
developers to use automatic ad campaign management tools. The
C. Nos. 07 & 30 of 2012 61
PUBLIC VERSION
AdWords API is one means that Google makes available for advertisers
to transfer ad campaigns to rival platforms. In order to obtain access to the
AdWords API, advertisers and developers must agree to certain terms and
conditions (“AdWords API T&Cs”).
Intermediation (syndication) agreements
169. Google’s intermediation agreements allow website owners (“publishers”)
to incorporate Google’s search and ad technologies on their websites.
Users can then conduct searches directly on the publisher’s site.
Publishers gain the opportunity to earn revenues from Google ads shown
on their websites’ pages. Google’s intermediation service is called
AdSense.
Distribution agreements
170. Google’s distribution agreements allow web browser providers (web
browsers are programs used to surf or navigate the internet) to incorporate
Google search functionality into their web browsers.
Issues under Consideration in the Inquiry
171. The Investigation Report states that Google biases its search results
because display of certain specialised search design is not strictly
determined by relevance. Not only that, Google provides insufficient
information to advertisers on performance of their ad campaigns. The
report further states that Google should block advertisers other than the
trademark owner from bidding for ads upon trademark keywords and
objects to Google’s intermediation and distribution agreements as being
restrictive. Besides, the Investigation Report takes exception to Google’s
conduct on certain counts which were not the subject matter of prima facie
orders, as detailed earlier.
C. Nos. 07 & 30 of 2012 62
PUBLIC VERSION
172. On a careful perusal of the Investigation Report , the Commission would
examine the issues in the present inquiry under the following broad heads:
(i) Whether Google biases its search results?
(ii) Whether Google imposes unfair conditions on its
advertisers?
(iii) Whether Google’s distribution agreements restrict
competition?
(iv) Whether Google’s intermediation agreements restrict
competition?
Search Bias
173. The DG objects to a number of Google’s specialised result designs –
Universal Results, OneBoxes, and Commercial Units – that Google shows
as part of its search results.
174. To begin with, it would be appropriate to preface the analysis with the
observations / findings of the DG in this regard:
… [i]t emerges that though determination of relevance and the
generation of search results to a large extent is automatic, it runs on
Algorithms which are computer processes and formulae, designed and
owned by Google and changed almost on a daily basis. Google being
in control of this algorithm which is pivotal to generate search results
is thus in a position to intervene in the automated process at any point
of time and impact the relevance and ranking of the results.
[Ref. Para 87, p. 304 Investigation Report]
Available information indicates that Google notifies publically only
selective algorithmic changes. Google has cited proprietary and
security reasons for not disclosing the same. While Google may not be
C. Nos. 07 & 30 of 2012 63
PUBLIC VERSION
in a position to publically disseminate complete details of algorithmic
changes, disclosure of limited information after safeguarding their
proprietary and security interest would enable the websites to make
timely bonafide changes and ensure level playing field.
[Ref. Para 92, p.306 Investigation Report]
While no website can claim any particular rank in the SERP;
considering the impact that these changes have on the ranking on the
websites in search results and their ability to attract traffic, there is a
need for greater transparency and timely dissemination of information
about algorithmic changes. This aspect assumes further significance
on account of the fact that Google competes with others in display and
ranking of results and has a major competitive advantage due to access
to information related to such changes.
[Ref. Para 95, p. 307 Investigation Report]
Due to the information asymmetry, non-transparency and considering
the fact that the algorithmic changes are not subject to external audit
or monitoring, Google is always in a position to alter the algorithm
and affect the search results discretely in a discriminatory manner.
Against this background it is found that despite Google’s algorithmic
search framework being largely automated there exist enough scope in
the process for manual intervention, manipulation of Results.
[Ref. Para 96, p.307 Investigation Report]
[A]ny manipulation by Google in the ranking of results especially in
the top ranks on the first page of SERP has potential to adversely affect
the ability of other web content providers to attract user eyeballs
particularly in light of the presumption of users that the top results are
the most relevant.
[Ref. Para 122, p.318 Investigation Report]
C. Nos. 07 & 30 of 2012 64
PUBLIC VERSION
Based on the examination of aforesaid facts is found that Google does
bring out algorithmic changes to effect ranks of the specialised search
results like One boxes, Knowledge Panel, Universal results in a pre-
determined manner. Rank of a result may thus not be strictly
determined by relevance (e.g. in the aforesaid instance local results
merged irrespective of relevance of individual results). The aforesaid
findings assume particular significance considering that throughout its
submissions Google has emphasised on its focus on delivering the most
relevant organic results to its users. Google’s algorithms are at times
changed/designed in a manner that provides high ranks to particular
search features introduced by it.
[Ref. Para 129, p.321 Investigation Report]
While the positioning of various kinds of results on the SERP in a
predetermined manner by itself is not a concern, the issue that merits
examination is whether all web content providers are treated
equivalently to appear on that position. Google is well within its rights
to innovate and modify its SERP, in view of its dominant position the
same should be done in a non-discriminatory and transparent manner
to ensure a level playing field for the web content providers.
[Ref. Para 130, p.321 Investigation Report]
Based on the analysis of the facts gathered during investigation it
stands established that under Universal/thematic Results that appear
in response to a query entered in Google’s general purpose search
engine, Google gives preferential treatment to its own web content/
specialized search/vertical search services like Google News and
Google Images, etc.
[Ref. Para 158, p.334 Investigation Report]
… [U]niversal results occupy prominent ranks on the SERP and
significant real estate on SERP. Thus, for any search queries for which
information categories like news, images etc. are relevant, links to
C. Nos. 07 & 30 of 2012 65
PUBLIC VERSION
Google specialised search options/ vertical search services appears at
high ranks on the SERP through this integration.
[Ref. Para 159, p.334 Investigation Report]
While there is no denying of the fact that the users to some extent
benefit from Universal Search results and other innovations discussed
above, the objectionable area is the manner in which Google, which is
a dominant search engine, is systematically and clandestinely favoring
its own specialized search services when providing such results as
opposed to its claimed provision of the best unbiased results to the
users, regardless of the source. There may exist third party specialized
search services that cater to the user query better.
[Ref. Para 160, pp. 334-335 Investigation Report]
… [i]t stands established that Google integrates and gives preferential
treatment to its own content, various products and services in the
Online General Web Search Services.
[Ref. Para 183. p. 343 Investigation Report]
… [G]oogle does favours its own products and services over others
while displaying results under its so-called concept of
Thematic/Universal results. Google’s Universal search allows Google
to leverage its search engine dominance into virtually any field that it
may choose (online mapping, travel search, financial search
etc).Merely because of the fact that under thematic results, related
information is presented covering various facets such as News, Places,
Maps and thus may have a component of utility for users cannot justify
compromise to the basic characteristic of an organic search engine of
providing neutral and unbiased results. Hence under the garb of
thematic results Google has been strategically promoting and
prioritizing its own properties and partners in its organic search
operations at the expense of other competing websites.
[Ref. Para 187, p. 344 Investigation Report]
C. Nos. 07 & 30 of 2012 66
PUBLIC VERSION
… [t]here is a conflict of interest in the present scheme of things where
Google’s own websites compete for visibility with third party websites
on search result displayed through Google’s Search engine. Complex
nature of process for generation of search results coupled with
information asymmetry distorts the level playing field and provides
room for preferential treatment of own websites in generation of search
results.
[Ref. Para 194, p. 347 Investigation Report]
Thus, the generation of OneBox results is being done in a
discriminatory manner and entails consumers harm by not necessarily
displaying the most relevant result which is further compounded by the
fact that they are not labelled which results in information asymmetry.
[Ref. Para 201, p. 350 Investigation Report]
It has therefore emerged that on Google.co.in, Commercial units for
content categories like Hotels and Flights contain triggers to their
respective Google Search Options/ Specialised search services. This
integration of Google’s own search options/specialised search services
in the commercial unit in a preferential and non-transparent manner
has anti-competitive effect …
[Ref. Para 221, p.361 Investigation Report]
… [G]oogle stands to benefit tremendously from the integration and
prominent placement of links to its specialized search options/services
in Universal Results and Commercial Units on its SERPs.
[Ref. Para 233, p. 366 Investigation Report]
Google’s general purpose search engine is used by a large consumer
base searching for information under the bonafide belief that the
results are unbiased. The consumer harm resulting from such
favouritism is that they are misled to believe that the links to Google’s
specialised search options/services that appear prominently in
response to their queries are purely driven by quality considerations
C. Nos. 07 & 30 of 2012 67
PUBLIC VERSION
even though they may not be most relevant and appear as such due to
Google’s practice of favouring its own results. The practice is thus
unfair to the users searching for information on Google in violation of
section 4(2)(a)(i) of the Act.
[Ref. Para 243, p.369 Investigation Report]
As a consequence of the said search bias, equally efficient
websites/specialised search service providers, due to reduced
visibility, may not be able to acquire a sufficient volume of business
(Minimum Efficiency Level) required to viably compete and survive
against Google’s own services and may thus be driven out of the
market thereby foreclosing competition and reducing alternatives to
consumers. This systematic exclusion leads to denial of market access
to competing websites/ specialised search service providers,
amounting to violation of section 4 (2) (c)of the Act. Such competing
specialised search services will have reduced incentive to innovate
resulting in consumer harm as well as limiting technological
development in violation of section 4(2) (b) (ii) of the Act.
[Ref. Para 244, pp. 369-370 Investigation Report]
… [i]n generation of several categories of results, Google doesn’t treat
third party websites/web content providers equivalently and steers
users to their own products and services, producing biased results.
Thus, the users may not receive the most relevant results. This also
adversely affects the competitive landscape in the markets for search
and search advertising as well as adjacent markets such as travel,
maps, social networking, e-commerce etc. Such actions deter
innovation in a wide-range of these online ancillary markets.
Consequently, users may be devoid of additional choices of results.
This shows that the competition is hampered in the market impeding
innovation, and thus, harming consumers. Therefore, Google’s
conduct is found to be anti-competitive in terms of Section 4(2)(a)(i),
4(2)(b)(ii) , Section 4(2(c) and 4(2)(e) of the Act.
[Ref. Para 253, pp. 384-385 Investigation Report]
C. Nos. 07 & 30 of 2012 68
PUBLIC VERSION
Google’s Response
175. Google challenged the findings of the DG on “search bias” as unfounded
in fact and in law. It was pointed out that the Investigation Report raises
objections against Google’s specialised result designs viz. Universal
Results, OneBoxes, and Commercial Units which Google shows as part
of its results. These designs are innovative features that improve the
quality of Google’s result pages and benefit the users.
176. It was pointed out by Google that the Investigation Report’s accusation
against its specialised result designs is based on three main claims:
(i) Google’s display of Universal Results, OneBoxes, and Commercial
Units involves “search bias” because the display of these designs
is not “strictly determined by relevance”;
(ii) The alleged bias misleads users, who believe that Google’s ranking
is “driven purely by quality considerations”; and
(iii) The alleged bias harms competition because it denies vertical
search services market access, reduces innovation, and impedes
choice.
177. It was stated that Google holds specialised result designs to the same
consistent and strict standards of relevance and quality that it applies to
all its results. Universal Results are groups of search results for a specific
information category, such as news, images, or local businesses. The
results shown in Universal Result groups are free search results – just like
generic blue link results. They are, however, designed in ways that make
them more relevant and useful for users.
C. Nos. 07 & 30 of 2012 69
PUBLIC VERSION
178. Google argued that the Investigation Report’s claim that Universal Results
are biased because their ranking “may […] not be strictly determined by
relevance”, is incorrect and reflects a misunderstanding of the
technologies involved. It was submitted that ranking of Universal Results
is determined by relevance and Google holds Universal Results to the
same standards of relevance as all its results. In particular, it was Google’s
systems which decide whether to show a Universal Result group in a given
position by comparing the relevance of that group with the relevance of
generic blue link results in that position. Google shows a Universal
Results group only if its technical systems conclude that the Universal
Results group is more relevant than other results in that same position.
179. On the DG’s finding with respect to “fixed positions” of Universal
Results, it was contended by Google that initially it limited the display of
Universal Results to certain “fixed” positions because its systems were
not advanced enough to conduct a relevance comparison for all the
positions on its SERP. Explaining further, it was stated that Google
preferred not to show Universal Results in positions for which it could not
conduct a reliable relevance assessment, rather than compromise its
relevance and quality.
180. Hence, it was sought to be canvassed that the DG has erred and confused
positions with relevance assessment. What matters is not in which
positions a particular result type, such as Universal Results, can appear in
principle. What matters is the conditions that these results must meet to be
shown in those positions. To appear in any given position, Universal
Results have to meet the same standards of relevance as generic blue link
results in that position.
181. Adverting to OneBoxes, it was argued that they provide short factual
answers in response to queries, such as a mathematical calculation, the
C. Nos. 07 & 30 of 2012 70
PUBLIC VERSION
local time, or cricket statistics. The Investigation Report acknowledges
that OneBoxes help users find the answers they seek. But the Report
claims that showing OneBoxes involves bias because the content
providers for OneBox information may not be the “most relevant”. This
contention is mistaken for two reasons: first, majority of OneBoxes
respond to factual queries for which there is only one right answer – the
result of a mathematical calculation, the time of day, or the outcome of a
cricket match. OneBoxes then do provide the correct answer. A different
content provider could not provide anything more relevant; second, for
OneBoxes where there can be different possible answers e.g., weather
forecasts, Google selects the content providers based on evaluation of
relevance, quality, and business terms. Google is not paid by content
providers and has, thus, no incentive to select an inferior content provider.
The Report provides no evidence that Google has ever actually selected
an inferior provider for its OneBoxes.
182. With respect to Commercial Units, it was pointed out by Google that such
result types are set apart in ad space and clearly distinguished from free
search results with a “sponsored” label. The display of Commercial Units
leaves the selection of free results untouched. In India, Google shows two
types of Commercial Units: Shopping Units, which display ads for
product offers of merchants; and Flight Units, which identify flight offers
for a given destination.
183. It was pointed out that Google’s search service is based on a two-sided
business model in which it provides its search service free to users, and
funds this service through the display of paid results (i.e., ads) for which
Google receives compensation from advertisers. The display of
Commercial Units is a normal and legitimate consequence of Google’s
two-sided business model and Google shows sponsored results or ads
together with free search results.
C. Nos. 07 & 30 of 2012 71
PUBLIC VERSION
184. In sum, it was argued that Google’s specialised result designs (Universal
Results, OneBoxes, and Commercial Results) do not mislead users.
185. It was also pointed out that Google’s specialised result designs do not
harm competition. The Investigation Report repeatedly recognises that
Google’s specialised result designs are product improvements that benefit
users. At the same time, the Report claims that these designs are abusive,
without any supporting analysis or evidence of competitive harm. If
Google’s conduct harmed competition, one would expect this to manifest
itself in an observable foreclosure of competition that is attributable to the
alleged abusive conduct. But since the time Google began showing the
allegedly abusive designs, referrals from Google to vertical search
services in India have skyrocketed:
(i) Between 2010 and 2015, free Google clicks to product search
services in India increased over 25-fold, from to over
.
(ii) Between 2007 and 2015, free Google clicks to local search services in India increased over 80-fold, from to over .
(iii) Between 2012 and 2015, free Google clicks to travel search services in India increased from around to over (116% growth in just three years).
PUBLIC VERSION Free Google Traffic to Vertical Search Services in India
186. It was submitted that these data understate the total traffic that vertical search sites receive because they only concern traffic from Google (Google does not have access to total traffic data for third parties). In fact, vertical search services receive of their total traffic from non- search sources.
187. Thus, it was submitted that hard data contradict any claim that Google has harmed competition in India by showing specialised result designs. Evidence on the file confirms this conclusion: 90% of the respondents to the DG’s questionnaires express no concerns of denial of market access in search. And a number of parties, including Google’s rivals such as Rediff and Flipkart, specifically confirm that Google does not deny market access to its rivals.
188. Referring to the provisions of Section 4 of the Act, it was submitted by Google that the Investigation Report failed to satisfy the requirements of PUBLIC VERSION Section 4 of the Act. Given that there is no search bias, no user deception, and no harm to competition, there can be no infringement of the Act. In addition, the specific provisions of Section 4 of the Act that the Report invokes do not apply here:
(i) Section 4(2)(a)(i) does not apply Section 4(2)(a)(i) applies only to “purchase and sale of goods and services”. But users do not purchase or sell anything when they see or click on Google’s search results or ads. Moreover, application of Section 4(2)(a)(i) is excluded where a practice is “adopted to meet the competition”. In the present case, Google developed its specialised result designs to improve the quality of its service, in competition with other search services, such as Yahoo! and Microsoft, which have also adopted similar designs.
(ii) Section 4(2)(b)(ii) does not apply The Investigation Report’s claim that Google’s conduct has led to a reduction in innovation in violation of Section 4(2)(b)(ii) is unsubstantiated. In reality, wide choice and innovation are the hallmarks of search in India.
(iii) Section 4(2)(c) does not apply The Commission confirmed in Kapoor Glass case that Section 4(2)(c) does not apply where alternative means of supply exist. The Investigation Report presents evidence showing that vertical search sites receive of their traffic from sources other than search, such as direct traffic and through mobile apps. Further, Hon’ble COMPAT also held in Fast Way Transmission that Section 4(2)(c) can only apply as between competitors. The Investigation Report notes in Chapter 2 that Google, as a general search service, does not compete with vertical search services.
PUBLIC VERSION
(iv) Section 4(2)(e) does not apply
Section 4(2)(e) requires two, distinct relevant markets, as confirmed by the Hon’ble COMPAT in National Stock Exchange of India: one market where a firm is dominant, and one market where this dominance is leveraged. But the Investigation Report does not define two separate markets nor does the Investigation Report identify any competitive harm caused by the alleged leveraging.
189. Summing up, it was contended by Google that if the Investigation Report was upheld, it would be forced to make changes that would diminish the quality of its service. For instance, Google could show maps from other service providers but this is technically not feasible without seriously damaging the quality of Google’s service – as the High Court of England and Wales recognised in its recent Streetmap judgment. Google cannot ensure the quality of results, and waiting for results of third-parties would cause significant delays in response time to the detriment of users. The only realistic option would, therefore, be for Google to discontinue its maps results and it would have to do this even when both the Taiwanese Fair Trade Commission and the High Court of England and Wales have recognised the indisputable benefits and lawfulness of Google’s map display. Thus, Google would be forced to retreat to pre-2000 levels of technology, limiting itself to showing merely blue links. That would mean a seriously inferior service for Indian users compared to that offered in the rest of the world.
190. The two screenshots below illustrate the contrasting designs that Google would be able to show in the rest of the world compared to India if the Investigation Report’s findings are upheld.
PUBLIC VERSION
191. An adverse order may risk chilling innovation in India. This would condemn product improvements and deny Indian users the benefits that innovation brings.
The Informants’ Response
192. Per contra, the Informant in Case No. 07 of 2012 supported the findings of the DG. The Informant submitted that Google’s shift to providing “universal results” by incorporating results from its own vertical properties onto its SERP, has pushed down the results of competing specialised search services from the SERP. This has resulted in a sharp reduction in traffic to these competing specialised search services. Google has combined the introduction of its own vertical properties on the SERP with demotion of competing vertical search services, through various “updates” to its algorithm, on the ground that they provide no original content. These two events, in combination, have had a disastrous effect on competitors of Google’s vertical properties.
193. The Informant further submitted that most vertical search services rely heavily on Google’s web search results for referrals, and a demotion on the SERP can lead to a significant decline in visitors (which has a knock- on effect on subscriptions as well as advertising revenues). Essentially, PUBLIC VERSION these websites depend on Google for their survival, and Google’s search bias threatens their very survival.
194. The Informant also stated that Google appears to recognise that this preference it gives to its own vertical properties has anti-competitive ramifications. Therefore, it has offered commitments in the EU that address this concern. To corroborate its position, the Informant pointed out that several companies that compete with Google in vertical search spaces have complained about manual demotions, which have caused them significant harm. One such company, Foundem, was the Informant in the recent comparison shopping case against Google in the EU.
The Commission’s Analysis
195. Having considered the DG’s findings and submission of the parties, the Commission notes that the Investigation Report claims that Google engages in “search bias” by showing specialised results designs. The Report specifically objects to three specialised result designs: Universal Results, OneBoxes, and Commercial Units. The Commission would deal with search bias under the following heads:
(a) Universal Results
(b) OneBoxes
(c) Commercial Units
196. Before delving into the issues, the Commission would like to emphasize the special responsibilities and obligation of a dominant undertaking under anti-trust law and with a specific reference to the digital markets.
197. Innovation cycles are progressing at a fast pace in the digital economy disrupting and reshuffling long-established positions. In this context, it is of paramount importance that public intervention in such markets should PUBLIC VERSION be targeted and proportionate. Such a calibrated approach in technological markets ensures that intervention remains effective; it does not restrain innovation and helps the market to regulate itself. This concern, which applies to all economic sectors, has undoubtedly more resonance in the digital economy.
198. In the digital economy, players with strong market position often enjoy a virtual hegemony, due to the “winner takes all” phenomenon. This often correlates with the existence of two-sided market or network effects. Network effects appear if the value of a product grows more than proportionately to the number of users of the product or compatible products. When additional consumers join the network of existing consumers, this has a positive external effect on its existing members. Network effects can be regarded as significant above a given sign-up rate, when a critical mass is attained. Beyond that point, more and more customers will be interested, since the benefits of the service increase with the number of subscribers.
