It is a truism to state that the Internet is powered by data. The world wide web is a communication protocol for the electronic exchange of data. This remarkably simple technology has given rise to a Cambrian explosion of information production and consumption—including your ability to read this text on your screen, in real-time without someone having to print and deliver it to you.
It goes without saying that websites and apps cost money to set up and run. There are bills for software developers, content producers, servers, electricity, interacting with customers, and so on. Website operators therefore need to figure out how to make ends meet. Many businesses existed offline-first and added an online offering in order to generate more sales. Other businesses are online-first and thus need to monetize exclusively through their website or mobile app.
One option is to charge users directly for access. This is only feasible if enough people are willing to pay for the content or service shown on the website, so one needs to offer a particularly useful or popular product for this to work. Another problem is that it shuts out the share of the population that lack sufficient income to pay for online services.
A second option is to sell goods and services through a website. This is a viable strategy for e-commerce businesses, though they need to ensure that they reach their potential market. Which brings us to the third option: charging others to buy screen real estate on a website on which users are shown ads. This helps businesses that place such ads to find new customers. It also helps websites that cannot charge for access and don’t sell anything directly, such as non-paywalled newspapers, most apps (around 80 percent of which are free to download), and other websites whose sole product is content.
Letting others pay money to websites in order to post ads thus forms the lifeblood of the Internet economy. This point cannot be stressed enough. Thirty years of web evolution is powered by this reality. This shouldn’t come as a surprise: since the beginning of the Internet when the optimal business model was still up for grabs, it became clear that the free, ad-supported model was the only one that could scale.
Since the first known ad in Thebes in 3000 BC, sellers have used advertising to get customers’ attention. Because the Internet is fundamentally a network of data, it opens up new possibilities for advertising hitherto unavailable in offline marketing. Advertising networks can build an understanding of what kind of person visits a website, what their interests are, and thus what they might want to buy. In the past, advertisers could only rely on a fuzzy picture of who would see their ads. A print ad in a car magazine would likely be read by car enthusiasts. A television ad during a tourism program should reach a viewership that is interested in travel. A billboard, or an ad on public transport, leaflet, or even a newspaper ad, on the other hand, targets the audience based on their general location and thus lacks precision: a lot of eyeballs looking at the ad won’t be interested in it.
This is critical: the less people seeing an ad are interested in its content, the worse is the ad’s return on investment. Money doesn’t grow on trees, and businesses don’t have limitless advertising budgets. They need to decide where to allocate their marketing spend. The better they can target their desired audience, the more sense it makes to spend money on ads. And the more effective each ad is in translating to a sale, the more money the ad seller can charge.
A key way to improve an understanding of a website’s audience is to use cookies: small text files saved on a user’s computer that collect data on their many journeys around the web. There are first-party cookies placed directly by the website users visit, and third-party cookies placed by ad networks to build profiles of users based on where they travel on the Internet. And that helps guide what ads the internet user sees.
As the use of cookies proliferated during the 1990s and 2000s, people grew concerned that there were no regulations in place to control this largely anonymized tracking. In the EU, lawmakers responded by passing the ePrivacy directive, and later, the General Data Protection Regulation. These are complex legal frameworks that set rules around the collection of personal data, based on the principle of informed user consent.
Now, in 2022, many in the EU are at a point in which they consider the very concept of personalized advertising a threat. For years, activists campaigned against the emotionally loaded term of “surveillance advertising,” which they claim violate an individual’s right to privacy. This very wording is intentionally misleading: surveillance, conventionally understood, involves spying on people for political reasons in order to manipulate or control their behavior, or for legal reasons in the context of law enforcement. Collecting information on people’s consumption and market behavior in order to give them the option to buy goods and services relevant to them, on the other hand, does not amount to “surveillance.” In almost all cases, the advertiser does not know the name or browsing history of the person seeing their ad, so there is no surveillance. Moreover, the EU created the GDPR to hold data collectors and processors accountable for keeping their operations secure, and to demarcate legitimate data collection against illegal use.
Privacy activists now routinely cast aspersions on all data-driven advertising, labelling it as “creepy”. Many point to the Cambridge Analytica scandal as evidence of its dangers in manipulating the public, ignoring that claims around the effectiveness and reach of the Cambridge Analytica operation have been widely debunked.
