iOS ATT Impact Calculator
Since iOS 14.5, App Tracking Transparency lets users opt out of cross-app tracking — and most do. The conversions those users drive still happen; your ad platform just never sees them. This tool estimates how many you are missing, so you stop trusting a dashboard that is structurally blind.
App Tracking Transparency (ATT) asks every iOS user to allow or deny cross-app tracking. When they deny, the post-click and post-view conversions they generate become invisible to ad platforms. This calculator models hidden conversions as iOS traffic share × (1 − opt-in rate) × attribution-loss factor, then grosses your reported number back up to an estimated true count and shows the gap between true and reported ROAS. It is a planning model, not a measured figure — the only way to confirm incrementality is a controlled experiment.
iOS ATT Impact Calculator inputs and result
| Attribution-loss factor | Hidden conversions | Under-reported % |
|---|
How to use this calculator
- Estimate your iOS traffic sharePull the share of sessions or conversions that come from iOS. App analytics and GA4 device reports both expose this; many consumer and DTC audiences sit in the 50-65% range.
- Enter your real ATT opt-in rateDo not guess. Your ATT prompt has its own analytics; read the actual Allow rate. The ~25% benchmark is a starting point, not your number.
- Add the conversions your platform reportsUse the platform-attributed conversion count for the period. The tool grosses it back up to the modeled true figure.
- Set an attribution-loss factor100% assumes every opted-out conversion vanishes. Because SKAdNetwork and modeled conversions recover some signal, 70-90% is often more realistic. The table shows the sensitivity.
- Read the gap, then validate itTreat the output as a planning estimate of how blind your dashboard is — then confirm true performance with blended measurement and geo holdout experiments.
RGM Expert Says
We reach for this calculator the moment a client says their paid social ‘stopped working’ after 2021. Usually it did not stop working — it stopped being visible. App Tracking Transparency knocked out the deterministic signal that platforms used to tie a conversion back to a click, and on a heavily-iOS audience the reported numbers fell off a cliff while the real business kept converting. Putting the iOS share, the actual opt-in rate and the loss factor into one view turns a vague panic into a sized estimate.
The number that moves the room is the ROAS gap. When a client sees that a 2.0× reported ROAS is plausibly a 2.6× true ROAS, the conversation shifts from ‘cut this channel’ to ‘measure this channel honestly.’ We deliberately label the output a model, because it is: the opt-in rate is the one input people get wrong, and the loss factor depends on how much SKAdNetwork and modeled conversions are recovering for them. We always run two or three loss factors before quoting a range.
Where this tool ends, experiments begin. The calculator tells you how much your single-platform dashboard is likely under-counting; it cannot tell you which conversions the ads actually caused. For that we pair it with blended measurement — total revenue over total spend — and geo holdout tests. The model sizes the blindness; the experiment proves the lift.
How it works
The model treats reported conversions as the visible slice of a larger true total. The hidden slice is the iOS traffic that opted out of tracking and whose activity the platform therefore cannot attribute:
- iOS share — portion of traffic on iOS, where ATT applies.
- Opt-in rate — share of iOS users who allow tracking; the rest are the at-risk slice.
- Attribution-loss factor — how much of an opted-out user's activity is truly unattributed (100% = strict worst case).
- Observed share — the fraction of true conversions the platform can still see.
This is an RGM planning model, not a measured value. It assumes the same conversion rate among opted-in and opted-out users and that platform reporting captures the observed share fully. Validate the estimate with blended measurement and a geo holdout test.
Why a single-platform dashboard now lies by design
Apple introduced App Tracking Transparency with iOS 14.5 in April 2021. Every app must now ask permission before tracking a user across other companies’ apps and sites, and when a user denies it, the device’s advertising identifier (IDFA) is withheld. The deterministic link between an ad click and a later conversion — the link platforms were built on — simply disappears for those users. The conversions still happen; the attribution does not.
The catch is that opt-in is low. Industry trackers such as AppsFlyer have reported consent rates clustering in the low-to-mid double digits, often cited around 25%, which means roughly three in four iOS users are invisible to cross-app tracking. On an audience that is half iOS, that is a large hole punched directly through your reported conversions — and it falls hardest on prospecting and upper-funnel campaigns, the ones last-click attribution already shortchanges.
The fix is not a better dashboard; it is a different method. Aggregated and modeled solutions (SKAdNetwork, modeled conversions, the Conversions API for server-side signal) recover some of the loss, which is why the loss factor here is adjustable. But the durable answer is to stop steering by one platform’s self-reported number and steer by blended measurement — total revenue over total spend — with periodic geo or holdout experiments to prove incremental lift.
ATT opt-in benchmarks worth knowing
Opt-in rate is the single input people most often get wrong. These are widely-cited ranges; always prefer your own ATT prompt analytics.
| Signal | Typical figure | Note |
|---|---|---|
| ATT launch | iOS 14.5, Apr 2021 | Tracking now opt-in by default |
| Global ATT opt-in | ~20-30% | Varies by vertical and prompt design |
| Common planning value | ~25% | Frequently cited industry benchmark |
| Heavily-iOS US consumer | 50-65% iOS share | Amplifies the hidden gap |
How practitioners talk about signal loss
After ATT, the conversions did not disappear — the visibility did. Steer by blended efficiency and prove lift with experiments, not by a platform's self-graded report card.
Data beats opinions, but only honest data. A dashboard that cannot see most iOS users is an opinion wearing a number's clothes.