The ATT opt-in rate is the share of iOS users who tap "Allow" on Apple's App Tracking Transparency prompt.
Among users who actually see the prompt, the industry average sits around 35% (Adjust, Q2 2025), with most apps landing between roughly 25% and 50%.
Measured across all devices at first open, the number drops to about 14% (Singular, Q2 2024), so always check which definition a benchmark uses.
Below are the published numbers by category, platform and country, what counts as a good rate for your app, and what our app test database says about asking for permission at the right moment.
The average ATT opt-in rate and the range
Three measurement companies publish ATT numbers regularly. They disagree, mostly because they divide by different things.
| Source | Period | Opt-in rate | What it measures |
|---|---|---|---|
| Adjust | Q2 2025 | 35% | Users shown the prompt who allowed tracking (34.5% in Q2 2024, 34% in Q2 2023) |
| AppsFlyer | 2024 data, released April 2025 | 50% | Users who consented, globally, in apps showing the prompt |
| Singular | Q2 2024 | 13.85% | Devices that allow tracking at first open (18.9% in Q4 2023) |
| AppsFlyer, early adopters | 2021 | 39% weighted, 29% median app | Authorized / (authorized + denied), about 550 early-adopter apps |
The gap between 14% and 50% is not noise.
Singular counts every device at first open, including users who never see a prompt or see it later.
Adjust and AppsFlyer count only users who were asked.
The 2021 AppsFlyer study also shows how much the average depends on weighting.
The weighted rate was 39%, but the average app saw 26% and the median app 29%.
A few big apps with high rates pull the headline number up.
For planning, a fair reading is this: if you show the prompt, expect somewhere around a third of asked users to say yes.
If you want to know what share of your total iOS installs you can track, expect a much smaller figure.
ATT opt-in rates by category
Category matters more than almost anything else you control.
Adjust's 2025 data ranks categories by the share of prompted users who allow tracking:
| Category | ATT opt-in rate (Adjust, 2025) |
|---|---|
| Gaming | 37% |
| Food and drink | 36% |
| Ecommerce | 34% |
| Fintech | 29% |
| Social | 26% |
| Travel | 25% |
| Education | 7% |
Adjust attributes the low education figure largely to rules around collecting data from minors, who make up much of that audience. Source: Adjust, ATT opt-in rates 2025.
Singular's first-open data shows the same shape at a lower level.
In Q2 2024, games averaged 18.59% and non-game apps 11.92%.
Weather apps led at 38%, while news and medical apps fell below 5%, and business, social and entertainment apps sat just above 5% (Singular).
Games lead in most datasets. AppsFlyer reported that in Q1 2024, 51% of gaming app users and 46% of non-gaming app users consented to tracking (App Developer Magazine, citing AppsFlyer).
The early 2021 sample was the exception, with non-gaming apps ahead at 42% versus 31% for games (AppsFlyer).
iOS vs Android: there is no Android ATT rate
ATT is an Apple framework. Apple requires it on iOS, iPadOS and tvOS 14.5 or later whenever an app links its user data with other companies' data for targeted ads or ad measurement, or shares it with data brokers.
Android has no equivalent system prompt. The advertising ID is available by default, and users have to opt out themselves.
According to Google's Play Console help, when a user deletes their advertising ID in Android Settings, apps get a string of zeros instead.
Apps targeting Android 13 or higher must also declare the AD_ID permission to read it.
So "Android opt-in rate" is the wrong comparison.
On Android, the question is how many users have deleted their ID.
Opting out takes a trip into Settings rather than a tap on a prompt.
The practical effect: on iOS you rely on the minority who opt in plus Apple's aggregated attribution.
Apple's AdAttributionKit, which builds on SKAdNetwork, attributes installs without the ATT prompt and without user-level identifiers.
By region and price point
Country shifts the rate by 20 points or more.
In Adjust's Q2 2025 data, Brazil led at 50%, followed by the United Arab Emirates at 49% and Turkey at 42% (Adjust).
AppsFlyer's 2024 data puts France just under 51% and Germany at 47% (AppsFlyer).
Rankings shift between sources and years, so compare against the newest data for your main market.
We could not find a public study that breaks ATT opt-in rates out by app price or subscription price, and our database has no data on it either.
We won't guess. If price is a factor for you, segment your own results by plan or paying status after launch.
What a "good" ATT opt-in rate means for you
A good rate is one that beats the benchmark for your category and main market, measured the same way the benchmark measures it.
By Adjust's numbers, a 30% rate is above average for a social app (26%) and below average for a game (37%).
Use this rough guide for prompted users:
- Below 20%: check your timing and pre-prompt screen first, unless you are in education or a low-rate category.
- 20% to 35%: around the cross-category average. Room to improve, but not a problem.
- Above 35%: at or above Adjust's category leaders. Further gains are usually small.
Before chasing a higher number, ask whether you need it.
ATT consent mainly helps ad attribution and ad revenue.
A subscription app that runs little paid acquisition and shows no ads may get more from a better onboarding flow than from a few extra points of opt-in.
Also watch what the prompt costs you. Every system dialog early in a session is a chance for a new user to leave.
Track onboarding completion and trial conversion alongside the opt-in rate, not just the rate on its own.
How apps in our database moved theirs
Our database does not yet hold a public A/B test of the ATT prompt itself.
Apps rarely publish them. What it does hold is a set of tests on the same underlying question: what happens when you change the order of screens and the moment you ask a new user for something.
Those results point the same way as Singular's advice to "build a relationship" and ask later (Singular). Here is what the closest tests show.
Earn the ask with onboarding first
Lose It! ran an A/B test on a longer onboarding with more questions and saw trial start rate up by double digits.
More screens before the ask did not cost them users; it raised commitment.
Applica rebuilt the Five Minute Journal onboarding around user research and reported +20% ARPU, about +10% trial starts and about +14% trial conversions.
This was a bundled change, so no single screen gets the credit.
Placement changes how many people even see the ask
Jake Mor reported that moving FitnessAI's paywall before onboarding, with video gave +50% users seeing the paywall.
A bundled change again, but it shows how much exposure depends on where a screen sits.
In a multivariate test shared by Rosie Hoggmascall, three placements of a paywall around onboarding, plus a carousel produced 2% (A) vs 8% (B) vs 15% (C) trial opt-in.
Same app, same offer, a 7x spread from sequencing.
The lesson for ATT is to test the prompt's position, not assume it belongs on first launch.
Apps often get more yes answers after the user has seen the core feature.
Be plain about what happens next
Marcus Burke reported an A/B test of Blinkist's trial timeline paywall, which spelled out exactly when users would be charged.
It brought a significant rise in trial opt-in, and trial cancellations fell 4%.
The ATT equivalent is an honest pre-prompt screen.
Apple allows you to explain why you want permission before the system alert, as long as you are transparent about how the data is used (Apple).
It does not allow incentives for allowing tracking, gating features on consent, or screens that imitate the system alert.
See our permission primer tests for how apps word these screens, and the ATT guide for the full rules.
A testing plan for your own prompt
- Measure your current rate as allowed / (allowed + denied), so you can compare it with Adjust and AppsFlyer.
- Test prompt timing: first launch vs after the first core action vs after onboarding.
- Test a pre-prompt screen with one or two plain sentences on what the data does, against no pre-prompt.
- Judge winners on onboarding completion and revenue too, not opt-in alone. A higher rate that loses new users is a bad trade.
The same approach works for notification permission. Compare with our push opt-in rate benchmarks, browse onboarding tests, and run the app onboarding checklist before you ship a new flow.