199. In multi-sided digital platforms, the network effects are more pronounced.
New users tend to choose platforms or networks that already have a large user base which can ultimately even lead to a dominance by a firm in the market. No doubt, network effects can also facilitate introduction of innovative products, yet it cannot be disputed that network effects can raise switching costs for users and barriers to entry for potential competitors. As a consequence, market entries become less likely and users switch less frequently to other suppliers, which has a market power enhancing effect.
200. While such effects may not necessarily lead to dominance of a single company, they entail potential anti-competitive risks because they may facilitate formation of a dominant position and curb market contestability, PUBLIC VERSION since the most appealing company for new customers is the one that already has the largest customer base.
201. To accurately assess whether a dominant enterprise in the digital space is abiding by special responsibility, it is important to take cognizance of fast- moving innovation, the novel products and services at issue, and the nature and extent of network effects that might exist. Dominant enterprises must not take unfair advantage of their position as such abusive conduct not only impacts the market as a whole but may also affect the entry and sustenance of other market participants into complementary markets associated with the platform.
202. In the present case, since Google is the gateway to the internet for a vast majority of internet users, due to its dominance in the online web search market, it is under an obligation to discharge its special responsibility. As Google has the ability and the incentive to abuse its dominant position, its “special responsibility” is critical in ensuring not only the fairness of the online web search and search advertising markets, but also the fairness of all online markets given that these are primarily accessed through search engines.
203. In view of the above, the Commission is cognizant of the fact that any intervention in technology markets has to be carefully crafted lest it stifles innovation and denies consumers the benefits that such innovation can offer. This can have a detrimental effect on economic welfare and economic growth, particularly in countries relying on high growth such as India.
204. The allegations and the findings against Google in respect of search results essentially centre around design of search engine result page, the Commission considers it appropriate to enter a caveat before embarking upon the scrutiny of the design to ascertain anti-trust violations. The PUBLIC VERSION Commission notes that product design is an important and integral dimension of competition and any undue intervention in designs of SERP may affect legitimate product improvements resulting in consumer harm.
205. Having said that, it is made clear that such regulatory forbearance from interfering with search design is only by way of self-imposed restraint but if in a given case the Commission finds the conduct to be egregious, appropriate remedies and directions shall be issued to correct such a distortion.
206. In the aforesaid backdrop, the Commission proceeds to examine the issues within the permissible parameters in technology markets.
Universal Results
207. Universal Results are groups of results for a specific type of information, such as news, images, local businesses etc. The results shown in Universal Result groups are stated to be free results – just like generic blue link results. According to Google, they are simply designed in ways that improve their relevance and usefulness for users.
208. Google states that it first introduced grouped results in 2001 (in the USA) for news articles. Over time, Google introduced other result groups, including for images (2004), and results for local businesses (2004). These result groups later came to be known as “Universal Results”.
209. It was stated that Universal Results improve the quality and relevance of Google’s results in three ways: (i) they group results for a specific information category; (ii) they select results within a group based on category-specific signals that allow Google to find more relevant results for that category of information; and (iii) they show results in formats tailored to the type of information at issue.
PUBLIC VERSION
210. To better appreciate the ecosystem of Universal Results, it would be apposite to note an illustration. For the query [Taj Mahal], for example, Google’s results on May 27, 2015, included a group of results for relevant news articles and a group of results for relevant images as illustrated below. Users who seek news articles about the mausoleum can find relevant results quickly in the news results group. Users who want other types of results can find them easily on other portions of the page:
211. It was pointed out that Google experienced the deficiency of relying solely on generic signals during the September 11, 2001 terrorist attacks. At that time, Google was unable to surface results for recent news articles on the attack. The generic signals that Google’s algorithms took into account, such as PageRank (which measures link relationships), were not suited for identifying relevant news articles. The relevance of news articles depends largely on how recent they are. Very recent news articles, however, will typically have few link relations. A search for a recent event, such as [IPL cricket news], on a search service that relies solely on link relationships as a signal would produce inferior results. Through this example, it was sought to be explained as to why Google developed mechanisms to take into account category-specific relevance signals for Universal Results. For PUBLIC VERSION news results, Google’s algorithms consider how recent a news article is. For local businesses, Google measures location. And for products, Google may consider things like price, merchant rating, or stock availability. For local businesses, Google may show a map indicating the location of the businesses. For image results, Google shows an image thumbnail that helps users evaluate the image at a glance. These formats help users find the type of information that they are looking for.
212. In short, Google has argued that specialised result designs were part of Google’s results almost from the beginning of Google’s operations. They represent an important element through which Google provides relevant and high-quality results. Rather than deviating from Google’s stated policy of providing search users with results that are most relevant and useful for their queries, Universal Results increase the relevance and usefulness of Google’s results.
213. The DG noted in the Investigation Report that the ranking of Universal Results may not be “strictly determined by relevance” because Universal Results in the past could appear only in “specific fixed positions”. The DG has pointed out in the Investigation Report that prior to October, 2010 Image Universal could only trigger in the 1st, 4th or 10th position on the SERP. Post-changes in algorithm in October, 2010 Image Universals were made free floating and, thus, were allowed to appear at any position on the SERP.
214. Google has submitted that Universal Results have to earn their place on the SERP based on the same consistent relevance standards that apply to generic blue link results. Universal Results appear at a particular position on the SERP only if their relevance when compared to the generic results, warrants that position.
PUBLIC VERSION
215. There is no doubt that Google initially limited the display of Universal Results to certain “fixed” positions. Prior to 2010 it limited the display of Universal Results to certain “fixed” (1st, 4th or 10th) positions. Google sought to ratiocinate the fixed positions being consistent with its relevance standards that it applied to generic blue link results. Yet, it gave no satisfactory reasons for limiting the display of Universal Results to the fixed positions. The contention of Google that “its systems were not sufficiently advanced to conduct a relevance comparison for all positions on the result page” cannot be accepted in the absence of concrete material in this regard. Google has not produced any material in support of its contention that its systems were not sufficiently advanced to conduct a relevance comparison for all positions on the SERP.
216. In light of this, the Commission holds that rankings of Universal Results prior to 2010 were not strictly determined by relevance, instead the rankings were pre-determined. The Commission holds that the aforesaid practice of Google displaying its Universal Results on fixed positions was unfair as it created a misleading façade that such search results appearing prominently in response to queries were algorithmically determined on the basis of relevance. Such a conduct falls foul of the provisions of Section 4(2)(a)(i) of the Act.
217. The Commission takes note of Google’s submission that it has introduced fully-floating ranking for news results, image results, and local results between September – October 2010. Thus, the conduct of Google in fixing position for its Universal Results (news, images and local) prior to this, is of historical nature and this has been addressed by Google by making positions of Universal Results on SERP free floating. The DG has not recorded any finding qua the fully-floating ranking of Universal Results that Google currently uses and as such, it is not necessary for the Commission to examine this aspect any further. Moreover, the PUBLIC VERSION Commission shall consider this aspect (of Google making positions of Universal Results on SERP free floating) while issuing remedy in this regard.
218. The DG has considered Universal Results as “biased” also because “everytime Universal Results appear in response to a search query, the link for ‘more results’ leads only to Google’s specialised/search options/services and not any other vertical search service”. The Commission is, however, of view that the DG’s conclusion in this regard is not well founded. First, the DG does not direct its criticism at the principal content of Universal Results; rather the DG takes objection to ‘more results’ link that leads to a page showing more results of the same type. The Commission observes that ‘more results’ link comprises part of Universal Results which enables Google to show additional results of the same category e.g., additional image results leading to third-party sites, or additional news results leading to third-party news articles.
219. Further, the Commission is of the opinion that that directing Google to replace these links with links to third-party vertical search services, may create confusion for the users. In such a scenario, Google’s algorithms would rank the first few results, while another provider’s entirely different ranking would rank the remaining set after the link. This may not be a workable proposition. Moreover, the Commission notes that no objective criterion can be laid down by way of a remedy to direct Google as to how it should choose between which third-party vertical search provider the more-results link should lead to. While reaching this conclusion, the Commission has also taken note of Google’s submission that it has no means to evaluate results generated by different search services and to select among them because it does not have information on their ranking functions.
PUBLIC VERSION
220. The DG also took objection to Google showing a map in response to queries for local businesses. The Investigation Report noted that it is not necessary that a map generated through Google Maps is the most relevant result among all service providers that display maps of those places. In addition, the Investigation Report notes that links to reviews from third parties are not displayed as prominently as reviews in local results.
221. The Commission has examined this issue in light of the DG’s observations and responses filed by the parties thereto. This issue needs to be seen and understood by taking into account the comparative relevance of maps provided by various service providers. The observation seems to be based essentially upon the potential for abuse by Google.
222. It is observed that the DG accepts that showing a map in response to queries for addresses and local businesses benefits users, when it notes: “the addition of a map result on the SERP can be regarded as a step towards delivering more relevant results to the user”. It cannot be denied that the map helps users understand the geographic location of the local results and makes it easier for them to evaluate and select results they are interested in. Thus, showing maps as part of local search results is an innovation and enhances user experience.
223. The DG’s observation that Google should include third-party maps directly on its pages in response to user queries, has to be seen in the context of Google’s business model as well as the technical feasibility of executing such a remedy. Google has made a submission that it does not have the technical systems to import data from third-party map sites and show them (within milliseconds) to users. Even otherwise, no justification has been made out by the DG or the Informants which would trigger such a course.
PUBLIC VERSION
224. The Commission is of opinion that requiring Google to show third-party maps may cause a delay in response time (“latency”) because these maps reside on third-party servers. Latency is an important quality parameter of a search service. The fast load times cannot be achieved by showing third- party maps. Further, requiring Google to show third-party maps may break the connection between Google’s local results and the map. Google may not accurately match its local results to locations on a third-party map. That being so, the Commission is of the view that no case of contravention of the provisions of the Act is made out in Google showing its own maps along with local search results. The Commission also holds that the same consideration would apply for not showing any other specialised result designs from third parties.
OneBoxes
225. The DG, while acknowledging that OneBoxes help users find the answers they seek, noted that showing OneBoxes involves bias because the content providers for OneBox information may not be the “most relevant”. For reference, the findings of the DG on this issue are reproduced below:
…[t]he underlying competition concern is whether in display of direct answers in response to search queries Google is discriminating against other websites/web content providers. Although no website can claim to appear at a particular position, for the reasons discussed earlier they should be treated in a fair and non-discriminatory manner by Google in the display of results. Further, Oneboxes are sourced as data feeds from content providers with whom Google enters into agreement. While display of direct answers may cater to the user queries better, it is not necessary that the entity with whom Google ties up for the information is the most relevant for the users. Presently it is not evident to the user that the supply of information under One Box is based on contractual arrangements and not through crawling the web content. Thus, the generation of OneBox results is being done PUBLIC VERSION in a discriminatory manner and entails consumers harm by not necessarily displaying the most relevant result which is further compounded by the fact that they are not labelled which results in information asymmetry.
[Ref. Para 201, p. 350 Investigation Report]
226. Google has submitted that large majority of OneBoxes respond to factual queries for which there is only one right answer and the OneBox provides that answer. Where there can be different possible answers, it was submitted that Google selects content providers based on an evaluation of relevance, quality, and business terms.
227. The Commission has examined the submissions and notes that OneBox provides short factual answer to users’ query and returns an answer. The queries about mathematics, stock quotes, local time, currency conversion, and weather have one possible answer which must be relevant and useful response to the query. The Commission notes that Google selects information for OneBoxes with focus on relevance. While the Investigation Report acknowledges the benefits of OneBoxes, it claims that there is bias in the sources that Google selects for OneBoxes’ content. The Commission finds that the DG’s observation is not backed up by evidence. Mere possibility that it may not select the most relevant provider, is not a substitute for actual evidence of bias.
228. The Investigation Report’s claim that crawling the web for providing OneBox information would be better than relying on licensed information, is also not well founded. The Commission notes that crawling generates unstructured data which is not suitable for displaying the structured, directly relevant information that is shown in OneBoxes. Relying only on crawled data would not allow to show users dynamic results in response. For example, mathematical calculations, currency conversions, or the time of day. If Google would rely only on crawled and not licensed data, users PUBLIC VERSION in India may receive a far-diminished service.
229. Further, the Commission observes that OneBoxes show authoritative information because large majority of OneBoxes respond to queries for which there is only one canonical or factual answer. In that case, different content provider may not a priori provide an answer that would be more relevant. The DG’s concern that Google’s OneBoxes do not select the most relevant answer is, therefore, mistaken.
230. The Commission is further of opinion that it would not matter if OneBoxes show authoritative information or non-authoritative information, for which there may be different possible answers, as long as the information provided is based on a careful evaluation of what is best for its users, relevance and quality as well as business terms. Google would have no reason to choose an inferior provider and more so, when it does not receive payment from content providers that it selects as an information source. In fact, if Google is paying for the content which appears in OneBoxes, it has every incentive to source the best possible content for the OneBoxes. Google has pointed out that it monitors and evaluates the content providers on an ongoing basis to decide whether to renew a content agreement with the same provider or select a different provider. If that is it, the Commission notes that the conclusions drawn by the DG qua OneBox partners are not made out and, more so, because no evidence has been presented by the DG to show that Google has selected an inferior information provider for any of its OneBoxes.
Commercial Units
231. Before coming to the issues involved, it may be appropriate to look at the product design and characteristics as submitted by Google.
PUBLIC VERSION
232. Commercial Units are result types that Google sets apart from free search results. It shows Commercial Units in the space utilised for showing ads which are above or at the right-hand-side of free search results. Google distinguishes Commercial Units from free search results with a label indicating Commercial Units are “Sponsored”. In India, Google presently shows Commercial Units for Shopping and Flights only and does not show the Hotels Commercial Unit any more.
233. The DG noted that Google treats Commercial Units in a “preferential”
manner because they are “based on mechanisms that do not apply in an equivalent manner to links to non-Google websites”. This conclusion of the DG has been denied by Google.
234. With reference to Shopping Units, Google has submitted that such Units are enhanced ad formats and their display involves no bias. Like all search services, paid advertising enables Google to offer free search results. Showing ads alongside free results is how Google monetises its search service which it offers free.
235. The Commission notes that Google’s display of Shopping Unit may not per se affect the ranking of free search results. However, this is not to suggest that commercial units displayed by Google would acquire any immunity from anti-trust investigation merely because such units have a commercial/ sponsored and paid element. It needs no reiteration that the product design may push down or altogether eliminate the market participants from competing in a fair and at a level playing field for that particular vertical category for which Google has placed its commercial units. This would appear evident, as discussed in succeeding paras, in the context of Google’s flight commercial unit.
236. Google launched Google Flights in India in 2015. The Commission notes that the results are indicated by the term “sponsored” that appears next to PUBLIC VERSION the Commercial Unit and clicking on the information icon next to that reveals that “Google may be compensated by some of these providers”. While the Flight Unit does not link to a separate and standalone search service, but it does lead to another page showing additional results of the same type.
237. Google has submitted that the development of Commercial Units represents Google’s innovation in ads: instead of showing plain text, Google’s ads in the Commercial Unit display design elements adapted to a particular category of advertisers, and this makes it easier for the user to find the information which they are looking for. The ranking in the Commercial Unit is based on a combination of relevance and bid price.
238. The Commission observes that Google monetizes Commercial Units based on advertisements and/ or by listing merchants/ travel agents etc. who can bid for inclusion and ranking on them. The DG has observed that on Google.co.in, Commercial Units for content categories such as Flights are integrated with Google’s own search options/ specialised search services. The Commission is of the view that by integrating/ linking specialized search result pages with the Commercial Units and placing them prominently on SERP, Google is able to drive traffic to its own pages and also generate revenues through advertisements/ sponsored results.
239. In this regard, it is observed that a user’s clicking behavior may also be influenced by Google’s public claim of ranking results based on relevance. Google’s claim of ranking results based on relevance is borne out from the following which have been summarised in the Investigation Report:
In its 2004 initial public offering (“IPO”) filing with the U.S. Securities and Exchange Commission, Google made the following statement:
PUBLIC VERSION “We will do our best to provide the most relevant and useful search results possible, independent of financial incentives. Our search results will be objective and we will not accept payment for inclusion or ranking in them.”
In their letter accompanying the IPO filing, Google’s founders similarly stated that its results are “unbiased and objective.”
“Google users trust our systems to help them with important decisions: medical, financial and many others, Our search result are the best we know how to produce. They are unbiased and objective, and we do not accept payment for them or for inclusion or more frequent updating. We also display advertising, which we work hard to make relevant, and we label it clearly. This is similar to a well-run newspaper, where the advertisements are clear and the articles are not influenced by the advertisers’ payments. We believe it is important for everyone to have access to the best information and research, not only to the information people pay for you to see.”
The “users first” principle is also part of Google’s official corporate philosophy that it summarised in its Ten things we know to be true statement. The first of these ten is the proclamation that Google places users above all:
“[w]hether we’re designing a new Internet browser or a new tweak to the look of the homepage, we take great care to ensure that they will ultimately serve you, rather than our own internal goal or bottom line.”
PUBLIC VERSION Google’s Executive Chairman, Eric Schmidt, has stated that:
“the natural search answers [are] completely unbiased with respect to economics.”
240. In view of such commitment to relevance by Google, it cannot be gainsaid that users would assume that ranking reflects the order of relevance. Hence, any deviance or departure from relevance in presenting search result would not only mislead the users of Google search services but may also hinder market access of the participants who are competing in particular verticals.
241. The implicit reliance of users upon rankings taking them to be based on relevance, is further highlighted from the various studies and reports collected by the DG during the course of the investigation.
242. In this regard, the Commission takes note of Microsoft’s submissions provided to the DG whereby it has provided an image of a “Heat map” showing how users scan a search page which highlights the areas of SERP where users focus their attention. It has been explained that the image was prepared by recording the movement of users’ eyes across SERP when search results are presented. Based on the analysis, it has been pointed out that the “top-left” section of SERP receives distinctly more attention than any other part of SERP.
243. Microsoft has also referred to various other internal studies conducted for the European Commission estimating the importance of rank of results on SERP in support of its contention that both demoting and promoting links yield significant differences in traffic, demonstrating the importance of rank itself (and not only site relevance) in generating clicks.
244. The Commission is of opinion that there can be no doubt that top ranks on SERP receive the maximum clicks and are thus crucial from the point of PUBLIC VERSION view of attracting Internet traffic. All the web content providers compete to appear at higher ranks. The Commission also agrees that the development of Search Engine Optimization (SEO) services for improving ranking in search results further highlights the significance of high ranks for online visibility. By not ranking its own pages as those of competitors and placing them in the top ranks on the first page of SERP has potential to adversely affect the ability of other web content providers to attract user eyeballs particularly in light of the presumption of users that the top results are the most relevant.
245. In this connection, it is also pertinent to take note of Microsoft’s letter dated 21.10.2013 submitted before the DG whereby it has stated that ” * * * * * * * * * *”
246. In terms of anti-competitive effect, the DG has found and the Commission notes that Google treats Commercial Units in a “preferential” manner because they are based on mechanisms that do not apply in an equivalent manner to links to non-Google websites. The DG has looked at screenshots for the content category “Flights” and noted that the Flights Icon/ search option in the top bar in the specialised search result page gets highlighted. The DG has further noted that clicking on “more Google flight results” in the Commercial Unit leads the user to the specialised search result page of Google Flights.
PUBLIC VERSION
247. Google has pointed out in its submission that Flight Units provide users with relevant information on flights directly on the page and like Shopping Units, Flight Units are set apart from free search results and are indicated with a “sponsored” label. It cannot generate directly relevant up-to-date flight information by crawling the web because crawled data is unstructured and can only be updated when the “crawler” can access the next page. Therefore, replacing this link with a link to third-party flight search services would damage the quality of Google’s service because it would create confusion.
248. The Commission observes that in case of Google’s Commercial Unit i.e. Flight Units in India, primary competition concern emanates out of its prominent placement of its commercial units on SERP in addition to providing disproportionate real estate thereof to such unit. Moreover, it contains a link to “Search flights”. Clicking on this link takes users to Google’s Flights Page and not to a third-party website such as MakeMyTrip.com or Yatra.com. Therefore, Google has given rise to a search bias by unduly giving prominent placement and real estate of the Flight Unit on its SERP for directing traffic to its own specialized search service. It is observed that many users start their vertical search on Google’s dominant general search engine, clicking on one or more of the verticals displayed on the Google SERP in response to their queries seeking specialist subject-specific content. It cannot be thus denied that Google’s dominant position in General Web Search is being leveraged to provide gateway for users to find relevant travel verticals. As a result, Google’s Search Engine is crucial for verticals that do not pay to appear in Google’s Flights Unit. The insertion of Google’s Flight Unit prominently above the blue link results in the SERP denying third-party travel verticals even the opportunity to be displayed on that key “real estate”. Google’s Flight Unit captures the traffic that would ordinarily PUBLIC VERSION flow to the top algorithmic listing. Further, Google’s Flight Unit pushes the algorithmic results down and in some cases, even off the first page. As a result of the displacement of algorithmic results, third-party travel verticals are driven to buy Google search advertising, since this is perhaps the only option left for them to re-acquire visibility and traffic, though, at a higher cost. In this situation, it cannot be ruled out that their advertising cost increases by intensified bidding on keywords.
