Media industry additional comments on the DMA proposal 2 Redacted.pdf

Additional comments from media industry associations on the Digital Markets Act
following the meeting with DG COMP representatives on 6 October 2021.

Inclusion of Digital Voice Assistants (DVAs)
We recommend including digital voice assistants in the scope of the proposal. This can be done either
by (i) clarifying expressly (in the recitals) that this service constitutes an online intermediation service,
or (ii) adding this service as a new core platform service.
The inclusion of DVAs would imply limited amendments to the current proposal. If DVAs were added
as standalone core platform services, the definition of ranking (article 2(18)) would need to be clearly
extended to voice assistant services to make sure that article 6(1)d is applicable to such services.
To the extent that DVAs constitute online intermediation services, their formal inclusion would not
bear  consequences  with  respect  to  the  scope  of  the  obligations  of  article  5  and  6.  Besides,  the
Commission has already gathered evidence of the practices in which those core platform services may
engage.1

Inclusion of web browsers
We would also recommend adding web browsers to the list of core platform services. In its Android
Decision, the Commission distinguished between “web browsers” and “operating systems”2.
Google’s Chrome web browser has a market share above 60% in Europe. The service forms a central
part of Google’s ecosystem. By determining the conditions for consumers’ access to websites, web
browsers  fulfil  traditional  gatekeeper  intermediation  functionalities.  As  acknowledged  by  the
Commission  Impact  Assessment,  “browsers  […],  provide  a  gateway  for  users  to  access  the  diverse
range of content available on the Internet”.3
The current investigation of the CMA into Google Chrome’s “Privacy Sandbox”4 illustrates the need of
subjecting  web  browsers  to  regulatory  oversight,  not  only  to  preserve  competition  in  digital
advertising markets but also to ensure a meaningful enforcement of certain provisions enshrined in
the current DMA proposal (e.g. Article 6(1)(i)).
We take the view that the inclusion of web browsers in the DMA proposal would not necessitate any
further adjustment to other obligations in articles 5 & 6. For our members and based on the recent
CMA investigation, the main harm pertains to their ability to hold back data. Therefore, web browsers
should at least be subject to article 5(a) and article 6(1)(i).

1   See e.g. on self-preferencing - DMA Impact Assessment support study, pp. 311-312.
2   Case AT.4099, 18/7/2018, section 5.2. vs section 5.7 – Google Android.
3   DMA Impact Assessment support study, p.7.
4   See  CMA,  Press  release,  8/01/2021,  CMA  to  investigate  Google’s  ‘Privacy  Sandbox’  browser  changes,
https://www.gov.uk/government/news/cma-to-investigate-google-s-privacy-sandbox-browser-changes.

Fair and non-discriminatory conditions of access (article 6(1)k)
We believe that an obligation for gatekeepers to afford business users FRAND terms to access all their
CPS is warranted and would help enact more meaningful and inclusive change in digital markets to
promote fairness and contestability.
In  addition,  ensuring  FRAND  terms  solely  for  app  stores  also  appears  to  constitute  an  arbitrary
approach in light of current and past concerns associated with other CPS. Below, we provide concrete
concerns and precedents to substantiate our arguments.
Advertising  services:  most  gatekeepers  hold  uniquely  strong  if  not  dominant  positions  in  online
advertising  markets  (Google  in  search  and  Facebook  in  display).  In  light  of  ongoing  and  systemic
concerns about the functioning of these markets, following major investigations by many competition
authorities globally, and ongoing (in FR, ES and DE) and past probes within the EU (AdSense – abuse
of exclusivity clauses), it seems warranted to introduce such terms to prevent harmful practices. This
is particularly important since some of these gatekeepers are very active both on the buy and sell side
of these markets.
Search engines: the strong position of Google in search has led to concerns about discrimination in
different contexts. One current example relates to the implementation of the publishers’ right where
Google initially imposed royalty-free licences as a condition for access, before attempting to impose
Google’s  News  Showcase  product,  a  search-related  product,  which  arbitrarily  determines  which
publishers should benefit from the publishers’ right and which should not. A pending case before the
Autorité de la Concurrence in France is examining whether Google acted in good faith when it asked
publishers to choose either to waive their right and to release their content without compensation or
to be penalised in the search results. A similar probe into the news platform Google News Showcase
was launched in June 2021 by the Bundeskartellamt in Germany, to examine whether the new product
could result in discrimination between individual publishers.
Another example is the Google Shopping case which highlighted unfair practices which FRAND terms
could address.
Social  networks:  Facebook’s  new  hyperlink  policy,  which  mimics  Google’s  initial  reaction  to  the
implementation of the publishers’ right in France, seeks to coerce publishers into providing their rights
for free or to risk effective exclusion from participation on the platform by way of removing the rich
content that ought to appear in snippets. It also appears that Facebook is striking selective agreements
with some publishers but not others. FRAND terms are therefore needed.
Voice assistants: the Commission’s interim report on the IoT sector inquiry clearly identifies the risks
associated with voice assistants and reflects the need for FRAND terms.
Browsers: the privacy sandbox investigation of the Commission and the shift of problematic conduct
and monopolization strategies by Google to its browser highlights the need to ensure fair and non-
discriminatory conditions for business users.
Intermediation  services  (marketplaces):  the  e-books  investigation  into  Amazon  showed  that
marketplaces can also prove to be problematic for media companies. Here the abuse of MFN clauses
and its discriminatory use was a source of unfairness. This should show the need for FRAND terms to
apply to such intermediation services.

