The online advertising market is dominated by the trio Meta, Amazon and Google. They represent 80% of the growth of the digital advertising market. Beyond the relationship between platforms and users, a whole ecosystem of stakeholders (advertisers, brands, SMEs...) is concerned by this market to "exist" online.
More than a month after the adoption of the "Digital Services Act" (DSA), targeted advertising is at the center of the European regulatory debate. The text partly regulates targeted advertising through the prohibition of targeting minors and the strengthening of transparency obligations. What does this text change? What are the alternative models for the future of the Internet? Can targeted advertising be both transparent and effective? When should it be regulated?
To shed light on these questions, the Human Technology Foundation invited Christine Balagué (Professor, Institut Mines-Télécom Business School and holder of the Good in Tech Chair), Paul Maher (Head of Performance, Coudac) and Grimaud Valat (Associate Lawyer, DTMV) to give us their insights during a webcast held on June 13, 2022
The analysis of targeted advertising raises many paradoxes:
The CNIL defines "targeted advertising" as "an advertising technique that aims to identify people individually in order to deliver specific advertising messages to them according to individual characteristics". This technique is based on massive data processing. Because it is impossible to process all this information manually, targeted advertising is almost exclusively programmatic.
With the possibility of tracing the desires and tastes of potential customers, targeted advertising has become very popular with companies. In 2021, the digital advertising market grew by 24% from 2020 in France, according to data from Oliver Wyman for the “Syndicat des régies Internet” (SRI).
However, the promise of targeted advertising seems challenged. In an interview with ADN, Tim Hwang (author of the book Le grand Krach de l'attention published in 2022 by C&F) talks about a "largely overvalued" phenomenon. In particular, he points out that "targeted advertising essentially works on people who would have bought the product no matter what".
Moreover, with the increase of content, our attention is more and more scattered. In this context, it seems that we are more sensitive to advertisements in traditional media such as television or newspapers.
Finally, is targeted advertising necessarily associated with sensitive data? Wouldn't tracking based on the processing of non-sensitive and anonymized data be less intrusive?
Advertising is often linked to its commercial purpose. Paul Maher even goes so far as to explain that "in the jargon of advertisers, we talk more about acquisition than advertising because the goal is to acquire customers".
But can advertising be used for other purposes? Couldn't it influence "virtuous behavior"? For example, what about advertising that promotes the adoption of safety precautions during covid, the encouragement of recycling or the sharing of data for the common good? Finally, what are we trying to regulate?
Let's keep in mind that "digital advertising is an architecture of massive persuasion" according to Christine Balagué.
Beyond the commercial objective, it can also be used for political purposes. Who can judge the virtuosity of a practice? The Cambridge Analytica affair is a perfect illustration of the use of advertising for political purposes and the risks it raises. In the end, isn't political persuasion even more harmful than commercial persuasion?
According to the DSA, advertising is defined as "information intended to promote the message of a legal or natural person, whether commercial or non-commercial, and displayed by an online platform on its online interface, in return for payment, for the specific purpose of promoting that information". Advertising thus represents an online representation in return for payment.
Christine Balagué explains that "through the use of programmatic advertising, some disinformation sites are financed by digital advertising. However, sometimes, some brands do not even know that they are displayed on disinformation sites". This type of practice can be detrimental to a company's image. Thus, advertising can both enhance or damage a company's image, especially if it is associated with misinformation.
The DSA is part of the European Union's desire for several years to protect personal data rights. In line with this, the United States also introduced the Banning Surveillance Advertising Act on January 18, 2022, which aims to ban all targeting based on sensitive data. In parallel to these regulatory advances, platforms such as Meta have also announced their willingness to regulate targeted advertising, particularly for minors.
Is this desire to regulate targeted advertising a good thing?
One of the key measures of the DSA, which has been praised on many occasions, is the ban on targeting minors. However, it should be noted that this provision is far from being in accordance with the European Parliament's original intention to abolish targeted advertising in general.
Paul Maher also points out that "minors are not the preferred targets of the online advertising ecosystem, which is primarily looking for people with purchasing power".
Christine Balagué nevertheless points out that "the impact of this provision varies depending on the platform. For example, Tiktok will be very concerned". In February 2021, Tiktok had paid 92 million dollars to stop being sued by the U.S. justice. The courts accused it of illegally collecting personal information on teenagers.
The DSA reinforces the transparency obligations of the General Data Protection Regulation, a regulation in EU law on data protection and privacy in the European Union and the European Economic Area. (hereafter GPDR).
However, despite the laudable objective of the GPDR, the phenomenon has led to "an exhaustion of consent by banners" according to Grimaud Valat.
Based on this observation, how to ensure the effectiveness of the transparency obligations defined in the DSA? How to deal with the existing asymmetry in terms of technical knowledge and financial means between digital companies and regulators? It is through the implementation of the text that effectiveness will be achieved.
The analysis of immersive virtual universes is part of a prospective approach, because the use and economic model of these technologies are still largely unknown. However, what would the development of metavers, which relies entirely on targeted advertising, look like?
How can we market our attention in the metaverse? Isn't it already saturated?
Christine Balagué states that "in the metaverse, all our turpitude and desires will be analyzed. As a result, data collection will be even deeper and more dangerous".
If each avatar perceives targeted advertisements, isn't there a risk that important perception biases will develop? Will everyone be confronted with shared universes?
According to Grimaud Valat, "The DSA is a world away from the developments of the targeted advertising model that could be developed in the metaverse. It is necessary to establish basic rules to govern people's behavior in these universes".
For Paul Maher "It is necessary to diversify the points of contact with customers. This cannot be done exclusively through targeted advertising".