249. The Commission also observes that such prominent placement of commercial units on SERP could give rise to yet another anti-trust concern. Capturing consumer eyeballs through search design by giving prominence to Google Flights and taking consumers to search for more Google Flight results may allow Google to collect more user data to reinforce its advantage in search advertising market. Additional specialized search data on users increases the value of the general search advertising as it provides search engines deeper data analysis of particular specialized search data. Extracting additional revenue from the advertisements alone by directing consumers to its specialized results pages may not be so much of an anti-trust concern, but the user data that it is able to collect may not allow other competing vertical search pages the same benefit and deteriorate their ability to further innovate on their products.
250. The Commission notes that the concerns raised by third parties regarding Google using its dominance to cross-sell its own products was captured by the DG by referring to a letter dated 09.10.2013 submitted by ‘MakeMyTrip’. The letter in addition to providing historical screenshots of the SERP in support of its assertions, stated that:
“While these products are unlikely to have a booking engine, the positioning of the product is likely to divert customers from Makemytrip.com at the stages of research. This reduction in PUBLIC VERSION customers is likely to have an impact on overall bookings. Over the long run, such rampant introduction of the new products can be detrimental to profitable existence of incumbents and will lead to reduction of choices for consumers. The only reason why Google is able to provide high visibility to its products is because of its dominant share in the search market – a platform that Google is using to cross-sell its own products.”
251. Before concluding, the Commission also deems it appropriate to mention the following public statement made by Ms. Marissa Mayer, the then Vice President of Search Products and User Experience of Google:
“[When] we roll[ed] out Google Finance, we did put the Google link first. It seems only fair right, we do all the work for the search page and all these other things, so we do put it first… That has actually been our policy, since then, because of Finance. So for Google Maps again, it’s the first link.”
252. Further, it is also important to point out that Consim has presented evidence against Google citing the Staff Report of the Federal Trade Commission dated 08.08.2012 (FTC Staff Report). From the FTC Staff Report, the Commission observes that it reveals in no uncertain terms that Google has manipulated SERP to its advantage and to the detriment of its competitors. No doubt, ultimately FTC concluded that Google’s search designs were not abusive, the statements made in the FTC Staff Report remained unchallenged. The Commission has only taken note of the FTC Staff Report without placing any reliance upon it in reaching the findings of contravention in the present case.
253. On a comprehensive examination of the matter, the Commission observes that given the most vertical search service providers have revenue generation models which are heavily dependent on user traffic, such an PUBLIC VERSION unfair diversion of traffic by Google may not allow third-party travel verticals to acquire sufficient volume of business. The Commission notes that they may be equally efficient websites/ specialised search service providers, but due to reduced visibility, they may not be able to sustain and survive in the market for flight search services. In fact, in the present ecosystem of various start-ups gaining traction, it would be travesty if such attempts made by keeper of search highway is countenanced. The Commission has no hesitation in holding that Google through its search design has not only placed its commercial units right at a prominent position on SERP, it has also allocated disproportionate real estate thereof to those units resulting into either pushing down or pushing out of the verticals who were trying to gain market access. To top it all, Google has provided link which leads users of Google Flight commercial unit to its specialized search result page (Google Flight). Consequently, users may be devoid of additional choices of results and therefore, such conduct amounts to an unfair imposition upon the users availing search services. Such conduct of Google, being an unfair imposition upon the users of general search services, is in contravention of Section 4(2)(a)(i) of the Act.
Unfair or Discriminatory Conditions on AdWords Advertisers
254. Findings of the DG in respect of Google’s advertising platform i.e. AdWords are examined under the following heads:
(i) Operation of Google’s advertising platform (AdWords)
(ii) Permitting advertisers by Google to bid on trademarked keywords (Trademark Issue)
(iii) Restrictions by Google upon advertisers from transferring ad campaign to other ad platforms (AdWords API T&Cs)
255. Before analysing the competition issues, it would be appropriate to PUBLIC VERSION reproduce the findings of the DG on this infringement:
Determination of the Quality Score is a complex process based on several parameters. Investigation has revealed that limited information is available in the public domain. Google also has access to information on the quality score that the system assigns to other websites including those of its competitors. Google being aware of these is in a position to ensure that its House Ads are assigned higher Quality Score than its competitors. Further, it has emerged that Google is not required to pay any monetary consideration for its House Ads which gives an additional competitive edge. The checks referred by Google are found to be inadequate to address and prevent preferential treatment of its own properties in paid results. Available information discussed above brings out that to appear at high ranks is critical for an ad to be clicked. The existing scheme of things provides enough scope for Google to ensure that invariably its House Ads appear in the top slots and above third party ads particularly of its competitors.
[Ref. Para 148, p. 450 Investigation Report] It is thus found that there does not exist a level playing field for the third parties competing with Google’s House ads to appear in paid results in response to user searches. The conduct amounts to imposition of unfair and discriminatory conditions on third party advertisers using AdWords and is found to be in violation of section 4(2)(a)(i) of the Act.
[Ref. Para 149, p. 451 Investigation Report] Google’s Response
256. Google has contended that it does not impose unfair or discriminatory conditions on AdWords Advertisers. It has submitted that just like a free newspaper or radio station, advertising is the way that Google monetises and funds its free search service. AdWords enables advertisers to bid on search terms (“keywords”) upon which ads would appear on Google’s PUBLIC VERSION SERP.
257. Google ranks the ads based on advertiser’s bid and auction-time predictions of the ad’s relevance and quality. Google shows ads that are relevant and useful to users, while offering an effective promotion opportunity for advertisers.
258. The Investigation Report claims that one of the reporting metrics that Google provides to its advertisers – a metric called the “1-10 Quality Score” – is “opaque”, of “very limited utility”, and “susceptible to manipulation”. Google has challenged this claim as wrong. It was pointed out that Google provides advertisers with extensive data and metrics to help them manage their ad campaigns. Google has no reason to provide insufficient data to advertisers because if advertisers’ ads perform poorly, Google earns less revenue.
259. The DG has noted in the Report that the 1-10 Quality Score provides advertisers with “an easy way to understand the relative quality of their ads”, because it “simplifies more complex metrics into a single one to ten scale”. Indian advertisers expressly confirm that AdWords is “transparent, fair plus very easy to use”. Therefore, the allegations that advertisers lack information on the performance of their ads has no basis.
260. DG’s objection to the 1-10 Quality Score rests on a single, minor error that occurred during a product test that the Informant Consim (an operator of a matchmaking website) voluntarily signed up for. An error in Google’s test systems caused Consim’s test ads to drop in ranking on one type of device. Using the reporting tools and metrics that Google makes available, Consim and other testers identified the error and alerted Google about the problem. Google promptly responded and fixed the error. This was, thus, an example of the successful resolution of an error in a test system, not a violation of the Act. The error had a negligible impact on advertisers and PUBLIC VERSION did not affect competition in India generally. Besides, errors and technical bugs do not violate competition law.
261. Google has further submitted that the 1-10 Quality Score is not “opaque”. It has explained the variables that comprise the 1-10 Quality Score. Further, the 1-10 Quality Score is not the only information Google provides to advertisers. It provides many other detailed metrics, such as average ad position, conversions, click-through-rates on ads, and average cost-per-click. Advertisers thus receive extensive data on their ad campaigns which, in fact, exceeds what Google’s rivals, such as Bing, provide to advertisers.
262. Further, Quality Score is automatically generated by algorithms. It is not subject to “manual manipulation”. The Investigation Report does not claim that Google has manipulated Quality Score but states that Google might be in a position to do so. The Report’s reasoning is purely hypothetical and abstract. Moreover, the Report fails to provide evidence of harm to competition or consumers in the relevant market.
263. Google has contended that it does not discriminate with House Ads. Google treats House Ads like third-party ads and strictly prohibits operators of Google’s House Ads accounts from accessing data or information that is not publicly available to all other AdWords advertisers.
264. The Investigation Report does not claim that Google has ever used non-
public information for House Ads and its claims are limited to hypothetical possibilities. Google’s House Ads have not harmed competition. The Report claims that Google hinders advertisers from transferring ad campaigns to other ad platforms. It was argued that Google makes it easy for advertisers to transfer ad campaigns and provides them with several mechanisms to do so. One of these is the AdWords API, which allows advertisers and third-party developers to use automatic PUBLIC VERSION campaign management tools.
The Informant’s Response
265. The Informant in Case No. 07 of 2012 while supporting the DG’s findings argued that Google provides only indicative quality scores which, even according to Google, provide only a direction to advertisers. The Informant submits that the DG’s findings are correct and there is lack of transparency associated with the quality score and the manual interventions made by the Google to cause demotions in the quality scores of advertisers. The complex nature of determination of quality score, coupled with non-disclosure of adequate information to advertisers renders the entire process opaque and susceptible to manipulation, amounting to imposition of unfair conditions on advertisers in violation of Section 4(2)(a)(i) of the Act.
The Commission’s Analysis
266. The Commission has examined the Investigation Report and the contentions of the parties. The Report alleges that the data Google provides to AdWords advertisers is “opaque”. It focuses on one performance metric i.e. – the 1-10 Quality Score – claiming that this is of “very limited utility”.
267. To appreciate the issue in its perspective, it would be necessary to take an overview of AdWords, the 1-10 Quality Score, Ad Ranking and Ad Performance Evaluation. The Commission notes advertisers create ad campaigns by creating ads, assigning keywords, and setting targeting metrics. Google shows the advertisers’ ads in response to users’ queries relating to selected keywords. In this process, Google ranks ads based on a combination of the ad’s relevance and the advertiser’s bid.
268. According to Google, it provides advertisers with data, tools and reports on the performance of their ads. These tools and metrics help advertisers PUBLIC VERSION understand how often their ads appear, how many times users click on them, and how frequently these clicks lead to purchases. This, in turn, enables advertisers to plan their bidding, improve the ad quality. In this process, advertisers can keep track of the performance of their ads in their AdWords account.
269. The Commission notes that Google makes information available to advertisers on the performance of their ads which includes not only the 1- 10 Quality Score, but also other metrics. The 1-10 Quality Score consists of following three components:
(i) Expected click-through rate (expected “CTR” or “eCTR”) eCTR estimates the likelihood that an ad will receive a click. The more likely users will click on a particular ad, the higher the eCTR.
(ii) Ad relevance Google analyses the language in an ad to determine how well it relates to a keyword. This helps ensure that ads shown will be relevant and useful to a user’s query.
(iii) Landing page experience Google evaluates the quality of the landing page i.e., the page to which an ad directs users if they click on it. The better the landing page, more likely the ad will provide a relevant and useful response to a user’s query.
270. Having perused the material on record, it appears that Google makes available the following information publicly:
(i) The AdWords Help Center contains over 200 articles – comprising nearly 600 pages of guidance for advertisers – on how to better understand their 1-10 Quality Scores, Google’s ranking of ads, and PUBLIC VERSION AdWords reports and analytical tools. These articles cover eight different categories of guidance related to ad quality and ad performance reporting. Google also provides YouTube videos, white papers and blogs to advertisers on the subject of ad quality.
(ii) Guidance on using and understanding reports for tracking ad performance, such as the Report Editor and Keywords Report. These articles include information on how to measure sales and conversions, return on investment, brand awareness, and paid and organic search results.
(iii) Guidance on how to improve the components of the 1-10 Quality Score.
(iv) Guidance to advertisers on how the AdWords auction functions and the various bid strategies that they may employ to meet their advertising goals.
(v) Articles covering AdWords account maintenance, troubleshooting, and frequently asked questions are also publicly available.
(vi) The Help Center provides advertisers with guidance on when and how to contact Google regarding any questions, needs for assistance, or complaints that an advertiser may have related to AdWords.
271. The Commission notes that Google does not use 1-10 Quality Score when calculating an ad’s position in an actual auction (“Ad Rank”). As the Investigation Report notes, an ad’s position “is recalculated each time an advertisement is eligible to appear” for a given auction. An ad’s position at auction is determined by Ad Rank, which is a function of an advertiser’s bid, auction-time predictions of the same factors that make up the 1-10 PUBLIC VERSION Quality Score (click through rate, ad relevance, and landing page experience), and the predicted impact of an ad’s ad extensions and ad formats. The components of Ad Rank take into account auction-specific information such as the time of day, the type of device (e.g., desktop or mobile), and the user’s location.
272. Further, the Commission notes that there is no merit in the finding of the DG that the information Google provides to advertisers severely restricts their ability to critically evaluate their campaigns and take corrective steps. The Commission is of the opinion that this finding of the DG emanates out of singular focus on the 1-10 Quality Score. While it is true that 1-10 Quality Score gives one estimate of ad quality but Google also provides other metrics and tools for assessing ad and campaign performance as well. These are:
(i) Click-through rate (CTR) CTR measures how often users click on an ad. Improving the quality or relevance of ads can lead to increase in CTR.
(ii) Bid estimates Google provides estimates for the bid necessary to appear on the first page, at the top of the page, or in the first position. The higher the quality of the ad relative to others’ ads, the lower is the bid amount necessary to appear in these positions.
(iii) Average position An advertiser can track where, on an average, its ad appears on the SERP on a daily basis. If an advertiser’s average position unexpectedly decreases and the advertiser did not change any settings, it may wish to investigate other metrics to see what might have caused the change.
PUBLIC VERSION
(iv) Conversions
This metric tracks events that are meaningful to an advertiser, such as when an ad click leads to a purchase, a phone call, or some other action. Tracking ad conversions allows an advertiser to see the immediate impact of its advertising campaign. Conversion tracking can help advertisers compare which clicks are more valuable.
(v) Time-of-day reporting This feature allows advertisers to run reports to analyse their performance metrics across different periods – ranging from hourly measurements, daily, quarterly, and beyond. The advertiser can use this information to adjust when it runs particular ads to maximise its return on investment.
(vi) Geographic targeting AdWords allows advertisers to target their ads to specific geographies. In India, this includes targeting at the city, state, and union territory level.
(vii) Bid simulator The bid simulator analytical tool helps advertisers test various bidding strategies before actually implementing them. If an advertiser is interested in knowing what impact, for example, increasing its bid might have on different metrics, it can use the simulator to test the results without actually increasing its bid.
(viii) Campaign drafts and experiments Advertisers seeking to test different ad creatives, keywords, bids, or ad placements can use Campaign Drafts and Experiments to apply experimental settings to a fraction of AdWords auctions. The PUBLIC VERSION advertiser can then track the outcome of its experimental set versus the original campaign to observe which changes were effective and which, perhaps, were not.
(ix) Auction Insights Report The Auction Insights Report allows advertisers to track their ad performance against that of competitors in ad auctions. The Auction Insights report shows an advertiser six different metrics for advertisers competing in the same auctions: impression share, average position, overlap rate, position above rate, and top of page rate.
273. In view of the above, the Commission is of the opinion that the DG’s concern regarding disclosure of advertiser performance data by Google does not appear to be well founded. In fact, Google provides sufficient data to advertisers on the performance of their ads.
274. The DG has given a finding that the 1-10 Quality Score is of limited utility because it is not reported for each auction and does not include “auction- time” information such as type of device or user location. The Commission, however, notes that the whole purpose of 1-10 Quality Score is to provide advertisers with a simple and easy-to understand metric on the performance of their ads. Changing the 1-10 Quality Score to reflect auction-time information might defeat the whole point of having a simplified 1-10 metric. Since Google provides many more granular metrics other than the 1-10 Quality Score that help advertisers understand the performance of their ads, reporting the 1-10 Quality Score for each auction and including “auction-time” information may lead to confusion. This is a practice being followed by Microsoft’s Bing. Bing’s quality score is similar to that of Google’s. In fact, Microsoft does not report quality scores for each auction and does not include auction-time information.
PUBLIC VERSION Rather, just like Google’s 1- 10 Quality Score, Bing’s quality score is reported once a day based on historical information.
275. In regard to the Investigation Report’s objection to the 1-10 Quality Score in the context of Consim’s allegations, the Commission notes that the same rests on a single technical error that occurred during a “beta test” for a new feature. In 2013, Consim volunteered to participate as a beta tester for an experimental AdWords format known as “image extensions”. Google has explained that the error was part of a legitimate test for a new feature, of which Consim was fully aware when it signed up to take part in the test. Explaining the operation of the test, Google pointed out that the test allowed beta test advertisers to add images to their text ads when they were shown in the first position on the page. Because these images took up additional space, in browsers with limited screen height, the system was designed to show the original text ad without the image extension. However, a technical error prevented the ad that was supposed to show in the first position from appearing where the user was using a browser with limited screen height. As a result, Consim’s beta test ads dropped in ranking in these browsers from position 1 to position 2 for nearly one month in 2013.
276. Google has pointed out that using the reporting tools and metrics that Google makes available, Consim and other beta testers identified the error and alerted Google about the problem. The error had nothing to do with the 1-10 Quality Score and no amount of additional reporting on the 1-10 Quality Score would have alerted Consim about the error. Instead, other metrics which Google provides to advertisers signaled the error and Consim (and others) quickly spotted the same and reported it to Google.
277. The Commission is of the opinion that it would not be safe to base a finding of competition law breach on errors in an experimental feature.
PUBLIC VERSION This may chill future experimentation and kill innovation in technology markets. The Commission is, thus, unable to lend its concurrence to such findings recorded by the DG. The same are set aside.
Trademark Issues
278. The DG has found Google to be abusing its dominance in the online search advertising market by imposing unfair condition on the trademark owners (particularly those who have notified their trademarks to Google) by allowing their trademarks to be bid as keywords by third parties in online search advertising. As per the AdWords mechanism, ads which appear first may not be the most relevant for the users and may appear at that position due to the higher bid of the entity. The competitors then get opportunity to free ride on the goodwill and brand value of the trademark owner, thereby hampering fair competition. Moreover, this practice creates a significant risk of causing confusion and deception in the minds of the users thereby causing consumer harm. Unsuspecting consumers may be misled to believe an association between the owner of the trademark and its competitors (whose ad appear in response to searches on their brands) and divert traffic. In such a scenario, the owner of the trademarks is compelled to participate and outbid competitors (for their ads to appear before them) thereby augmenting their advertising budgets. The DG has further noted that there is scope for Google to use the system in a discriminatory manner and ensure that its own trademarks are not subject to the same unfair conditions as those of third parties, while bidding under AdWords.
279. Lastly, the DG has noted that though Google’s AdWords policy restricts usage of notified trademarks in the Ad text of competitors, on various occasions, ads of competitors appeared using Consim’s trademarks in Ad text in response to searches on these trademark terms despite notification. Google allowed usage of minor variations of Consim’s notified PUBLIC VERSION trademarks in the Ad text of competitors under its AdWords program despite repeated complaints and this led to unfair bidding between the trademark owner and other advertisers. This appears to be driven by Google’s commercial interests. Consim’s own ads were blocked for searches on its trademark terms even after it complied with the requisite procedures suggested by Google. Google not only abused its dominant position and imposed unfair condition on Consim, such conduct also resulted in unfair business gains to its competitors as well as to Google itself.
Google’s Response
280. Google has contested the DG’s findings that permitting advertisers to bid on trademarked keyword is anti-competitive. According to Google, it actually increases competition. For example, in response to a search for [Ford cars] on google.co.in, Google may show ads from the official Ford website in India. At the same time, Google may also show ads from other advertisers, such as autoportal.com and cartrade.com, two Indian websites or other service providers of Ford. Google does not prevent advertisers other than Ford (the owner of the Ford trademark) from bidding and there is no good reason to do so. It does not violate trademark law and it is beneficial to competition. It enhances user choice and enables Indian websites to compete against the trademark owner. If Google blocks advertisers other than the trademark owner from bidding for ads for trademarked keywords, trademark owners would have a monopoly over ad space for their trademarked terms, reducing competition and user choice.
281. Google has argued that the Investigation Report cites no evidence that Google’s trademark policies have diminished competition in any relevant market, harmed consumers, prevented consumers from finding Consim’s websites, or otherwise caused Consim competitive injury. According to PUBLIC VERSION Google, the Report resorts to speculation about the “possibility” of events, what may be “presumed”, “in all likelihood”, and what “may not be evident” from the text of any given ad. These “conjectures and imaginations” cannot establish an abuse.
282. Lastly, Google pointed out that the Investigation Report does not identify an unfair condition that it imposes on advertisers. Rather than remove an unfair condition, the Report suggests that Google should impose restrictions on bidding for ads (i.e., preclude advertisers from bidding on trademarked terms). This condition would restrict, not promote, competition in India.
The Informant’s Response
283. The Informant in Case No. 07 of 2012 submitted that far from protecting the intellectual property rights of its AdWords and AdSense users, Google allows advertisers to bid on trademarked keywords of their competitors. By doing so, Google supports a bidding war between trademark owners and their competitors as trademark owners are forced to outbid their rivals in paying Google to protect their brand. This results in a situation where first, competitors place high bids to be able to have their ads on the SERP for a trademarked term, and then the trademark owners bid even higher to ensure that their own ads appear on the SERP above those of competitors. Google benefits from such higher bids, and is effectively able to monetise such search results only by facilitating trademark violations.