Personal data combination (article 5(a))
Article 5 (a) prohibits the combination of personal data from various services of the gatekeeper only
if the user does not consent to such combination (opt-in).
By definition, given the lack of alternatives, end users are dependent on a gatekeeper as described in
Article 3. Thus, Gatekeepers will find it easier than any other platform to obtain end user consent for
data usage as a pre-condition for the use of their platforms.
Consequently, for the DMA to reduce barriers to entry vis-à-vis digital gatekeepers, Article 5 (a) shall
not allow gatekeepers to circumvent systematically this obligation by nudging end users into declining
consent thereby depriving this provision of its “effet utile”. This is why we believe that the obligation
should  not  be  attached to  end  users’ consent.  Besides,  the  co-legislators  should  underline  that  by
default, gatekeepers shall separate their data pools into separate data silos.
There must be a high standard for combining such data. This could be based upon an end user’s explicit
request to combine data in order to obtain a significantly improved service. In such a case, the provider
should  bear  the  burden  of  proof  that  the  combination  of  data  indeed  (i)  improves  or  offers  an
additional service, (ii) allows the end user a fair share of the resulting benefits, (iii) is indispensable
for such improvement and (iv) does not eliminate competition.
It is also important to maintain the prohibition on gatekeepers to combine, for their own purposes,
data  sourced  from  their  core  platform  services  with  personal  data  collected  from  sources  and/or
services where they are present  as third parties, like  a publisher’s or broadcaster’s site. This is to
ensure that the DMA will put an end to gatekeepers’ practices that oblige not only end-users, but also
business users to agree to any practices as a precondition for the use of the gatekeepers’ core platform
services.

Access to data (article 6(1)i and article 11(2))
At  present,  when  distributing  content  through  Gatekeeper  platforms,  broadcasters  and  publishers
find that Gatekeepers refuse to share with them end-user data to better understand the use of their
services.  Gatekeepers  typically  refuse  to  share  any  data  regarding  the  number  of  views  or  any
interactions of end users with relevant content.
As  a  result,  while  Gatekeepers  can  use  such  data  to further  improve  their  services  and  enter  new
markets (often in competition with their business users), business users cannot benefit from the same
datasets  to  improve  their  own  services  (and  to  compete  with  the  Gatekeeper).  Given  that  the
Gatekeeper is unavoidable in reaching audiences, broadcasters and publishers have no other choice
but to accept this one-sided and unfair situation and to make their data available to the Gatekeeper
while accepting its conditions as regards the free combination and use of such data.
To make matters worse, to enhance their superior access to data, Gatekeepers increasingly use privacy
laws  as  an  excuse  to  further  deprive  business  users  and  competing  operators  of  relevant  data.
Meanwhile, given end users’ dependency on Gatekeeper services, the providers of such service find it
easy  to  collect  and  bundle  data  in  a  GDPR-compliant  manner.  They  reject  sharing  such  data  with
business users based on alleged GDPR-non-compliance. Since business users cannot equally leverage
market  power  to  acquire data,  the  data supremacy of Gatekeepers  vis-à-vis  their  competitors  and
business customers will only increase.

