What a Reinstall Is
A reinstall happens when someone who had your app, removed it, installs it again. People reinstall for many reasons: a new phone, freeing up storage and coming back, a retargeting ad, a friend's recommendation, a seasonal need such as travel or tax time.
For measurement, the key question is whether the reinstall is recognized. A recognized reinstall can be reported as a returning user and credited to the campaign that brought them back. An unrecognized one looks like a brand-new user, which inflates acquisition numbers.
How Reinstalls Are Detected
Attribution platforms recognize returning users through identifiers that survive an uninstall:
- Advertising IDs. The GAID on Android, and the IDFA on iOS when the user granted ATT consent, stay the same across reinstalls unless the user resets them.
- Backup data. Apps can restore some data from the device's backup on reinstall, including identifiers the SDK stored earlier, when backup is configured.
- Customer user ID. When the user logs in, the advertiser's own ID links the new install to the existing account.
- Apple redownload flag. SKAdNetwork postbacks mark installs where the same Apple account had downloaded the app before.
On iOS without ATT consent, there is often no device-level way to recognize a reinstall before login, so some reinstalls are counted as new installs. That is a known limit, and it is worth checking login data to estimate how many.
Reinstalls and Attribution
When a recognized reinstall happens, the platform checks whether a campaign brought the user back:
- Was the user inactive for at least the inactivity window?
- Did the reinstall happen within the re-attribution window after the uninstall?
- Is there an eligible click or impression from a campaign inside the lookback window?
If all three are true, the reinstall is credited to that campaign as a re-attribution, and it appears in retargeting reports. If not, it stays with the user's original source or is reported as an organic reinstall.
Why Reinstalls Matter
Clean acquisition metrics. New installs, CPI and new-user cohorts should exclude returning users. Mixing them in makes campaigns look cheaper than they are.
Win-back measurement. Reinstall counts by campaign show whether spend on lapsed users works. Combined with uninstall tracking, they show the full loop: who left, when, and who came back.
Fraud signals. Fraudsters sometimes reset device IDs or reinstall repeatedly on the same hardware to fake new installs. Unusual reinstall patterns from one source are worth reviewing.
Reinstalls in Linkrunner
Linkrunner detects reinstalls with GAID on Android and IDFA on iOS, with Android backup configuration set up in the SDK. On iOS, ATT permission is needed for reinstall re-attribution. A reinstall is re-attributed to a campaign when the user meets the inactivity window, which defaults to 0 days based on the last app open, and reinstalls within the 90-day re-attribution window.
The Retargeting view on the Campaigns page shows reinstalls and re-engagements per campaign, and the reporting API returns the same numbers. Fraud Control's dashboard library includes a template for reinstalls from fraudulent clicks. Re-attributed installs are billed like any attributed install; re-engagements are free. See retargeting and pricing.