If you spent the last three years learning SKAdNetwork, here is the uncomfortable question: is it already being replaced? AdAttributionKit is Apple's answer, and it is not just SKAN with a new name.
The framing matters because most teams have just finished getting SKAN 4.0 working. The good news is that AdAttributionKit builds on the same foundation rather than tearing it down. The less convenient news is that it adds capabilities SKAN never had, and the apps that understand it early will design cleaner iOS measurement than the ones that wait. This is a practitioner read on what it is, how it differs, and what to actually do now.
What Is AdAttributionKit?
AdAttributionKit: Apple's privacy-preserving attribution framework, introduced as the successor to SKAdNetwork, that uses on-device, postback-based measurement and adds support for re-engagement, reinstalls, custom creative, and alternative app marketplaces while preserving SKAN-style crowd anonymity.
The mental model is straightforward. SKAN answered one question: which campaign drove this fresh install, reported in aggregate so no individual user is exposed. AdAttributionKit keeps that privacy model and widens the set of questions it can answer.
A few things to hold onto before the detail:
-
It is a successor, not an overnight replacement. Both frameworks coexist, and your app can support each at the same time.
-
It is still postback-based and still privacy-first. There is no return to device-level identifiers here.
-
It first shipped with iOS 17.4 as the attribution path for alternative app marketplaces in the EU, then expanded.
If you want the broader context for why Apple moved measurement on-device in the first place, our explainer on mobile attribution in today's privacy landscape covers the post-ATT shift that led here.
How AdAttributionKit Differs From SKAdNetwork
The differences are practical, not cosmetic. The ones that change how you plan campaigns:
-
Re-engagement attribution. SKAN could not cleanly attribute a returning user. AdAttributionKit can. An ad network marks an impression as eligible for re-engagement, and when an existing user taps it, the app opens and a re-engagement signal is passed through. This is the single biggest gap SKAN left open.
-
Reinstall and multi-store coverage. SKAN only worked with the App Store. AdAttributionKit also covers installs from alternative marketplaces, with the storefront country code carried into the postback.
-
Custom creative support. It works across static images, video, audio, and interactive ad formats, while still limiting what the postback exposes.
-
Developer mode for testing. SKAN testing was painful because of randomised timers and long conversion windows. AdAttributionKit adds a developer mode that shortens windows and speeds up postbacks so you can validate conversion-value logic in a sane timeframe.
What stays the same is just as important: the crowd-anonymity thresholds, the postback structure, and the conversion-value windows all carry over from the model documented in our SKAN 4.0 decoding guide.
Why Apple Built a Successor at All
SKAN solved the install-attribution problem under ATT, but it left real gaps that marketers worked around with deep links and guesswork.
-
Re-engagement and retargeting had no privacy-safe measurement path, so iOS reengagement spend was effectively flying blind.
-
Third-party app marketplaces, which Apple was required to allow in the EU, needed an attribution model of their own.
-
Testing was slow enough that misconfigured conversion values often shipped to production undetected.
AdAttributionKit is best read as Apple closing those gaps while keeping the direction of travel intact: postback-based, aggregated, on-device. It is evolution, not a clean break. For teams still reconstructing full journeys under these constraints, our piece on tracking user journeys in a post-IDFA world pairs well with this shift.
What Postback Behaviour Looks Like Today
Before you plan for AdAttributionKit, it helps to know the baseline it is improving on.
Across roughly 67k iOS installs and about 114k SKAN postbacks processed on Linkrunner over the last 30 days, spanning around 20 iOS-active apps, the typical app received roughly 1.7 postbacks per install. (Methodology: 30-day window ending 8 June 2026, aggregate counts only, no project-level data; postbacks are not individually matched to installs, so this is a coverage-intensity figure, not a per-user rate.)
That number sits below the three-postback ceiling SKAN 4.0 allows, and the gap is the point:
-
Each later postback is conditional on the user staying active into the next window, so the second and third fire less often than the first.
-
Some installs never generate a postback at all (limited-ad-tracking users, redownloads without a fresh ad touch).
AdAttributionKit does not magically lift that ratio, but its re-engagement postbacks add a new category of signal on top of the install postbacks you already measure, which is where the extra value comes from.
What App Teams Should Do Now
You do not need to rebuild anything this quarter. You do need a clean starting position.
-
Audit your current SKAN setup first. A messy conversion-value schema will carry its problems straight into AdAttributionKit. Fix mapping gaps now.
-
Design conversion values to be portable. Keep your value logic framework-agnostic so the same schema serves both SKAN and AdAttributionKit.
-
Plan for re-engagement as a measured channel. If you run iOS retargeting, decide now what a re-engagement conversion means for your app.
-
Confirm your MMP roadmap covers AdAttributionKit ingestion. This is a question to put to your provider directly.
-
Do not rip out SKAN. Both frameworks will coexist through the transition, so run them in parallel.
Our SKAN 4.0 configuration strategies by app type and the postback setup guide are the two practical references to work through while you do this.
How This Looks Inside an MMP
The framework choice should be largely invisible to the marketer. The decoding work, mapping conversion values, receiving postbacks, and reconciling them into an iOS view, is conceptually identical across SKAN and AdAttributionKit.
That is the case for letting your measurement layer absorb the transition rather than handling it in spreadsheets. Tools like Linkrunner already decode SKAN postbacks and conversion values into a single iOS dashboard, and the same plumbing extends to AdAttributionKit, so the goal is one reconciled iOS picture regardless of which framework delivered the signal. You can read Apple's own reference in the AdAttributionKit documentation, and if you want to see how the postback and conversion-value cuts are surfaced in practice, the Linkrunner SKAN dashboard overview and conversion values docs are the place to start.
FAQ
What is the difference between AdAttributionKit and SKAdNetwork?
AdAttributionKit is Apple's successor to SKAN. It keeps the same privacy-preserving, postback-based model but adds re-engagement and reinstall attribution, alternative app marketplace coverage, custom creative support, and a developer testing mode.
Does AdAttributionKit support re-engagement and reinstall attribution?
Yes. Re-engagement support arrived with iOS 18. An ad network marks an impression as eligible for re-engagement, and a returning user who taps it generates a signal SKAN could never capture.
Do I have to migrate from SKAN to AdAttributionKit right now?
No. The two coexist, and you can support both at once. The sensible move is to clean up your SKAN setup now so the eventual transition is smooth.
How many postbacks does AdAttributionKit send per install?
It follows the same multi-postback window model as SKAN 4.0, with up to three install postbacks. In practice apps receive fewer than three because later postbacks depend on continued activity, with re-engagement postbacks adding signal on top.
Does AdAttributionKit work with Meta, Google, and TikTok campaigns?
Yes, as networks adopt it. Your postback and conversion-value configuration for those networks stays conceptually the same, which is why a clean SKAN setup makes adoption easier.
Closing
AdAttributionKit is not a reason to panic, and it is not a marketing rebrand of SKAN either. It is Apple closing the gaps SKAN left open, especially around re-engagement, while keeping the privacy model intact. The teams that benefit first are the ones that tidy their conversion-value schema now and treat the framework choice as something their measurement layer handles, not something they reconcile by hand.
If you want a clean iOS view that decodes SKAN today and is built to absorb AdAttributionKit as networks adopt it, request a Linkrunner demo. A good next step is to run the SKAN audit in the configuration guide above, confirm your conversion values are portable, and you will be ready for whichever framework delivers the next postback.