ATT Opt-in Rates in 2026: When You Get the IDFA, and the New EU Consent Prompt (iOS 27.2)
ATT and the IDFA, the starting point of iOS app advertising, explained from Apple's official documentation. Why 2026 opt-in benchmarks differ by vendor, the rules for raising opt-in, Germany's competition settlement and the EU-only prompt in iOS 27.2.
In iOS app marketing, "What's our ATT opt-in rate?" really means "How many users can we measure precisely with the IDFA?" Based on Apple's official developer documentation, this guide explains how ATT and the IDFA relate, then covers the opt-in benchmarks published in 2026 and the consent prompt changing in the EU. SKAN conversion values and postbacks are covered in our SKAdNetwork guide, and measurement vendors in our MMP comparison.
ATT and the IDFA: the basics
According to Apple's developer documentation (User privacy and data use), on iOS 14.5, iPadOS 14.5 and tvOS 14.5 or later you must get the user's permission through the App Tracking Transparency (ATT) framework to track them or access the device's advertising identifier (IDFA).
What Apple means by "tracking"
Apple defines tracking as "linking user or device data collected from your app with user or device data collected from other companies' apps, websites or offline properties for targeted advertising or advertising measurement", and as "sharing user or device data with data brokers". The same document gives these examples:
- Showing targeted ads in your app based on data collected from other companies' apps and websites
- Sharing a list of emails or advertising IDs with an ad network to retarget users or find look-alikes
- Including a third-party SDK that combines your app's data with data from other developers' apps (even if you don't use it for that purpose)
By contrast, it states that combining data only on the device without sending it off in identifiable form, or using a data broker solely for fraud prevention or security, is not tracking.
Without consent, the IDFA is all zeros
Apple's documentation says that without the user's permission the advertising identifier value is all zeros. Adjust's help center says the same, adding that ATT requests are made per app and the default state is "not tracking". If the user turns off "Allow Apps to Request to Track" in device settings, the prompt never appears and a "Denied" status is returned.
| Situation | Result |
|---|---|
| Allows in the prompt | IDFA available; the MMP attributes with the IDFA |
| Declines in the prompt | IDFA is all zeros; no tracking |
| Turned off tracking requests in device settings | No prompt; "Denied" status returned |
| Analytics across apps from the same company | The IDFV (identifier for vendor) works without ATT, but must not be combined with other companies' data to track |
Both Apple's documentation and Adjust's help center explain the IDFV. Adjust notes that because the IDFV is generated differently for each app vendor, it can't be used to track across other companies' apps, which is why Apple allows it regardless of ATT status.
ATT opt-in rates in 2026: the actual numbers
Apple doesn't publish an official opt-in rate, so measurement companies (MMPs) each publish figures based on their own customer samples. These are the main figures cited in a September 2026 roundup by PrivacyTerms.io (source).
| Source | Opt-in rate | Basis |
|---|---|---|
| Adjust 2026 Mobile App Trends | 38% (up from 35% in Q1 2025) | Q1 2026, all verticals |
| Adjust 2026 Gaming App Insights | Games 39% | Q1 2026 |
| AppsFlyer 2026 State of Mobile | 29% | 2026, worldwide |
| Singular Quarterly Trends | 13.85% | Q2 2024, users who allowed at the first prompt |
| Right after ATT launched (Adjust historical data) | 11–15% | April 2021 |
According to the same roundup, differences by vertical and country are large too.
- Verticals: sports games 54%, action games 50%, e-commerce 34%, publications 26% (up from 18% in Q1 2025)
- Countries: Indonesia 60%, India 51% (up from 44% in Q1 2025)
Why the numbers differ by vendor
According to PrivacyTerms.io, Adjust counts the share of users who allowed among those who saw the prompt, while Singular counts only users who allowed immediately at the first prompt. With different denominators and timing, numbers for the same market can differ by more than double. In practice, rather than targeting a single industry average, it's more meaningful to track your own app's rate consistently and compare it with benchmarks for your vertical.
Raising opt-in: what Apple allows and forbids
Improving opt-in is ultimately about the experience around the prompt, but crossing Apple's rules can get your app rejected in review. The table below follows Apple's FAQ.
| Allowed | Not allowed |
|---|---|
| Showing a screen before the prompt explaining why you need permission (if it's transparent about data use) | Gating features behind tracking consent |
| Showing the prompt at a time you choose | Offering rewards for consenting |
| Explaining data use in the prompt's purpose string | Tracking without consent through other identifiers such as hashed emails or phone numbers |
| Linking to Settings so users can change per-app tracking settings | Fingerprinting users or devices from device signals |
Apple treats restricting features or offering incentives as a violation of App Review Guideline 5.1.2(i), and manipulation or coercion as a violation of 5.1.1(iv). It also states that developers are responsible for data collection and tracking by third-party SDKs in their apps.
