Table of Contents
App retargeting is the practice of serving paid ads to people who already have your app installed, with the goal of bringing dormant users back, reactivating lapsed trials, or moving existing users toward a purchase or subscription renewal. App re-engagement is the broader discipline that surrounds it: every paid and owned touchpoint used to win back attention after the first install. The question every growth team asks in 2026 is simple: does app retargeting still work after Apple’s App Tracking Transparency (ATT) framework gutted device-level identifiers? The short answer is yes, but only if you rebuild it around addressability, owned channels, and privacy-safe measurement rather than the IDFA-everywhere playbook that died in 2021.
Admiral Media has spent years rebuilding re-engagement for app advertisers in exactly this environment. Based on the Admiral Media team’s work managing over €500M in mobile ad spend across 150+ app brands, the agencies that still get retargeting to pay back are the ones that treat it as a measurement and segmentation problem first and a media-buying problem second. This guide lays out what is and is not addressable on iOS today, the Admiral Media framework for structuring re-engagement, a decision matrix for choosing channels, and how to measure it all without leaning on signals Apple no longer provides.
Does App Retargeting Still Work After ATT?
Yes, app retargeting still works after ATT, but the addressable audience is smaller and the measurement is fundamentally different. ATT did not ban retargeting. It removed the default, silent access to the IDFA (Apple’s Identifier for Advertisers) that made deterministic, one-to-one retargeting trivial. After a user declines the ATT prompt, advertisers cannot read that device’s IDFA, which means the classic flow of dropping a user into a retargeting audience and chasing them across apps no longer applies to most of your iOS base.
What remains is a layered reality. A portion of your users opt in to tracking and stay fully addressable through the IDFA. A much larger portion are reachable only through channels you own, such as push notifications and email, or through contextual and modeled approaches that do not depend on a stable cross-app identifier. Admiral Media’s position is that the winning re-engagement strategy in 2026 is not a single tactic but a portfolio: owned channels carry the volume, opted-in paid retargeting carries the precision, and privacy-safe measurement decides where the next euro goes. According to Apple’s App Tracking Transparency documentation, any access to the device identifier for cross-app tracking requires explicit user permission, which is why owned audiences have become the structural backbone of post-ATT re-engagement.
What ATT and SKAdNetwork Actually Changed
ATT changed who you can reach with paid ads; SKAdNetwork changed what you can measure about the people you reach. These are two separate constraints, and conflating them is the single most common mistake the Admiral Media team sees when auditing a stalled retargeting program.
On the targeting side, declining ATT removes the IDFA, so deterministic audience building shrinks to the opted-in slice of your users. On the measurement side, SKAdNetwork, Apple’s privacy-preserving attribution framework, was built for installs. It reports campaign performance through delayed, aggregated, and privacy-thresholded postbacks, and historically it did not measure what happened when an existing user returned to your app after seeing an ad. In practical terms, that meant re-engagement actions were invisible to the campaign optimization loop unless you measured them yourself.
Apple’s newer framework changes part of this picture. AdAttributionKit can attribute installs and re-engagement from the same ad, and its conversion-type signal distinguishes a fresh download, a redownload, and a re-engagement. As of iOS 18.4, an app can run multiple active re-engagement conversion windows at the same time, which gives lifecycle marketers far more room to measure distinct win-back moments rather than a single coarse event. Critically, AdAttributionKit reports that users returned after an ad, but it does not let you target named individuals. As Apple’s AdAttributionKit and SKAdNetwork interoperability documentation describes, the two frameworks are designed to coexist, with AdAttributionKit extending attribution to re-engagement and alternative marketplaces while SKAdNetwork continues to serve App Store installs. The takeaway: you can now measure privacy-safe re-engagement, but you still cannot rebuild old-school individual-level retargeting at scale.
The Admiral Media Addressability-First Re-Engagement Framework
The Admiral Media Addressability-First Re-Engagement Framework is a six-step sequence for rebuilding retargeting around what iOS still permits, instead of fighting the platform for signals it no longer shares. The Admiral Media team applies this framework when restructuring re-engagement for subscription, dating, fintech, and health apps, and the order matters: each step removes a failure mode before the next one can compound it.
- Segment by addressability, not just behavior. Before choosing a tactic, split your user base into three buckets: fully addressable opted-in users, owned-channel-reachable users (those who accepted push or gave you an email), and unaddressable users. Most re-engagement budgets are wasted because they assume the first bucket is far larger than it is.
- Exhaust owned channels first. Push notifications and email reach a large share of lapsed users at near-zero marginal cost and require no third-party identifier. Owned channels should carry the bulk of routine win-back volume before a single euro of paid retargeting is spent.
- Build the measurement architecture before the media plan. Decide how each re-engagement action will be measured: AdAttributionKit re-engagement conversions for paid iOS, deterministic measurement for opted-in users, and incrementality testing as the source of truth. If you cannot measure a channel honestly, you cannot scale it responsibly.
