Which timers shape Apple Ads framework reporting

Privacy delay comes first#

Every framework postback waits out a privacy delay of hours to days before delivery, deliberately breaking the link between a tap and its report. Intraday framework dashboards are therefore always incomplete, and the freshest rows are the least trustworthy.

Plan around the delay instead of resenting it: no framework based decision inside 72 hours of the traffic, ever. The SKAdNetwork timing reference describes the delay mechanics, and the Apple Ads hub places them in the wider measurement map.

Console data paces the day; framework data judges the week. Confuse the two and every test reads wrong.

Tap and view windows decide credit#

Apple Ads credits taps for 30 days and views for 1 day, and framework streams inherit the same boundaries. Behaviour inside the window can earn credit; behaviour outside it cannot, however loyal the user eventually becomes.

Separate tap cohorts from view cohorts in every report rather than blending them into one install line. Our Attribution API guide covers the token side of the same windows, and our customer type guide shows how new and returning splits interact with them.

When credit cliffs at day 31, that is configuration working as designed, not a tracking failure to escalate.

Pre-orders stretch the calendar#

Pre-order campaigns run on 90 day tap and 61 day view windows, matching the long runway between announcement and release. Early buzz banked months ahead can still earn credit when release day converts it.

Keep pre-order cohorts labelled through launch so the release burst reads as delayed fulfilment. Coordinate announcement inventory with Today tab and search results coverage for release week.

Report pre-order and launch windows separately: blended calendars hide whether announcement spend or release spend did the work.

Conversion value updates extend the clock#

Each qualifying conversion value update can extend the reporting window, giving engaged users more time to prove worth. Well instrumented apps thus earn longer observation than silent ones, which is another reason update discipline matters.

Design event firing so meaningful behaviour lands early: first session depth, trial starts and first purchases should all update values promptly. The AdAttributionKit timing reference documents the extension rules. Detail the mapping in our conversion values guide.

Apps whose values never move get the shortest clocks and the thinnest reads. Instrumentation buys observation time.

Build a weekly reading rhythm#

Read framework data on a fixed weekly cadence with cohorts anchored on tap date, and ignore intraday framework wobbles entirely. Same weekday, same cohort definitions, same null annotations: routine beats brilliance in delayed data.

The layout in our reconciliation guide standardises the weekly table, while our null rates guide keeps suppression visible. Share the pack with stakeholders on schedule so nobody improvises reads from live dashboards.

Three calm weekly packs teach more than thirty anxious daily checks. Cadence is a measurement instrument.

When windows mislead and what to do#

Windows mislead whenever campaigns change faster than delays resolve: relaunches, budget jumps and audience swaps all render in framework data days late. Annotate every change date on framework charts so delayed echoes are never mistaken for fresh response.

Hold budgets steady for a full window cycle when testing, per our learning budget guide, and confirm anomalies with structured campaign splits before reacting. If timing puzzles persist, talk to folkfox and we will untangle delay from signal.

Patience is not passivity here; it is the only correct instrument setting.

Frequently asked questions#