Seller-Defined Audiences (SDA)

Seller-Defined Audiences (SDA) is an IAB Tech Lab standard that lets publishers attach standardized, privacy-safe audience segment IDs to a bid request, built from their own first-party data, so buyers can target audience segments without relying on cookies or other cross-site identifiers.

How SDA works#

A publisher, using its own logged-in data, subscription information, or on-site behavior, classifies its audience into standard IAB Tech Lab segment categories (things like broad demographic or interest groupings) rather than proprietary, publisher-specific labels. Those standardized segment IDs get attached to the bid request the publisher's supply-side platform sends out over OpenRTB, sitting alongside the usual context and inventory data.

Because the taxonomy is standardized across the industry rather than each publisher inventing its own labels, a demand-side platform can target the same audience segment consistently across many different publishers, without needing a custom data-mapping integration with each one individually.

Why SDA exists#

It's a direct response to the decline of third-party cookies and other cross-site identifiers that programmatic targeting leaned on for years. Instead of trying to replace that tracking with a new device- or user-level identifier, SDA moves the audience signal to the seller's side: the publisher already knows things about its own logged-in or subscribed audience, and SDA gives it a standardized way to share that knowledge with buyers at the bid-request level, without exposing individual user data.

This puts SDA in the same general category as contextual targeting and publisher-supplied Data Management Platform segments, as a privacy-conscious alternative to identity-based targeting, but it's distinct in being standardized specifically for the bid request itself rather than requiring a separate data-onboarding relationship between buyer and seller.

Why it matters for buyers#

SDA gives buyers a consistent way to target audience interest and demographic segments across publishers who've adopted the standard, even as third-party cookies and mobile ad IDs become less reliable. The catch is coverage: SDA only works where publishers have actually implemented it and classified their audiences accurately, so its usefulness depends heavily on adoption across the specific inventory a buyer cares about. For buyers evaluating whether a given publisher or network is worth targeting at all, seeing what's actually running there, rather than relying purely on audience segment labels, is a useful cross-check. OpenAdLibrary's ad intelligence index shows live creative activity by network and vertical that complements audience-segment targeting decisions.

FAQ#

Is Seller-Defined Audiences the same as third-party audience data?

No. Third-party audience data comes from an external data provider, often stitched together from cross-site tracking. SDA comes directly from the publisher's own first-party data, standardized into a common taxonomy the publisher classifies its own audience against.

Do all publishers support SDA?

No. Adoption depends on whether a publisher's SSP and ad stack support the standard and whether the publisher has done the work of classifying its audience into the IAB taxonomy. Coverage varies significantly across the industry.

Does SDA replace cookies or mobile ad IDs?

Not directly. It's an alternative targeting signal that doesn't depend on cross-site identifiers at all, rather than a replacement mechanism that mimics what cookies used to do.

Is SDA more privacy-safe than cookie-based targeting?

Generally, yes. Since the audience classification happens on the publisher's side using its own first-party relationship with users, no individual-level identifier needs to travel with the bid request the way a cookie or device ID historically did.