iOS 16.1 Ships SKAdNetwork 4

Advertisers could now receive up to three postbacks per install, covering up to 35 days. How much detail each one carries depends on "crowd anonymity": small campaigns get coarse low, medium or high values, and large ones get more. It gave back some signal while staying aggregated and delayed.

iOSAttributionApp Measurement

In short: Some aggregate signal returned: three postbacks over 35 days

What Changed

SKAdNetwork 4, shipped with iOS 16.1 in October 2022, was the largest update to the attribution framework Apple offered advertisers after App Tracking Transparency.

Three postbacks instead of one. Earlier versions reported a single conversion value within a day or two of the install. Ad networks can now receive up to three reports per install:

  • Postback 1, days 0–2: a fine-grained value (one of 64) for large campaigns, or a coarse value (low, medium or high) for smaller ones.
  • Postback 2, days 3–7: a coarse value only.
  • Postback 3, days 8–35: a coarse value only.

Crowd anonymity. How much detail a postback contains depends on how many installs a campaign drives. Large campaigns can receive a fine-grained conversion value (one of 64) in the first postback. Smaller ones get only a coarse value, low, medium or high, or nothing at all. The campaign identifier works the same way: a hierarchical source ID of up to four digits, with more digits revealed at higher volume. Apple suggested treating it as three nested numbers: the first two digits for the campaign, for example, the third for a location bucket and the fourth for the ad placement, so smaller campaigns still report the campaign even when the finer detail is withheld.1

Web-to-app attribution. For the first time, ads shown on websites in Safari could be attributed to app installs.1

Why It Mattered

SKAdNetwork 4 gave back some of the signal that ATT took away, particularly about what users do after the first day.2 But all of it stays aggregated and delayed, and the most useful detail goes to the biggest campaigns, which favours large advertisers.

Adoption was slow. Ad networks and measurement partners had to rebuild their systems, and many advertisers kept relying on modelled results from the major platforms.3 Eric Seufert later asked why SKAdNetwork 4 failed to win over the industry.4 In 2024 Apple folded SKAdNetwork into a successor framework, AdAttributionKit.

Notes

  1. “What’s New with SKAdNetwork,” Apple Developer (WWDC22 video), June 2022, https://developer.apple.com/videos/play/wwdc2022/10038/. ab
  2. Eric Benjamin Seufert, “SKAdNetwork 4.0: What Problems Does It Solve?” Mobile Dev Memo, September 28, 2022, https://mobiledevmemo.com/what-problems-with-skadnetwork-4-0-solve/. ↩
  3. “Why Apple’s SKAdNetwork 4.0 Release Could Affect Mobile App Measurement,” Digiday, October 31, 2022, https://digiday.com/marketing/how-apples-skadnetwork-4-0-release-could-affect-mobile-app-measurement/. ↩
  4. Eric Benjamin Seufert, “Why Did SKAdNetwork Fail?” Mobile Dev Memo, May 29, 2024, https://mobiledevmemo.com/why-did-skadnetwork-4-0-fail/. ↩

Where This Fits

Read the Full Timeline

26 of 49 changes, 2012–2026

You’re reading · iOS 16.1 Ships SKAdNetwork 4
  1. iOS · Apr 26, 2021 · 1.5 years earlieriOS 14.5 Enforces App Tracking TransparencyCross-app tracking and the IDFA need opt-in consent
  2. You’re here · Oct 2022
  3. iOS · Mar 2024 · 1.5 years lateriOS 17.4 Introduces AdAttributionKitApple's attribution framework extended beyond the App Store

Measurement That Keeps Up with Every Tracking Change

Bconic combines attribution, incrementality and marketing mix modelling, so you can see what your marketing drives on every platform, whatever changes next.