284. Specifically, for the Informant, Google’s AdWords allowed its competitors’ ads to appear above the Informant’s own ads on the SERP for searches for the Informant’s trademarked keywords, even though the Informant’s ads are more relevant to the search term. According to the Informant, this resulted in a loss of traffic that was meant for the Informant, as well as increased bids that the Informant had to make to ensure that its ads remained on top of the SERP.
PUBLIC VERSION
285. The fact that the Informant, despite raising its bid price to an astronomical INR 1000 per click, was unable to bid on its own trademarked terms as keywords, during a time (October- November 2013) when the Informant was heavily investing in advertising, shows that Google’s conduct is discriminatory. Google benefits immensely from this conduct as it can monetize pages that would otherwise have no advertisements on them (as the keyword is trademark protected).
286. Further, adverting to use of trademarks as ad text and minor variations of trademarks, it was submitted that the Informant suffered significant financial loss on account of its competitors being allowed to use its trademarks in the Ad text of their ads. While such ads have been removed by Google on request, Google did not set up a preventive mechanism to disallow such use in the first place. Further, despite notification to Google of the Informant’s trademarks, these trademarks have repeatedly been allowed to appear in the ads of competitors. In several cases the Informant noted that competitors used minor variations of its trademark protected words in their Ad text, even though such ads would be considered deceptive and confuse ordinary consumers. By not acting on the notifications and allowing repeated violations, Google’s conduct amounts to imposition of an unfair and discriminatory conditions.
287. In the context of matrimonial websites, only the Informant uses the word “matrimony” and, therefore, Google’s claim that this is a generic word must be rejected. Further, the phrase “bharat matrimony” is deceptively similar to “Bharatmatrimony”. As per the Informant, none of the Informant’s competitors use these so called “generic” words or phrases in any of their usual promotional material, except in the AdWords program. This adds credence to the fact that the purpose of using these terms in their ads is merely to divert traffic intended for the Informant, and therefore, Google’s support and encouragement of such conduct violates the Act.
PUBLIC VERSION
288. Lastly, it was submitted that the Informant noticed a change in Google’s conduct after the Informant filed a complaint in the Hon’ble Madras High Court against Google. In particular, Google no longer provided the Informant with favourable credit terms, which forced the Informant to suspend some of its advertising campaigns. Further, the Quality Score of its various advertisements reduced drastically, and no explanation was provided by Google for the same. Therefore, Google’s conduct is opaque, leaving the Informant, and other advertisers, open to retaliation from Google and this may be the case with several advertisers who may not be able to identify and assess similar changes in Google’s conduct.
The Commission’s Analysis
289. The Commission has examined the DG’s findings and the submission of the parties thereon.
290. The Commission notes that Google’s Keyword Bidding Policy is part of AdWords i.e. Google’s advertising service. AdWords enables advertisers to bid for keywords in ads to appear on Google’s pages. When a user enters a query, Google shows ads in an area separate from free search results. To use AdWords, advertisers bid on keywords. When a user enters a query on Google, AdWords select candidate ads from those advertisers that bid on keywords and which correspond to the user query. The candidate ads participate in auctions, that rank the ads based on many factors, such as ad quality and bid amount, to determine which ads to show. The degree of similarity between the keyword and the query required for the ad to participate in the auction depends on the matching options the advertiser chooses, such as exact match, phrase match, or broad match.
291. The Commission finds that Google’s Keyword Bidding Policy does not PUBLIC VERSION prohibit advertisers from bidding on trademarked keywords. Google applies this policy universally, and permits advertisers to bid on Google’s own trademarks as well. Depending on the query and a variety of other factors, free results may include links for a trademark owner’s rivals in response to queries that include a trademarked term. Prohibiting advertisers from bidding on queries that include trademarked terms might result in a perverse situation where Google cannot return ads for competitive or complementary products even when users are searching for them. Therefore by allowing bidding on trademarked terms, it increases the relevance of Google’s ads which benefits users also.
292. Google has pointed out that advertisers want their ads to be seen by the same consumers who are looking at their competitors’ ads and using their competitors’ products. This proximity helps consumers, both in reviewing the ads and locating the products. Advertising based on keyword bidding is another way that competitors can target their ads to users who have mentioned a rival, and may be interested in viewings its goods and services. This targeting strategy offers similar consumer benefits to traditional advertising.
293. The Commission finds it logical and notes that the DG’s finding of contravention based upon Google’s Keyword Bidding Policy allowing bidding upon third party trademarked terms as amounting to “imposition” of unfair condition, is stretched. A plain reading of Section 4(2)(a)(i) of the Act makes it clear that it requires imposition of an unfair or discriminatory “condition in purchase or sale of goods or service” to violate the Act. Thus, Google could violate Section 4(2)(a)(i) if it placed an unfair or discriminatory condition on the sale of (AdWords) keywords to the advertisers. The Investigation Report does not point out any unfair condition which has been imposed by Google upon the users or any condition it seeks to impose for blocking competitors of trademark owners PUBLIC VERSION from bidding on trademarked keywords.
294. The Commission is of the view that Google’s Keyword Bidding Policy enables a user to include a trademarked keyword in its query and, consequently, the user is not only presented with ads from the trademark owner but will also see a broader range of ads, including from the trademark owner’s competitors. This promotes competition and enhances user choice.
295. The Commission does not agree with the findings of the DG that Google did not enforce its Ad Text Policy in respect of Consim’s trademarks properly. The Investigation Report notes that Google failed to stop Consim’s rivals from: (i) using Consim’s trademarks in ad text, consistent with Google’s Ad Text Policy to investigate complaints against such use; and (ii) using “minor variations” of Consim’s trademarks in their ad text, consistent with Google’s Ad Text Policy i.e. Google failed to stop Shaadi.com, which offers matrimonial services in India/ Bharat, from using the words “Bharat Matrimony” in the text of its ads to describe its services because Consim has a trademark on the combined term (without a space) “BharatMatrimony.” The DG does not find fault with lawfulness of Google’s Ad Text Policy per se, but its challenge is primarily directed against its enforcement by Google in respect of Consim’s trademarks.
296. As to the claim that Google did not follow its own Ad Text Policy to stop Consim’s rivals from using Consim’s trademarks in ad text, the Commission notes that the evidence showed that Consim did not take necessary steps to trigger Google to investigate and monitor its complaints. Google’s Ad Text Policy is clear on the information a trademark complaint must contain and to whom it must be sent to be processed. Trademark complaints must be addressed to the Trademark Operations Team within the legal department, which has the knowledge, experience, training, and authority to take action consistent with Google’s PUBLIC VERSION policies.
297. Consim did not comply with Google’s established process for submitting complaints under its Ad Text Policy and failed to direct the complaints to the right place as well as provide the information necessary to process them. That being so, the Investigation Report should have rejected Consim’s allegation in this respect.
298. Adverting to the allegations that Google failed to stop Consim’s rivals using “minor variations” in their ad text with spaces between words which, when combined, make up Consim’s trademarks (e.g., “Bharat matrimony”), Google submitted that processing those complaints would not have been consistent with trademark law or Google’s Ad Text Policy. The Trade Marks Act, 1999 allows anyone to use words to convey their plain meaning, even if those words are registered trademarks. In this case, no trademark infringement would arise if Google refuses to block Consim’s competitors from using combinations of words such as “Bharat” and “matrimony” in the text of Google AdWords to describe the services they offer – i.e., matrimonial services for those who live in Bharat.
299. Google’s Ad Text Policy specifically exempts from investigation ad text that uses the term descriptively in its ordinary meaning rather than in reference to the trademark. This exception, according to Google, is consistent with the Trade Marks Act, 1999 which establishes that descriptive uses do not constitute trademark infringement. Since Consim’s competitor’s use of “Bharat matrimony” to describe its matrimony services in Bharat is a descriptive use, Google’s Ad Text Policy is not to investigate such uses. Therefore, the DG’s claim that Google failed to follow its Ad Text Policy regarding such use is factually incorrect.
300. Google has submitted that the allegations that Google did not successfully block every use of Consim’s trademark before it appeared in ad texts of competitors do not raise competition issues. The DG has not defined the PUBLIC VERSION affected markets, and has not explained as to how delays in disabling isolated ads according to its Ad Text Policy, restricted competition in any such market.
301. Finally, joining issue with the DG holding that a delay by Google to “whitelist” a single Consim account following notification of Consim’s September, 2009 litigation constituted a competition law violation, it was contended that any delay in whitelisting one account from a single website does not raise a competition issue.
302. The Commission has examined the issue in light of the DG’s findings and the responses of the parties thereupon. The Commission notes that the DG found Google not enforcing its Ad Text Policy with respect to the use of Consim’s trademark in ad text of its competitors. Under Google’s Ad Text Policy, Google is to block trademark uses in the text of an ad of another in response to valid complaints from trademark owner. It is observed that Consim has not challenged the terms of the Ad Text Policy per se, but its allegation is essentially confined to the fact that Google did not implement its Ad Text Policy in response to Consim’s complaints to its advertising account representatives at Google and that Consim’s competitors were using Consim’s trademark and “space variations” of those trademark in their ad text.
303. For reasons detailed below, the Commission is of opinion that the allegations laid by Consim do not appear to be well founded.
304. It would be relevant to outline Google’s Ad Text Policy which provides that, “Google will investigate and may restrict the use of a trademark within ad text. Ads using restricted trademarks [i.e., trademarks for which valid complaints have been processed] in their ad text may not be allowed to run.” A few stated exceptions to the Ad Text Policy apply, including:
PUBLIC VERSION An ad can use a trademarked term in its text if either of these conditions is true:
(i) the ad text uses the term descriptively in its ordinary meaning rather than in reference to the trademark
(ii) the ad is not in reference to the goods or services corresponding to the trademarked term.
305. Thus, it can be seen that the descriptive use condition allows uses of trademarked words in their ordinary, plain meaning consistent with the legal principle of fair use. The Commission notes that the same principle is embodied and codified in Section 30(2)(a) of the Trade Marks Act, 1999, which states: “A registered trade mark is not infringed where … the use in relation to goods or services indicates the kind, quality, quantity, intended purpose, value, geographical origin, the time of production of goods or of rendering of services or other characteristics of goods or services.” Furthermore, Section 35 of the Trade Marks Act, 1999 also provides: “Nothing in this Act shall entitle the proprietor or a registered user of a registered trade mark to interfere with any bona fide use by a person … of any bona fide description of the character or quality of his goods or services.”
306. The Commission observes that Google does not automatically block all uses of all trademarked words in ad text worldwide. As trademarks are territorial and apply only to certain goods and services, Google needs to know specific information about the scope of the trademark owner’s rights (e.g., geographic area and product class) and the scope of the trademark owner’s complaint (i.e., against a particular advertiser or against all advertisers), before Google will monitor and restrict use of a particular trademark in ad text.
PUBLIC VERSION
307. To process a trademark complaint and put into place monitoring restrictions, the trademark owner needs to identify certain information, including:
(i) which trademarks it wants monitored;
(ii) whether the marks are word or design [i.e., device] marks;
(iii) the registration status of the marks;
(iv) the registration numbers of the marks;
(v) for which regions the trademark owner has rights;
(vi) for which products or services the trademark owner has rights; and
(vii) whether the trademark owner wants to block specific advertisers or all uses of the trademark within the region and product category.
308. Further, a detailed procedure has been outlined for trademark owners to communicate the aforesaid information to Google. In this regard, the Commission notes that Google provides an online complaint form. Once the requested information is typed in, the trademark owner can press the submit button. Google also provides instructions for emailing, faxing, or sending the requested information by regular mail.
309. The Commission is of the opinion that no fault can be found in Google insisting that a trademark complaint needs to be sent to its Trademark Operations Team, using the contact information Google provides. This ensures uniform application of trademark policy by Google.
310. The Commission also notes from the Ad Text Policy as detailed by Google in its submissions that once its Trademark Operations Team receives a valid trademark complaint i.e., one that provides all of the necessary information, the Trademark Operations Team verifies the submitted information and then adds the term in question to Google’s monitoring system. The monitoring system automatically flags submitted ad text which uses a monitored trademark in the registered territory. Using a PUBLIC VERSION combination of automated and manual procedures, Google evaluates the proposed use to determine if it satisfies one of the exceptions to the general rule of not using third-party trademarks in ad text. If Google determines that the submitted ad text contains an impermissible trademark use, it disallows the ad from being shown. Upon it disallowing the ad for trademark reasons, the advertiser will receive an error message informing it that its ad is disapproved for trademark reasons.
311. Coming to the issue as to whether Consim provided valid notification of its trademark complaints under the Ad Text Policy to Google, the Commission notes that the DG appears to have relied on certain communications Consim sent to various Google employees raising concerns about its trademark use in ad text of shadi.com and seems to have assumed that these communications triggered the investigation and restriction of Google’s Ad Text Policy.
312. In this regard, the Commission notes that in January, 2008, Consim wrote to its Google customer account representative about Shadi.com’s use of “BharatMatrimony” in its ad text. In February 2008, Consim wrote again to its Google customer account representative about certain competitors using Consim’s trademarks as keywords, and in March 2008, it wrote to follow up on its February email. These Consim communications neither contained the information required by Google to process trademark complaints nor were they addressed to the right group to process trademark complaints. Yet, Google representatives for Consim’s account are stated to have contacted Consim’s competitors to request that they withdraw the challenged ads, and the competitors complied. According to Google, Consim’s account representatives at Google took this action to assist it. The same does not, however, change the fact that Consim failed to follow Google’s trademark policy procedure for notifying it of trademark complaints nor did it provide sufficient notice for Google to PUBLIC VERSION implement its monitoring and blocking procedures. The DG’s conclusion that Consim’s incomplete and misdirected emails constituted “valid notification” of Consim’s trademarks complaint as of 2008, is factually erroneous.
313. The Commission also notes that more than a year after the last email Consim sent in 2008 raising trademark concerns, it raised additional trademark concerns with various Google employees in 2009 before Consim filed its trademark lawsuit in the Hon’ble Madras High Court near the end of September 2009. As with its 2008 communications, Consim’s pre-lawsuit 2009 communications neither contained all the information required by Google to process trademark complaints nor were they addressed to the right group to process trademark complaints.
314. Having considered all this, the Commission is of opinion that Consim did not comply with Google’s laid out procedure, as adumbrated supra, for notifying complaints under its Ad Text Policy. It failed to direct such complaints to the designated Trademark Operations Team containing the requisite information as required under the Policy. The question of Google imposing any unfair or discriminatory condition upon Consim does not arise. Consim has failed to provide evidence of imposition of any unfair or discriminatory condition upon it by Google.
315. In the result, the Commission holds that the purported claims made by Consim against Google for allegedly not enforcing its Ad Text Policy are not only not made out from the evidence presented by the DG, but the same cannot be said to raise any competition issue.
316. The Commission further finds no merit in the claim that Google failed to stop rivals using “minor variations” with spaces between words that, when combined, make up Consim’s trademarks (for e.g. “Bharat matrimony’).
PUBLIC VERSION The Commission is of the opinion that it is not within the domain of the Competition Agency to pronounce upon trademark infringement issues simpliciter unless the same raise any competition issue. Moreover, it appears that Consim’s trademark claims in relation to Google’s Keyword Bidding Policy are already pending before the Hon’ble Madras High Court. In these circumstances, it was inappropriate for Consim to have canvassed these issues projecting them as competition claims before this forum.
317. The Commission does not also approve of the approach adopted by the DG in embarking upon itself the task of examining trademark infringement issues, particularly, when the same were pending adjudication before competent courts. The Commission notes that such an approach has the potential to create chaos in the regulatory cosmos if divergent rulings are handed out by the Competition Agency and the Trademark Authority/ Civil Court. The Commission notes that the approach by the DG in addressing the instant issue was misdirected in law. There was no need to have gone into the issues of Trade Marks Act, 1999 in such detail and project isolated transactional imperfections as competition infringement. This was particularly so when Consim never triggered Google’s trademark policy as it failed to adopt Google’s trademark notification procedure.
318. In this backdrop, the Commission finds it unnecessary to dilate upon the use in ad text of “space variations” of Consim’s trademarks by Consim’s competitors.
319. The Commission, however, notes that the DG has rejected Consim’s allegation of retaliation by Google for the 2009 lawsuit. Nonetheless, the DG has recorded that a delay by Google to “whitelist” a single Consim account following notification of Consim’s September 2009 litigation PUBLIC VERSION constituted a competition law violation.
320. The Commission is unable to lend concurrence to this finding of the DG. The delay in whitelisting one account from a single website does not constitute a competition law infringement and the same in any event stands belied from the material on record. From the records, it does not appear that Google’s delay in obtaining whitelisting of a single Consim account was intentional. The Commission is satisfied with the explanation provided by Google in this regard.
321. The Commission notes that Google’s Ad Text Policy sets out that “once a trademark complaint has been processed, all relevant ads (including the trademark owner’s ads), will be rejected if they contain the trademark(s) listed in the complaint. The trademark owner must notify Google if it needs to exempt itself from the complaint by use of the AdWords trademark authorization process”. Thus, after Google processed Consim’s litigation trademark complaints, all uses of Consim’s trademark (as trademarks) in ad text were blocked, including in Consim’s own ads, until Consim authorised its accounts to use those trademarks by “whitelisting” those accounts.
322. However, the DG found that because whitelisting of one of Consim’s “child” accounts was not automatically processed with the whitelisting of its “parent” account, and instead took a couple of weeks to sort out, the conduct of Google amounted to imposition of unfair condition on Consim. This conclusion of the DG does not seem to be borne out from the material on the record.
323. The Commission is of opinion that every transactional dispute cannot be made a subject matter of anti-trust inquiry. It is the abusive conduct of the dominant undertaking as provided under the law that is proscribed.
PUBLIC VERSION AdWords API T&Cs
324. The DG has found that Google hinders advertisers from transferring ad campaigns to other ad platforms. The DG noted that the AdWords API Terms & Conditions (“T&Cs”) prevent advertisers from transferring campaigns between platforms.
325. The DG has observed that Google enters into agreements with customers licensing AdWords API from it. Google’s AdWords API agreements with third party tool developer entities contain certain restrictive clauses which have anti-competitive effects in violation of the provisions of Section 4(2)(a)(i), Section 4(2)(b)(ii) and Section 4(2)(c) of the Act. These restrictions have the potential of being used as a tool for discouraging advertisers from multi-homing, thereby resulting in denial of market access to competitors and causing other anti-competitive effects. Further, DG has noted that inclusion of a provision on termination without reason in the AdWords API terms amounts to imposition of an unfair condition on AdWords API users in violation of Section 4(2)(a)(i) of the Act.
Google’s Response
326. Google has challenged the finding of the DG stating that it makes it easy for advertisers to transfer ad campaigns and provides them with several mechanisms to do so. One of them is the AdWords API, which allows advertisers and third-party developers to use automatic campaign management tools.
327. It has pointed out through data that Indian advertisers frequently use multiple ad platforms and that they report no constraint in data interoperability. The data and the statements from Indian advertisers in the DG’s file also contradict the DG’s speculation that the AdWords API T&Cs have harmed competition.
PUBLIC VERSION The Informant’s Response
328. The Informant in Case No. 07 of 2012 has submitted that being the gateway to the internet, Google is an essential trading partner for advertisers who wish to reach out to potential customers online. While competing platforms (such as Bing Ads) exist, by virtue of its scale in general web search, Google is a necessary partner for any advertiser.
329. The Informant has further submitted that as the dominant search engine in the market and with a market share which is by any measure several times more than its nearest competitors’ in India, Google automatically receives a lion share of all search advertising (as advertisers have a strong preference for the platform that provides them with more eyeballs). However, Google requires all advertisers to sign up for the AdWords terms and conditions, which restrict advertising platform interoperability, making it prohibitively expensive for advertisers to use competing platforms along with Google’s AdWords program.
330. Advertising campaign data is an important tool for measuring the effectiveness of campaigns. In view of that, the Informant has submitted that Google erects barriers to prevent advertisers from using their own advertisement data on other advertising platforms. Since Google is the unavoidable trading partner for all advertisers participating in the search advertising market, each advertiser would first create and manage its campaign on Google’s AdWords program. Therefore, any restriction on data interoperability has the effect of creating a barrier to entry for competitors, as they limit advertisers to using only Google’s services.
331. The option of placing the same ads on multiple platforms, such as Google website and Bing, has facilitated the emergence of tool developers offering advertisers and agencies the ability to manage online advertising campaigns across platforms. Global providers of campaign management PUBLIC VERSION software include Kenshoo, Marin Software, Search Force, and DoubleClick SearchV3 (Google’s own search advertising campaign management tool).
332. In this regard, the Informant submitted that the proliferation of such tools has been restricted by the terms in Google’s AdWords API License agreement which is entered into by Google with its customers for licensing the AdWords API. It stipulates terms and conditions for access to mainly big advertisers and third-party tool developers. The Informant submitted that several terms and conditions contained in the agreement are extremely restrictive, and effectively prohibit data portability between different advertising platforms. While Google voluntarily revised some of these restrictions in 2013 for a period of 5 years, pursuant to the FTC investigation, and this revision is being voluntarily applied by Google worldwide, it is required to implement the same only in the United States of America (U.S.) and it can withdraw these changes elsewhere anytime. But, even after these changes, the agreement remains restrictive in several ways and this been examined by the DG pre-and-post 2013.
