Jon Fletcher 2021-01-29

Universal ID solutions: The cookie killer

Screenshot_2021_02_02_at_09.33.59.png
Google announcing the blocking of third-party cookies was a bit like hearing an asteroid was set to skim by Earth in the near future. Worrying? Sure. But, the likelihood would be that someone in NASA would silently develop an elegant solution that would save us from really having to deal with it. 

For some, Universal IDs are the unsung NASA hero for the cookie asteroid. 

For publishers that are unable to build their own bespoke first-party data marketplace, or don't want to rely on Google's privacy Sandbox proposals, Universal ID solutions seem to be the closest replica of the current programmatic system. 

Generally, systematic change means upheaval and loss for all but the biggest players. But, third-party cookies have always been an imperfect model. There are many reasons why publishers may be looking forward to the development of a Universal ID solution to replace third-party cookie-based tracking. 

In this post we're going to cover the basic principles of Universal IDs and why the forced change could be a blessing in disguise for some publishers. 

How the Cookie system leaks revenue 


In the current system every website has its own way of identifying users. Every click on a publisher site is dragging with it hundreds of different third-party cookie-based IDs. These all have to be matched by SSPs and DSPs to perform the ad targeting they were meant for.

When a user visits an ad-supported site, a wrapper like Prebid is activated. It calls its monetization partners who organize auctions with their buys, represented by the DSPs. These DSPs in turn, rely on some form of data from data exchanges. This is the final link between the publisher and the brand. This all relies on individual user data. 

Each SSP will have access to its user id stored on its cookie. They will then issue the bid request on the server-side. But, incomplete matching tables mean the user info is not always transferred. 

For the DSP to understand the user they need to match the user. In some cases, they will have a match. The same then happens from the data point. You may have data available for some of your users, but not all of them. 

Each SSP will have access to its user id stored on its cookie. They will then issue the bid request on the server-side. But, incomplete matching tables mean the user info is not always transferred. For the DSP to understand the user they need to match the user. In some cases, they will have a match. The same then happens from the data point. You may have data available for some of your users, but not all of them. In the end, you have a poor auction that doesn't maximize the potential of the bidders or the user data because the people involved are all using a different system. That costs the publisher revenue and affects the brand’s campaign.


For a brand targeting a specific audience on a publisher site, the match rates and losses along the way mean the current cookie system does well to get close to a 40% match rate. 

For a brand targeting a specific audience on a publisher site, the match rates and losses along the way mean the current cookie system does well to get close to a 40% match rate. A system with this many leaks in it quickly becomes expensive. It can also slow page-load times and opens numerous vulnerabilities for data leakage. A universal ID solution means that publishers can transact on one shared ID per user. This makes the process much simpler and makes it more transparent for the end-user.


How do Universal ID solutions work? 



Where the cookie system is based on probabilistic matching, Universal IDs use a deterministic matching system. This means if you run the model with the same initial conditions you will get the same results. 

With a Universal ID, when a user visits a publisher's site, they can give their consent via the CMP for the publisher and the publisher's ad partners to use their data for ad targeting. Once the information is read by the CMP it gets stored in a first-party cookie and passed to the ID provider's API. 

This is not actually the ID at this point, it is just the permission to create an ID for that user. Publishers can then retrieve the ID, store it on a first-party cookie and pass it to their demand partners via a simple on-page configuration. The Bid adapters in Prebid can listen to the ID and send it to their RTB demand partners on 100% of impressions


What are the advantages of a universal ID solution? 


As well as enabling publishers to serve programmatic ads using user data, Universal IDs bring with a lot of advantages across the entire ecosystem. 

When implemented properly, universal IDs will enable: 
  • Consumers can see more relevant advertising without slower page loads
  • Advertisers will reach more valuable users and have a more comprehensive view of their customers 
  • Publishers will match more of their valuable audience data and earn more for their inventory
  • SSPs will reduce operational costs from network syncs and data storage
  • DSPs will get more auction participation and make it easier for clients to hit campaign goals
  • Data providers will make more matches between supply and demand partners – and earn more from their data. 

Publishers that have aimed their strategy at gaining subscribers can also use a universal ID to match users across all of their sites.



The main issue with a universal ID solution is that it requires everyone to agree on the identifier to use. So far, a single choice hasn’t been singled out as the favoured or dominant approach across the industry. Achieving an agreeable consensus is needed in order to realize the objective and stop advertisers moving to walled-garden networks.

Latest Articles

‹ Back to Blog Home

Get the headlines

Sign up to get the best headlines direct to your inbox

Your name
Your email