The first identity-focused solution to come out of the IAB Tech Lab’s Project Rearc, Seller Defined Audiences (SDA) seeks to empower publishers with more control over their first-party data strategy to reduce reliance on third-party IDs.
Quick Takeaways
- In late February 2022, the IAB Tech Lab released a new addressability specification known as “Seller Defined Audiences”, or SDA.
- Seller Defined Audiences enables publishers to segment their first-party audience data into relevant cohorts, then pass the IDs for these cohorts in an OpenRTB bid request to DSPs for decisioning.
- SDA reduces the reliance on third-party user identifiers or broad-brush contextual solutions, instead empowering publishers to segment their own first-party audiences into relevant cohorts based on user activity across their multi-platform properties.
What is Seller Defined Audiences?
On February 24th, 2022, the IAB Tech Lab released a new addressability specification known as “Seller Defined Audiences”, or SDA. This is the first initiative to sprout from Project Rearc, the IAB’s response to Google’s upcoming deprecation of third-party cookies, currently set for 2023.
Seller Defined Audiences enable publishers to leverage their valuable first-party data—and, more importantly, scale it—without the risk of data leaks or any reliance on third-party identifiers. It also doesn’t rely on PMP or deals-based mechanics, instead operating within the bidstream itself as OpenRTB bid request metadata. This means that SDA can operate seamlessly across oRTB supported browser, app, and CTV environments.
SDA requires publishers to segment their users into one or more of the IAB’s Audience Taxonomy categories. There are currently almost 1700 available categories based on three classifications: Purchase Intent, Interest, and Demographic.
When broken down to its bare essentials, Seller Defined Audiences relies on three basic steps:
- The publisher assesses their first-party audiences and, alongside their DMP, maps the users to one of the IAB’s ~1700 Audience Taxonomy categories. They do this using a standardized transparency schema known as the Data Transparency Standard, or DTS.
- The publisher includes the segment ID for the relevant taxonomy in the bid request using the ORTB 2.6 specification. The IAB currently recommends the use of Prebid.org header bidding integrations to pass SDA values, but there are other (and presumably less friction-free) ways to achieve this.
- The DSP receives the bid request along with the SDA metadata (i.e. the taxonomy’s Unique ID) and decides whether or not to bid.
The anatomy of an SDA cohort
Audience cohorts are not a novel concept, but the actual mechanics behind Seller Defined Audiences are new.
Unlike algorithmically generated solutions like Google Topics, SDA doesn’t rely on machine learning or any probabilistic logic to make a best guess about user attributes. Instead, SDA empowers publishers to bring to bear their trove of first-party data and audience insights across all of their properties. In addition, Google Topics relies on a far broader scope of cohort groupings—just 350 at present—as opposed to the granularity provided by the database of 1700 SDA cohorts.
This means that, rather than having an automated platform decide that a user simply likes “Sports”, publishers can assign a highly specific SDA cohort such as Interest > Sports > Equine Sports > Horse Racing and help advertisers be far more granular in their targeting. The result is more ‘yes bids’ and improved incremental revenue.
What does Seller Defined Audiences mean for media owners?
Perhaps the most interesting aspect of SDA from the perspective of a publisher is that it offers a way to monetize granular first-party data at scale without the need to rely on a third-party cookie, ID, or tech partner. It can also do this outside of any single technical environment, meaning users across browsers, apps, and CTV can be included in SDA cohorts. What’s more, publishers retain complete control over their data, opting instead to assign relevant segment IDs on their side, then passing them via bid requests for demand-side decisioning.
Importantly, SDA utilizes multiple pre-existing programmatic technologies—OpenRTB, Prebid, DTS (Data Transparency Standard), and so on—to achieve a privacy-first cohort-based targeting solution. This mirrors a trend in the industry right now that there’ll be no single solution to preserve addressability in the cookieless world, but rather a suite of solutions unique to each programmatic player.
The IAB put it quite succinctly in their SDA announcement: “SDA makes no claims of being a silver-bullet for the industry, but should be a valuable tool in our toolbox”. And so, just like the many other cookieless solutions the industry is currently mulling over, SDA might just be worth a look to see if it’s a good fit for your business.
Are you a publisher trying to make the most of your data and optimize yield? Contact the Criteo Commerce Grid team today for expert help supercharging your first-party data strategy.