333. The Informant submitted that Google is the first platform on which the Informant creates and manages its search advertising campaigns, as its (Google) search engine receives significantly more traffic than other search engines. The cost associated with managing advertising campaign is significant. The restrictions on data portability make it time consuming for the Informant to manage its campaigns on different platforms together. The reason being, it cannot transfer data from one platform to another. Other advertising platforms such as Bing and Yahoo have a simple process of porting the advertising campaign because they are aware that advertisers will advertise on their platforms only if they can transfer campaign data from Google’s AdWords (where they continuously update and manage their campaign) without employing significant amount of PUBLIC VERSION additional time and resources.
334. During investigation, the Informant demonstrated the time and cost involved, and that use of competitors’ tools such as Bing Ads and Bing Editor, which use AdWords API, are not suited for advertisers running multiple campaigns. The Informant also demonstrated the difficulty faced in achieving interoperability from Google’s AdWords to Bing. For porting 1 ad account from Google’s AdWords to Bing Ads, the Informant required 22 steps which took more than 33 minutes to complete. This 1 account of the Informant ran 56 campaigns with 117 ad groups and 37,964 key- words. As of December 2013, the Informant had 32 accounts, running 3,808 campaigns with 4,09,873 ad groups bearing 11,74,468 text advertisements and 35,47,574 keywords. The Informant has estimated that it would take 2,142 minutes or 35.7 hours or 4.5 working days to move all of its own campaign data from Google’s platform to Bing – and that too once.
335. The Informant submitted that whilst, easier methods are technically feasible, they have been contractually disallowed by Google, which causes advertisers harm. It also causes Google’s competitors significant harm, as they lose out on potential customers who do not have the time or resources to manage a separate campaign on, for example, Bing Ads.
336. The Informant went on to submit that Google’s argument that most advertisers have accounts on multiple search advertising platforms, is fallacious. The appropriate test for competition in the market is not whether Google’s competitors have advertiser accounts set up on their platforms, but rather whether advertisers regularly update and manage such accounts, and how much of their total search advertising revenues are spent on such accounts. Given Google’s pre-eminence in the online general web search market in India and, therefore, the search advertising PUBLIC VERSION market in India, lack of interoperability between Google and its competitors truly chokes Google’s competitors of traffic.
The Commission’s Analysis
337. The Commission notes that the DG has raised objections against the T&C for the use of Google’s AdWords API. As discussed earlier, AdWords is Google’s advertising platform. Through AdWords, advertisers bid on specific keywords entered by users. They can also define other targeting variables, such as time of day, location, and language. Many advertisers manage their AdWords campaigns through the AdWords website.
338. Google has submitted that it licenses (for free) a proprietary API for AdWords. The AdWords API allows advertisers and third-party developers to create tools that advertisers can use to manage their AdWords campaigns automatically. Because the AdWords API involves use of Google proprietary technologies, tools using AdWords API require a license from it (Google) to do so. The AdWords API T&Cs govern the use of the AdWords API. The AdWords API T&Cs leave advertisers free to use their AdWords data as they deem fit. They permit third-party developers to create tools that allow advertisers to transfer campaigns and run campaigns on multiple platforms (known as “multihoming”). Thus, advertisers can run campaigns simultaneously on Google, Bing, Facebook, Twitter, and Baidu, with tools that make use of the AdWords API.
339. Google has pointed out that the DG’s market investigation showed that no advertiser or ad agency had conveyed that Google’s AdWords API T&Cs restricted multihoming across ad platforms or data interoperability. Every advertiser questioned by the DG confirmed that they faced no such restrictions. MakeMyTrip’s response is typical, noting that it had not faced “any difficulties” related to “interoperability of data across different PUBLIC VERSION advertising platforms” and that it had not “come across any issues related to the Google [AdWords] API”.
340. Yet, the Investigation Report notes that the AdWords API T&Cs “restrict the ability of advertisers to manage the ad campaigns across search platforms.” Specifically, the Investigation Report objects to two groups of AdWords API T&C provisions:
(i) Sections III(2)(c)(i)-(ii) (the “Input and Copying Clauses”), which Google removed from the AdWords API T&Cs in January 2013; and
(ii) Certain clauses that remained in place or were added to the AdWords API T&Cs when Google removed the Input and Copying Clauses (the “Post-2013 Clauses”).
341. The DG notes that these provisions had the consequence of limiting the ability of the developers to design tools for efficient management of online campaigns across search platforms. And it alleges that Google restricted the advertisers relying on third party tool […] from multihoming.
342. Google has argued that AdWords API T&Cs have not harmed competition in India, Indian advertisers, or any third-party tools developers. To the contrary, challenged provisions of the AdWords API T&Cs serve important pro-competitive purposes. It was pointed out that the Input and Copying Clauses ensured that AdWords’ innovative features were available to advertisers. The Input and Copying Clauses were license limitations that defined the scope of the license that Google granted for its AdWords API. These limitations applied only to ad management tools developed by third parties. They did not apply to tools developed by advertisers.
PUBLIC VERSION
343. Further, that the Input and Copying Clauses protected the functionality and user experience of AdWords. When an advertiser uses a tool developed by a third party to manage its AdWords campaigns, that tool controls the advertiser’s experience of AdWords. The tool affects the advertisers’ perception of the functionality and performance of AdWords.
344. The data at issue involve the settings and targeting parameters that an advertiser chooses for its advertising campaign. But different advertising platforms do not offer the same functionality and targeting parameters. For example, AdWords allows advertisers to specify granular time settings in 15 minute intervals. Other platforms offer much less granular time settings. As a result, if a third-party tool were to use common input fields for the settings of different ad services or automatically copied settings of one ad service to another, this could have adverse consequences for the advertiser and the advertiser’s perception of AdWords quality.
345. The DG’s conclusion that AdWords API T&Cs have harmed competition is incorrect. Impugning the finding of the DG that the Input and Copying Clause restricted advertisers relying on third party tools […] from multihoming, Google has pointed out that the DG’s own market investigation found that the advertisers [DG] wrote to have not expressed any constraint in data interoperability. The Brazilian antitrust authority, CADE, conducted a similar inquiry with advertisers and found that advertisers can easily multihome and that AdWords API T&Cs raise no concerns.
346. Google has pointed out that major developers, such as Kenshoo and Marin Software, provide ad campaign management tools with multihoming functionality, including cross-platform synchronisation and optimization, and actively encourage advertisers to multihome. This contradicts the DG’s finding that AdWords API restricts advertisers relying on third-
PUBLIC VERSION party tools from multihoming.
347. The Commission notes that the DG does not take into account the fact that one of the major reasons why developers license AdWords API is to create tools that help advertisers port their campaign data. The AdWords API T&Cs define the scope of the license for the use of AdWords API and thus, makes multihoming easier.
348. The DG has taken exception to a number of clauses Google introduced since 2013 (the Post-2013 Clauses) in AdWords API without specifying as to how the Post-2013 clauses restrict competition. DG’s objections to the Post-2013 Clauses largely seem to rest on the assumption that there exists certain non-clarity in the clauses. Before examining the issue in detail, the Commission is of opinion that “non-clarity” does not in itself raise a competition law concern much less amount to contravention of the provisions of the Act.
349. As regards the Post-2013 Clauses, Google has explained that these clauses and new AdWords Policies serve legitimate business purposes related to protecting advertisers and the AdWords platform. Specifically,
(i) Bi-Directionality Clause in Section III(2)(f)(iv) requires that third-
party tools provide reciprocal campaign data transfer functionality and this promotes advertiser multihoming.
(ii) Termination and Inspection Clauses in Sections IV(12) and IV(3)(A) are standard provisions which allow Google to determine whether licensees comply with the AdWords API T&Cs and to terminate their access to the API if they do not. In reality, Google has not used to terminate an API license arbitrarily or for illegitimate, anticompetitive purposes.
PUBLIC VERSION
(iii) Automated Use Clause in Section III(2)(m) Automated Use Clause requires licensees do not permit third parties to gain automated access to the AdWords platform without using their own API access credentials.
(iv) Transfer of AdWords API Report Data Clause in Section II(3)(a) protects advertisers by preventing licensees from transferring advertisers’ data related to campaign performance to third parties.
(v) Reporting Clause in Section III(2)(c)(iii) allows AdWords campaign performance data to be reported either (i) separate from performance data from other platforms, or (ii) aggregated with performance data from other platforms. This benefits advertisers and protects AdWords by allowing advertisers to understand how AdWords campaign performance compared to performance on other platforms.
350. The Commission takes note of Google’s submissions as well as the fact that in February 2015, Google moved the Post-2013 Clauses addressed in the Report, except the Inspection and Termination Clauses, from AdWords API T&Cs to the AdWords API Policies. At that time Google clarified some of the clauses, which eliminate the possibilities of any “non-clarity” as is alleged by the DG. After considering the matter in proper perspective, the Commission is of the opinion that the findings of the DG holding contravention of the provisions of Section 4 of the Act in respect of the AdWords API T&Cs are not sustainable. The Commission holds that advertisers are provided with different mechanisms to manage the parameters of their ad campaigns For Example the keywords for which they want to bid, the bid amount, the overall budget that they want to spend, the time period during which their ads should run, and the locations where they should show as well as the export parameter data from AdWords to rival ad platforms. The classic means for doing this – used PUBLIC VERSION by many AdWords advertisers – are the AdWords WebUI and the AdWords Editor:
(i) AdWords WebUI AdWords WebUI (formerly the AdWords Frontend) is a free website that allows advertisers to create and modify AdWords campaigns. Advertisers can export campaign parameter data directly from the AdWords WebUI into machine-readable, industry-standard formats and easily transfer the data into other ad platforms. ]
(ii) AdWords Editor Introduced in 2006, AdWords Editor is a free downloadable application that manages AdWords campaigns. The AdWords Editor allows advertisers to export a complete AdWords account with a single action.
351. Google introduced the AdWords API in 2005 as an additional means for advertisers to manage campaigns and export their data from AdWords. The AdWords API is a set of technologies that enables advertisers and third-party developers to build and use tools that manage AdWords campaigns automatically Third-party developers have used the AdWords API to build tools that allow advertisers to manage their AdWords campaign automatically and export their AdWords data to other ad platforms.
352. The Commission notes that the DG also recognises that “constant innovation and new development” characterise the advertising campaign management tool industry. Both third-party developers and rival ad platforms have used the AdWords API to create sophisticated ad campaign tools that enable campaign management and multihoming across different platforms. In addition, Google provides advertisers with PUBLIC VERSION several different tools that allow them to export their AdWords campaign data.
353. The Commission notes that the DG’s market investigation has revealed that advertisers face no barriers to multihoming across different ad platforms. This is evident from the Report which acknowledges that “[t]he advertisers [the DG] wrote to have not expressed any constraint in data interoperability”. The Commission finds merit in the contention of Google that if advertisers believed Google’s T&Cs imposed obstacles to using other platforms, advertisers surely would have lodged objections. Instead, they attest to the opposite:
(i) MakeMyTrip reported that it has not faced “any difficulties” related to “interoperability of data across different advertising platforms”. MakeMyTrip further clarified that it has not “come across any issues related to the Google API”.
(ii) Yatra.com stated that it has not encountered “any major technical or other constraint in data interoperability while working on such different platforms”.
(iii) Flipkart and JustDial reported that they have not encountered any issues related to unfair conditions imposed through the AdWords API T&Cs.
354. In the result, the Commission is of the opinion that the AdWords API T&Cs do not impair data interoperability between search advertising platforms. Hence, the findings recorded by the DG on this issue are legally unsustainable and are set aside.
Distribution Agreements
355. The DG found Google to have contravened Section 4(2)(c) of the Act on the ground that two of its distribution agreements (i.e. Google’s agreement PUBLIC VERSION with Apple for its Safari browser and Google’s agreement with Mozilla for its Firefox browser) set Google as default search engine.
356. For appreciation of competition concern highlighted by the DG, it would be apposite to excerpt the finding of the DG from the Investigation Report itself:
…[i]t is found that the appearance of a default search engine of a dominant entity like Google through long term contractual arrangements also have the potential to strengthen its market position in online general web search and search advertising by denying access to others. It acts as an additional tool through which Google’s dominance in entrenched. This amounts to a denial of market access to competing search engines in markets for in violation of Section 4(2)(c) of the Act…
Google’s Response
357. Google has pointed out that the Report objects to two search distribution agreements – Google’s agreement with Apple (for its Safari browser) and Mozilla (for its Firefox browser). The Report does not claim that these agreements are exclusive. It only alleges that because Google is set as the default option, this “amounts to a denial of market access to competing search engines”.
358. According to Google, a default setting does not deny market access to competitors. Defaults simply provide a convenient way for users to access a preferred search service. Users can easily switch away from the default if they so choose to. In fact, Microsoft Bing’s website explains that it is “super easy” to use Bing from any web browser. Microsoft explains that in a few simple clicks Firefox users can set Bing as their search provider.
359. Google has contended that the Investigation Report identifies no evidence to suggest that the two distribution agreements have denied market access to rivals. In fact, Google lost one of these two distribution deals i.e. the PUBLIC VERSION Mozilla agreement, in 2014. Other search services – including Yahoo!, Yandex, and Baidu – are now the default providers on Mozilla’s Firefox browser in some countries. For other countries, Mozilla was not contractually obliged to set any particular search service as default. The Mozilla experience demonstrates that there is ample competition for distribution deals. It was highlighted that Microsoft controls search distribution deals with all major PC OEMs and sets Bing as the default search service (via Internet Explorer and Edge).
The Commission’s Analysis
360. The Commission has examined the DG findings and the material available on record. The Commission notes that search services use various distribution channels to provide their services to users. For example, a web browser or a PC OEM may set a particular search service as default for search access points. If a user types a search query into the address bar of the Apple Safari browser, the results will be provided by Google. The Investigation Report objects to two search distribution agreements – Google’s agreement with Apple (for its Safari browser) and Mozilla (for its Firefox browser). The Commission notes that the DG seems to have been swayed by the fact that Google is set as the default option under these agreements that this amounts to a denial of market access to competing search engines Thus, the charge against Google is not that these agreements create exclusivity for Google. Rather, the conclusion of the DG is based upon the fact that such contractual arrangements by Google have the potential to strengthen its market position in online general web search and search advertising by denying access to others.
361. Thus, the entire finding is based upon the supposition that through such contractual arrangements, Google has the “potential” to strengthen its market position to the exclusion of other search engines. The Commission is of opinion that a default setting does not deny market access to PUBLIC VERSION competitors and users are free to switch away from the default if they so choose to. The DG has presented no evidence to show that these two distribution agreements have denied market access to rivals.
362. Moreover, the distribution agreements are contestable and Google is stated to have lost one of the two distribution deals namely, the Mozilla agreement in 2014. Other search services – including Yahoo!, Yandex, and Baidu, are now the default providers on Mozilla’s Firefox browser in some countries. Besides, the Commission notes that Microsoft controls search distribution deals with all major PC OEMs and sets Bing as the default search service (via Internet Explorer and Edge).
363. In view of the foregoing, the Commission is of the opinion that Google’s distribution agreements are neither exclusive nor has it been established that such arrangements have denied market access to any of the competing search engines. The two browser distribution deals with Mozilla’s Firefox and Apple’s Safari, are not exclusive. They merely specify that Google should be the default search service on these browsers. The user, however, is not obliged to use that search service. In other words, default arrangements do not hamper a user’s ability to access any other search service, such as Yahoo!, Flipkart, Jabong, or MakeMyTrip. The Commission notes that the DG found the default settings to create competitive problems because the process for selecting another search service is not apparent for ordinary internet users. The Commission notes that such a finding of the DG does not appear to be based on records of any survey or evidence to that effect. Moreover, the Commission holds that default setting cannot be equated with exclusivity because default arrangements leave partners free to provide users with other search service options as well. Thus, browsers typically list several other search options directly within the browser.
PUBLIC VERSION
364. This is borne out from an illustration given from Mozilla Firefox where it lists other search options directly within the browser:
365. Thus, it can be seen that there is a drop-down menu on the Firefox browser that allows a user to choose a search service with just two clicks. Similarly, Google has pointed out that Safari also makes it easy for users to choose a general search service. iOS Safari’s standard home screen shows the icons of several search engines, including Bing and Yahoo!, which users only need to tap to enter their query as shown below:
366. The Commission has considered the submission of Google that default setting arrangements leave other search service access options entirely free. Without a search service set as default, a browser could not support PUBLIC VERSION all of its features “out-of-the box”, such as the ability for users to conduct a search simply by entering a query in the browser search box. This would severely diminish user experience. Google has urged that its distribution agreements are pro-competitive.
367. Having considered the submissions and reasons therein, the Commission is of the opinion that no case of contravention of the provisions of Section 4(2)(c) of the Act is made out against Google with regard to distribution agreements with browsers which neither create any exclusivity nor do they deny market access to competing search engines. The users are free to switch search engines in their browsers and it is intrinsic for default settings to have only one search engine. In the absence of any restriction in switching, the “potential” concern highlighted by the DG can in itself be no ground to hold Google in contravention. Accordingly, the findings of the DG cannot be sustained and are set aside on this count.
Intermediation Agreements
368. The DG has noted that apart from online web search services, Google also offers online search and advertising services on other websites through Syndication/ Intermediation services. With regard to advertising, intermediation can take place for both search and non-search advertising. Google offers Syndication services under its AdSense program. The Online Search and Advertising Syndicate Services constitute distinct relevant markets. By virtue of its position of strength in the relevant markets of Online General Web Search Services and Online Search Advertising Services, Google is also a preferred Syndicate service provider for publishers wanting to offer search and advertising services on their websites. Google is using its dominant position in Online General Web Search Service and Online Search Advertising Service to impose certain restrictive conditions in its agreements for syndicate search and advertising services in violation of Section 4(2)(e) and Section 4(2)(c) of the Act. The nature of restrictions varies across types of agreements.
PUBLIC VERSION While in some cases partners are prohibited from using competing services, in other there are restrictions related to the manner of placement of ads of competitors. These restrictions prevent competing service providers from achieving the necessary scale which results in creation of entry barriers for them.
369. Further, it was noted by the DG that the policy and conduct of Google, prior to May, 2010, for not disclosing AdSense Revenue sharing with online AdSense partners, amounted to imposition of unfair conditions on them amounting to infringement of Section 4(2)(a)(i) of the Act. Lastly, it was noted by the DG that such agreements are one-sided and provide enough scope for arbitrary conduct without a fair opportunity to the other party. This amounts to imposition of unfair conditions by Google within the meaning of Section 4(2)(a)(i) of the Act.
Google’s Response
370. Google has in response argued that Google offers a range of different types of intermediation agreements in India. Broadly speaking, these agreements enable website publishers to show Google search results or ads on their websites. Publishers choose to enter into these agreements because they create value both for publishers themselves and for users of their websites. It was submitted that these agreements do not create exclusivity nor do they harm competition.
371. Further, Google’s online intermediation agreements are non-exclusive. The Investigation Report objects to a clause in Google’s online intermediation agreements requiring publishers to distinguish between Google and non-Google search services and ads, so that they cannot reasonably be confused by users. The Report claims this non-confusion clause provides some scope for Google to interpret these provisions in a manner that in effect imposes exclusivity. The Investigation Report, PUBLIC VERSION according to Google does not claim that the non-confusion clause in fact amounts to exclusivity; that Google has ever actually interpreted it in this way; or that any publisher has ever felt bound by exclusivity. In fact, Maps of India confirmed that there is nothing in Google’s intermediation agreements that prevents it from displaying non-Google ads.
372. The Investigation Report claims that Google might hypothetically interpret its agreements to impose exclusivity, is a mere speculation. It does not mean that Google does in fact do so. Under the Act, the Report needs to show that an abuse has actually occurred; not merely that an abuse is a hypothetical possibility.
373. The non-confusion clause could not, in any event, impose exclusivity as it does not preclude publishers from showing third-party search functionalities and ads alongside Google’s. All that the clause requires is that publishers appropriately label or distinguish third-party features so that users can tell the difference between Google and non-Google functionalities. Elaborating further, it was submitted that Google’s direct intermediation agreements are non-exclusive and immaterial for India. Direct intermediation agreements are individually negotiated agreements.
374. The Investigation Report identifies three aspects of the template clause in direct intermediation agreements which Google uses as a basis for negotiation and which are objectionable :
(i) Google’s direct search intermediation agreements that asks publishers not to show search functionalities that are “same or substantially similar” to Google’s on the same site.
(ii) Template minimum ad and ad placement clauses in Google’s direct search ad intermediation agreements.
(iii) Historic clause in Google’s direct display ad intermediation agreements that asked publishers not to show non-Google display PUBLIC VERSION ads on their sites that were “the same as or substantially similar” to Google’s display ads.
375. Referring to the Investigation Report’s claims that, as a result of these provisions, Google’s direct intermediation agreements deny intermediation rivals access to publishers; it was submitted by Google that these concerns are not correct because:
(i) The publishers remain free to include non-Google search functionalities that are not the same or substantially similar as Google’s.
(ii) The contested minimum ad and ad placement clauses do not prevent publishers from showing non-Google ads in addition to Google ads.
