Reasons and consequences of the impending cookie calypse

Firstly, website operators must inform about the use of cookies at the latest after the GDPR comes into force in 2018 and the European ePrivacy Directive, and website users must give their active consent (Consent) for the use of marketing cookies in particular. Users who refuse their consent – unlike an activated ad blocker, for example – may not be excluded from visiting the website.

The same applies to the use of the Identifier for Advertisers or Advertising IDs on smartphones, although the respective app operator has previously obtained approval when an app is used for the first time. With the iOS update 14, however, Apple has obliged the app operators to obtain consent from the user using a central Apple consent tool (Tracking Transparancy Prompt). This change was officially carried out to ensure greater transparency for iOS users with regard to their data usage. However, the app and advertising industry fears that only a fraction of iOS users will give their consent to tracking in the future.

In addition to the need for users to actively consent to cookies or IDs, they have always been able to delete cookies in their browsers at any time. Mobile IDs, which are not discussed in detail here, can also be reset by the user in the smartphone, which also means that the information collected so far can no longer be used. And more and more browsers, such as Firefox from Mozilla or Safari from Apple, prevent a so-called Tracking-Prevention the storage of cookies, first and foremost third-party cookies, is now even standard or automatically deleted after a short time, even if a user has consented to the storage of a cookie on a website.

At the latest, however, after Alphabet (“Google”) announced in 2020 that their Chrome browser, which has a global market share of almost two thirds across all end devices in terms of page views (StatCounter 2021), will not have third-party cookies from 2023 will allow more, the entire online marketing industry is in turmoil (“cookie calypse”) and is feverishly looking for alternative tracking and targeting approaches.

Large platforms such as Facebook or Amazon will continue to be able to intensively analyze the usage behavior of their users in order to use this first-party data to continue to generate detailed targeting features for advertising companies in addition to optimizing their own offers. Apart from these large, but closed ecosystems (walled gardens), most website operators should not have enough first-party data to offer advertisers attractive targeting.

As a cross-website solution are currently in particular ID solutions, such as the NetID of Mediengruppe RTL Deutschland, ProSiebenSat.1 and United Internet, discussed. The basic approach of ID solutions is to make website visitors an offer for which they are willing to register with their email address (or some other unique identifier). This identifier is cryptographically encrypted (“hashed”). The ID generated in this way is unique, but cannot be converted back into the original e-mail address. The ID thus fulfills all anonymity requirements, so that it can, for example, be used across websites and devices within an advertising network (Priebe 2021a).

In addition to the various ID solutions that are currently being promoted, Google is also working on its own solution called Federated Learning of Cohorts (FLoC), which is intended to enable data protection-compliant targeting in the Chrome browser after the third-party cookies are no longer available. According to this, users should be assigned cohorts directly in the browser based on their surfing behavior, which are made available to advertisers as anonymous target group segments with several thousand members each as targeting (Priebe 2021b).

Another alternative is always a digitales Fingerprinting discussed, which technically has existed for a long time, but – at least in the online marketing context – has so far not been widely accepted. With this method, an individual signature of the device used is created. This contains specific information on features and settings of the respective device, for example the computer processor, the graphics card and the screen resolution, the operating system, the browser version used, including any installed plug-ins and the language set. With this “digital fingerprint”, an end device can be clearly identified and, in turn, the browsing behavior can be traced across websites. However, the prerequisite is that all websites have implemented the tracking method.

Comprehensive tracking is also a basic requirement for determining the effect of advertising campaigns or, in general, the contribution to success of individual touchpoints along the customer journey. The future elimination of third-party cookies is therefore likely to have a massive impact on attribution modeling. The analysis will probably have to be limited to the touchpoints at which there was an advertising interaction (click). This is likely to lead to considerable inaccuracies in the attribution modeling, which can, however, possibly be cushioned with supplementary media mix modeling, in which the effect of advertising budgets on company sales is determined using statistical time series analyzes.

As an (interim) conclusion, it can be said that the digital industry is currently in turmoil due to the massive restrictions of third-party cookies and the associated changes in tracking and targeting. It is unclear whether one of the currently discussed ID approaches will establish itself as the central solution in the digital advertising market in the future. So far, it has been assumed that it will result in a mix of different ID solutions, cohorts and a renaissance of contextual targeting for unknown users (Priebe 2021c). In any case, the digital industry is – once again – in a state of upheaval.

Leave a Reply

Your email address will not be published. Required fields are marked *