- Fix deep-linking infrastructure. Re-engagement fails silently when an ad sends a returning user to a generic home screen instead of the exact in-app destination tied to the offer. Deferred deep links and reliable routing are the difference between a re-engaged user and an immediate re-bounce.
- Tailor creative to lifecycle stage, not acquisition. A lapsed subscriber, a trial that never converted, and an active free user need different messages. Re-engagement creative that recycles install ads underperforms because it ignores what the user already knows about the product.
- Validate with incrementality, then reallocate. Because retargeting is structurally prone to taking credit for organic returns, every paid re-engagement channel must pass an incrementality test before it earns more budget. The Admiral Media team treats unvalidated retargeting spend as a liability, not an asset.
This framework pairs naturally with a retention-first user acquisition approach: when acquisition is optimized for users who are worth retaining, re-engagement has higher-quality cohorts to work with and pays back faster.
The Re-Engagement Channel Decision Matrix
Choosing a re-engagement channel post-ATT comes down to four questions: is the audience addressable, can the outcome be measured, how much control do you have, and what lifecycle moment does it serve best. The matrix below is the decision tool the Admiral Media team uses to sequence channels for an app’s win-back program.
| Channel | Addressable on iOS? | Measurement | Marketer control | Best for |
|---|---|---|---|---|
| Push notifications (owned) | Yes, for users who accepted push | First-party, deterministic | High | Routine win-back, dormant-user nudges |
| Email and CRM (owned) | Yes, for users who shared email | First-party, deterministic | High | Lapsed-subscriber reactivation, offers |
| Paid retargeting, opted-in IDFA | Partial, opted-in users only | Deterministic for opted-in cohort | Medium | High-value, identifiable segments |
| Contextual and modeled re-engagement | Yes, no individual ID required | AdAttributionKit re-engagement + incrementality | Medium | Scaling beyond the opted-in slice |
| Web-to-app remarketing | Yes, web audiences plus deep links | Web analytics + deferred deep-link tracking | Medium to high | Subscription apps with a web presence |
No single row wins. The Admiral Media team sequences from the top down: owned channels first for volume and margin, opted-in paid retargeting for precision on high-value segments, then contextual and web-to-app layers to extend reach past the addressable ceiling.
Owned Channels: The New Retargeting Backbone
Owned channels are the most reliable form of post-ATT re-engagement because they require no third-party identifier and no platform permission beyond the one the user already granted. Push notifications reach users who opted in to notifications, and email reaches anyone who shared an address during onboarding or checkout. Neither depends on the IDFA, which makes them immune to the addressability collapse that hit paid retargeting.
The strategic shift is to stop treating owned channels as a CRM afterthought and start treating them as the primary re-engagement engine. In Admiral Media’s experience managing lifecycle programs for subscription apps, the apps that recover the most lapsed revenue are the ones that capture an owned-channel permission early, then orchestrate push and email around real product milestones rather than generic promotional blasts. Deep-linking is what makes this work at the mechanism level: a push that opens the exact paywall, streak, or saved item the user abandoned converts far better than one that dumps them on a home screen, because it removes the friction between intent and action. For subscription products specifically, pairing owned re-engagement with web-to-app funnels lets you reach users in an environment where web identifiers and email still function, then route them back into the app through a reliable deep link.
Paid Re-Engagement That Still Works
Paid re-engagement still works on iOS, but it now operates on a smaller addressable audience and a different measurement model. There are three durable paid approaches. The first is deterministic retargeting against opted-in users, where the IDFA is still available and one-to-one targeting behaves much as it did before ATT. This slice is smaller than it once was, so the Admiral Media team reserves it for the highest-value segments where precision justifies the cost.
The second is contextual and modeled re-engagement, which reaches users based on the app environment and signals that do not require a stable cross-app identifier, then measures returns through AdAttributionKit re-engagement conversions backed by incrementality. This is how you scale past the opted-in ceiling without pretending you can still see every device. The third is web-to-app remarketing, which leans on web audiences and deep links to pull existing users back into the app, a tactic that is especially strong for subscription apps with a meaningful web footprint. Across all three, the mechanism that protects performance is honest measurement: because re-engagement ads are shown to people already likely to return, naive last-touch attribution systematically overstates their impact. That is why Admiral Media gates paid re-engagement behind incrementality testing before scaling spend.
Measuring Re-Engagement Without the Signals You Used to Have
Measuring re-engagement post-ATT requires combining three lenses, because no single source tells the whole truth anymore. The first lens is AdAttributionKit re-engagement attribution, which reports privacy-safe, Apple-signed evidence that users returned after seeing an ad, with conversion windows you can now run in parallel as of iOS 18.4. The second is deterministic first-party measurement for owned channels and opted-in users, where you control the event stream end to end. The third, and the one Admiral Media treats as the source of truth, is incrementality: controlled holdout testing that answers whether the re-engagement actually caused the return or merely took credit for it.