(iii) Google’s direct display ad agreements do not create exclusivity.
376. More importantly, Google’s intermediation agreements do not harm competition. The Investigation Report does not define any markets where intermediation agreements restrict competition nor does it provide evidence of competitive harm in any such putative markets. Further, Google’s intermediation agreements are not a significant channel for generating search and ad business and therefore, cannot deprive rivals of query scale in the manner alleged by Microsoft.
377. The Report claims that Google’s terms of trade entered into with its intermediation partners constitute an abuse of dominance. But the Report does not claim or establish that Google is dominant in search intermediation, search ad intermediation, or display ad intermediation. It is also submitted that the specific provisions of Section 4 of the Act that the Investigation Report invokes do not apply.
PUBLIC VERSION The Informants’ Response
378. The Informants have, however, supported the DG’s conclusions.
The Commission’s Analysis
379. To appreciate the issues, it would be appropriate to first understand the ecosystem surrounding Google’s intermediation agreements which allow website publishers to benefit from Google’s technologies. In this regard, it is observed that Google’s intermediation agreements allow website owners (“publishers”) to incorporate Google’s search technology and ads on their websites. The users can then conduct searches directly on the publisher’s site and publishers earn revenues from Google ads shown on their websites’ pages.
380. Google claims that its intermediation agreements provide at least five distinct benefits to Indian publishers, advertisers, and users:
(i) First, publishers benefit from an incremental revenue stream which they might not get otherwise.
(ii) Second, they allow the same publishers to benefit from the opportunity to make their websites’ content more easily discoverable through the power of Google’s search technology.
(iii) Third, they allow publishers these benefits without having to make significant investments.
(iv) Fourth, they give users the ability to conduct searches directly on the publisher’s website.
(v) Fifth, they allow advertisers broader access to relevant advertising opportunities.
PUBLIC VERSION
381. The Commission notes that Google offers different types of intermediation agreements, namely:
(i) Search intermediation agreements Google’s search intermediation agreements enable publishers to add a Google search bar to their website. The search bar allows users to search the pages of that particular site or the web at large (at the discretion of the publisher). Google’s algorithms generate the search results shown.
(ii) Search ad intermediation agreements – AFS Google’s search ad intermediation agreements, called AdSense for Search (“AFS”) enable publishers to show Google search ads when users enter search queries on their sites. Advertisers pay when users click on these ads. Google then shares the revenue from these clicks with publishers
(iii) Display ad intermediation agreements – AFC Google’s display ad intermediation agreements, called AdSense for Content (“AFC”) enable publishers to show Google display ads that relate to the content of a page or the user’s interest based on his or her past browsing history. These ads are not targeted based on a search query, but based on the content of the relevant page. As under AFS agreements, Google shares the revenue from display ads with the publisher.
PUBLIC VERSION
382. These intermediation agreements come in two forms:
(i) Online agreements Online agreements are standard contracts that interested publishers can sign-up to through an online form. The large majority of Google’s intermediation agreements are online agreements.
(ii) Direct agreements Direct intermediation agreements are agreements that Google negotiates individually, typically with some larger publishers. Google has just direct intermediation agreements in force with Indian publishers. All cover display ad and search ad intermediation, and one also covers search intermediation.
383. Google has explained as to how its different intermediation agreements interrelate:
384. The DG held Google’s intermediation agreements being violative of the provisions of Section 4 of the Act essentially on the ground that Google’s terms and conditions have the potential to create exclusivity. The findings PUBLIC VERSION of the DG in this regard are:
These prohibitions imposed under negotiated Search Agreements are therefore found to be unfair, as they restrict the choice of these partners and disallow them from using the search services provided by competing search engines along with Google, amounting to violation of section 4(2)(a)(i) of the Act.
[Ref. Para 44, p. 588 Investigation Report] Based on above analysis it is thus found that Google is using its dominance in market for Online General Web Search to impose restrictive conditions in online syndicate search agreements to strengthen its position in the market for online syndicate search services in violation of section 4(2)(e) of the Act. Its competitors in the search syndicate business have been denied market access to the online search syndicate market is violation of Section 4(2) (c) of the Act. Such syndicate agreements constitute an important source of traffic for Google’s competitors.
[Ref. Para 51, p. 590 Investigation Report] Google has thus used its dominance in the market for online General web search and online search advertising to impose restrictive conditions in AFS Agreements to enhance its position in online search advertising syndication market in violation of section 4(2) (e) of the Act. The conduct also amounts to denial of access to competing search engines to the search advertising syndicate market in violation of section 4(2)(c) of the Act. Through this conduct it is further able to strengthen its strong hold in online search advertising market.
[Ref. Para 80, pp. 601-602 Investigation Report]
385. The Investigation Report claims that Google’s direct search intermediation agreements violate Section 4(2)(a)(i) of the Act because they involve imposition of unfair conditions on publishers. However according to Google, its direct search intermediation agreements do not PUBLIC VERSION contain any terms that amount to exclusivity. Direct search intermediation agreements are negotiated and if a publisher does not want to enter into such an agreement, it can always choose to enter into an online agreement which does not contain the aforementioned allegedly unfair terms. Further, the conduct that imposes exclusivity obligations that affect only a small portion of demand and therefore cannot deny market access.
386. As for the violation of Section 4(2)(e) of the Act, Google has submitted that Section 4(2)(e) is predicated on the existence of two distinct, relevant antitrust markets. The Investigation Report claims that Google leverages into a market for search and ad intermediation but it does not properly define such markets. Moreover, Google cannot leverage its strength in general search and online ads to impose restrictive conditions in online search intermediation and AFS agreements. It extensively negotiates clauses in direct AFS agreements with large and sophisticated publishers and if publishers do not want to agree to the template clauses in direct agreements, they are free to sign up to the online agreement.
387. The Commission has perused the findings of the DG as well as Google’s submissions carefully. While dealing with Google’s Online Customs Search Agreement, the DG examined the following clauses contained therein :
“Clause 1.5 During the Term, if You provide non-Google search services on any Site, You will ensure that such non-Google search services cannot reasonably be confused with or mistaken for those provided by Google. You further understand that Google will provide the Service to You on a nonexclusive basis, and that Google will continue to customize and provide its services to other parties for use in connection with a variety of applications, including search engine applications.
PUBLIC VERSION Clause 1.6: The Results Page or associated elements provided by Google through the Service may contain advertising, which You agree to display. The Service is compatible with, and You may apply for, the Google AdSense program (www.google.com/adsense), subject to the AdSense Terms and Conditions. If You provide Your own advertisements or third party advertisements in connection with the Results provided by the Service, You will ensure that such advertisements cannot reasonably be confused with or mistaken for those provided by Google.”
[Ref. Para 24, p.579 Investigation Report]
388. Further, the DG examined Google’s Online Site Search Agreement. The relevant clauses pertaining to quoted below:
Clause 1.5
1. From Google. Customer may choose to display Ads on the Results Page in its sole discretion via the Admin Console. If Customer elects to display Ads on the Results Page, Customer must register for an AdSense Account and be subject to Google’s terms and conditions as they relate to the placement of Ads.
2. From Customer or Third Parties. Customer may display its own advertisements, or third party advertisements, on the Results Page. If Customer chooses to do so, it must ensure that these advertisements cannot be confused with Ads.
[Ref. Para 26, pp. 580-581 Investigation Report]
389. From the above, the DG deduced that in so far as Google Custom Search and the online agreement for Google Site Search are concerned, there does not appear to be any prohibition on the use of third party search services or display of third party ads in the search results. The requirement being put in place by Google pertaining to third party search/ advertising PUBLIC VERSION services is that the display of search results or ads must not be done in a manner that can reasonably be confused with or mistaken for Google’s services.
390. Accordingly, the DG returned the following finding:
On the face of it these restrictions on the use of search services or display of ads that might be confused with Google appear to provide a reasonable check that Google would like to have to guard their bonafide interests. However, the provisions are drafted in a broad manner without clarifying what might be construed as something that can “reasonably be confused with or mistaken” for a service provided by Google. Therefore, considering the broad nature of the conditions there is some scope for Google to interpret these provisions in a manner that in effect imposes exclusivity.
[Ref. Para 29, p. 582 Investigation Report]
391. The Commission is of the opinion that the observations of the DG, as reproduced/ quoted hereinabove, cannot be termed as a finding. The Commission finds it egregious when the DG notes that “there is some scope for Google to interpret these provisions in a manner that in effect imposes exclusivity”. Such an observation can find a place in the realm of speculation and the Commission has no hesitation in holding that no exclusivity, de jure or de facto can be said to flow from Google’s online search intermediation agreements, as cited above.
392. That takes the Commission to Google’s negotiated search intermediation agreements. On a perusal of the Investigation Report, it appears that in case of Google’s direct (i.e. negotiated) search intermediation agreements, the Investigation Report finds one template clause for direct search intermediation agreements which asks publishers not to implement search technologies on their sites that are “same or substantially similar” to that of Google (the “substantially similar search clause”), as objectionable.
PUBLIC VERSION
393. With regard to this clause, the DG has noted that by virtue of substantially similar search clause in the agreements, Google imposed prohibitions under direct such intermediation agreements which are unfair. They restrict the choice of these partners and disallow them from using the search services provided by competing search engines along with Google.
394. A perusal of the stipulation evidences that Google, in abuse of its market power, has imposed restrictive conditions upon the publishers in negotiated search agreements. The Commission does not find merit in the contention of Google that publishers remain free to include non-Google search functionalities that are not the same or substantially similar as Google’s. The Commission finds that Google prevented partners with whom it entered into negotiated search agreements from implementing on their websites any search services which are the same or substantially similar to Google’s search service. The Commission observes that while Google is a crucial partner for publishers seeking web search syndicate services for their websites, they may wish to offer search services from websites other than Google. Therefore, the Commission holds that the prohibitions imposed under the negotiated search agreements are evidently unfair and they restrict the choice of the partners and prevent them from using the search services provided by competing search engines. Thus, Google has imposed unfair conditions on publishers which amounts to a violation of Section 4(2)(a)(i) of the Act.
395. The Commission further notes that by restricting websites from partnering with competing search services, Google was denying its competitors access to the search business and further marginalizing competitors and endangering their viability, while strengthening its own position. These restrictions amount to a de-facto imposition of online search exclusivity which is a contravention of Section 4(2)(c) of the Act.
PUBLIC VERSION
396. The Commission does not find merit in Google’s argument that it has only web search syndication deals with Indian partners. The Commission notes that the volume of business generated through these agreements is substantial as can be seen from the table containing data on the revenue earned by Google from negotiated intermediation agreements:
Data of Syndicate Web Search Agreements Total number of queries Fee collected from Web Search Direct partners 2011: 2011:
(negotiated
agreements) 2012: 2012:
2013: 2013:
397. The exclusive agreement needs to be examined in light of the importance of scale and network effects in online search and online search advertising and Google’s substantial market shares in these markets. Viewed in this context, the exclusive search intermediation agreement ranging for 2-3 years reinforces the network effects enjoyed by Google, leaving no room for the competitors to attain sufficient scale to be seen by publishers as credible intermediation partners. This conduct creates conditions for extending and preserving Google’s dominance in search intermediation in perpetuity. Google was using its dominance in the market for online general web search to impose restrictive conditions in online syndicate search agreement, in violation of Section 4(2)(e) of the Act. Further, as mentioned above, competitors were denied access to the online search syndication services market, a contravention of Section 4(2)(c) of the Act is also made out.
PUBLIC VERSION
398. This brings the Commission to Google’s ad intermediation agreements. In case of Google’s direct ad intermediation agreements, the Investigation Report’s objections are in respect of the following clauses:
(i) Two template clauses for direct search ad intermediation agreements that asks publishers to request a minimum number of Google’s ads per search query and define the placement of Google and non-Google ads (the “minimum ad and ad placement clauses”). Specifically, this clause implies that ads which are the same as or substantially similar to Google’s AFS ads should not appear above or directly adjacent to Google’s AFS ads; and AFS ads should be displayed in a single continuous block, not interspersed with other ads or content.
(ii) A historic template clause for direct display ad intermediation agreements that asked publishers not to show display ads on their sites that were the “same or substantially similar” to Google’s display ads (the “substantially similar display ad clause”).
399. With regard to the aforementioned clauses, the DG noted that Google has used its dominance in the market for online general web search and online search advertising to impose restrictive conditions in AFS Agreements and to enhance its position in online search advertising syndication market in violation of Section 4(2)(e) of the Act. It was also noted that the conduct amounted also to denial of access to competing search engines to the search advertising syndicate market in violation of Section 4(2)(c) of the Act. It was further noted that through this conduct, Google is further able to strengthen its strong hold in the online search advertising market.
400. Similarly, in respect of negotiated AFC agreements, it was concluded by the DG that on account of advantage of scale and large advertiser base Google is an attractive syndicate service provider for display advertising.
PUBLIC VERSION Further, it was noted that several publishers opt for Google for various intermediation services like search, search advertising and display advertising simultaneously. Thus, Google was found to be using its dominant position in online general web search and online search advertising to impose restrictive conditions in its agreements for syndicate services for display advertising in violation of Section 4(2)(e) and 4(2)(c) of the Act.
401. Before proceeding to examine the direct (negotiated) AdSense agreements (AFC and AFS), it would be appropriate to first deal with the objections raised by the DG with regard to online AdSense agreements. Though, the DG has noted that the terms of such agreements do not contain any prohibition on the use of competing services for display of ads, however, it was also noted that according to Clause 1.5 of the online AdSense terms and conditions, if an AdSense partner chooses to opt for the Custom Search Engine service of Google, it is also required to accept Google’s Custom Search Engine Terms of Service. Therefore, it was noted by the DG that Clause 1.6 of the Custom Search Engine Terms (quoted supra in Para 389), would be applicable to AdSense partners who opt for the Custom Search Engine service of Google, as it requires them to ensure that their own or third party advertisements cannot reasonably be confused with or mistaken for those provided by Google.
402. While the DG found this clause to be a reasonable check that was needed by Google to safeguard its interests; considering the broad nature of the condition, the DG noted that there was ample scope and incentive for Google to impose exclusivity through this condition.
403. The Commission is of the view that for the reasons given in the context of online search intermediation agreements, the aforesaid observations of the DG are insufficient to return a finding of contravention against Google.
PUBLIC VERSION
404. Coming now to the direct (negotiated) AdSense agreements (AFC and AFS), Google challenged the findings recorded by the DG which have been noticed hereinabove.
405. Google argued that with regard to negotiated AFS agreements, the contested minimum ad and ad placement clauses do not prevent publishers from showing non-Google ads in addition to Google ads. Google pointed out that the DG’s Report itself recognises that these clauses “do not prohibit the display of online search ads from competing search engines”. Moreover, with regard to negotiated AFC agreements, substantially similar display ad clause does not create exclusivity either and again, publishers can show non-Google display ads in addition to Google display ads.
406. The Commission has examined the DG’s findings and the material available on record. The Commission notes that none of the clauses noted by the DG prevents publishers from accessing other ad intermediation operators. They leave publishers free to show third-party ads. The clauses at issue are, moreover, template clauses for agreements that Google negotiates with large and sophisticated publishers. Moreover, publishers have always the option to conclude online ad intermediation agreements, which do not include the challenged clauses. The Commission further notes that these clauses are legitimate means to protect its brand by Google and monetise the free results and technologies provided by it (Google) to third-party websites.
407. With regard to negotiated AFS agreements, the Commission notes that the DG’s Report objects to the minimum ad and ad placement clauses in these agreements on the ground that these deny rival operators access to publishers’ ad space. The DG in its Report also claims that even if third- party ads appear, the clauses ensure that “their chances of being clicked PUBLIC VERSION remain significantly low”.
408. The minimum ad and ad placement clauses leave publishers free to display third-party ads, including in prominent space. The clauses only apply to ads that are the same or substantially similar to Google’s. But publishers can place non-Google ads that are not the same or substantially similar to Google search ads above or alongside Google search ads. This includes:
(i) ads of a different type to Google’s ads (e.g., display ads as opposed to Google search ads); and
(ii) ads of a different format than Google ads (e.g., non-text search ads where the agreement with Google only provides for text ads).
409. As an example, Indian direct partner Info Edge (Naukri.com) is showing a non-Google display ad above the Google AFS ads on its SERP, without breaching the terms of its direct agreement with Google:
PUBLIC VERSION
410. The Commission further notes that publishers can choose to include as many search ads on their search results pages as they like. Even if a publisher shows Google ads in response to a query, it can still show non- Google ads on the same page. As an example, direct partner RightHealth is using both Google and Bing search ads at the same time. The publisher shows four Bing ads together with two Google ad, this is not held to be breaching the terms of its direct agreement with Google.
411. The Commission also notes that for direct intermediation partners, Google makes specific investments and commits dedicated resources, including offering higher level of implementation and technical support, enhanced reporting tools, and optimisation assistance to publishers. The Commission is of the opinion that minimum ad and ad placement clauses are in a way a trade-off for the relationship-specific investments that Google makes in direct ad intermediation agreements. Thus, the Commission does not see any objection to Google seeking minimal assurances from publishers in terms of the number of ads they will show and their placement.
PUBLIC VERSION
412. With regard to negotiated AFC agreements, the DG’s objections are against direct display ad agreements which relate to a historic substantially similar display ad clause. The Report recognises that this clause no longer applies to whole sites, and the requirement concerns only pages showing Google’s display ads (AFC ads). The Report goes on further to claim that the substantially similar display ad clause denies to its rivals’ access to the advertising space of publishers.
413. The Commission notes that the Report has not given evidence to support this conclusion. Substantially similar display ad clause leaves publishers free to show non-Google display ads and the clause according to Google, now applies only to pages that show Google AFC ads. Other pages on a partner’s site are not covered. Even on those pages to which the clause applies, publishers can place non-Google display ads, provided that they are not the same or substantially similar to Google AFC ads.
414. As in the case of AFS ads, the Commission notes that ads of different type or format can be shown alongside Google AFC ads. The clause is thus meant to prevent confusion about Google and non-Google ads, and this protects legitimate interest of each company in preserving its brand. However, it is relevant to point out that the direct intermediation clauses to which the Investigation Report objects, are template clauses (i.e., model clauses), and are in reality subject to negotiation. Google has submitted that it concludes direct intermediation agreements with large and sophisticated publishers who are free to choose the option. The publishers generally do not seek to multihome search functionality or search ads on the same website. Those which do it, can use the template clauses as a starting position for negotiations. The clauses are subject to amendment during the course of negotiations. Google has argued that the negotiation process negates concerns that direct intermediation clauses involve imposition of “unfair” conditions from the outset. The publishers who PUBLIC VERSION want to deviate from the same or substantially similar search clause can and do negotiate amendments to the template clause. Likewise, publishers who want to deviate from the minimum ad and ad placement clauses can and do negotiate amendments to the AFS template agreement. Not only that, publishers can also negotiate amendments to the substantially similar display ad clause to provide even greater flexibility. In fact, publishers who have negotiated such amendments in the past include .
415. In light of this, the Commission holds that as publishers who do not want to enter into direct ad intermediation agreements have the option to enter into online ad intermediation agreements. Those who enter into direct ad intermediation agreement, do so as a result of free choice because the agreements accord with their business needs and commercial interests. This thus excludes the suggestion that Google “imposes” terms on publishers with its direct agreements.
416. With regard to the DG’s finding that Google denied market access to competing search engines in violation of Section 4(2)(c) of the Act, the Commission finds that an enterprise violates Section 4(2)(c) of the Act only if it limits the ability of competitors to effectively compete in a market. Google’s intermediation agreements do not deny market access within this meaning. They do not impose exclusivity, and certainly not in a way that limits the ability of rivals to compete. The Commission hence holds that the DG’s Report does not present evidence to the contrary. Therefore, the Commission is unable to accept the DG’s finding with respect to Section 4(2)(c) of the Act.
417. Similarly, with regard to the DG’s finding that Google leverages its dominant position in online general web search and online search advertising to impose restrictive conditions in its agreements for syndicate services for search and display advertising, the Commission finds that PUBLIC VERSION Google’s intermediation agreements do not violate Section 4(2)(e) of the Act as the Investigation Report does not establish it nor does the Report explain as to how Google’s strength in general search and online ads resulted in imposition of restrictive conditions in AFC and AFS agreements.
418. Before concluding on this aspect, the Commission notes that the DG has returned a number of the findings about Google’s agreements. One of them is that Google imposes unfair conditions on publishers by not disclosing the revenue shares for its online ad intermediation agreements. Though Google has denied this claim as incorrect, the Commission is of the view that such issues raise no competition concern whatsoever. Further, it could not be established that Google’s non-disclosure policy has restricted competition.
419. The DG’s objection to the arbitration clause contained in Google’s online AdSense Terms and Conditions with Indian publishers providing for arbitration in the United States on the ground that such framework may entail significant legal and financial costs for Indian partners, is not tenable. An arbitration clause in commercial agreements is a standard and legitimate business practice. It is beyond comprehension as to how such a clause in commercial agreements by itself can restrict competition without presenting evidence to support and sustain such finding of contravention of anti-trust law.
