What is AdAttributionKit?

What it adds over SKAdNetwork

AdAttributionKit keeps the architecture of SKAdNetwork — operating-system-reported, aggregated, delayed, threshold-gated — and extends it in the directions SKAN could not reach.

The most significant additions are support for alternative app marketplaces, which regulatory change made necessary in some jurisdictions, and measurement of re-engagement rather than only first install. It also improves developer ergonomics around testing, which sounds minor and removes a real source of misconfigured campaigns.

What stays the same

• No user-level identifiers. Reporting is aggregate by construction.
• Randomised delay between the event and the postback.
• Privacy thresholds that withhold or coarsen detail on low-volume campaigns.
• A constrained conversion signal, so post-install outcomes must be prioritised in advance.
• No join key to your own data.

What to do about it

The practical advice is unchanged from SKAN, because the constraints are the same. Keep campaign structures consolidated enough to clear anonymity thresholds. Decide deliberately which post-install outcomes are worth the limited signal. Accept that optimisation cycles are slower and stop trying to reconstruct user-level truth from aggregate output.

And use it for what it is good for — a directional read on which campaigns produce installs — while running incrementality tests for the question of whether the spend is causing anything.