Comment: Ad tech reimagined post third party cookie

400x260-Scott.jpg

Writing for WARC, Ozone CTO Scott Switzer reimagines ad tech post-third-party cookie, with 3PC’s demise being the single-most impactful change to online advertising in over 20 years

The death of the third-party cookie is the single-most impactful change to online advertising since cross-site advertising began over 20 years ago.

Thursday 25th March 2021 - first published on WARC

The third-party cookie is the raw data signal on which almost all audience databases are written, and is also the method advertisers use to measure whether a campaign performs well. These and other features, like frequency capping and campaign optimisation, need to be considered as we rethink the future of cookieless advertising.

The absence of the cookie means the entire architecture of the ad tech ecosystem must be rewritten. This presents risks to companies big and small, as well as a once-in-a-generation opportunity to shape the rules in a way that resets where value is created.

Some companies have tried to rebuild the functionality of third-party cookies. A new deterministic ID would be formed with similar features as the cookie, and all of the existing systems of the last two decades would continue to operate with minimal disruption.

These efforts have not gained much traction with publishers, likely because the existing third-party cookie extracts publishers’ valuable data without adequate compensation. Google clarified its position on the matter, and stated it would not participate in any effort to replace the third-party cookie as a data transport, thus marking third-party cookie replacements as dead on arrival.

Why is the cookie so problematic?

A deterministic ID like a third-party cookie is used by ad tech platforms to build audience databases. Three data points from a publisher – an ID, URL, and IP address – can be translated into a database that contains years of information about a person (identified by an ID), their interests and behaviours (by parsing content from the URL), and their demographic and geographic make-up (by translating geography and demography using IP address).

The power of this audience database is that it can be used on any website, not just where the data is collected. This means the publisher that built an audience database by contributing its data may not receive the benefits of the database, because advertisers may want to buy cheaper inventory on lower-priced publishers. Publishers refer to this as “data leakage”. Data leakage is the reason why publisher media inventory is commoditised, while ad platforms are significantly more valuable.

The far-reaching impact of cookie deprecation

The death of the third-party cookie affects everybody in the entire ad tech ecosystem, in different ways:

  • Ad tech platforms without a significant, direct relationship with advertisers or publishers will be challenged in collecting data needed to maintain their audience databases. Audience DMPs have already had to take a significantly liberal view of their place in the ad tech ecosystem to get around GDPR regulation. They now face another task of convincing consumers, publishers and advertisers that the moral high ground that Apple (and now Google) is taking is incorrect.

  • DSPs that have also built audience databases from their connections with publishers will have a much harder time as cookies become scarcer and publishers stop leaking data in preference to their own audience databases. The efforts that DSPs have built to replace the third-party cookie will likely be scrapped. While DSPs have been a heavily relied-upon partner for advertisers, they are at risk of becoming more commoditised connectors between advertisers and publishers.

  • Large publishers with the audience scale as well as the technical expertise and budget will benefit from taking steps now to build their own databases. Smaller publishers are at risk of losing their advertisers because it will be harder for an advertiser to reach the long tail of publishers. All publishers would benefit by partnering together to bring scale to an advertiser campaign.

  • Advertisers and their agencies will have to re-evaluate their technology partnerships, to make sure they have the tools they need to run their media campaigns. Their existing DSP and DMP relationships may need to be supplemented by audience platforms that facilitate first party relationships with publishers, like InfoSum.

  • Last but not least, consumers will benefit from a newly architected ad tech ecosystem. Instead of providing consent for their publishers to send their data to dozens or hundreds of ad platforms, there will be a much stronger and understandable relationship with the publisher, without the feeling that the publisher is treating the customer data in a way that is against their wishes.

The publisher’s role in reimagining ad tech

When thinking about the publisher’s role in this reengineering, it’s important to remember that the creation of a new ad tech ecosystem needs to work for both brands and publishers as they are the primary stakeholders and the controllers of customer data in the advertising exchange. For publishers, this means taking a number of actions to maximise value for advertisers, while maintaining greater control over their most valuable assets:

  • Publishers should double-down on their efforts to build audience databases themselves. While third-party cookies are dying, publishers’ highly valued first-party data can be used to build audience segments with the same or more richness as third-party platforms. 


  • Publisher-built databases should not allow any IDs to be transferred outside of the publisher’s ecosystem. This gives advertisers the opportunity to purchase differentiated audiences without publishers losing control of their data, all while giving consumers a better relationship with the publishers that they give their data to.


  • Individual publishers should consider whether they have sufficient scale to build their own audience databases. It remains highly inefficient for advertisers to individually manage campaigns with even the largest publishers’ audience databases.

  • Publishers should try to engage with brands directly to bring first-party data to market. The more direct a connection that can be created between both parties, the more tuned the data and media can be to the advertiser’s needs. In addition, the less primary data that is sent to platforms that can build databases on behalf of others, the more valuable the publisher’s data is as a unique, differentiated asset.

  • For publishers not big enough to work with brands directly, they should be very selective about the tech partners they engage with and aim to truly understand the value these partners provide. It’s critical for these publishers to remember that they have the audience, data and inventory that fuels these ad tech businesses.


  • At Ozone, we believe collaboration between publishers is essential in forging this new future. As advertisers and their agencies seek easier routes to market without compromising on effectiveness, we’re seeing more publishers around the world build scale by combining assets via M&A, or by building publisher consortiums like our own. 

It’s time to rebuild


Third-party cookies are currently forcing everyone in the ad tech ecosystem to re-evaluate their business models. While there are those who would like things to remain the same, there is a real opportunity for advertisers and publishers to come together to develop a solution that meets both needs in the coming years.

We should not let this opportunity go to waste.