Retargeting
User acquisition finds new people. Retargeting goes back to the ones you already paid to acquire, and for subscription apps that often means lapsed trials and cancelled subscribers.
Updated
Most users who install a subscription app do not pay, and many who pay eventually cancel. Retargeting, also called re-engagement, spends ad budget on these known users instead of strangers. The appeal is obvious: they already know the product. The risk is that you pay for people who would have come back on their own.
How app retargeting works
- Define a segment from your own data, for example users who started a trial but did not convert, or subscribers whose plan expired in the last 60 days.
- Send that segment to ad platforms as an audience, or describe it with signals the platform can match.
- Show ads that give a reason to return, usually with a deep link that opens the relevant screen in the app rather than the home screen.
- Attribute the return, the purchase or the resubscription, and compare it with users who were not retargeted.
ID-based vs contextual retargeting after ATT
Classic retargeting matches users by device identifier. On Android the GAID still makes this broadly possible. On iOS, App Tracking Transparency limits the IDFA to users who opt in, so device-ID audiences for iOS are much smaller and match rates drop.
- ID-based and first-party audiences. Device IDs where consent allows, or hashed emails and phone numbers collected with permission, uploaded to platforms that support customer lists.
- Contextual retargeting. Instead of targeting a person, you target placements and moments where your lapsed users are likely to be, such as keywords, content categories or apps they use.
- Owned channels. Email, push notifications and in-app messages reach known users without any ad identifier, and they cost far less. Use paid retargeting for the users those channels cannot reach.
On the App Store, Apple Ads can exclude or target users who already have your app through its audience settings, so you can decide whether a campaign aims at new users, returning users or both.
Win-back offers for lapsed subscribers
An ad alone rarely changes a cancellation decision. What brings subscribers back is a reason: a discount, a new feature, a better plan. Apple supports win-back offers for auto-renewable subscriptions, and discounted offers can be shown on a paywall that matches the message of the ad.
Measurement caveats
- Overcounting. Retargeted users are already engaged, so attribution credits ads for returns that would have happened anyway. Use a holdout group to measure incrementality.
- Overlap with owned channels. A user who got an email and saw an ad may be counted twice. Deduplicate before comparing channels.
- Wrong success metric. Opens and sessions are cheap signals. Judge retargeting on resubscriptions, revenue and how long returning subscribers stay.
- Cannibalizing renewals. Discounts shown to subscribers who were not going to leave reduce revenue. Keep active subscribers out of win-back audiences.
Related terms: churn, push notifications and deep linking are covered in the adapty.io glossary; incrementality and GAID have their own entries here.