This article is part of a three-part series investigating the role of “privacy capitalism” in shaping the EU proposal to ban targeted advertising:
- Introduction: How “Privacy Capitalism” Is Taking Over European Policy Debates
- Part 1: No, Contextual Advertising Is Not a Substitute for Targeted Advertising
- Part 2: The Political and Corporate Interests Behind the Push to Ban Targeted Ads
- Part 3: Oops, the Groups Pushing to Ban Targeted Ads Use Targeted Ads Themselves
Online advertising pays for the vast array of free online apps and services. Last year, online ads generated nearly €70 billion in revenue last year in Europe, with a significant portion coming from targeted ads. Those pushing to ban targeted ads in the EU have begun arguing that targeted ads are unnecessary because contextual advertising, which deliver ads based on the content of the page rather than information about users, can generate the same amount of revenue without any of the alleged downsides. They know that if they can convince policymakers that a ban on targeted ads comes with no costs, their task will be much easier.
While a number of advocates have echoed this claim, the evidence they cite is dubious at best. Let’s consider the three primary claims from a recent report by the Irish Council for Civil Liberties (ICCL), a European advocacy group, pushing for a ban.
Claim #1: “Advertising revenue increased by 149% when Dutch publisher NPO Group replaced tracking-based ads with contextual-based ads.” (ICCL, 2021)
Facts: At first glance, these results may seem impressive. But notably, it omits a crucial detail: after the EU implemented the GDPR in 2018, NPO began asking users for consent to show targeted ads. NPO reported that 90 percent of users did not opt in, which means it could only effectively deliver targeted ads to 10 percent of its users. So comparisons to revenue in 2019 are highly misleading, as they do not reflect a true head-to-head comparison between targeted ads and contextual ads. Notably, ICCL has not provided comparisons to 2018 or earlier.
Also, consider the source. The chief policy officer for Brave, a company that sells contextual advertising, provided the analysis (1, 2) cited by this study. Perhaps even more curiously, Brave’s chief policy officer moved to ICCL and authored the ICCL’s report. So, in short, both the report and the evidence cited by the report, all appear to have come from the same individual, and that individual has not released any of the underlying data that would allow for meaningful analysis.
Claim #2: “A Norwegian news publishing group earned an average of 391% more for contextual ads than tracking based ads over 12 months.” (ICCL, 2021)
Facts: Once again, the context matters. The analysis refers to a period of time between July 2020 and May 2021, again, after the GDPR went into effect. So it is not clear how many of these publications may have already implemented changes to their website that significantly reduced their use of third-party data that could allow for targeted advertising. Undoubtedly, targeted ads are not going to be effective, and thus will earn less revenue, if there is no data available to personalize the ads to users. In these cases, contextual advertising will provide more revenue than targeted ads, but that is not a reasonable comparison.
Again, it is also worth noting the source. Buried in the endnotes to this report, ICCL discloses that it “uses data provided by Kobler.” Kobler is a Norwegian company that sells contextual ads to publishers, and thus has a significant financial stake in banning targeted ads. While that does not mean the evidence it provides is wrong, it suggests that it should be independently validated and verified.
Claim #3: “TV2, a major Norwegian news website, reports that ads sold through Kobler’s contextual targeting return a 210% higher average price than competing tracking-based ad targeting.” (ICCL, 2021)
Facts: While this claim at least acknowledges it comes from Kobler, it provides no details again to make the comparison meaningful, such as by acknowledging what data, if any, the news website used for its targeted ads, or when this comparison occurred. But perhaps what makes the claim the most suspect is that, at least as of late November 2021, TV2 still uses targeted ads. Why would a major news website choose to use targeted ads that earn significantly less revenue if contextual ads earned it much more? There appears to be more to the story than what the ICCL reports.
Finally, ICCL’s report concludes that “practical evidence from European publishers now shows that publishers’ ad revenue can increase when tracking-based advertising is switched off.” In short, it is claiming that ad publishers stand to make more money if they would use contextual ads instead of targeted ads. Not only does ICCL fail to back up this claim with any solid evidence, it also defies common sense—if businesses could make more money with contextual ads over targeted ads they would switch voluntarily. The primary reason online apps and services use targeted ads is because they are more effective and thus generate more revenue.
Notably, those defending targeted ads are not trying to ban contextual ads. If contextual ads are better, ad publishers should use them. There is nothing wrong with them and they certainly have many important uses.
Those calling for a ban on targeted ads may believe sacrificing free online apps and services for consumers is a price they are willing to pay (and that the average consumer agrees with them), but they shouldn’t pretend such a ban will come at no cost, or even more outlandishly, generate a boost in ad revenue for European app developers, news publishers, or content creators.
Image credit: Joshua Earle on Unsplash.