A 571-page study commissioned by the European Commission concludes that the Digital Markets Act's advertising transparency rules have not given advertisers and publishers meaningful access to the pricing and performance data that Google, Meta and Amazon hold. The report, titled "Online advertising at the crossroads of different regulatory frameworks", was released on the EU Publications Office website on October 29, 2026. It was prepared by the research consultancy AWO for the Commission's Directorate-General for Communications Networks, Content and Technology (DG CNECT), and written by Nick Botton, Lex Zard and Oana Goga. The manuscript was completed in April 2026, and the opinions in it are those of the authors, not the Commission's official position, according to the document's legal notice.

In Short

The European Union asked researchers to check whether its new rules for the biggest tech companies had made online advertising easier to understand for the businesses that buy and sell ads. The researchers spoke to 60 organisations and surveyed 50 advertisers, and found that Google, Meta and Amazon still share too little detail about ad prices, auctions and where ads appear for most businesses to check those numbers independently. The report is advice, not law, but it gives the Commission a list of options, from tighter rules under existing laws to splitting up companies that run both sides of the ad market.

What the study is and how it was built

The study answers four questions set by the Commission in a 2024 call for tenders: what has changed in the online advertising industry, how those changes affect consumers and businesses, what the Digital Services Act (DSA) and the Digital Markets Act (DMA) have done for users, publishers and advertisers, and where EU law leaves gaps.

The authors describe the report's central contribution plainly. "The study's most important contribution is its findings on the impacts of Articles 5(9), 5(10) and 6(8) DMA on the transparency of the provision of online advertising services by gatekeepers," the document states, adding that "this is the first study to analyse the impact of these provisions." Articles 5(9) and 5(10) require gatekeepers to tell advertisers and publishers about prices, fees and remuneration. Article 6(8) requires them to give advertisers, publishers and their chosen third parties access to the data and tools needed to measure and verify ad performance.

Evidence came from four sources. The researchers interviewed 60 stakeholders, among them 14 advertisers, 10 publishers and publisher groups, 7 adtech intermediaries, 7 civil society organisations and 4 gatekeepers: Amazon, Apple, Google and Meta. Named participants include the World Federation of Advertisers, ISBA, the European Publishers Council, Schibsted, Criteo, ID5, IAB Europe and Check My Ads Institute; many others asked for anonymity.

A survey run with the market research agency RONIN reached 50 advertisers in six countries, 20 of them in Germany and 13 in France. Three quarters of respondents (74%) worked at companies with 250 or more employees. A September 2025 workshop drew 33 participants in person and 50 online, and the researchers reviewed the gatekeepers' public DMA compliance reports.

The market under review

The report opens with the size of the business. Advertisers were expected to spend just over €790bn on online advertising globally in 2025, according to a WPP Media forecast cited in the study. The EU is the third-largest single market after the US and China, with €63bn spent to reach EU consumers online in 2024, based on IAB Europe AdEx data that excludes Cyprus, Luxembourg and Malta.

Per head, the gap is wide. Ad spend per capita in the EU amounts to €147, against €704 in the US and €640 in the UK, according to the study's own calculation. Concentration is the other defining feature: approximately 60% of global ad spend flowed to Google, Meta and Amazon in 2023, the document states.

The authors frame the problems in three groups: privacy and data protection, user safety, and contestability and fairness. "Much market innovation has grown misaligned with the objectives of EU law linked to privacy, user safety, contestability and fairness," the abstract reads.

Article 5(2): consent prompts and no visible change

Article 5(2) of the DMA bars gatekeepers from combining personal data across their services for advertising without user consent. On paper, it is one of the regulation's central rules. In practice, according to the study, the article and the compliance measures gatekeepers built around it "do not appear to have significantly affected the extent to which gatekeepers process, combine and cross-use data for online advertising purposes."

The researchers offer two explanations. The first is consent design: gatekeepers have been documented relying on a single consent prompt covering several services, and those prompts may be built to maximise opt-ins. The second is technical. Gatekeepers may be using privacy-enhancing technologies to pseudonymise data across services, and may then treat that data as falling outside "personal data", or treat insights derived this way as something other than "combining" data. "When questioned on this matter, none of the gatekeepers we interviewed provided information with sufficient clarity or detail to enable us to assess with reasonable assurance how such data is handled in practice," the report states.

Industry interviewees were blunt. One intermediary association called the article a "great disappointment" because "nothing has changed." A publisher said they "haven't seen changes."

