Almost all playable creative is written for someone who has never seen the product. Retargeting turns that around: the viewer has installed the app, played it, and stopped. They do not need the mechanic explained, they will not be impressed by a tutorial, and the one thing your ad has that a UA ad does not is context. You know something about them.
Running a UA playable at a lapsed audience is the most common mistake here, and it fails in a specific way. Being taught the basics of a game you already played is mildly insulting, and the creative reads as though the advertiser has forgotten who you are.
What to cut
The hook survives, because attention still has to be won. Nearly everything else changes.
- The tutorial goes. No hand, no highlight, no cue. This audience knows what a match is.
- The explanation of the genre goes. You are not establishing what kind of game this is.
- The easy first board goes. Assists tuned for a stranger read as condescension to a returning player; you can start at a level that looks like the game they left.
- The install framing goes, and it is the one people forget. Copy is written for install intent (get it now, play free) and none of it applies to somebody who already has the app on their phone.
The three jobs a re-engagement playable can do
Remind. The simplest and often the most effective: a few seconds of the loop they enjoyed, played competently, with the CTA reopening the app. Most lapse is drift rather than rejection, and drift responds to a reminder that the thing exists.
Show what changed. If the product has shipped something since they left (a mode, a season, an event, a rebalanced economy) the ad's job is to show it, because the reason they lapsed may well have been addressed. This is the strongest re-engagement creative available and the most under-produced, because it requires the creative to keep pace with the release schedule.
Deliver the incentive. Where the campaign carries an offer (returning-player currency, a free continue, a limited event entry) the interaction should be the act of claiming it. The playable becomes the thing that hands over the reward rather than the thing that argues for it, which is a much easier ask.
The CTA is a different button
This is the change teams most often miss, and missing it wastes the campaign. A UA playable exits to a store listing. A re-engagement playable that sends an existing user to the store sends them to a page with an Open button on it: an extra step, an extra decision, and a measurable share of the audience lost at it.
The re-engagement CTA should deep link into the app, and where possible to the specific destination the ad was about: the new mode, the event screen, the offer. A user brought back to a generic home screen has to find the thing they were promised, and often does not. If the ad demonstrated a choice, the choice should be carried through the link so the app opens on it.
Deferred deep linking still matters, because part of any lapsed audience has uninstalled. The link should route an installed user into the app and an uninstalled one to the store, preserving the context so it survives the reinstall. That is a tracking-link configuration rather than something the creative does, but the creative has to be built knowing which behaviour it is triggering.
Measurement changed on iOS, in a way that helps here
SKAdNetwork was built around install attribution and handled re-engagement poorly, which is a large part of why iOS retargeting was hard to justify for several years. AdAttributionKit, Apple's successor framework introduced with iOS 17.4, explicitly includes re-engagement measurement alongside configurable attribution windows, and it can operate alongside SKAdNetwork rather than replacing it outright.
The practical implication for a creative team is that re-engagement results on iOS are becoming attributable in a way they were not, so the creative decisions in this post are increasingly testable rather than assumed. It is worth confirming what your measurement partner currently supports before designing a test around it.
Audience construction is half the creative decision
A re-engagement audience is not one audience. Somebody who churned after one session and somebody who played for three months and stopped need different ads, and pointing the same creative at both wastes the format's main advantage.
- Early churn: they never reached the interesting part. Show the part they never reached.
- Deep lapse: they exhausted the content. Show what is new; a reminder of what they finished is a reason not to return.
- Recent lapse: they are still in the habit's shadow. A reminder and an incentive is usually enough.
- Payers who stopped: the highest value and the most sensitive. This audience notices being marketed at, and generic creative reads worse to them than to anyone else.
Exclusions matter as much as inclusions. An active player seeing a returning-player offer is an unhappy active player, and an offer visible to everyone is a reason to lapse deliberately.
A worked example: one UA creative, three re-engagement cuts
Take a merge game's UA playable: a hand shows the first merge, the player makes three merges, a big item appears, and the CTA invites an install. For a recent-lapse audience, cut the hand, start on a board with a high-tier item one merge away, and change the CTA to a line about carrying on where they left off, deep linked into the game. For a deep-lapse audience, replace the board with the new season's items and let the single interaction reveal the season's headline reward. For a returning-player incentive campaign, make the merge produce the offer itself and send the CTA to the screen where it is claimed.
All three share the original's art, mechanic and hook. What differs is the starting state, the copy and the destination: exactly the three things the audience segment should decide.
Producing them without doubling the workload
The economics work because a re-engagement variant is a small edit rather than a new build. Duplicate the UA project, remove the tutorial beat, change the CTA copy and destination, and swap the end card. In Playable Ads Maker that is a project duplicate plus two scene edits, exported to the same networks, which is what makes it feasible to keep a re-engagement creative current with a release schedule rather than shipping one and leaving it for a year.
Checks before a re-engagement creative ships
- Does the deep link open the app on the screen the ad promised, on both platforms?
- Does a user without the app reach the store, and does the context survive the install?
- Is every line of install-intent copy gone, including on the end card and the button?
- Does the audience exclude active players?
- Does any offer shown in the ad match what the app actually grants that user?
Create playable end cards in minutes—no code required.
Open Playable Ads Maker