Identity and Addressability in an Anonymous-by-Default Internet
February 1, 2021 | 4 min read
Third-party cookies have long been used as standard practice in the adtech world, implemented to identify audiences, attribute the success of certain campaigns, cap ad frequency, and enable users to “opt-out” of tracking.
Over time though, the number of cookies exploded into the thousands. Some users felt they had no control over their data and some types of ads started to raise questions about the use of data behind them. This led to a crackdown by regulators and privacy advocates and the creation of standards such as the GDPR and CCPA. While third-party cookies were already living on borrowed time, Google dealt them a death blow this year by announcing it would phase them out completely in Chrome by 2022 — or at some point.
With privacy concerns center-focus and the death of the third-party cookie, the future of the internet will be anonymous-by-default. This new normal has left marketers and publishers scrambling for new solutions to addressability. How will buyers and sellers identify audiences in a privacy-first world?
The Way Forward
The way forward will in some ways be a return to digital advertising’s beginnings: a renewed focus on grouping audiences by interest and putting users first. To achieve this, we’ll need across-the-board collaboration because the future of privacy-first, web-based digital advertising will center on two things: first-party signaling and the creation of privacy-conscious audience segments.
1. Publisher and/or End-User Elected First-Party Signaling
Looking forward, a proportion of audience tracking will be transparently set and controlled by the users themselves. Tracking and addressability could be based on “first party signals” (by logged-in users), or on a first-party identifier as provided by the publisher through an open-source platform such as Prebid.
For this model to be viable, we’ll need to make it clear to users that first-party data isn’t being used for nefarious purposes, but instead, in order to track user interests — such as music taste or hobbies — to improve ad quality. With interest-based data, ads become more relevant. We’ll also need to clarify that users can opt-out of tracking, as well as convey the inherent value that digital advertising brings to the open internet— ads keep the internet free and open for consumers.
Audiences may also need to be encouraged to volunteer information for a better experience or tangible reward. For publishers, this means addressing the question — How can you make identification seem palatable or desirable, instead of just something you’re willing to give up?
While this model feels like it upends identity as we know it, it could actually make tracking more effective because it’s consensual. A consumer is more likely to want to engage with a brand they trust, on sites they trust with their personal information.
2. Audience Segments Grouped by Privacy-Differential Cohorts
Regarding addressability, we’ll also depend on audiences grouped by interest signals. These cohorts must be large enough to be identifiable by buyers without violating user privacy and/or preferences.
Without third-party cookies, the buy-side will no longer be able to create and transmit such audience segments. Consequently, we’ll likely instead rely on publisher-controlled first-party data and IDs, as well as some percentage of user logins from each site, to probabilistically link activity across the internet. Through a trusted, transparent, and open platform we could conceivably create segments that are large enough to be valuable to advertisers. In such an anonymous-by-default system, the focus will be put on modeling audiences, and letting buyers buy either a) users b) categories or c) private marketplace deals (PMPs).
Eventually, the third-party cookie will be dead, leaving a white space for a new approach to identity. The new identity model that emerges will be all about control and consent — and will need to work with an anonymous-by-default internet. To navigate such a world, buyers, publishers, and exchanges will need to collaborate to make identity a shared community asset instead of something that individual companies use for competitive advantage. The first step towards this is rebuilding trust with consumers and with each other, so we can connect with users in an open, transparent, and privacy-conscience way.