This three-lens model is the same discipline behind the AdAttributionKit iOS measurement playbook the Admiral Media team uses on live accounts. The economic logic is straightforward: re-engagement that fails an incrementality test is not re-engagement, it is a tax on organic returns. Tying measured returns to downstream value, the way a predictive LTV model does for acquisition, keeps the win-back budget pointed at users whose return is actually worth paying for. The structural facts to anchor any measurement plan: SKAdNetwork attributes installs, AdAttributionKit extends attribution to re-engagement, and only incrementality testing isolates true causal lift.
What Admiral Media’s Results Reveal About Lifecycle Value
Re-engagement only pays back when the underlying acquisition and monetization economics are sound, and Admiral Media’s case studies show what that foundation looks like in practice. The two examples below come from full-funnel acquisition and scaling work, not isolated retargeting campaigns, but they illustrate the lifecycle value that disciplined re-engagement is built to protect.
Admiral Media managed PURE’s US Android user acquisition by testing a programmatic DSP channel against an established self-attributing network, achieving a 74% lower cost per install and exceeding the app’s seven-day ROAS goals, which opened the door to new market entries. The efficiency gap was stark: the programmatic channel delivered a CPI of $2.44 against the self-attributing network’s $9.43, roughly four times lower. That early-LTV efficiency is exactly the cohort quality that makes downstream re-engagement economical.
The second example shows the monetization ceiling that re-engagement exists to defend. In Admiral Media’s work with Fastic, the intermittent fasting app, full-funnel scaling from near-zero toward one million users produced a 1655% increase in purchases, a 439% increase in revenue, a 639% increase in installs, and a 952% increase in monthly active users, while cutting cost per purchase in half, comparing May 2020 with December 2019. When monthly active users and purchases grow at that rate, the value of keeping each cohort active, the core job of re-engagement, compounds accordingly.
The lesson the Admiral Media team draws from both accounts is that re-engagement is a multiplier on acquisition quality, never a substitute for it. Efficient acquisition produces cohorts worth winning back; strong monetization makes winning them back profitable. For a fuller view of which signals to watch across that lifecycle, the Admiral Media guide to app marketing metrics that matter maps the measurement layer that ties acquisition and re-engagement together.
Frequently Asked Questions
Does app retargeting still work after ATT?
Yes, app retargeting still works after ATT, but on a smaller addressable audience and with different measurement. ATT removed silent IDFA access, so deterministic one-to-one retargeting now applies mainly to users who opted in to tracking. The rest of your base is reached through owned channels like push and email, or through contextual and modeled re-engagement measured with AdAttributionKit and incrementality testing. The tactic is not dead; the IDFA-everywhere version of it is.
What is the difference between app retargeting and app re-engagement?
App retargeting usually refers to paid ads served to existing users to bring them back or drive a purchase, while app re-engagement is the broader set of paid and owned activities used to recover attention after the first install. Retargeting is one channel inside a re-engagement strategy. In practice, owned channels carry most re-engagement volume and paid retargeting adds precision on high-value segments.
Can you retarget iOS users who declined the ATT prompt?
You cannot use the IDFA to retarget users who declined the ATT prompt, because Apple blocks cross-app identifier access without permission. You can still reach those users through channels you own, such as push notifications and email, and through contextual or modeled paid re-engagement that does not rely on a stable cross-app ID. Returns from paid ads can then be measured with AdAttributionKit re-engagement conversions rather than device-level tracking.
How do you measure re-engagement campaigns on iOS in 2026?
Measure iOS re-engagement with three lenses: AdAttributionKit re-engagement attribution for privacy-safe paid signals, first-party deterministic measurement for owned channels and opted-in users, and incrementality testing as the source of truth for causal lift. SKAdNetwork attributes installs, while AdAttributionKit extends attribution to re-engagement and, as of iOS 18.4, supports multiple parallel re-engagement conversion windows. Incrementality is what separates real lift from credit for organic returns.
Why does deep-linking matter for app retargeting?
Deep-linking matters because re-engagement only converts when the user lands on the exact in-app destination tied to the ad or message, not a generic home screen. A push or ad that opens the abandoned paywall, cart, or saved item removes friction between intent and action, which lifts conversion. Without reliable deferred deep links, re-engagement spend leaks at the final step even when targeting and creative are correct.
Should owned channels or paid ads lead a post-ATT re-engagement strategy?
Owned channels should lead, because push and email reach a large share of lapsed users at near-zero marginal cost and require no third-party identifier. Paid retargeting should follow as a precision layer for high-value, identifiable segments and to extend reach past the opted-in ceiling through contextual and web-to-app tactics. Admiral Media sequences owned channels first, then validates each paid layer with incrementality before scaling it.