Revenue numbers point the same way. Meta's revenue grew 22% in 2024, Google's 14% and Amazon's 11%, according to annual reports cited in the study. Amazon said it "did not have data to share" on any revenue effect, but described DMA compliance costs involving "hundreds of teams" that "vastly exceed the European Commission's estimate of €2m per company." During a 2025 Commission workshop, Google reportedly said Article 5(2) compliance had required 3,000 engineers working full-time for two years, according to an account of that workshop cited by the authors. Even so, the study judges it "unlikely" that these costs had a material effect on either company's finances.

Meta did not consent to having quotes attributed to it in the published study. Google declined to share non-public data on the question and pointed the researchers to an economic study funded by a trade association of which it is a member.

Pricing transparency: one toggle for every counterparty

Articles 5(9) and 5(10) were meant to let advertisers and publishers see what they pay, what intermediaries take and what publishers receive. Amazon, Google and Meta each told the Commission in their compliance reports that they already complied before the rules took effect, while adding new tools.

Advertisers split down the middle. Asked whether they had enough information to understand the price of Google, Meta and Amazon services and why prices fluctuate, 48% of surveyed advertisers said yes, 48% said no and 4% were unsure. Most interviewed advertisers were dissatisfied. One said that "while gatekeepers typically provide some level of pricing data, insights into the factors driving price fluctuations are not enough." Another argued that Google's information was getting "harder to understand," pointing to an increased "bundling of services and ad types" that produced "less clarity on why we're paying what we're paying."

Publishers were unanimous. "All publishers and publisher industry groups interviewed for this study argued that Article 5(10) DMA and gatekeepers' relevant compliance measures had failed so far to improve ad pricing transparency," the study reports. One publisher industry group said there had been "no discernible change." Another said there was still "no transparency on the price paid by advertisers, the fees charged by gatekeepers, and rules on monetisation."

The identifier problem

Much of the criticism focused on Google. A publisher industry group told the researchers that the unique identifier attached to bids in Google's Regional Data Access files differs from the one used in impression files, so it is "not possible to link data related to bids with data related to impressions." One publisher called the files "worthless." That publisher said that "for privacy reasons, they have removed the identifier in the files that they are obliged to give us," without an explanation of why an identifier unrelated to individual users posed a privacy threat.

Google does offer a separate file, called "NetworkBackfillBidsJoinable", that can be joined with other data transfer files. Publishers must buy it, which, according to them, suggested that comprehensive pricing transparency may only be available at an additional cost. Volume is a separate complaint. One publisher industry group said Google's data transfer files are "so big and hefty" that they "cannot be opened on a normal computer." A representative of a large publishing group said that "not even a handful of publishers have the time to go through all this data."

The design of the consent mechanism for sharing prices draws some of the study's sharpest criticism. Under Articles 5(9) and 5(10), granular price data flows between an advertiser and a publisher only if both consent. Google's advertiser prompt, reproduced in the study, asks for consent to share price data "with any publisher located in the EEA on whose property that advertisement was served."

One publisher described the result as "double consent" and "blanket consent." A publisher industry group said the "global toggle" in Google Ad Manager settings "discourages the sharing of information thereby defeating the purpose / effectiveness of Article 5(10)," and expected that most advertisers and publishers "just say no." Another group accused Google of "misusing consent" as a "defensive mechanism to prevent transparency."

On the buy side, the prompt seems to have gone largely unseen. An advertiser industry group said it had "received no feedback from [their] 250 members" when it asked about the feature. A representative of one of the six largest holding companies, according to the group, "said they had never heard of it." The group eventually traced the prompt to a member's demand-side platform, meaning advertisers would learn of it only if their DSP told their agency and the agency told them.

What did advertisers want instead? Minimum bid prices, the number of advertisers in an auction, real-time data on whether competitors win the impressions they bid for, and, for Google Search, more detail on why some search terms cost more.

Performance data: satisfied buyers, unverifiable numbers

Article 6(8) produced the most mixed picture. The survey shows broad contentment: 76% of advertisers were satisfied or very satisfied with their measurement tools, and 96% relied on the built-in tools of Google, Meta or Amazon. Third-party measurement providers were used by 34%, and agencies by 30%. Asked whether the amount or quality of performance data from the three gatekeepers had changed since 2022, 18% said it had improved significantly, 50% said it had improved somewhat, 6% saw no change and 26% said it had declined.

