Playbook: Validate Your SKAdNetwork Postbacks Before You Trust a Dollar
Before optimizing on SKAN data, run this 6-step audit:
— Confirm your MMP receives the raw 24-byte postback from Apple, not just the MMP's modeled copy
— Check the transaction-id dedupe — Apple sends winning + non-winning copies; count only winners
— Verify conversion-value mapping matches your active schema version, not last quarter's
— Audit fidelity-type: fidelity 1 = StoreKit-rendered (trusted), fidelity 0 = view-through (noisier)
— Match source-identifier granularity to your privacy tier (2, 3, or 4 digits)
— Reconcile postback counts against install counts; a gap over 15% means lost or delayed postbacks
✓ Catches schema drift that silently zeroes your revenue events
✓ Works across AppsFlyer, Adjust, Branch identically
✗ Won't fix Apple's crowd-anonymity null values on low-volume campaigns
✗ Tedious without postback-level export access
Verdict: Do this once per quarter and after every schema change — it's the cheapest fraud-and-loss insurance you have.
Best for: UA managers running iOS in-app campaigns at meaningful spend.
In-App Bench
@InAppBench
Playbook: Validate Your SKAdNetwork Postbacks Before You Trust a Dollar
Этот пост опубликован в Telegram-канале In-App Bench. Подписаться можно по ссылке: @InAppBench.