Advertising Attention

Get More Control with Seller Defined Audiences

Sovrn Advertising Team // November 8, 2022

When brands and ad buyers create digital ad campaigns, they typically target audiences by selecting from a list of IAB-defined categories. Depending on the content of the ad, an auto parts retailer might choose a category like Auto Parts, Auto Repair, Electric Vehicle, Performance Vehicle, and the like. Category information is then attached to the bid request and used to match the ad to the most relevant inventory.

While the list of IAB categories is extensive, there are still plenty of gaps and overlap. They may be useful for approximating an audience segment, but they aren’t granular enough to guarantee an engaged audience. However, a new protocol called seller-defined audiences (SDAs) is emerging in the ad ecosystem that has the potential to give publishers more control in targeting by balancing targeting options between both buyer and seller.

Understanding SDAs

As a publisher, you know your site visitors better than anyone else. You have first-hand insight into their interests and demographics, the content they consume, and how much time they spend on your site. SDAs provide a way for publishers to package these first-party data insights into relevant audience segments and broadcast them to demand partners. On the buyer side, SDAs make it easier to target audiences based on demographics, interests, and purchase intent without the use of third-party cookies.

For example, a health and fitness publisher might know they have an audience of 15,000 parents who enjoy running with their kids. SDAs would let that publisher offer this audience segment to brands that want to target this unique group — like retailers who sell running shoes, jogging strollers, and more.

The Sovrn Ad Exchange was the first SSP to support SDAs for programmatic ads through Prebid. And while the approach is still gaining traction, ad giant Google recently announced that SDAs would be supported in Google Ad Manager.

Engagement takes SDAs to the next level

While alignment with the buyer’s target criteria is important, the most valuable audience segments are those who are also paying attention. To help capture these attention signals, Sovrn has devised a better measure of attention — called “engaged time” — which analyzes more than 50 distinct on-page actions like clicks, swipes, scrolls, mouse movements, and more. Sovrn measures not only when an ad is in view, but also when a person is actively engaged with the content. This ability to analyze and act on user engagement is the foundation of our data product, Sovrn Signal

Layering engagement data from Signal onto the demographic and interest data via SDAs lets you deliver more targeted (and more valuable) inventory to ad buyers. It’s a win-win: buyers can confidently direct their ad budget to target the right audience, and you’ll drive more revenue from engaged site visitors.

Now is a great time to test SDAs

As with any new technology, there are limitations facing SDAs today; currently, there isn’t a way to regulate SDAs or for advertisers to validate the data they receive from sellers. Without enough historical evidence to support using SDAs over the long term, the ad ecosystem requires more publishers and advertisers to start testing SDAs to establish a track record.

These gaps are making many advertisers hesitant to adopt SDAs while third-party cookies are still available. As Sovrn’s John Rosendahl said in a recent AdWeek article, “A small subset of publishers will put whatever information in their bid requests makes them the most valuable. Third-party cookies used to verify that, but right now it’s an open question.”

The best way to counter buyer uncertainty is to run experiments in browsers that have already eliminated cookies (like Safari and Firefox). Using either your own SDAs or any number of UIDs (universal IDs), you can add your audience data to the bidstream, where buyers can use it to bid on your ad inventory. You can then compare the performance of cookie-free ad placements to those in a standard cookie-based auction, and fine-tune your SDA data to improve results over time. Because these tests mimic a future where cookies no longer exist, they provide a reliable forecast of ad performance in a cookieless environment.

As buyers experience positive results with SDA-targeted ad placements, they’ll become more receptive to embracing this protocol as a viable replacement for third-party cookies. Perhaps the most important thing to remember is that you don’t need to dive head-first into the SDA pool — start by dipping a toe in the water, monitor results, and see what works.

Get started with Sovrn today

If you’re not already working with Sovrn’s Ad Exchange, or if you need to add Signal to your current ad ops tech stack, just email us at Our team will be happy to provide more information and help you get started.

Want to learn more?

Share this article