Interviews told a different story. Most advertisers the researchers spoke to disagreed that the DMA had improved performance transparency. "We've just sort of accepted that this is just how these walled gardens operate, and we haven't felt that we're in the position that we can challenge that," one large advertiser said. Some reported that requests for data had been refused on privacy grounds since the DMA took effect. "They started using privacy as a reason not to share data," one advertiser said, citing Google Search: "we used to be able to see full search queries and how many people used them, but now Google only reports examples of query searches, and approximations of their usage."

The granularity question is where the study's verdict lands hardest. Only 58% of surveyed advertisers said Google, Meta or Amazon provide raw data, such as log-level data, that would allow their own verification. Access varies sharply by metric: 82% said they "always" had access to impression data, against 46% for conversions, 45% for frequency and just 18% for attribution data. Amazon's DMA Clean Room API, introduced for Article 6(8), "only produces aggregate reports, without providing access to the raw exposure data processed within it," the study notes. Gatekeepers increasingly rely on data clean rooms to provide this kind of access.

The study does credit some changes. In 2025, Google introduced placement transparency for Performance Max and the Search Partner Network, the report notes. PPC Land covered both: Google announced channel performance reporting for Performance Max on April 30, 2025, and enabled full placement reporting for the Search Partner Network across Search, Shopping and App campaigns in August 2025. Meta, according to the study, has not introduced similar placement transparency for the Meta Audience Network.

Publishers saw nothing new. "We haven't gotten anything that we didn't get before the DMA," one said. A publisher industry group described what they do receive as "topline, like socio-demographic, what device was used, how many impressions, etc. [...] it's the basic stuff they used to receive pre-DMA."

What the compliance reports admit

The study compares these accounts with what the gatekeepers say publicly. Google's 2025 DMA compliance report states that the section on business user feedback is "not applicable because Google has not implemented new measures since 6 March 2024" to comply with Articles 5(9), 5(10) and 6(8), and redacts its effectiveness data. Meta's 2024 report says it "regularly enhances those tools in response to the feedback that it receives from its advertisers" and shares no effectiveness data.

Performance Max, Advantage+ and the "black box"

Algorithmic targeting gets its own scrutiny. In the survey, 60% of advertisers had used Google Performance Max, 48% Meta Advantage+ and 12% Amazon Performance+. Results were broadly positive: 12% reported a large increase in performance from AI targeting and 44% a small increase, while 8% saw a small decrease. Yet 62% worried about having less control over targeting, 46% about performance being harder to verify, 36% about privacy risks and 28% about rising costs. Only 8% had no concerns.

Interviewees rarely complained about performance itself. Many described the tools as "black boxes," and their complaints centred on placement and control. "The problem with some of the more AI-driven systems is that they deliver anywhere," one agency said. Another agency said its largest clients "don't want to be on Gmail, because we consider that placement to be weak." One advertiser does not use Performance Max at all: "lack of transparency means that we don't use it."

Some went further, attributing the design to commercial motives. "Profit-maximisation is the real explanation. The less restrictions there are, the more impressions," one advertiser said. Another described the conflict as "giving the auctioneer all our money and hoping they give us the most effectiveness."

The 2025 changes to Performance Max reporting were described by one agency as "encouraging." An advertiser industry group was less impressed: "it's still not perfect."

The DSA: repositories used for competitive research, not policing

Under Article 39 of the DSA, Very Large Online Platforms and search engines must maintain public ad repositories. Advertisers are barely using them for their intended purpose. Less than half of surveyed advertisers (48%) knew the repositories existed. Some 26% had been affected by IP-infringing or scam ads, such as ads impersonating their brand, yet none of those advertisers reported using repositories to identify them. Those who did use repositories did so "to gain inspiration from competitors."

There is also a scope gap. None of the designated platforms use their repositories to disclose ads served through their open-web intermediary services, such as the Google Display Network or Meta Audience Network, reflecting their reading that the DSA covers only ads on the platform's own interface. "Under this interpretation, the open web, and thus a large proportion of the risks associated with online advertising, are not tackled by Article 39 DSA," the study states.

Survey respondents named Facebook as the platform with the worst brand safety problem (30.6% of those answering), followed by Google Ads on the open web (18.4%) and X (12.2%). On ad fraud, Facebook (32.7%) and Google Ads on the open web (30.6%) led again.

