Understanding cohorts

Posted on: Tuesday 24 August 2021

What are cohorts? And what do advertisers looking to pursue cohort-based advertising need to know?

A cohort is a group of users who share common characteristics or interests and have been pooled together accordingly. They offer advertisers a 1:many view of the consumer. This means that advertisers can't identify any individual user, but they can see what broader interest group or cohort they came from.

Cohort targeting is often compared to interest-based targeting, however the main difference is that advertisers lose the ability to understand what consumers are doing at an individual level. Targeting may not be impacted that much, but cohorts will have a big impact on things like frequency capping, conversion measurement and creative personalisation. 

Google’s Federated Learning of Cohorts (FLoC) and First Locally-Executed Decision over Groups Experiment (FLEDGE), as well as Apple’s SKAdNetwork, are some of the most renowned cohort solutions that are being developed in market at the moment. These solutions aim to deliver advertisers a 1:many view of consumers based on data at a browser or operating system level.


What can advertisers expect in terms of targeting & measurement?

Data will be pseudonymised and based on web browsing history, app downloads and offline or panel data indexed by postcodes.

Targeting capabilities and the segments available will vary by browser or operating system. So far, these solutions appear to only offer more upper funnel segments such as demographics, interests and spending habits. 

Measurement will be based on which group of cohort a conversion comes from, rather than which individual. We therefore expect that optimisations will be based on conversion modelling. Attribution will be facilitated and controlled by the browsers and operating systems and, over time, we expect that the metrics being used to measure effectiveness from cohort targeting will evolve.


What should advertisers do to prepare?

Advertisers planning to utilise cohort-based targeting should do the following: 

  • Ensure a thorough understanding of their digital audiences now via analytics and other tools - test how effective media performance and creative personalisation are against these different groups, in order to provide a baseline for when they are replaced with cohorts
  • Investigate and test different ways to measure media performance and consider metrics that are possible without cookies, for example attention
  • Use Safari and Firefox as a testing ground before Chrome makes its changes
Finger print icon in blue

User ID Solutions from IAB UK Members

From alternate personal identifiers to contextual targeting, it can be hard to stay on top of all that is happening in the User ID space and to know what will work best for your company. Here we explore the three types of solutions that are available - user enabled ID solutions, cohorts and contextual solutions.


Related content

cogs icon

Understanding user-enabled ID solutions

Discover the key considerations for advertisers looking to utilise user-enabled IDs

Learn more
quotation mark icon

User ID: The story so far

Find out more about what steps advertisers should be taking to prepare for a future without third-party cookies

Learn more
Question bubble

User ID Q&A: What’s next?

Why are third-party cookies being phased out? Is it a good thing for digital advertising? And how should you be preparing? Get your head around the key...

Learn more
Magnifying glass icon

Understanding contextual targeting

What are the pros and cons of contextual targeting? We weigh up the opportunities for advertisers in a post-cookie world

Learn more

Why digital advertising works

Discover why digital advertising is effective for reaching your customers and building brands.