SKAdNetwork is Apple's privacy-preserving framework for crediting app installs to ad campaigns without handing you any user-level data, currently at version 4, with the newer AdAttributionKit built on top of the same approach.
You run iOS install campaigns and you want to know which ad drove which install, but after App Tracking Transparency most of your users never handed over an identifier. SKAN is Apple's answer. The ad network registers with it, Apple's own system watches the install and the post-install activity, and then Apple, not you and not the network, sends back a signed postback saying a campaign got some installs and roughly what those users did.
SKAN 4, current since iOS 16.1, gives you up to three postback windows stretched across the first weeks, hierarchical source identifiers, and crowd-anonymity thresholds that decide how much detail you are allowed to see. Apple's newer AdAttributionKit, introduced in 2024, builds directly on the same model, keeps the postbacks and conversion values, and adds re-engagement and deep-link support. Right now the two run side by side, and Apple frames AdAttributionKit as the direction of travel.
Understand what you are giving up. The data comes back aggregated, delayed by Apple's random timers, and stripped of anyone's identity, with only a tiny conversion value to describe behavior. If a campaign does not clear the anonymity threshold, the detail collapses to coarse buckets or nothing at all, so you plan your reporting around Apple's clock rather than your own.
Stop fighting SKAN and start designing around it, because on iOS it is the measurement system, not an obstacle to one.
Sources
- SKAN 4 remains current; AdAttributionKit (2024) builds on SKAdNetwork and can run alongside it, keeping aggregated, delayed, anonymity-gated attribution. adjust.com · verified 9th August 2026
Last checked 9th August 2026. Next check 15th August 2026.
