Short answer: this is expected. Apple adds a random delay of 24 to 48 hours after the first window and 24 to 144 hours after the second and third. Locking a window early starts the delay sooner, but you cannot remove it.
Why later SKAN postbacks are slow
Random delay: a deliberate wait Apple adds before sending each SKAN postback so that its timing cannot be linked to a specific person's install.
The windows themselves are long, too. The second covers days 3 to 7 after first launch and the third days 8 to 35, so a third postback can arrive weeks after install.
When should I expect each postback?
Expect the first two to four days after install, the second around 8 to 13 days, and the third around 36 to 41 days, if no window is locked early.
Likely causes
- Apple's random delay. Apple documents 24 to 48 hours for the first postback and 24 to 144 hours for the second and third.
- Long windows. Each delay only starts when the window ends, at day 2, day 7 and day 35.
- Windows not locked. If the value is final early but the window is not locked, the postback waits for the full window.
How to confirm it
- List postbacks by postback sequence and received date.
- Work out when each window ended for the install cohort.
- Check that the gap falls within Apple's documented range.
- Check your code for lockWindow calls.
How to fix it
- Accept the delay and report SKAN cohorts by install week, not received date.
- Lock a window early with lockWindow when the value for that window is final.
- Read early optimisation signals from the first postback and use later postbacks for longer-term quality.
- Label dashboards so that recent cohorts are clearly incomplete.
| Postback | Window | Delay after window ends | Typical arrival after first launch |
|---|---|---|---|
| First | Days 0 to 2 | 24 to 48 hours | About days 3 to 4 |
| Second | Days 3 to 7 | 24 to 144 hours | About days 8 to 13 |
| Third | Days 8 to 35 | 24 to 144 hours | About days 36 to 41 |
These ranges assume no early lock. AdAttributionKit uses the same three windows and delays, according to Apple.
The SKAN 4 postbacks guide covers windows and locking in more detail. If the later postbacks never arrive, see only the first SKAN postback, and for general reporting lag see attribution data delayed. The SKAN troubleshooter helps if the postbacks are empty.