AdAttributionKit (AAK)

Think of SKAN as version 1.0 of privacy-safe attribution. AAK is version 2.0. It retains the anonymous, crowd-anonymity postback system but adds tools that marketers have been begging for since the launch of App Tracking Transparency (ATT).

Why AAK Matters: The 4 Breakthrough Features

While SKAN provided a baseline, AAK introduces features that bring iOS measurement closer to the flexibility of traditional MMP attribution. Here are the four game-changers:

1. Re-engagement Measurement

The single biggest complaint about SKAN was its inability to measure retargeting or re-engagement campaigns. If a user already had your app, you had no way to know which ad brought them back.

AAK fixes this. The framework now supports re-engagement measurement natively. Ad networks can flag an ad as “eligible-for-reengagement.” When an existing user clicks it, AK sends a dedicated re-engagement postback, complete with its own conversion value .

For Affise MMP users: This means you can finally optimize your win-back campaigns on iOS with the same precision you use on Android.

2. Configurable Attribution Windows (No More One-Size-Fits-All)

SKAN forced everyone into a rigid 30-day click window and a 1-day view-through window. AAK changes that.

With iOS 18.4 and later, you can now configure attribution windows based on your specific business model:

• Per ad network: Set different windows for Meta vs. TikTok.

• Per interaction type: Give clicks a longer window than views.

• Global settings: Define your own rules in the Info.plist .

Why this matters: A short window (e.g., 7 days) works for fast-converting gaming apps, while a subscription-based fitness app might need a full 30 days. You decide.

3. Country Codes in Postbacks

One of the most painful limitations of SKAN was the absence of geographic data. Marketers had to guess where their users came from or waste precious conversion bits encoding location.

AAK introduces optional country codes in postbacks. Derived from the App Store itself (and subject to crowd anonymity thresholds), this allows you to see, for example, that your campaign is crushing it in Germany but underperforming in Brazil .

Bottom line: Better geo-targeting, smarter budget allocation, and cleaner reporting inside your MMP.

4. Easier Testing with “Developer Mode”

Testing SKAN integrations was notoriously painful—you needed a live ad flow and had to wait through random delays. AAK introduces a development postbacks tool right in iOS Settings.

You can now manually generate install or re-engagement postbacks, configure test parameters (country code, granularity), and send them directly to your test server. No publisher app required .

AAK vs. SKAdNetwork: A Quick Comparison

Ecosystem Adoption

Here is the reality check. AAK was announced in March 2024, yet as of mid-2025, adoption remains slow. Major players like Meta, Google, and Snap are still primarily on SKAN 3, not even SKAN 4, let alone AAK .

Why? Many large ad networks have built their own probabilistic modeling solutions (e.g., Meta’s Advanced AEM, Google’s Integrated Conversion Measurement). They are not rushing to adopt Apple’s framework.

What this means for you: While AAK is the future, you cannot abandon SKAN tomorrow. The smart strategy is dual implementation. Affise MMP supports both frameworks, ensuring you lose no data regardless of which version an ad network or user device supports .

The Privacy-First Future Is Here

AdAttributionKit is not a revolution – it is an evolution. Apple is not backtracking on privacy. Instead, it is giving marketers better tools to operate within the privacy sandbox.

The brands that win in 2026 will be those that master AAK’s configurable windows, leverage re-engagement data, and combine it with their MMP’s modeling capabilities.

Don’t get left behind on SKAN 3. Embrace AdAttributionKit.