A practical checklist:
- Is the prompt shown after users experience the app's value, rather than right at first launch?
- Does the pre-prompt screen describe the real data use accurately, such as "personalised ads"?
- Are your ad and analytics SDKs avoiding sending identifiers for users who declined ATT?
- Are you receiving ATT status (allowed, denied and so on) in MMP raw data and analysing it by cohort?
What's changing in 2026: a new consent prompt in the EU
Germany's competition settlement
According to Inimino (source), Germany's Federal Cartel Office made Apple's ATT commitments binding on August 17, 2026. The proceedings, open since 2022, closed without a fine or a finding of infringement. The key points:
- The wording and order of choices in Apple's own ad-consent prompt and the ATT prompt for third-party apps become more alike
- Symbols and wording that could discourage consent are removed from the third-party prompt
- Publishers get up to 4,000 characters to explain why they are asking
- Implementation within four months of the decision being served, then seven years of oversight by an independent monitor
The same report says France and Italy previously fined Apple €150 million and €98.6 million respectively over ATT.
The EU-only alternative prompt in iOS 27.2
In its developer documentation, Apple says that following agreements with some European competition authorities, an alternative version of the ATT prompt can be used in the EU from iOS 27.2 and iPadOS 27.2. In summary:
| Item | Details |
|---|---|
| Scope | Apps in the EU. Apps distributed in Germany, France, Italy, Poland and Romania may use only the alternative version |
| What changes | New format and wording, with an "Additional Information" text button linking to more explanation and granular consent screens |
| Re-asking | EU users can be shown the prompt again one year after their previous choice (allow or deny) |
| Settings name | In the EU, "Allow Apps to Request to Track" becomes "Allow Apps to Request to Link Your Activity Across Companies" |
| What stays the same | The requirements for when you must ask for tracking consent |
Apple's documentation states that if the user has turned off tracking requests in device settings, you can't re-ask after a year either. And whatever you show behind the "Additional Information" button, you can't access the IDFA or collect tracking data until the user allows tracking in the ATT prompt.
According to The Mac Observer (source), Apple released the first iOS 27.2 beta on September 16 and hadn't announced a release date at the time of the report. We couldn't confirm a release date as of writing either.
What it means for app marketers outside the EU
These changes target the EU, and Apple's documentation lists no changes for other regions such as South Korea. If you distribute apps in Europe, though, prepare the following.
- Implement the alternative prompt and its wording for users in Germany, France, Italy, Poland and Romania
- Decide whether to use the one-year re-ask, and how to combine it with your existing GDPR consent management platform (CMP)
- Flag in reports that opt-in trends by EU country may break before and after the new prompt
Summary
Before it's a "higher is better" number, the ATT opt-in rate shows how much your measurement depends on the IDFA. With 2026 industry averages at roughly 29% to 38% depending on the vendor, it's realistic to treat consenting-user data as a bonus and build on aggregated measurement such as SKAN and AdAttributionKit.
Frequently asked questions
Q1. Does Apple publish an official ATT opt-in rate?
No. Apple doesn't publish one. Most figures come from measurement companies such as Adjust, AppsFlyer and Singular, based on their own data.
Q2. What happens to the IDFA if a user declines?
According to Apple, without permission the advertising identifier returns all zeros and can't be used for tracking.
Q3. Can I track users who declined ATT with hashed emails?
No. Apple's FAQ states that ATT permission is required even when tracking with identifiers other than the IDFA, such as hashed emails or phone numbers.
Q4. Can I run a promotion that rewards users for consenting?
No. Apple treats restricting features or offering rewards conditional on tracking consent as a violation of App Review Guideline 5.1.2(i). A screen before the prompt that transparently explains data use is allowed.
Q5. Does the new EU prompt apply outside the EU?
According to Apple's documentation, the alternative prompt and the one-year re-ask rule apply to EU users. The documentation lists no changes for other regions.
Sources: Apple Developer "User privacy and data use", Adjust Help Center "App Tracking Transparency (ATT) framework", PrivacyTerms.io (2026-09-03), Inimino (2026-08-18), The Mac Observer (2026-09-17).