ORDER
420. In view of the discussion in the preceding paras, the Commission holds that Google enjoys dominant position in Online General Web Search and Web Search Advertising Services markets in India. The Commission further holds Google to have abused its dominant position on the PUBLIC VERSION following three counts:
(a) Ranking of Universal Results prior to 2010 which was not strictly determined by relevance. Rather the rankings were pre-determined to trigger at the 1st, 4th or 10th position on the SERP. Such practice of Google was unfair to the users and was in contravention of the provisions of Section 4(2)(a)(i) of the Act.
(b) Prominent display and placement of Commercial Flight Unit with link to Google’s specialised search options/ services (Flight) amounts to an unfair imposition upon users of search services as it deprives them of additional choices and thereby such conduct is in contravention of the provisions of Section 4(2)(a)(i) of the Act.
(c) The prohibitions imposed under the negotiated search intermediation agreements upon the publishers are unfair as they restrict the choice of these partners and prevent them from using the search services provided by competing search engines. Imposing of unfair conditions on such publishers by Google amounts to violation of the provisions of Section 4(2)(a)(i) of the Act. Google is doing so because it has dominance in the market for online general web search to strengthen its position in the market for online syndicate search services. This amounts to violation of the provisions of Section 4(2)(e) of the Act. Further, as competitors were denied access to the online search syndication services market, contravention of Section 4(2)(c) of the Act is also made out.
421. Coming to the remedies, the Commission notes that so far as display of Universal Results at fixed positions is concerned, it has been submitted that since October, 2010, Google has made display of such results on free floating basis. As such, the contravention remains confined to the period PUBLIC VERSION from May, 2009 (i.e. when the provisions of the Act relating to Abuse of Dominant Position came into effect) to October, 2010 and that is no longer subsisting. Accordingly, the Commission takes Google’s submission on record and refrains from issuing any cease order. In this regard, however, the Commission issues a desist order and directs Google not to resort to such fixing of position in future.
422. So far as the contravention noted by the Commission in respect of Flight Commercial Unit is concerned, the Commission directs Google to display a disclaimer in the commercial flight unit box indicating clearly that the “search flights” link placed at the bottom leads to Google’s Flights page, and not the results aggregated by any other third party service provider, so that users are not misled.
423. Lastly, the Commission orders Google to not enforce the restrictive clauses with immediate effect, as is found in the order, in its negotiated direct search intermediation agreements with Indian partners.
424. The Commission has also considered the issue of imposition of monetary penalty upon Google and has bestowed its thoughtful consideration thereon.
425. Under the provisions contained in Section 27(b) of the Act, the Commission may impose such penalty upon the contravening parties as it may deem fit which shall be not more than ten per cent of the average of the turnover for the last three preceding financial years, upon each of such person or enterprises which are parties to such agreement or abuse.
426. It is evident that the legislature has conferred wide discretion upon the Commission in the matter of imposition of penalty. It may be noted that the twin objectives behind imposition of penalties are: (a) to reflect the seriousness of the infringement; and (b) to ensure that the threat of PUBLIC VERSION penalties will deter the infringing undertakings. Therefore, the quantum of penalties imposed must correspond with the gravity of the offence and the same must be determined after having due regard to the mitigating and aggravating circumstances of the case.
427. In this connection, it would also be apposite to refer to a recent decision of the Hon’ble Supreme Court of India in Excel Crop Care Limited v. Competition Commission of India & Anr., Civil Appeal No. 2480 of 2014 decided on 08.05.2017. One of the issues which fell for consideration before the Hon’ble Supreme Court in this case was as to whether penalty under Section 27(b) of the Act should be imposed on the total/ entire turnover of the offending company or only on “relevant turnover” i.e. relating to the product in question.
428. After referring to the statutory scheme as engrafted in Section 27 of the Act and analysing the case law at length, the Hon’ble Supreme Court opined that adopting the criteria of ‘relevant turnover’ for the purpose of imposition of penalty will be more in tune with the ethos of the Act and the legal principles which surround matters pertaining to imposition of penalties. While reaching this conclusion, the Hon’ble Supreme Court recorded the following reasons:
When the agreement leading to contravention of Section 3 involves one product, there seems to be no justification for including other products of an enterprise for the purpose of imposing penalty. This is also clear from the opening words of Section 27 read with Section 3 which relate to one or more specified products. It also defies common sense that though penalty would be imposed in respect of the infringing product, the ‘maximum penalty’ imposed in all cases be prescribed on the basis of ‘all the products’ and the ‘total turnover’ of the enterprise. It would be more so when total turnover of an enterprise may involve activities besides production and sale PUBLIC VERSION of products, like rendering of services etc. It, therefore, leads to the conclusion that the turnover has to be of the infringing products and when that is the proper yardstick, it brings home the concept of ‘relevant turnover’.
429. Thus, the starting point of determination of appropriate penalty should be to determine relevant turnover and thereafter, to calculate appropriate percentage of penalty based on facts and circumstances of the case.
430. Before examining this aspect, it would be first appropriate to note the contention of Google that calculation of penalty, if any, would have to be based on the relevant revenues generated by it in India, rather than globally, and would have to be limited to the revenues from the allegedly affected markets. It was argued that the Hon’ble Supreme Court has confirmed that only the revenues generated from the allegedly infringing product should be taken into account when determining the amount of the fine. It was pointed out that the DG has not identified any such affected markets.
431. The Commission has carefully considered the submissions made by Google on the issue of relevant turnover. To begin with, it is observed that the Commission had directed Google to provide the revenue generated from its India operations only. Hence, the Commission would consider the revenue generated from India operations and not the revenue generated by Google through its global operations. Having said that, the Commission is of opinion that the concept of relevant turnover cannot be applied to a technological platform such as Google as it is applied in the context of a conventional multi-product company. In a two sided market, the search side is free whereas the other side is monetized through advertisements. It would stultify the very object and intendment of the Act if Google were to be allowed to contend that search being free, no penalty can be levied as there is no revenue stream from this side of the market. Such a pedantic PUBLIC VERSION approach, which is more appropriate for conventional market, would virtually allow two sided platforms to abuse one side and accord a virtual immunity from monetary penalty for anti-trust violations. It needs no reiteration that the two sides of such platforms are intricately interwoven with each other as one side cannot operate without the other. In such case, the entire platform has to be taken as one unit and revenue generated therefrom has to be seen as a whole. Any other interpretation or approach would render the deterrence exerted by the Statute as redundant and nugatory. That could not have been the intent of the legislature or the judicial pronouncements.
432. No doubt, the Commission directed Google to provide its segmental revenue but that was with a view to quantify the penalty in a proportionate measure so that the penalty, which is to be based after taking the whole revenue from the platform, does not become disproportionate vis-à-vis the revenues generated from the infringed segments. Thus, the Commission would consider the sum total of the revenues generated by Google as provided by it from its different segments of its India operations to reach a quantum of monetary penalty which is proportionate and commensurate to the infringing conduct.
433. In this connection, it is observed that the Commission vide its order dated 20.12.2017 directed Google to provide the details of revenue generated from its India operations in respect of the services specified therein by 05.01.2018. Subsequently, Google moved an application dated 29.12.2017 seeking extension of time by at least 3 weeks to comply with the Commission’s directions.
434. Having considered the application and the reasons cited therein, the Commission vide its order dated 02.01.2018 allowed Google to submit the financial details sought for by the Commission by 15.01.2018. The Commission further permitted Google to submit the said details duly PUBLIC VERSION certified by an internal Chartered Accountant, instead of an external auditor, as prayed for. Google, however, moved another application on 15.01.2018 seeking extension upto 18.01.2018 again citing various reasons including religious holidays and logistical challenges in complying with the directions of the Commission. The request was acceded to by the Commission and Google was allowed to submit the financial details by 18.01.2018 i.e. the date requested by Google itself.
435. On perusal of the financial details submitted by Google, the Commission is constrained to note that notwithstanding a categorical and clean direction issued by the Commission to Google to furnish “…details of revenue generated from its India operations…” in respect of the specified services, Google has given financial details in which it has indicated at the top in Annex 1 to the following effect: “Relevant Turnover from Direct Sales in India (INR, Millions)”. It is indeed perplexing as to what is meant by Direct Sales and what is left out by way of Indirect Sales. Further, Google has qualified every data its various footnotes and has indicated various aspects of its revenue which it does not “retain”. Moreover, Google has also stated in the submissions that “…it was not possible to collate revenue data by user location in the time accorded by the Hon’ble Commission”.
436. The Commission is dismayed at such disclaimers and caveats in respect of a figure which should have presented no difficulty in collecting and collating after availing sufficient time and certified by an internal auditor.
437. On penalty, Google has contended that even if the Commission finds a violation of Section 4 of the Act by Google, it must take into account the uncontested consumer benefits arising from Google’s conduct, lack of evidence of anti-competitive intent or secretive behaviour. It was pointed out that the Investigation Report also acknowledges the procompetitive effect of Google’s conduct in India.
PUBLIC VERSION
441. The Secretary is directed to provide copies of the public version of this order to the parties through their respective counsel(s) and inform them accordingly.
Sd/-
(Devender Kumar Sikri) Chairperson Sd/-
(S. L. Bunker) Member Sd/-
(Augustine Peter) Member Sd/-
(U.C. Nahta) Member New Delhi Date: 31/01/2018 [Dissent Note dated 08.02.2018 by Member (SM) and Member (GPM) at pp. 167-190] PUBLIC VERSION DISSENT NOTE PER Mr. Sudhir Mital Member Mr. Justice G. P. Mittal Member
1. In this opinion, we record our dissent to the order of the Commission that disposes of Case No. 07 of 2012 and Case No. 30 of 2012 filed under Section 19 (1) (a) of the Competition Act, 2002 (‘the Act’) by Consim Info Private Limited (‘the Informant/ Consim’, now known as Matrimony.com Limited) and by Consumer Unity & Trust Society (CUTS) (‘the Informant’) respectively against Google Inc. (now Google LLC) and Google India Private Limited alleging contravention of the provisions of Section 4 of the Act.
2. Besides perusing the material on record, we have had the benefit of going through the Majority order of the Commission. We are in agreement with the Majority order on the two relevant markets as have been defined and the assessment of Google’s dominance in the same. We also agree with the Majority order’s analysis relating to the procedural objections raised by Google as also with the conclusions of the Majority order on the issues of More Universal Results, Maps, One Boxes, Commercial Shopping Unit, AdWords, Trademarks, AdWords API T&Cs, Distribution Agreements and Direct Ad Intermediation Agreements. However, we are unable to persuade ourselves to subscribe to the Majority view in respect of the alleged contravention of the Act by Google in respect of Flights Unit, Direct (negotiated) Search Intermediation Agreements and historic use of Fixed Positions for Universal Results for the reasons recorded in the subsequent paras.
PUBLIC VERSION
3. For the sake of brevity, we shall not again recapitulate the background and facts of the matter (s) at hand which have already been dealt with in detail in the Majority order, and shall confine to the reasons for disagreeing with the Majority view in respect of Flights Unit, Direct (negotiated) Search Intermediation Agreements and historic use of Fixed Positions for Universal Results.
Flights Unit
4. Google Flights Unit is a Commercial Unit, which is distinguished from the free, organic search results with a label indicating the Unit as ‘Sponsored’. The Unit appears in the space for display of ads on the Google Search Engine Results Page (‘SERP’) – above the free, organic search results. The Flights Unit provides users with flight information on routes specific to their search query. It contains a link ‘Search Flights’, clicking on which leads the users to Google’s Flights page showing additional results of the same type. According to the Majority order, “Google through its search design has not only placed its commercial units right at a prominent position on SERP, it has also allocated disproportionate real estate thereof to those units resulting into either pushing down or pushing out of the verticals who were trying to gain market access. To top it all, Google has provided link which leads users of Google Flight commercial unit to its specialised search result page (Google Flights). Consequently, users may be devoid of additional choices of results and therefore, such conduct amounts to an unfair imposition upon the users availing search services.” Such conduct of Google, as per the Majority, “being an unfair imposition upon the users of general search services, is in contravention of Section 4 (2) (a) (i) of the Act” [Para 253].
PUBLIC VERSION
5. We are unable to concur with the Majority’s assessment of the issue at hand and accordingly with their conclusion of infringement of the provisions of the Act. We are of the view that determination of infringement, in this case contravention of Section 4 (2) (a) (i) of the Act, cannot be in the abstract. Section 4 (2) (a) (i) of the Act deprecates imposition of unfair condition by a dominant enterprise in purchase or sale of goods or services. However, the onus is on the Commission to establish, based on facts and evidence on record, as to whether and how an impugned practice of a dominant enterprise amounts to a conduct of unfair imposition on its consumers.
6. The issue under scrutiny here is the sponsored Flights Unit of Google.
To arrive at a conclusion that the ‘prominent’ placement of the Flights Unit is a conduct of unfair imposition on the users of Google’s general search services, it needs to be established with evidence that user-clicks are guided solely or largely by the position at which a particular result is placed on Google SERP irrespective of whether it is an advertisement, Commercial Unit or generic result and that the users find the Flights Unit of Google and its ‘prominent’ placement on SERP to be confusing or misleading. As elaborated in the subsequent paragraphs, we could not find any relevant, reliable or cogent evidence that could aid a reasoned assessment of whether and how the mere presence of Google Flights Unit amounts to a conduct of imposition of unfair condition on the users.
7. The two metrics that would enable an objective assessment of user click behaviour are (i) the user click-through rates (CTR), calculated as clicks divided by impressions for users in India, of different ad positions, of Commercial Units including the Flights Unit and of different generic search results positions and (ii) the actual traffic flow to Google Flights via the Flights Unit on Google SERP vis-à-vis the traffic flow to other travel verticals via Google in India. The DG investigation did not bring PUBLIC VERSION the necessary data on these two metrics on record. During the course of hearings, Google was directed by the Commission to provide data in respect of CTR for different ad positions and free search result positions from 2009 to 2015 for users in India. Google submitted global data for desktop users for top ads, right hand ads and free results for the said period. The data submitted by Google for top ads and generic results, as tabulated in Table 1 below, shows that the CTR for the first ad appearing on the topmost position of the SERP was in while the corresponding figure for first free/ generic result was . It further reveals that intra-format, i.e. within the ads, the CTR goes down from the first to the second to the third ad position and the same holds true for generic results as well. While we note that the data is not specific to India and the CTR for the Commercial Units including the Flights Unit is still not on record, what can reasonably be inferred from the data is that globally higher position or prominent placement on SERP is a crucial determinant of CTR only within a particular category of results, i.e. generic results or ads. It further indicates that the top ad, despite having been placed on the most prominent real estate of the SERP could not attract the maximum clicks and the top free/generic result despite having been placed below the ads enjoyed the highest CTRs. This shows that users can and do distinguish between free results and ads and they have revealed their preference for top free/ generic results over ads.
Table 1: CTR 2009-2015
CTR (%age)
Position
2009 2010 2011 2012 2013 2014 2015
Top Ad 1 * * *
Top Ad 2 * * *
Top Ad 3 * * *
Free Result 1 * * *
Free Result 2 * * *
PUBLIC VERSION
CTR (%age)
Position
2009 2010 2011 2012 2013 2014 2015
Free Result 3 * * *
Free Result 4 * * * *
Free Result 5 * * * *
Free Result 6 * * * *
Free Result 7 * * * *
Free Result 8 * * * *
Free Result 9 * * * *
Free Result 10 * * * *
Source:
8. Here we would like to reiterate that the Flights Unit does not appear on the SERP as a free/ generic search result. It is neither surreptitiously embedded in the free/ generic search results, nor does it affect the ranking of free search results. It comes within a box and that it is a sponsored link is made explicit thereby making it visually distinguishable from generic blue link results. Thus, while there can be no ambiguity that the top free/ generic search result would get the maximum clicks amongst the free results, the same cannot be transposed to the Flights Unit, without supporting data. It may be the case that users distinguish between the commercial Flights Unit and the free/generic results and actually scroll down to the relevant generic results for what they might be looking for, especially if those results appear on the first few positions of the free generic results. It can also be the case that users do click on the Commercial Unit and do so consciously as they find the feature useful. Be that as it may, without any empirical underpinnings, a conclusion that users click on the Flights Unit just because it appears above the free results presupposes what the consumers i.e., the web-searchers, prefer and actually do. Not only that, it is also a presumption that the consumers are not sufficiently intelligent to distinguish an enhanced ad format from PUBLIC VERSION a free search result and that they will be lured by the ‘prominent’ placement and the ‘disproportionate’ size of the Flight Unit in the SERP.
9. The DG has referred to and the Majority has relied upon a few studies and reports conducted in other jurisdictions to highlight the importance of visibility in attracting user clicks. We have reviewed them carefully. The Microsoft ‘Heat Map’ study highlights the areas of the SERP where users focus their attention by recording the movement of users’ eyes and points out that the ‘top-left’ section of the SERP receives distinctly more attention than any other part. We observe that the study is limited to eye movement and user attention. That the highest user attention may not translate into concomitant maximum actual clicks, is corroborated by the CTR data of top ads, which typically appear on the ‘top-left’ section of the SERP. The other external studies referred to by the DG highlight the importance of ranking of results within the generic free results, without providing a comparative picture of user click behaviour vis-à-vis the Commercial Units, the Flights Unit in particular. The only finding that has been relied upon by the DG and the Majority vis-à-vis the Commercial Units is of the submitted to the by Microsoft with regard to . The finding referred to by the Majority only shows that the CTR for the top algorithmic result dropped considerably with the , without providing any CTR data for the other ads and search result positions including that of the . Any advertisement in any format is bound to affect the CTR of the top algorithmic result. Nevertheless, as the global data on CTRs submitted by Google shows, despite ads the top algorithmic result continues to enjoy the highest CTR. Moreover, to conclude that it is an unfair imposition on the users, as concluded by the Majority, it needs to be ascertained that the users did not actually find the to be useful but clicked on the same PUBLIC VERSION only because of its placement above the generic results deceptively believing the same to be the most relevant result, which the fails to do. The finding relied upon by the Majority does not throw any light on user preferences but simply shows that the CTR of top algorithmic results are affected by advertisements/ Commercial Units.
10. The Majority, in Para 239, observes “… a user’s clicking behaviour may also be influenced by Google’s public claim of ranking results based on relevance.” The Majority further extracts certain statements of Google to bring forth Google’s claim of ranking results based on relevance. We observe that in the statements of Google relating to online search, ranking of results in terms of relevance has been claimed for intra-format search results, i.e. relevance within free/ generic search results and within advertisements separately and not inter-format as construed by the Majority. For instance, the letter of the founders of Google accompanying the IPO filing (as quoted in Para 239 of the Majority), clearly distinguishes free/ generic search results from advertising:
“… Google users trust our systems to help them with important decisions: medical, financial and many others. Our search results are the best we know how to produce. They are unbiased and objective, and we do not accept payment for them or for inclusion or more frequent updating. We also display advertising, which we work hard to make relevant, and we label it clearly. This is similar to a well-run newspaper, where the advertisements are clear and the articles are not influenced by the advertisers’ payments.”
[emphasis supplied] PUBLIC VERSION Google’s Executive Chairman, Eric Schmidt’s statement, “the natural search answers [are] completely unbiased with respect to economics” [emphasis supplied] further indicates that Google’s claim of relevance is with respect to its free/ generic search results. The Flights Units are an enhanced ad format that appear in the ad space and are marked as such by Google with the ‘Sponsored’ label. They need to be viewed as a means of monetisation of free search services through paid advertisements and not as free search results which Google claims to rank based on relevance. Google is a two-sided platform where search is free on one side and Google monetises the other side through advertising.
It is axiomatic that like all search services, paid advertising enables Google to offer free search results. To conclude ‘prominent’ placement of a publicly declared sponsored result as a deviance or departure from relevance in presenting search results is to strike at the core of the two- sided business model of Google or any other search engine. In fact, the Flights Unit typically appears below the AdWords ads and not on the topmost section of the SERP. Thus, taking the argument to its logical conclusion would mean that the premium real estate on the SERP should be bereft of any advertisements, lest they mislead the users, even though the CTR data on record corroborates that users are discerning and they do not blindly click on the advertisements that appear on top.
11. In Para 238, the Majority has opined, “The Commission is of the view that by integrating/ linking specialised search result pages with the Commercial Units and placing them prominently on SERP, Google is able to drive traffic to its own pages and also generate revenues through advertisements/sponsored results” [emphasis supplied]. We respectfully differ from this Majority finding in absence of any evidence or data on record that corroborates significant traffic flow to Google Flights page via the Flights Unit. This brings to fore the other important metric of PUBLIC VERSION traffic flow that could have aided the assessment in an objective manner. While the DG report does not provide any data or systematic analysis of traffic flow to Google Flights via the Flights Unit, the data submitted by Google on the direction of the Commission shows that in , Google travel pages accounted for only of Google traffic to all travel pages.
