How to Be Data-Driven Without Third-Party Cookies

Published on 15 Dec 2020

Data-driven digital strategies are the starting point of every personalized advertising campaign since their performance needs to be measured constantly. These strategies, which place user data at the center of media activation, are made possible with navigational data. The regulation is changing, and third-party cookies are slowing disappearing from major browsers and as a result, the ecosystem is facing a great challenge in terms of ad targeting.

Navigating Towards a Cookieless World

So far, the main purpose of third-party cookies has been to target Internet users. These cookies are collected while they are navigating the web and can track them during their browsing and observe their buying journey. This leads to companies having access to a large volume of data and being able to implement personalized digital strategies.

However, the CNIL’s new guidelines on the use of third-party cookies published last October and the implementation of TCFv2 will have an important impact on data accessibility. From now on, without the explicit consent of the Internet user, it will no longer be possible to collect their personal data and this will prevent any use of cookies for ad targeting purposes.

In January 2020, following Firefox and Safari’s leads, Google, publisher of Chrome, announced its decision to gradually ban third-party cookies. This is a new blow for advertising players who will no longer be able to rely on these cookies to provide them with the data necessary to personalize their services and target Internet users. According to a study by Ad-Exchange published in 2020, following the blocking of cookies on Apple and Firefox, 40% of third-party cookies have already disappeared.

Emerging Alternatives

These changes in the digital advertising ecosystem mean that ad targeting strategies need to evolve too. Several possibilities such as logged-in environments, the implementation of a universal ID or semantic analysis and targeting have already been discussed.

Tracking your audience based on registered users requires so much data to be relevant. Logged universes are a good solution when the number of users is important as with Google, Facebook or even Verizon with Yahoo. But this may not work for publishers with smaller audiences. Few users resort to login outside e-commerce or large audience platforms which are inherently logged in. Here, we are confronted to limited initiatives. Asking publishers to combine all their logged audiences could be a possibility, but this would remain complicated because they do not all have the same technologies or means. After TFCv2 and CMPs, it would still be up to the publishers to bear the costs associated with the implementation of logged-in environments without seeing an actual improvement on their revenues.

While the implementation of a universal ID is interesting, it seems to be once again promised to the “Walled Gardens”. The risk is seeing several different universal IDs emerging without any agreement on a single nomenclature between all the players.

These two possibilities remain centered on user data which can only be shared with their consent. Another disadvantage is that, unless it is a high-traffic area, the target population will unfortunately end up decreasing. It will therefore be necessary to use other methods to keep delivering targeted ads without cookies. Semantic targeting is one of them.

The Future of Targeting Will Be Shaped by Semantics

Third-party cookies are disappearing, but it does not mean that data and performance are too. On the contrary, this evolution will allow the emergence of a new kind of data, called “contextual” or “semantic” data. This data will be able to target the content read by a user in real time. No more targeting the intention, it is now all about targeting the attention. The ad delivered will have a direct link with the content read by this user. Machine Learning algorithms will permit the analysis of reading contexts, expressions, meanings and sentiments associated with this content and thus “semantically” classify each URL suggested to them. This method of classification is an extremely efficient ad targeting since the personalized ad will be displayed when the Internet user’s full attention is on a specific topic. As a result it will immediately spark their interest, in real time.

The appeal of semantic targeting is that unlike logged-in environments and universal ID it does not require user data but rather attention data. The user is reachable regardless of their device (mobile or computer) or browser. In this case, the data is the editorial content of the web page that draws their attention at this precise moment. This technology can target the Internet users’ topics of attention. Semantic targeting ultimately restores meaning to advertising, by linking an advertising offer made of values and ideas to be passed, thanks to a website’s editorial content. At the same time, the advertiser’s product regains its meaning, and the editorial content, its value.

Ad targeting is undergoing profound changes and with the upcoming disappearance of third-party cookies, players in the advertising industry are faced with many questions on the evaluation of advertising effectiveness. These are the reasons why exploring new alternatives and reducing the market’s dependence on user data is necessary. Cookieless and consentless targeting made possible with semantic data offers new performance perspectives and illustrates the possibility to maintain a data-driven approach while avoiding the use of third-party cookies.

Want to know more about cookieless targeting?

Geoffrey Berthon
CEO & Founder @Qwarry