Exposing Ad Trade Secrets, and Unintended Consequences of the EU’s Digital Services Act

The European Parliament has passed into law the Digital Services Act, a regulation that aims to rein in big tech by curbing the spread of illegal content and disinformation on social media platforms in Europe. The DSA aims to limit how big tech collects data, as well as ban ads targeted at children and based on sensitive data. Noncompliance with this law costs 6% of the platform's total global turnover, while fines for repeat infringement may be up to 10%, reaching billions of dollars in fines. Although the law aims to regulate the digital space, privacy experts are concerned, specifically about how the DSA's transparency rules could change how advertisers invest ad spend on platforms. Sources are also worried that disclosing campaign algorithm details would potentially spill trade secrets. The DSA could kick in as early as mid-2023, giving platforms in the U.S. a taste of tightening regulation to come. 4 Charts Reveal How Facebook Traffic to Publishers Has Changed The DSA refers to online platforms with 45 million or more users. This could include networks, online marketplaces, app stores, online travel and accommodation sites, and content-sharing websites, according to Greg Mroczkowski, director, public policy at IAB Europe. But “not every publisher would be effectively an online platform,” he notes. “The ethos of this [law] is great, but there's going to be unintended consequences,” said Elizabeth Harding, privacy attorney and vice chair, technology transactions and data privacy group at Polsinelli. Buy-side must disclose campaign details The DSA prioritizes ad transparency: online platforms will be required to create ad libraries where anyone can search details on ad campaigns and compare. For example, under the DSA, an advertiser can search campaigns run by particular brands like Coca-Cola in certain regions or targeted to specific audiences. And a particular algorithm used for advertising purposes by that brand will be public knowledge, according to Mroczkowski. “This [search] has potential to be very granular and raises some questions on the buy side, specifically about advertisers at brands and in agencies,” he said. "It could potentially lead to unintentionally uncovering marketing strategies.” “Any public access to those algorithms can promote spammer content while creating issues around balancing trade secrets against the law,” said Harding. How brands target audiences is under the microscope The DSA’s transparency requirements ask online platforms to share information regarding the parameters used to display the ad to an individual. "With greater transparency for users around the targeting parameters used to deliver an ad to them, brands should contemplate whether the individual would find these attributes or parameters used in serving them an ad to be fair and responsible," said Arielle Garcia, chief privacy officer at UM Worldwide. For example, targeting people for a low-calorie chocolate ad campaign based on their interaction with similar brands or showing interest in chocolate is one thing, points out Garcia. "But targeting individuals interested in low-calorie snacks with an ad for a weight loss product," can be perceived as intrusive by people. “The [platforms'] knock-on effects have implications for brands and agencies in the industry at large,” Garcia added. Additionally, real-time programmatic tech companies like Xander or the Trade Desk, will need to co-participate in this process “because the data will need to come from somewhere to pass on the information in the supply chain,” said Mroczkowski. This means, “there would need to be a high level of integration between the buy side and sell side.” Content recommendation and ad effectiveness Similarly, under DSA, content recommendation algorithm transparency requirements need platforms to offer a non-profiling-based option to users for the content viewed, similar to Meta's chronological feed. “The purpose of having an engagement-based algorithm is to keep users engaged with the platform,” said Garcia. “But the option to not have something that's tailored to [a person’s] interests and prior engagement is what the DSA is going to require.” A platform may create an alternative content recommendation system that encourages as many likes and shares. However, core to platforms’ business models are systems that encourage audience interactions, such as likes and comments. More people using an alternative option may degrade the experience of the profiling-based option, "and that in turn, can impact ad effectiveness and ad cost [invested on the platform],” said Garcia. Platforms might be tempted to create a more negative user experience for an alternative viewing experience to dissuade people from choosing it. “I don't think that will be well received and potentially seen by regulators as akin to a dark pattern,” she said.