Scam ads receive extended treatment. The study cites Gen Threat Labs research finding that 30.9% of 14.5m ads shown on Meta platforms in the EU and UK "pointed to a scam, phishing or malware link," and Meta documents reported by Reuters projecting that 10% of its 2024 revenue would come from scam ads and ads for banned products. PPC Land reported in November 2025 that this amounted to roughly $16bn.

The study's chapter on the two laws ends with a qualified judgement. The DMA and DSA "have had a positive, albeit limited, impact" on publisher and advertiser concerns, but that impact "has generally fallen short of achieving the broader objectives of the DMA and DSA."

Ad fraud: estimates without peer review

The authors are candid about the evidence base. "Peer-reviewed studies about the prevalence of ad fraud are lacking, and industry estimates vary greatly," the report states. IAB Europe cited a World Federation of Advertisers estimate in 2020 that ad fraud losses would reach €45bn in 2025. Juniper Research estimated that 22% of global ad spend, or €76bn, was lost to fraud in 2023. Verification providers interviewed for the study hypothesised that as much as 25% of global ad spend could be going to fraudsters.

Applied to the EU, the authors calculate that a 25% loss rate would amount to over €15bn in 2024, which they describe as more than three times the revenue of Axel Springer, the EU's largest publishing group. That is an illustration built on a hypothesis, not a measurement. Augustine Fou, an independent ad fraud researcher listed among the study's interviewees, has argued separately that headline figures undercount the problem: PPC Land reported on October 21, 2026 that Fou says a $63bn invalid traffic estimate undercounts fraud on CPM-priced ads.

Privacy Sandbox and ATT: privacy gains, publisher losses

The study spends considerable space on the two biggest industry privacy initiatives. Both, it concludes, reduced or would have reduced publisher revenues.

For Google's Privacy Sandbox, the numbers come mostly from tests run before the project's cancellation. Google's 2024 study found that removing cookies had a 33% negative impact on revenues through Google Ad Manager and 21% through AdSense; with the Sandbox APIs switched on, the drops narrowed to 20% and 18%. Raptive found ads on Chrome without third-party cookies generated 30% less revenue. Index Exchange found a 33% decline in CPMs. Criteo found a 60% drop in publisher revenues. Google itself cautioned, in a written statement for the study, that "[t]he impact of the APIs in a world where cookies are maintained [...] is likely to be very different."

Publishers called the process "shambolic," "a joke" and "a mess." PPC Land reported in October 2025 that Chrome retired most Privacy Sandbox technologies, including Topics and Protected Audience, after adoption failed. UK Competition and Markets Authority oversight of the project ended the same month, the study notes.

For Apple's App Tracking Transparency, the study cites research by Kraft and others showing that the price of ads delivered to users who rejected an app's tracking request was 51% lower in France, 39% lower in Germany and 39% lower in the Netherlands. Three app publishers told the CMA their ad revenue fell by 27% to 40%. France's competition authority fined Apple €150 million over the framework in March 2025, a decision the study quotes as finding that the ATT framework "is not necessary, insofar as the consent obtained is not valid under the applicable laws."

AI, search traffic and the chatbot gap

On search, the study gathers several data points. Digital Content Next found in May and June 2025 that year-on-year traffic from Google Search to premium publishers fell 10% on average, and 14% for non-news publishers. Referrals from AI platforms other than Google's grew sharply, by 99.4% for the New York Times between 2024 and 2025, but still represented less than 1% of total referrals, according to Similarweb data cited in the study. Perplexity's programme to share 80% of Comet subscription revenue with publishers was cancelled in February 2026.

OpenAI has begun testing ads in ChatGPT, including through Criteo, the study notes; PPC Land reported Criteo's entry into the ChatGPT ad pilot on March 2, 2026. The authors see a legal gap. If AI chatbots are classified as search engines under the DSA, their ads would fall outside Article 26 on ad transparency and Article 28 on the protection of minors, which do not apply to search engines. A user planning a holiday with such a system "could be presented with information and links provided by advertisers (e.g., hotels and tours) that are simply worked into the output of the system," the study warns, and "it is not clear whether users would be able to identify the sponsored information."

The study describes ChatGPT as "approaching the 45m monthly active users threshold" for designation as a Very Large Online Search Engine. That description has since been overtaken: PPC Land reported in August 2026 that ChatGPT faces EU risk rules after declaring 159.1 million average monthly users, a figure self-declared by the provider.

Separately, the Commission is investigating Google's use of publisher content to power AI Overviews and AI Mode, the study notes. PPC Land covered the opening of that formal antitrust investigation on December 9, 2025.

