Why householding is the best solution to cookieless… for now


Learn more about Lineate’s custom solutions for AdTech

We are all on a path to moving away from third-party cookies in all browsers. Most providers in the ad tech industry have spent the past few years trying to figure out how to evolve their businesses to the new reality. We have worked with many ad tech companies who are navigating this switch and trying different approaches. The problem is that none of the post-cookie approaches that exist today are as effective as cookies have been. While we see promising technologies and improved results on the horizon, there is an opportunity to minimize existing challenges over the next three months, after which Google is planning to turn third-party cookies off in Chrome and use householding approaches to bridge the gap to the cookieless future. Moreover, householding may stay useful even past this point, because even if Google will really turn third-party cookies off in Chrome when it says, there are some other browsers where the householding identification approach remains useful. Indeed, it looks as if householding, which makes use of cookie information gathered from some (but not all) members of a household, is currently garnering the most traction among our clients, as opposed to identity fingerprinting or other cookieless user identification approaches.

Once third-party cookies are fully disabled, cookies will no longer be used by ad tech providers to identify users across multiple systems. The single biggest challenge coming up is Google’s plan to eliminate all cookies in the Chrome browser by Q3 2024. To facilitate testing, Google has already restricted third-party cookies for 1 percent of users effective January 4, 2024, and plans to ramp up through Q3, subject to addressing any remaining competition concerns of the UK’s Competition and Markets Authority (CMA). Google, during research, turned off third-party cookies for a small group of its Ad Manager customers and found that ad revenue dropped by 52 percent for these customers. Studies of the App Tracking Transparency introduced by Apple in iOS 14.5 show similar results. We have seen similar disparities between cookieless and cookied traffic.

Our customers have been using various technologies, such as fingerprinting or increasing focus on contextual targeting, to mitigate the impact of the switch so far, and the results for those methods have been much better than the results for generic cookieless traffic. That being said, those numbers have still been significantly below those for traffic with cookies. This means there is still an opportunity to maximize value during the period when cookies are still partially implemented and for the near future beyond that time. In short, an investment in interim “partial cookie” solutions such as householding should play an important part in any ad tech company’s transitional strategy.

Longer-term solutions 

Of course, any ultimate solution to this problem needs to work in a world where third-party cookies don’t exist at all. Ad tech companies have been trying different approaches for years, with varying degrees of success. And promising new technologies continue to come to the market.

When news first appeared about turning off third-party cookies, the initial solutions were based around user fingerprinting. The fingerprint is an attempt to uniquely identify a user, as best we can, by using an aggregated set of information that we can collect from different technical fields of the cookieless request. We have done various custom implementations of this approach for our clients, who identified this solution as important and valuable.

We have also implemented various commercial identity solutions, such as 33Across’s Lexicon, DigiTrust’s Universal ID, and The Trade Desk’s Unified ID. These, combined with the extended identifiers functionality added to OpenRTB, have shown clear and measurable differentiation versus unfiltered cookieless traffic.

Our experience with these fully cookieless approaches to targeting is that they are better than no attempt at user identification whatsoever, but they fall far short of the potential revenue gains that can be realized when even partial cookie information is used for targeting. 

The upcoming Google Privacy Sandbox offers a vision for building on that performance by providing a safe way to target without revealing shared identification. It is still in prerelease, and there is much that we still don’t know, but the service promises more aggressive measures to prevent some of the identity workarounds that have proven partially effective today. All this leads to ongoing concerns as to how well Google Privacy Sandbox will supplant lost revenue, along with the obvious concerns of concentrating too much of this information with Google. The clients we have worked with have plans for how to use it and are hopeful that it will work, but they don’t yet have enough information to make good predictions. What is clear is that companies should take whatever measures they can today to preserve revenue while this transition is still in progress.

Managing the transition

Given the limitations and uncertainty around an ultimate solution, ad tech companies are faced with managing a rapidly declining share of cookie-matched traffic. Householding has emerged as a more viable approach than device fingerprinting, and it is being successfully adapted by many of our customers.

The concept of household targeting in ad tech has been around for years, and it has often been used for things like frequency capping of ads across all users or devices in the same household. The current focus of householding, in contrast, refers to the practice of targeting ads not to individual users, but to entire households where one or more users (but not all) continue to use browsers with third-party cookies enabled.  This approach can provide the ability to transform cookieless traffic to identified traffic by mapping household IP addresses to cookies that come in from the devices that still have cookies enabled. These enriched IP addresses can provide a reasonable facsimile of cookie traffic and maintain a tail of (declining) efficacy over time, even after cookies are switched off.

Implicit in this approach is the assumption that households, as a unit, share certain characteristics, behaviors, or interests. Real-world results thus far show that this assumption is good in practice; for most cases, we have clients showing double the revenue with householding compared to pure cookieless traffic.


Years into the cookieless transition, the future is still uncertain. Many methods are in place that provide significant lift over regular cookieless traffic. However, increasingly stringent privacy protections are decreasing the efficacy of such methods, leading to the next generation of solutions, such as those from 33Across and Google, which show promise. However, there is money to be made even in this interim period and beyond by using methods like householding to increase revenue. In addition to providing immediate revenue benefits, the same techniques can be used on an ongoing basis to enhance whatever solutions come next.  

Learn more about Lineate’s custom solutions for AdTech


Got a project?

Harness the power of your data with the help of our tailored data-centric expertise.

Contact us
Recent Posts