Cookieless Environments Robustly Support Advanced Behavioral Targeting

The deprecation of third-party cookies has been one of the most talked about topics within digital advertising for the past few years. However, the reality is that cookieless environments are already here, providing challenges for advertisers.

Google Chrome’s loss of third-party cookies may mean a definite ending to decades of cookie-based targeting. However, other browsers waved goodbye to this tracking method long ago, and many Chrome users already opted out of being followed around the web for a while now.

As a result, advertisers urgently need to put solutions in place to continue reaching consumers with ads that are relevant to them. But what tools are out there to help advertisers deliver?

Established Tools

As with developments in any field, some methods are more mature than others. Within digital advertising, contextual tools and external identifiers (EIDs) are well-established answers to the cookieless problem.

Contextual targeting has existed for decades and was once the leading way that online advertising was delivered. Advances in the area have meant the mechanisms are again garnering a lot of attention from advertisers.

Contextual is at its core privacy-friendly, basing the ads served on the content featured on a web page, rather than the user’s data. This approach has been leveled up by the emergence of artificial intelligence, in particular natural language processing.

More similar to third-party cookies, EIDs enable cross-site tracking by transforming the personally identifiable information of users into a deterministic identifier, or a set of device-oriented signals – such as IP address – to a probabilistic identifier.

However, due to their similarity to third-party cookies, EIDs have come under plenty of scrutiny. Regulators aren’t keen on the idea of replacing one privacy-intrusive tool with another, while industry experts have questions over the scalability of that method. Additionally, browsers have introduced technical limitations to allowing these IDs to work by cutting down the information required to build probabilistic identifiers through mechanisms like User-Agent Reduction or Private Relay.

Breakthrough Tools

Alongside these more established mechanisms are a group of emerging tools looking to prove they are the best options for advertisers in cookieless environments, including the Protected Audience API, the Topics API, and publisher-partitioned identifiers (PPIs) and Seller-Defined Audiences (SDAs).

The Protected Audience API (PAAPI), part of Google’s Privacy Sandbox, replaces cross-site tracking by placing users into interest groups that are passed in the bidstream. Any data needed for improved personalization and bid evaluation is stored in the user’s browser or on an isolated, trusted bidding server.

What makes PAAPI stand out is how it can work with, and be strengthened by, other cookieless tools, making it a promising tool for both retargeting and branding campaigns. It can be combined with other methods – such as PPIs, SDAs, Topics API, and contextual targeting – to enable more precise creation of interest groups, and the more accurate activation of these interest groups.

The Topics API is a less sophisticated targeting API from the Privacy Sandbox. The tool enables the Chrome browser to form a short list of recognizable categories based on the user’s recent browsing history, which is then passed to DSPs to inform the ads that are served. The API is predominantly for upper-funnel campaigns, as its taxonomy of around 470 topics leaves it unable to hyper-personalize.

PPIs and SDAs are favored by publishers because they’re provided with control over how their inventory and audiences are labeled. However, these methods have been slow to gain traction, as publishers are wary of sending them in parallel to cross-site identifiers for fear that the valuable information will be monetized externally.

Data Modeling Imperative for Success

Whichever solutions advertisers begin to explore, they have to understand that not all cookieless environments are the same, so different tools will be needed for varied circumstances – whether that’s between different browsers or when considering same-site or cross-site scenarios.

Different browsers will allow certain practices, while others won’t and not all users will be identifiable. In such cases, it is critical to properly understand those who are and properly extrapolate their characteristics on those who aren’t.

In cross-site circumstances, it’s a little bit more complicated. Marketers may see partitioned environments where each publisher or advertiser’s website sits in a separate silo.

This can be tackled by building lookalike audiences through advanced data modeling, enabling the proper execution of targeting and ad frequency management. Advanced data modeling combined with a selection of cookieless tools will ensure that advertisers can reach audiences across all types of inventory.

A Suite of Solutions

Ultimately, succeeding in the cookie-free world will require solutions consisting of more than one method. There is no silver bullet, but there are such combined solutions in the market that are privacy-friendly. Advertisers need to identify the needs of their business and establish which mix of tools works best for them.