The recommendations

The final chapter sorts its proposals into three categories: building on existing law, new initiatives that would need further study and a full impact assessment, and simplification.

Using powers the Commission already has

For performance data, the study points to Article 8(2) of the DMA, which allows the Commission to adopt an implementing act specifying how a gatekeeper must comply with Article 6. Such an act could require:

  • Ad placement data, including the sub-domain and URL of each ad, and the content and creators it appeared next to on social media, connected TV and video on demand, potentially extended to conversational AI.
  • A transaction ID, such as the IAB Tech Lab's Shared Campaign Identifier, so that gatekeeper data can be matched with advertisers' and publishers' own records.
  • Independent collection by verification providers through their own technical means, rather than gatekeepers supplying the data themselves.

For pricing, Article 12 of the DMA allows delegated acts specifying how Articles 5(9) and 5(10) are performed. The study proposes auction transparency covering minimum bid prices, the number of advertisers in an auction and the kinds of companies winning bids, and, for publishers, auction type, bid amounts and bid parameters. It also proposes rewriting the consent framework so that advertisers and publishers can "pick and choose which partners they would like to share information with."

Other proposals include tightening Article 5(2) consent through Article 12 or the Joint Guidelines on the DMA and GDPR, due to be finalised in 2026, and checking whether connected TV operating systems such as Android TV, Samsung and Amazon Fire qualify as core platform services. The study also recommends designating providers of generative AI tools as a distinct core platform service, warning that early entrants combining AI, browsers and ad platforms "could entrench themselves in ways that are difficult to address ex post."

Under the DSA, the authors suggest Article 35(3) guidelines requiring platforms to publish more data on ad-related risks and monetisation. Ad repositories would gain targeting parameters, including whether algorithmic targeting such as Performance Max was used, and placement information.

New initiatives

The most consequential proposal is structural separation. "Many of the problems in the online advertising market identified in this study are at least partially linked to vertical integration, especially in the case of Google," the report states, recalling the Commission's preliminary view in its 2023 Statement of Objections that "only the mandatory divestment by Google of part of its services would address its competition concerns." PPC Land covered the €2.95 billion fine the Commission imposed on Google on September 5, 2025 for self-preferencing in adtech, and Google's November 2025 response rejecting an asset sale. The study cites a Vermont bill, H.390, that would bar companies above a certain size from owning both buy-side and sell-side platforms. Short of separation, it suggests conflict-of-interest rules modelled on the MiFID II financial markets directive.

The second is de-risking RTB. Supply-side platforms and ad exchanges would be required to fully anonymise bid requests in open auctions where the security and confidentiality of personal data cannot be guaranteed. The proposal targets real-time bidding, which on the open web "generally sees users' personal data shared with thousands of other market participants," according to the study. Gatekeepers would also face a dedicated framework for collecting advertising consent, and a legal obligation would make browser privacy signals such as Global Privacy Control, Do Not Track and the Advanced Data Protection Control valid expressions of data protection rights.

The third is a due diligence regime for adtech intermediaries, borrowing from the Anti-Money Laundering Directive and MiFID II. It would include risk assessments, a "best execution" duty, a ban on inducements that bias routing decisions, and a Know Your Business Customer requirement under which "an SSP could be required to verify the identity of each publisher that it processes bid requests for." Sellers.json and the SupplyChain Object could serve as compliance tools.

The fourth is an EU Ad Fraud Taskforce, covering threat sharing, wider adoption of the SupplyChain Object and disclosure rules for verification vendors. Vendors do not always disclose the sample size behind reported fraud rates, the study notes, and "if only 10% of impressions are tested, the reported fraud rate may differ substantially from what would be observed if 100% of impressions were analysed."

Simplification

The simplification package centres on universal control tools: a browser or operating system interface letting users set advertising preferences once across the market, switch off behavioural advertising under the GDPR's right to object, and include or exclude specific publishers and intermediaries. The study warns that gatekeepers' role in building them "may thus need to be limited," given its findings on Privacy Sandbox and ATT.

The authors name Article 88b of the Commission's Digital Omnibus proposal as a possible legal basis. That route has since narrowed. PPC Land reported in June 2026 that the Council position published on June 18, 2026 removed Article 88b, the provision for an automated browser signal, from the Digital Omnibus text. Research published in May 2026 had also found that Global Privacy Control could reduce EU consent banners only partially without legal change.

