Based on comments from activists groups and legal experts, I now believe there is a high probability (maybe 70%) that the European Commission will reject the less personalized ads or ask for another option. Legal scholars note that the Commission is leaning heavily on the EDPB’s strict interpretation of “consent or pay” under the GDPR. GDPR states that consent must be given voluntarily and that this sense of voluntariness may be undermined if users are only given a choice between accepting data processing or paying a fee, without being offered a genuinely free option that avoids such data use. The less personalized ads misses this mark.
Alba Ribera Martínez, a lecturer in competition law, recently wrote that “if the EC is going to take the case with all seriousness and make the EDPB’s reasoning good, there is a great chance that Meta will face further challenges from the enforcer.” Similarly, Max Schrems—founder of NOYB and the litigant behind the landmark Schrems CJEU cases—has described the model as “less illegal,” suggesting the Commission is unlikely to accept it. His assessments tend to precede regulatory action.
Apple’s App Tracking Transparency (ATT) which took effect in 2021 can be used as proxy to model the impact of no personalized ads model on Meta’s revenue since both have some similarities such as causing customer opt-outs.
Meta had said the ATT would lead to an headwind of $10 billion on Meta Platforms. Given the DMA affected region contributes 16% of Meta’s revenue, that would be an headwind of $1.6 billion. However, ATT applied only to Apple users (roughly 40% of users) and only limited third-party data. In contrast, the DMA affects all users and restricts *first-party data—such as user behavior on Facebook and Instagram, making the expected impact materially larger.
Since the alternative to providing personal ads is introducing unskippable ads such as those in YouTube. Given the annoyance factor, many users may prefer personalized ads—especially since most people don’t consciously notice personalization. Therefore, I am giving more weight to the fact that around 30% of Meta users will opt out of the ads, leading to a decline in CPM by around 20%. The impact on CPM is based on estimated impact of Apple ATT on Meta’s CPM in 2022. I estimate this would translate into a weighted revenue headwind of:
- $3.7 billion in 2025 (2% of total revenue), starting in Q3
- $8.3 billion in 2026 (4% of projected revenue)
These estimates assume ad volume remains constant—lower CPMs could encourage advertisers to maintain, increase spend or shift spend to other platforms, but this isn’t guaranteed and is difficult to quantify.
Bank of America estimates a 4% ($8.45 billion) revenue hit in 2026, based on a 30% opt-out rate and 40% CPM decline. Goldman Sachs sees a 5–6% impact by 2027 under a 40% opt-out rate and 50% CPM decline.
Given that Meta’s core AI team executes well, as seen during Apple’s ATT, I expect them to come up with workarounds that could offset a large part of the losses from 2026. This is probably why analysts are not giving much attention to the impact of the DMA.