The Identity landscape: introduction and different approaches to identification

  • Posted by Valbona Gjini
  • On Mar 09, 2022

The identity space started off as a small and technical niche within the digital advertising ecosystem. Just three years ago, the main goal of identity solution providers was to make the cookie matching process more efficient for publishers and their technology partners.

The increasing restrictions on traditional identification methods, such as the third-party cookie and the MAID, and the implementation of new privacy regulations, have ignited the growth of the identity landscape and made it one of the hottest areas in the industry when it comes to innovation, developments and media attention. As a consequence, lots of companies have jumped on the opportunity to provide identification solutions of all sorts, solving different use cases and leveraging various methods and technologies. Today, we find ourselves in a landscape which is incredibly crowded and overly complicated.

Time to bring some clarity

At ID5 we have witnessed first-hand all the developments that have taken place within thein space and have realized that it’s time to shed some light on the identity landscape. This blog post is the first of a series that aims to explain, in the clearest way possible, the difference between all the identity solutions available in the market, the use cases they solve, as well as the role that different companies play in the identity landscape.

Different approaches to identification

There are various ways to approach identification that serve different purposes and that we believe will coexist in the future. Most of the cookieless identity strategies in the space are built on top of these approaches.

The cohort approach

Google’s Privacy Sandbox is the best known cohort-based approach currently being developed in the industry. Privacy Sandbox aims to provide the industry with an API to collect aggregated data about user profiles and aggregated campaign performance data covering use cases such as targeting (with Topics), retargeting and measurement (with Fledge). The initiative’s goal is to prevent adtech platforms from tracking users’ journeys between websites and devices and avoid privacy regulators’ scrutiny. 

The Privacy Sandbox approach presents several challenges including performance (it doesn’t provide the same level of granularity that advertisers expect and have in Google’s advertising platforms) and differentiation (by using a common aggregated dataset, platforms and their clients won’t be able to build their competitive advantage). It also makes advertising in the Walled Gardens’ platforms more efficient and effective, potentially driving more market share to big tech.

Fragmentation could become an additional challenge. We don’t know if other browsers will use Privacy Sandbox (Mozilla has explicitly said they have no plans of implementing it into Firefox). If Privacy Sandbox won’t be adopted by other browsers, Chrome and Chromium audiences will need to be addressed differently than those using Safari or Firefox, causing even more work for publishers, ad tech platforms and marketers.

And finally, is it fair that Google gets to define the rules of user qualification, targeting, and performance measurement?

The walled garden approach

First-party data is often praised as the best option available for the cookieless world. Publishers, in particular, are encouraged to collect as much data as possible and build product strategies to increase the amount of data they can collect to make their audience more valuable to advertisers in the post-cookie era. By creating their mini walled gardens, publishers could try to mimic the strategy adopted by the most famous walled gardens par excellence: Google and Facebook.

Increasing first-party data collection efforts is always a good strategy as it enables media owners to better understand customers. But considering them the optimal solution to the disappearance of cookies and MAIDs is not realistic for many reasons. The most evident reason is that first-party data belongs to the domain that has collected it. It cannot be shared without relying on a third-party mechanism. This is exactly why we have been using third-party cookies for such a long time: to enable advertising players to share data with each other.

Another reason that makes the creation of mini walled gardens a suboptimal solution (on its own) is that it creates fragmentation. Sell-side first-party data is a valuable way for publishers to help brands find their audience and reduce media wastage but few can do this at the necessary scale. It can be resource-intensive for brands to aggregate enough sell-side data segments to meet their scale and performance requirements.

There are only a few players that can afford to survive on first-party data alone: Google and Facebook and maybe some of the largest publishers. Everyone else needs to find ways to share their data and leverage third-party data providers to supplement their audience strategy. 

The shared ID approach

Shared or universal identifiers have been created to identify users and share information within the advertising value chain for targeting, frequency capping and measurement purposes.

Identity solution providers collect signals from publishers to generate an ID that media owners share in the bid stream to monetize their traffic and enable advertisers to reach their target audience and measure the performance of their campaigns. The main challenge that shared IDs face is adoption. The value of an ID is strictly correlated to their adoption rate. Today there is a large variety of IDs available in the market but just a few of them have achieved the scale needed to prove their value.

Furthermore, the sharing of users’ data is strictly regulated in regions such as Europe, meaning that universal identity providers need to ensure that the signals they collect to generate an ID are gathered with the user’s consent and that they are only shared with authorized platforms. It is, therefore, crucial for ID vendors to implement technologies and mechanisms that ensure that users’ information is collected and shared legitimately.

What about contextual? 

No matter what you read about contextual targeting being the best alternative for the post-cookie world, it can’t be considered an identification method. Contextual is a reasonable approach to help infer some information about a user and can help brands reach the right people in the right context to achieve their performance goals. It’s a tactic that’s been around since the early days of advertising. However, it’s not a scalable replacement for more conventional identification solutions. It’s a targeting tactic that doesn’t replace any of the capabilities of the third-party cookie and the MAID.

In the post-cookie era, contextual will be a valid option to serve ads to those users that want to remain anonymous and to complement more elaborated targeting strategies.

Coming up next

In the next post of the identity landscape series, we will explore different types of identifiers and how they work in more detail. If you are keen to learn more about how universal identifiers work and how they differentiate from each other, watch this space.

To read Part Two: first-party IDs and identity resolution methods explained, click here.