Other measures include guidance on overlaps between the GDPR, DSA, DMA, the Unfair Commercial Practices Directive and the AI Act; harmonising consent collection beyond the existing EDPB opinion on Consent or Pay models; narrowing divergence between national data protection authorities, with regulators in France and Italy differing from others on anonymous statistics; and exploring opt-out regimes for contextual and cohort-based targeting and for conversion measurement. Exemptions "should only be permitted under strict, clearly defined criteria," the study cautions.

What the evidence does not establish

The study is open about its limits, and several matter for how its findings are read. The survey covers 50 advertisers in six countries, recruited through a research agency's networks and past panels. It is not a representative sample of EU advertisers, and the authors do not present it as one. The interviews favoured advertisers by design "because we identified a large variety in their perspectives," and many participants spoke anonymously.

The two data sources also point in different directions. "While our interviews suggest that the DMA has had a limited impact on improving transparency, our survey suggests that less than half of advertisers are concerned about this," the authors write. Advertisers with Joint Business Plans with gatekeepers "had few complaints," they found. A majority of surveyed advertisers reported improved performance data since 2022, even as interviewees said little had changed.

On the DSA, the authors say they could not assess the effectiveness of ad screening or monetisation policies "based on publicly available data." The ad fraud figures rest on industry estimates the authors describe as varying greatly. And the April 2026 manuscript predates the Council's June 2026 position on the Digital Omnibus and the Commission's July 2026 DMA fines on Google.

Why this matters for the marketing community

For advertisers running Performance Max, Advantage+ or Amazon DSP campaigns in Europe, the study puts on the record of a Commission-funded report several complaints that industry publications have also documented: aggregated reporting, missing join keys, no independent verification and pricing tools few buyers know exist. Its proposals are specific enough to translate into implementing and delegated acts, which the Commission can adopt without new legislation. That is a shorter path than the structural remedies that dominate headlines.

The timing adds weight. PPC Land reported that the Commission fined Google €890 million on July 23, 2026 in two DMA non-compliance decisions, after fining Apple €500 million and Meta €200 million in April 2025. The Joint Guidelines on the DMA and the GDPR, which the study names as a vehicle for tightening gatekeeper consent, drew more than 100 public submissions published in March 2026, including a submission from Check My Ads Institute on Article 5(2) consent, Performance Max and Advantage+.

For publishers, the study documents a failure that is technical rather than legal. The rules exist; the data arrives; but without a common identifier, without consent rates that allow price sharing, and with files too large to open on a normal computer, it does little. One intermediary interviewed argued that a transaction ID alone would be enough to make the data usable: "we don't need user identifiers that we could re-use anywhere else."

The report's closing argument is broader than the DMA. The market's reliance on personal data, the authors conclude, is "the source of problems linked to privacy and data protection, the dissemination of content associated with systemic risks, and exploitative targeting practices," as well as the basis of the network effects that concentrate it. They describe reducing that reliance as "akin to reducing its reliance on an unsustainable fuel." Whether the Commission adopts that view, which the document says is not its official position, is a question the study leaves open.

Timeline

Summary

Who: The research consultancy AWO, with authors Nick Botton, Lex Zard and Oana Goga, for the European Commission's DG CNECT. The study draws on 60 stakeholder interviews, including Amazon, Apple, Google and Meta, and a survey of 50 advertisers in six EU countries.

What: A 571-page study, "Online advertising at the crossroads of different regulatory frameworks", which finds that the DMA's advertising rules (Articles 5(2), 5(9), 5(10) and 6(8)) have not meaningfully improved data combination limits, pricing transparency or performance data access with Google, Meta and Amazon. It recommends implementing and delegated acts under the DMA, structural separation of buy-side and sell-side businesses, anonymised bid requests, a due diligence regime for adtech intermediaries, an EU Ad Fraud Taskforce and certified universal control tools for users.

When: The study was released on the EU Publications Office website on October 29, 2026. The manuscript was completed in April 2026.

Where: The European Union, with the survey covering Germany, France, Ireland, the Netherlands, Belgium and Austria.

Why: The Commission asked for an overview of industry developments and an assessment of whether the DMA and DSA give advertisers and publishers enough transparency, choice and fairness. The answer, according to the authors, is that improvements have been "piecemeal, and insufficient for advertisers and publishers to make informed decisions about where to spend their resources." The study states the opinions of its authors, not the Commission's official position.