Let’s be clear: privacy activists are not debating in good faith. Any attempt to point out the fact that personalized ads boost revenue for websites and provide value to advertisers—why use them otherwise, given the wide availability of non-targeted advertising methods?—is simply dismissed. An employee of the European Commission says the utility of such ads is “unproven.” A TechCrunch journalist believes the claim that businesses benefit from personalized ads to be “grade-A BS” because such ads only benefit “U.S.-based big tech.” One of the most prominent privacy activists speaks of “SME-washing” when defenders of targeted ads point out how they help European SMEs.
These are not arguments, they are unsubstantiated claims which simply pretend there are no benefits whatsoever to personalized ads. Instead, critics dismiss targeted ads wholesale as both intrusive surveillance, and completely ineffective: Alexandra Geese, an MEP vocally opposed to targeted ads, argues that “many small and medium-sized businesses suffer heavily from the lack of alternatives in digital advertising,” and that “a ban of targeted advertising is a first important step to give customers a real choice and small and medium-sized business a real chance to compete in the digital realm.” The opposite is true, given that SMEs find personalized ads cheap and effective.
The EU now finds itself in a situation where the Digital Services Act (DSA), which is supposed to create rules around providing and monitoring online content, has had amendments added to it by MEPs with the explicit aim of bringing about a de facto ban on personalized ads—a policy nowhere to be found in the European Commission’s original proposal for the DSA, where the Commission wrote that its proposed “measures concerning advertising on online platforms complement but do not amend existing rules on consent and the right to object to processing of personal data.” Margrethe Vestagher, Executive Vice President of the Commission, has previously stated the basic case for targeted ads: “It is legitimate to advertise to try to find the people with whom you want to communicate.”
The consequences of a ban on personalized ads are of course ignored by privacy activists. Study after study has shown that cookie-based targeted ads are up to 70 percent more valuable to advertisers than non-targeted ads. It is obvious why that is the case: a company with a small advertising budget prefers spending it on ads that are likelier to reach people who might act on them. If someone has previously searched for a new keyboard, seeing a subsequent ad for keyboards whilst browsing a news website is far likelier to elicit a click than an ad for a toilet brush. The practice of “re-targeting” users based on their searches is highly effective and explains why two-thirds of advertisers on Facebook have customer acquisition costs of $10 or less–orders of magnitude cheaper than alternative forms of marketing. A study of online advertising for the European Parliament’s Internal Market Committee confirms, “An advantage for SMEs is that online advertising is usually cheaper than placing traditional/offline ads.” When Google announced a plan to phase out third-party cookies, a group of German publishers urged the European Commission to intervene—not the behavior you’d expect if targeted ads were an uneconomical waste of money that publishers need to be saved from. Two-thirds of UK SMEs use online ads, and 63 percent of those believe they are effective. When Apple introduced a change to the iPhone that made tracking more difficult, some advertisers saw their customer acquisition cost shoot up three-fold.
It is no wonder that activists pretend these facts don’t exist: they completely rebut their claims. There are valid concerns around the lack of competition in the online advertising market—ironic, given research that finds GDPR ended up increasing market concentration in ads. But to willfully deny the reality that personalized ads are valuable, especially to SMEs who struggle to afford the costlier alternatives, is a bad-faith effort to place anti-corporate ideology, disguised as protecting privacy, over empirics.
The Digital Services Act is now in trilogue between the European Commission, Council, and Parliament. If EU legislators give in to privacy activists who pursue their goal of banning personalized ads with a quasi-religious zeal, they will hurt the millions of small businesses in Europe for whom the Internet is a chance to find new customers without requiring the bulky budgets of the corporate behemoths that dominate the economy. And of course, internet users will experience an internet ecosystem with less revenue, meaning either lower quality or the introduction of prices. New EU startups trying to build world-class digital unicorns will compete with American and British startups with one hand tied behind their back. And finally, consumer privacy will not be meaningfully improved beyond the GDPR. Other than all that, what’s not to like?
Photo by Mohamed Essawy on Unsplash