Gatekeepers are in a unique position to facilitate the obtaining of consent.  Where a user accesses a
third-party service through a Gatekeeper platform, the platform is used to facilitate the obtention of
consent for both the  use of data  by itself and the third-party service.  The Gatekeeper  should not
restrict the obtention of consent to its own uses alone.
Against this background, parallel to the concerns expressed against the current version of Article 5 (a)
(see above), we fear that the “consent-requirement” in Article 6 (1)(i) creates the opposite incentive
for Gatekeepers to nudge their end users into denying consent for the sharing of data with business
users. This will lead to imbalance. In order to reduce barriers to entry by creating a level playing field
for the collection of data, the commercial interests of the Gatekeeper in obtaining consent for the use
of data should be aligned with the equivalent interests of its business users.
More importantly, the wording of the anti-circumvention clause (article 11(2)) is unlikely to prevent
the above-mentioned loophole. Indeed, “where consent for collecting and processing of personal data
is required to ensure compliance with this Regulation” gatekeepers have the option to comply with
GDPR by other ways including providing anonymised data to business users (in lieu of requesting end
user consent).To address the above and align Gatekeepers and business users interests we suggest
introducing the following principle:
wherever the Gatekeeper fails to obtain consent from end users to share personal data that
has  been  generated  in  the  context  of  the  use  of  the  core  platform  service  with  the
intermediated business user pursuant to Article 6 (1) (i)), the Gatekeeper should be prohibited
from using such data for any purpose other than the original intermediation.

Scope of the self-preferencing obligation and enforcement
Article 6 (1) (d) exclusively addresses favourable “ranking”. Article 2 (18) of the Commission’s proposal
for a DMA defines “ranking” as the “relative prominence given to goods or services offered through
online intermediation services or online networking services, or the relevance given to search results
by online search engines”. Yet, this narrow approach fails to sufficiently address self-preferencing:
(i)
by  any  other  core  platform  service  such  as  video-sharing  platforms,  voice  assistants,
operating systems, cloud computing services, advertising systems or ancillary services, even though
such practices will be equally harmful;
(ii)
practices that go beyond ranking, such as a preferential crawling, indexing or other access of
content to the intermediation service (e.g. through settings);
(iii)
by search engines that provide their services directly within general search results pages or
exchange unbiased intermediation (via organic results) for biased intermediation (via paid results or
the display of own content).
The DMA proposal likewise should prohibit giving preferential treatment to selected third parties. We
believe  that  gatekeeper  platforms  are  able  to  circumvent  the  prohibition  of  self-preferencing  by
favouring  selected  services  and  partners,  thus  creating  the  same  anticompetitive  effects  for
competitors and undermining the free choice of the user. We therefore recommend that the ban on
self-preferencing  is  extended  to  selected  third  parties.  As  such,  “belonging  to  the  same  provider”
should be deleted.
In  addition  to the  required  adjustments of the  ban on  self-preferencing, the DMA  should  also  pay
more attention to the instruments available for the enforcement authorities to detect, evaluate and

monitor such conduct. The complex algorithmic and data-driven ranking of the various types of results
in search engines, app stores or marketplaces can only be effectively monitored by a highly specialised
and sufficiently equipped enforcement unit, ideally on a day-to-day basis.
Such units must be able to audit and test any mechanisms underlying an intermediation process so
that  discrimination  becomes  detectable  and  alternative  solutions  can  be  evaluated.  Against  this
background,  we  are  highly  skeptical  of  presuming  that  “[t]he  burden  on  the  Commission  for
implementing this [DMA] initiative is low (mainly redeployment of existing job positions)”.5 Despite a
high-profile investigation of now over eleven years into Google’s favouring of various own services in
general search results, we still have not seen any end to this process. We do not see how the current
institutional set-up for the DMA would allow a significantly more effective approach to this.

