On both sides of the Atlantic, legislators are considering restrictions on targeted advertising, such as the Banning Surveillance Advertising Act in the United States and the Digital Services Act in the European Union. These proposals have never been grounded in facts and logic—despite claims that targeted ads are a massive intrusion on consumer privacy, advertisers rarely obtain consumer data (instead they pay ad platforms to deliver ads to a specific audience). And critics of targeted advertising refuse to acknowledge the ample benefits of personalization to advertisers, publishers, and consumers alike, especially how these ads fund the Internet economy. But as the most popular platforms continue to introduce new innovations in online advertising, the rationale for banning targeted ads makes even less sense now than ever before.
Targeted online ads form an essential part of the digital economy: advertisers can link consumers to specific queries and interests, and then show them relevant ads as they visit different websites. This has three positive effects: first, consumers see ads for items that are likelier to be relevant to them than the non-targeted ads they encounter in traditional media such as magazines and television. Second, advertisers spend their marketing budget on ads that are likelier to generate a response from the audience, which makes their ad spend more cost-effective and affordable than traditional forms of marketing. This is why personalized ads have been a godsend to small businesses: millions of enterprises benefit from the ability to show their wares to interested customers, rather than wasting money on ads shown to uninterested audiences. Third, websites and app publishers can sell inventory on their sites to advertisers, earning them valuable income and allowing them to offer content and services to users for free.
The assault on targeted ads is thus entirely misguided. Yet critics routinely point to targeted advertising in an attempt to vilify Big Tech, claiming that they are the architects of “surveillance capitalism” that is undermining personal autonomy and eroding democracy. While their critiques were wrong in the past, they are increasingly nonsensical given the many technological innovations that major online advertising platforms are introducing.
Google has set a 2023 deadline to end cross-site identity tracking and eliminate third-party cookies on its Chrome browser, and will similarly phase out ad identifiers on its mobile Android platform. The company plans to replace existing methods with a suite of privacy-preserving online advertising solutions, whilst trying to safeguard the basic principle—confirmed many times over—that personalized ads increase revenues for publishers and make marketing significantly more cost-effective for advertisers.
Beginning this year, Google will move Chrome-based ads over to its new “Topics” system. The basic idea is that Chrome assigns users five interests (out of a set of 350 topics, none of which include sensitive information like sexuality or religion) based on the websites they visit. After allocating these interests, Topics shares them with publishers and ad platforms, which enables them to provide personalized ads without access to granular user data or identifiers.
Figure 1: Google Topics
Source: The Drum
In this system, ad personalization takes place on the user’s device, no personal information is shared on external servers, and Google deletes all data after a maximum of three weeks. Google’s “Topics” service will preserve user anonymity, as advertisers and publishers receive no personally identifying data when they serve ads.
Meta (formerly Facebook) is also innovating in the adtech space. It has set up its Conversions API, a tool for advertisers to more accurately identify which of their ad campaigns helped bring about new customers. Meta designed this system to gradually replace the “Pixel” tool that was previously used to measure users’ interactions with ads shown on the Facebook platform. Another strategy Meta is pursuing is to encourage advertisers to integrate its direct messaging services (Facebook Messenger, WhatsApp, and Instagram) into the ads shown to users. This is a win-win for advertisers and customers: it allows users to click on an ad and open an instant conversation on Messenger or WhatsApp with an advertiser’s sales representative. This makes it easier for advertisers to make sales directly within Meta’s platforms and enables a quick and interactive service experience for users.
Another advertising tool Meta is testing is called Private Conversion Lift. With Private Conversion Lift, advertisers can request target audiences, which Facebook randomly separates into a test and a control group, serving the ad to the former only. By measuring the subsequent sales uplift to the test group, advertisers can more accurately track the attribution of sales to specific ads.
Figure 2: Facebook Private Conversion LiftSource: Facebook
The theme across both Google and Meta’s ad personalization innovation efforts is helping advertisers know which of their ads are working without sharing user data in the process. Both companies are using probabilistic attribution, where sales generated by personalized ads cannot be linked to ad engagements with uniquely identifying user characteristics. Meta is collaborating with Mozilla, a web browser company, to develop a tool called Interoperable Private Attribution (IPA), which aggregates advertising outcomes, such that they cannot be traced to individuals.
The irony is that banning targeted advertising today would not only hurt consumers, businesses, and publishers, but given these recent innovations in online advertising, would likely have the unintended consequence of giving an artificial boost to the same Big Tech companies that critics say have too much power by reducing competition from traditional targeted advertising platforms. Rather than pushing for restrictions on targeted advertising, policymakers and civil society should allow the private sector to do what it does best: innovating and developing novel technologies that improve welfare for everyone, including publishers, (who can continue to earn billions in advertising income), consumers (who can obtain the benefits of free, ad-supported apps and websites, plus prefer to see ads tailored to their needs rather than being blanketed with irrelevant messages), and advertisers (who can continue to access affordable, effective ads, instead of relying on the kinds of pre-digital marketing that only helps large brands).