12. In Para 253, the Majority concludes, “To top it all, Google has provided link which leads users of Google Flight commercial unit to its specialised search result page (Google Flight). Consequently, users may be devoid of additional choices of results …” It is not apparent to us as to how the ‘search flights’ link to Google’s flight page takes away the choice of results from the users when the AdWords ads that appear above the Flights Unit on SERP and the free/ generic results that appear below the Flights Unit continue to provide links to airlines and other travel verticals. There is no evidence on record that shows that the competing travel sites are demoted to such positions or pages that users cannot easily find and access them. If the conclusion of the Majority is to suggest that the impugned link should ideally have led the users to other third party travel verticals, instead of Google’s own page, that would be incongruent with the Majority’s own stance on ‘more results’ link in the case of Universal Results. As the Majority opines in Para 219 of its order and we agree, “… directing Google to replace these links with links to third-party vertical search services, may create confusion for the users. In such a scenario, Google’s algorithms would rank the first few results, while another provider’s entirely different ranking would rank the remaining set after the link. This may not be a workable proposition.”
13. Finally, the views of the users of Google, on whom the Google Flights Unit is allegedly an unfair imposition, have not been accounted for in any manner in the investigation. Leave alone a reasonable sample of users, not even a single user/ searcher’s view or even anecdotal account PUBLIC VERSION is on record for us to understand how users actually perceive the Google Flights Unit. Analysis of a design feature such as the Flights Unit with complete disregard for user preferences would be mere speculation about a possibility of user confusion, which in our view, may lead to an outcome that few users may possibly favour or require. Internet search has transitioned from the archetypal ‘ten blue links’ model to a model where search engines take on the function of providing end information or interactive interfaces to website information. Thereby, finding objections with a particular design feature without having gauged the users’ experience may result in imposing regulatory preferences to the detriment of consumer preferences. Any conclusion based on imprecise perceptions runs the risk of misjudging user behaviour, thus disapproving legitimate, competitively benign product designs and harming efforts at innovation. We are of the considered view that unless supported by evidence that indicates confusion or difficulty caused to any user of Google by a particular design feature, such second-guessing may lead to inferior outcomes for all market participants and chill innovation.
14. We shall now turn to examine the issue of the Flights Unit vis-à-vis the other stakeholder group, i.e. the competing travel sites. The Majority, while not finding the impugned conduct to be an unfair imposition on and thus an abuse of dominance vis-à-vis these websites, has deliberated at length on how the ‘prominent’ placement of the Flights Unit and the ‘disproportionate’ real estate allocated to it “may not” allow equally efficient websites to sustain and survive in the market for flights search services. The Majority, by using phrases such as “may also hinder market access”, “has potential”, “perhaps”, “may not allow”, “may be equally efficient”, “may not be able to” has woven a narrative of hypothetical anti-competitive foreclosure. However, the hypothesis has not been tested.
PUBLIC VERSION
15. As per the Majority order, “Google’s Flight Unit captures the traffic that would ordinarily flow to the top algorithmic listing” [Para 248] [emphasis supplied]. However, we note that there is no traffic data analysis on record that corroborates the diversion of traffic from competing sites to Google’s flights pages. The Majority conclusion rests on the premise that “… such an unfair diversion of traffic by Google may not allow third-party travel verticals to acquire sufficient volume of business” [Para 253] [emphasis supplied]. Clearly, the first test to arrive at this conclusion should have been to examine and establish the diversion of traffic through an analysis of CTRs and traffic data of the competing flight services and the ‘wrongful’ gain made by Google on account of diversion of traffic by analysing their data. However, we note that no such analysis is available in the DG Report.
16. Google, in its reply to the DG Report, has presented an empirical analysis to show that ‘prominent’ placement of the Flights Unit did not affect traffic flow to competing sites in India. As per the data provided, clicks to Google Flights and Hotel result pages were minor, accounting in for only around total clicks. By contrast, MakeMyTrip, Goibibo, and Cleartrip have all substantially increased their clicks by at least over : for MakeMyTrip, from in to in (an increase of over ); for Goibibo, from around in to in (increasing by factor of over ); and for Cleartrip, from clicks in , to in (an increase of over ). Any counter-claim needs supporting evidence and we were presented with none in the instant case by the DG.
17. In this context, we further note that the Flights Unit as also the Google Flights Page offer a tool for the users to compare different flight offers from different airlines by various parameters such as price, duration, PUBLIC VERSION schedule etc. They do not offer the possibility for flights to be booked directly on Google Flights. By contrast, travel verticals such as MakeMyTrip, Cleartrip etc. not only have the search function but also allow for booking of flights on their sites. This distinction is reflected in the letter dated 09.10.2013 submitted to the DG by MakeMyTrip (quoted in the Majority order Para 250).
“While these products are unlikely to have a booking engine, the positioning of the product is likely to divert customers from Makemytrip.com at the stages of research.”
[emphasis supplied] In view of the difference between a pure online flight comparison site such as Google Flights and the online flight booking sites, it was all the more necessary for the DG to empirically examine as to how the high visibility of the Flights Unit affected the traffic share of the third-party travel verticals. Apart from some conjectural assertions made by MakeMyTrip on the likelihood of an impact of the Flights Unit on its overall bookings, no other material has been brought on record by the DG that systematically analyses the traffic flow to the competing sites vis-à-vis Google Flights in India. For a complete assessment of alleged ‘unfair’ treatment meted out by Google to its rival sites in the flights domain, it was also necessary to examine who these competing sites are and at what positions on the SERP do they typically appear. However, such analysis was also not attempted.
18. It is our considered view that without any relevant data on or analysis of user click behaviour in India vis-à-vis the Commercial Units or actual traffic flow to these units, diversion of traffic by Google to the extent that would prevent third-party verticals to acquire sufficient volume of business, as claimed by the Majority, is not substantiated.
PUBLIC VERSION
19. We would like to further deliberate on our reservations on the Majority stance on certain fundamental issues. The Majority in Para 248 states:
“The Commission observes that in case of Google’s Commercial Unit, i.e., Flight Units in India, primary competition concern emanates out of its prominent placement of its commercial units on SERP in addition to providing disproportionate real estate thereof to such unit.” [underlined for emphasis] Further, in the same Para, the Majority states:
“the insertion of Google’s Flights Unit prominently above the blue link results in the SERP denying third-
party travel verticals even the opportunity to be displayed on that key “real estate” Google’s Flights Unit pushes the algorithmic results down and in some cases, even off the first page.” [underlined for emphasis] While it is tautological that even a centimeter of “real estate” that Google uses for its enhanced format sponsored units denies other competing verticals that “key real estate”, but to conclude that it actually drives traffic to Google’s flights page and affects the competing websites’ ability to effectively compete is a presumption. Why do we say this is a presumption? As discussed in the previous para, the supporting evidence in the form of either CTR of the Flights Unit or traffic data for the Flights Unit has not been placed before us. In any case, the Flights Unit does not typically appear on top of the Google SERP for search terms related to flights. On the top of the results page, the AdWords ads appear, and the Flights Unit appears below these ads. Ads in any format on the top of the SERP would push the algorithmic results down. This is not to say that there should be no antitrust liability on a fact-specific basis for acts of exclusion against rival websites. However, without identifying who the allegedly affected websites are and how they would have received substantially more traffic in absence of the Flights Unit PUBLIC VERSION despite being ‘equally efficient’, alleged denial of opportunity to be displayed on “key real estate” to an unknown and the entire universe of third party websites, cannot be a sound basis to establish abuse of dominance. Moreover, it would be insightful to build a counterfactual scenario where the ‘finding’ of contravention by the Majority on account of ‘prominent’ placement of and ‘disproportionate’ real estate allocated to the Flights Unit is taken to its logical conclusion. If the size of the Flight Units is ‘disproportionate’, then what will be a proportionate size? If the ad format is prominently placed, then what will be an innocuous place such that the “misled” consumers do not click it? The answers to these questions have been avoided in the Majority order.
20. The remedy directed by the Commission, “a disclaimer in the commercial flight unit box indicating that the “search flights” link placed at the bottom leads to Google Flights page and not the results aggregated by any other third party service provider, so that users are not misled” [Para 422] does not address the purported harm caused or likely to be caused to the third party websites and presumes that users click on the ‘search flights’ link on the Flights Unit expecting it to lead to third party sites, without any evidence on record to that effect. Further, the concern raised by the Majority that the users cannot distinguish between sponsored and generic results despite the ‘Sponsored’ disclaimer of Google, does not get addressed by the proposed remedy. Given the premise of the Majority, it is equally likely that despite the insertion of the proposed disclaimer, the ‘undiscerning’ users might not take note of it and still be ‘misled’ to Google flights.
21. To conclude, in our opinion, the investigation has not brought on record the evidence necessary for a reasoned assessment of whether and how the ‘prominent’ placement of Flights Unit amounts to a conduct of imposition of unfair condition on the users. A conclusion of infringement PUBLIC VERSION of the Act needs to be premised on clear, unambiguous evidence covering various important aspects of user preference, the impact of higher visibility of the Flights Unit on CTRs, the flow of user traffic to Google flights via the Flights Unit and Google traffic to the competing websites, the positions at which such websites are ranked on the Google SERP etc. in India. The provisions of the Act relating to abuse of dominant position warrant ex post determination of abuse by a dominant enterprise, supported by facts and evidence. The application of the law is not amenable to hypothetical frameworks built on perceived premises. In view of the same and based purely on putative contentions, we are unable to persuade ourselves to concur with the Majority conclusion that by placing its Flights Unit on the SERP, Google has contravened Section 4 (2) (a) (i) of the Act.
Search Intermediation
22. In abuse of dominance cases, the first step is to delineate the relevant market followed by an assessment of dominance of the enterprise under scrutiny in that relevant market. The DG has not defined ‘online search intermediation/ syndication services’ as a relevant market in terms of the provisions of the Act and Google’s dominance in this market too has not been assessed in terms of Section 19 (4) factors. The DG and the Majority have inferred dominance in this market from Google’s dominance in online general web search and online search advertising. The DG, in the investigation report, has stated that Google is found to be dominant in the relevant markets of online general web search in India and online search advertising in India. As per the DG, “By virtue of this position of strength in these relevant markets, Google is also a preferred syndicate service provider for publishers wanting to offer search and advertising services on their websites.” Google, in its reply to the DG report, has averred, “the Report claims that Google’s terms of trade PUBLIC VERSION entered into with its intermediation partners constitute an abuse of dominance. But the Report does not claim – let alone prove – that Google is dominant in search intermediation, search ad intermediation, or display ad intermediation. The Report thus fails to satisfy the first step in an abuse of dominance case. Its claims must fail as a result.”
23. Online search intermediation is a service provided by Google and other search engines to web publishers in order to facilitate search on their websites, by providing search toolbars directly on the websites. Both the DG and the majority have considered online search intermediation/syndication services as a distinct market, which has allegedly been foreclosed from competitors and leveraged into by Google using its dominance in the online general search market. The Majority, in Para 397 of its order, observes, “Google was using its dominance in the market for online general web search to impose restrictive conditions in online syndicate search agreement, in violation of Section 4 (2) (e) of the Act. Further, as mentioned above, competitors were denied access to the online search syndication services market, a contravention of Section 4 (2) (c) of the Act is also made out” [emphasis supplied]. However, this market has not been defined as a relevant market and Google’s dominance has not been established in the same as per the provisions of the Act. As per Section 4 (2) (e) of the Act, there will be an abuse of dominant position, if a dominant group or enterprise “uses its dominant position in one relevant market to enter into, or protect, other relevant market” [emphasis supplied]. Thus, the relevant provision of the Act warrants delineation of the relevant market where the enterprise is dominant as well as the relevant market that is being leveraged. In the instant case, as mentioned above, the second market of ‘online search intermediation/ syndication services’ has not been defined as a relevant market in accordance with the provisions of the Act. In view of the same, we are of the opinion that the legal requirement of defining PUBLIC VERSION ‘online search intermediation/ syndication services’ as a relevant market has not been met.
24. The impugned conduct is that in Google’s direct search intermediation agreements, it stipulates a template clause that asks publishers not to implement search technologies on their sites that are “same or substantially similar” to that of Google. This, according to the DG and the Majority, is a restrictive condition imposed upon the publishers in negotiated search agreements since they may wish to offer search services from websites in addition to Google. Further, these restrictions, allegedly, amount to de-facto imposition of online search exclusivity in agreements of 2-3 years duration, thereby depriving competitors from acquiring the requisite scale.
25. It is important to note that the impugned clause is not there in the online standard contracts, which are available to all publishers. These are stipulated only in the negotiated contracts. As per Google, the impugned clause is a template clause subject to negotiation. The DG investigation has not brought forth any material that reveals the direct partners’ views on the agreement and the negotiation mechanism. The publishers questioned by the DG have not suggested that they feel restricted by the agreements at issue.
26. Google, on the other hand, has contended that its experience of search intermediation agreements is that partners generally do not wish to multi-source search services as this adds to the transaction and integration cost without attendant improvement in end-user experience. Google has further averred that the DG report provides no evidence that a publisher would want to implement two search functionalities that are the “same or substantially similar”. Including two substantially similar PUBLIC VERSION search functionalities on the same site would likely confuse and frustrate users.
27. We are of the opinion that the direct partners’ views were critical in forming a conclusive view on the desirability of multiple search functionalities on the same website. The investigation does not adduce any evidence to establish that absent the clause, the direct partners would have opted for multiple search functionalities on their websites. Moreover, publishers that do not want to enter into direct intermediation agreements have the option to enter into online agreements, which do not contain the clause at issue. As per the DG Report, under the direct negotiated agreement, Google provides enterprise level support to its direct partners. The Majority, in Para 411, notes, “… for direct intermediation partners, Google makes specific investments and commits dedicated resources, including offering higher level of implementation and technical support, enhanced reporting tools, and optimisation assistance to publishers.” Thus, it seems that the direct contracts bring certain additional benefits to the publishers, which incentivise them to enter into the same with Google rather than opting for online contracts. In view of the above, one can reasonably infer that the publishers that enter into the direct agreements do so in exercise of their free choice, driven by their business interest. Given that these are websites which generate larger volumes of search queries, they are presumably in a position to exercise some degree of countervailing power while negotiating the terms of agreement. Google, in its reply to the DG Report, has claimed, “no direct partner has ever expressed a desire to implement two search functionalities that are “the same or substantially similar”, either before or after entering a negotiated search intermediation agreement.” Absent any evidence on record to the contrary and proper investigation into the issue, the impugned clause in agreements entered into by the direct partners of their own volition, PUBLIC VERSION cannot simply be presumed to be an ‘imposition’ by Google. When publishers opt to adopt only one search intermediation service provider, they may be exercising their preference. Thus, we are of the considered view that the investigation provides no basis for holding the impugned clause to be either ‘unfair’ to or an ‘imposition’ on the direct partners. Moreover, in absence of the determination of unilateral imposition by Google and without ‘online search intermediation/ syndication services’ having been defined as a relevant market in terms of the provisions of the Act, we are unable to conclude that Google leveraged its dominance in the search and search advertising markets to enter into or protect the market of online search intermediation/ syndication services. Thus, we respectfully disagree with the Majority conclusion that the impugned clause is in contravention of Section 4 (2) (a) (i) and Section 4 (2) (e) of the Act.
28. On the question of whether the impugned clause results in denial of market access to the competing search intermediation service providers, we observe that the DG Report does not provide a competitive analysis of the online search intermediation market in India, neither does it identify intermediation operators active in the country who are allegedly affected by the clause. The internet landscape in India is thriving with the ever-growing number of websites. Only publishers ( ), during the period of investigation, were found to be subject to the impugned clause. The entire set of websites, barring the ones owned by these publishers, was open to the rival operators. Moreover, the direct intermediation agreements expire periodically, thereby offering the intermediation service providers the opportunity to compete for the direct contracts as well. However, there is no evidence or analysis on record that systematically analyses the intermediation market in India and the search intermediation contracts entered into by publishers in the PUBLIC VERSION country over time for a holistic assessment of switching cost or lock-in effects caused by the impugned clause.
29. In absence of the testimony of the direct partners with whom Google had the negotiated search agreements in India during the period of investigation, without any data or analysis on record pertaining to the search intermediation market in India, the players therein and the effect of the impugned clause on competition, we are of the considered view that no contravention of the Act is made out against Google on account of the impugned clause in the negotiated search intermediation agreements.
Universal Results
30. The DG investigation has revealed that prior to October 2010, Google limited the display of its grouped search results (i.e. image, news and local), known as ‘Universal Results’, to certain fixed (1st, 4th or 10th) positions on the SERP. The reason proffered by Google for having such fixed positions for Universal Results on the SERP is that “its systems were not sufficiently advanced to conduct a relevance comparison for all positions on the result page”. The Majority order has found this contention of Google to be unacceptable “in the absence of concrete material in this regard.” While taking note of the fact that Google introduced fully-floating ranking for the Universal Results between September and October 2010, the Majority found the historic use of fixed positions to be an unfair imposition on users, in contravention of Section 4 (2) (a) (i) of the Act.
31. We are unable to agree with the Majority on this issue. Google has submitted that prior to 2010, it limited the display of Universal Results to certain fixed positions (1st, 4th and 10th) because its systems were not sufficiently advanced to conduct a reliable relevance comparison for all PUBLIC VERSION positions on the SERP. Limiting Universal Results to fixed positions, as per Google, demonstrates its exacting relevance standards. Google preferred not to show Universal Results in positions for which it could not conduct a relevance assessment, rather than compromise its relevance and quality. Limiting the ranking position in which Universal Results could appear in principle does not mean that Google held Universal Results to a lower relevance standard. We would not like to comment on the contention raised by Google regarding technical infeasibility. However, it is an admitted position that Google shifted to the fully-floating regime, i.e. to display Universal Results at any position on the SERP depending on the same relevance screen applicable to other generic results, well before the present informations were filed before the Commission in June 2012 and the proceedings were initiated. In our view, the change in the system brought about by Google, on its own, obviates the need for any regulatory intervention in this aspect, especially when the new fully-floating regime appropriately addresses the concern. Hence, Google cannot be held liable for imposition of unfair condition on account of its historic use of fixed positions for the Universal Results on the SERP in contravention of Section 4 (2) (a) (i) of the Act.
32. Before we conclude, we may extract Section 4 of the Act for ready reference:
(1) No enterprise or group shall abuse its dominant position.
(2)There shall be an abuse of dominant position under sub-Section (1), if an enterprise, –
(a) directly or indirectly, imposes unfair or
discriminatory–
(i) condition in purchase or sale of goods or
services; or
PUBLIC VERSION
(ii) price in purchase or sale (including predatory price) of goods or service; or Explanation.– For the purposes of this clause, the unfair or discriminatory condition in purchase or sale of goods or services referred to in sub-clause
(i) and unfair or discriminatory price in purchase or sale of goods (including predatory price) or service referred to in sub- clause (ii) shall not include such discriminatory conditions or prices which may be adopted to meet the competition; or
(b) limits or restricts –
(i) production of goods or provision of services or market therefor; or
(ii) technical or scientific development relating to goods or services to the prejudice of consumers; or
(c) indulges in practice or practices resulting in denial of market access; or
(d) makes conclusion of contracts subject to acceptance by other parties of supplementary obligations which, by their nature or according to commercial usage, have no connection with the subject of such contracts; or
(e) uses its dominant position in one relevant market to enter into, or protect, other relevant market.
[emphasis supplied] It is important to note that each and every clause of sub-section (2) of Section 4 of the Act uses words or operatives to reflect abuse. For instance, Section 4 (2) (a) (i) uses “imposes unfair or discriminatory condition in purchase or sale of goods or services”. Similarly, clauses 4 (2) (b), (c), (d) and (e) emphasise on other abuses with operatives such as “limits or restricts”, “indulges in practice or practices resulting in denial of market access”, “makes conclusion of contract”, “uses its dominant position … to enter into, or protect …”. Thus, a dominant PUBLIC VERSION player will be guilty of abuse only in the presence of proof of such behaviour as emphasised in the operatives used in these clauses. It goes without saying that the onus is on the Commission to establish from the evidence on record that there is either an imposition of unfair or discriminatory condition in purchase or sale of goods or services or there is a restriction of production of goods or provision of services or market, technical or scientific development or indulgence in practice or practices which result in denial of market access to some player (s) in the relevant market. Unfortunately, as detailed in the preceding paras, we do not find any evidence on record to establish abuse as indicated by the operatives used in Section 4 of the Act.
33. In conclusion, we note that with exponential growth of the internet, online markets now cover an ever-increasing spectrum of commercial activities. What we are also witnessing is creation of large online platforms that can wield substantial power over all market participants. By virtue of their access to the entire internet landscape as also to large volumes of personal data, they may be in a position to deter new innovation or dampen consumer welfare. However, market power or dominance in itself is not an antitrust concern; it is the conduct of such players that warrants careful competition scrutiny. It is when the evidence shows that the dominant firm uses its market power to stifle innovation and/ or competition or exploits the market power to the detriment of its consumers that a competition agency should intervene. Intervention can no longer revolve primarily around the creation or the strengthening of market power, but it should focus on the conduct of the dominant players and its implications for competition and consumers. We are of the view that regulatory interventions should be evidence- based as opposed to perception-based. In the instant case, the investigation has not brought on record the evidence and competitive analysis necessary to have a complete understanding of either the PUBLIC VERSION markets concerned or the conduct. Hence, we hesitate to use the instrumentality of this law to correct perceptions at the expense of the consumers who according to us are the constituency of this law. In view of the same, we do not find Google to be in contravention of Section 4 of the Act.
Sd/-
(Sudhir Mital) Member Sd/-
(Justice G. P. Mittal) Member New Delhi Date: 08/02/2018