Bundling and tying
In addition to any need to “subscribe” or “register” with any core platform service, the need to “use”
such a service as a pre-condition to access, sign-up, register or use another service must equally be
prohibited in article 5(f). This would help cover situations where media services are forced to provide
their content through the gatekeeper’s subscription service in order to be able to use the gatekeeper’s
platform free service (e.g. YouTube premium vs YouTube).
Additionally,  under  the  current  wording,  a  gatekeeper  could  make  the  use  of  its  identified  core
platform service dependent on the subscription to its own video on demand platform - but not to its
own video sharing platform. There is no plausible reason for such differentiation. Put differently: the
reasons for limiting the scope of core platform services to multi-sided intermediation services do not
apply  when  it  comes  to  identifying  the  markets/services  that  should  be  protected  from  unfair
leveraging or trading practices. To cover such practices, the scope of unfairly tied services should be
extended  to  any  core  platform  service  (not  just  those  with  a  gatekeeper  status).  Since  there  is  no
justification for making the use of a gatekeeper service dependent on the use of any other service the
provision should be extended to the tying to any digital service.
Furthermore, such prohibition should also cover cases of a mixed bundling, often referred to as multi-
product  rebates.  While  in  the  case  of  a  ‘pure’  bundling  the  products  are  only  sold  jointly  in  fixed
proportions, in the case of mixed bundling, the products are also made available separately, but the
sum of the prices when sold separately is higher than the bundled price. The Commission’s “Guidance
on the [..] enforcement priorities in applying Article [102] TFEU to abusive exclusionary conduct” treats
pure and mixed bundling equally. Accordingly, a multi-product rebate is seen as abuse of dominance,
“if  it  is  so  large  that  equally  efficient  competitors  offering  only  some  of  the  components  cannot
compete  against the  discounted  bundle”.  The DMA  should  follow  the  same  holistic  approach.  This
would prevent circumventions, such as the current practice of Amazon to grant significant discounts
if an end user purchases goods on the Amazon Marketplace and agrees to also subscribe to Amazon’s
video-on-demand service Amazon Prime.

Audience measurement (article 6(1)g)
Measuring  audiences  across  media  is  increasingly  important  to  grasp  the  extent  of  multi-platform
reach of both content and advertising for the entire media and advertising industry. Key elements,
namely (1) independently verified information by trusted governance structures and (2) industry-wide

5 Commission proposal for a Digital Markets Act, explanatory memorandum, p.11.

agreed  measurement  standards,  have  long  been  essential  prerequisites  for  the  measurement  of
‘traditional’ media, such as TV, radio, print, etc. However, global online platforms have generally never
adhered to the same measurement practices of data transparency and independent data verification.
A significant recent development has been a high-profile push by the advertising industry to create a
cross-media measurement system to allow advertisers to track and optimise cross-media campaigns.
Part  of  this  process  has  been  to  gather  the  industry  around  a  transparent  and  globally  applicable
measurement solution, with the support of Google, Facebook and other global digital platforms. This
is a positive step towards greater platform data transparency and sharing, but there are concerns that
these platforms are playing too dominant a role in the design and development of the measurement
solution being built – especially as the blueprint was designed by Google and Facebook. Their ad-based
business models prompt questions about whether the measurement standards they have conceived
are ultimately designed primarily to serve their interests and further entrench their dominant market
positions.
We  therefore  welcome  the  provision  outlined  in  article  6  (1)(g)  that  imposes  a  new  European
obligation for audience measurement to ensure a minimum European-wide access to performance
measuring  tools  and  information  so  that  relevant  actors  can  carry  out  their  own  independent
verification  of  the  ad  inventory.  To  strengthen  this  provision,  we  strongly  recommend  that  it  is
amended  to  explicitly  empower  publishers  and  advertisers  to  receive  granular,  reliable  and
transparent  data  on  a  continuous  and  real  time  basis  that  has  been  independently  verified  by
trusted, approved and neutral third parties.

Contact details

European Broadcasting Union (EBU)
@ebu.ch
+

www.ebu.ch

EMMA/ENPA
@enpa.eu

www.enpa.eu
www.magazinemedia.eu

European Publishers Council (EPC)
@epceurope.eu
+
  www.epceurope.eu

Association of European Radios (AER)
@aereurope.org
+
| www.aereurope.org

EGTA
@egta.com
+

http://www.egta.com/

Association of Commercial Television in Europe (ACT)
@acte.be
+

www.acte.be

News Media Europe
@newsmediaeurope.eu
+

www.newsmediaeurope.eu