Around 3 per cent of the installs flowing through Linkrunner in the last 30 days are not new installs at all. They are reattributions: existing users whose app reopen was credited to a fresh ad touch because their previous attribution window had closed.
That number is up from the 1.6 per cent we measured in April. It is also large enough that any team running active retention spend should treat it as a line item in their CAC math, not as a footnote on the install report.
This post lays out the benchmark, explains what a reattribution is and is not, and shows you the dashboard cuts that separate genuine new installs from credited returnees.
What a Reattribution Actually Is
A reattribution is an install event credited to a new marketing campaign when an existing user returns to the app after their previous attribution window has closed. It is distinct from a brand new install (a first-time user) and from a reengagement (an in-app action by an active user).
The mechanics:
- A user installs the app on day 1 attributed to a Meta campaign
- The attribution window on that install closes after 7 days
- 60 days later the user clicks a Google retargeting ad and reopens the app
- That reopen registers as a fresh install event with the Google campaign attached
Two things follow from this. First, the install column on your dashboard mixes net new users with credited returnees by default. Second, the reattributed user already counted toward your CAC the first time around. Counting them again inflates total installs and deflates effective CAC.
For context on how attribution windows shape this whole picture, our attribution windows guide walks through the click and view window mechanics that decide when a reattribution kicks in.
The Benchmark: Around 3 Per Cent of Installs Are Reattributions
Headline figures from the rolling 30-day window ending 29 June 2026:
- Total install events processed: around 2.3 million
- Reattribution events: roughly 73k
- Reattribution rate: 3.22 per cent
- Projects represented: around 145
Methodology note: every install event in the 30-day window is included. The is_reattribution flag on the install record is the source of truth. No filtering by vertical or scale.
Three observations the headline does not capture:
- The figure is up roughly 2x from the April baseline of 1.61 per cent. Re-engagement spend is rising across the customer base, not flat.
- The distribution is not even. Apps running active reactivation budgets contribute disproportionately, while apps without retargeting see near-zero reattribution rates.
- Three per cent is the population average, not the target. Any individual app should compare against the band that matches their re-engagement spend, not the headline.
Where the Reattribution Number Sits in Your Reports
Most MMPs expose reattributions through one of three surfaces. None of them are on by default.
- A toggle on the install report that splits new installs from reattributions
- A column called something like Re-Engagement or Reattributed Installs
- A separate report dedicated to retargeting campaigns
The reporting trap to watch for:
- Install totals on the executive dashboard usually blend both
- Channel-level CAC calculations often divide spend by blended installs
- Cohort retention curves can double-count users when reattribution is counted as a new cohort entry
Platforms like Linkrunner expose the is_reattribution flag directly inline with install totals, so the split is visible in the same view rather than buried inside a separate retargeting report. If you live in raw exports, the hidden cost of inaccurate attribution covers the downstream impact when blended installs flow into budget reallocation decisions.
What 3 Per Cent Means for Different App Types
The blended 3 per cent figure hides large differences across app types. Use the bands below as a sanity check against your own number.
- Subscription and trial-led apps: typically 5 to 8 per cent. Long renewal cycles plus active win-back campaigns push the rate up. Treat anything above 10 per cent as worth investigating.
- High-frequency utility apps: typically below 2 per cent. The reattribution window rarely closes because users keep the app active. A rate above 3 per cent often signals an over-long reattribution window.
- Gaming and casual apps: typically 3 to 6 per cent depending on retention curves. Hardcore apps with multi-month sessions sit lower; hyper-casual apps with frequent churn sit higher.
- eCommerce apps with seasonal re-engagement: spike-shaped distributions, not flat. Expect 6 to 10 per cent during sale windows and 1 to 2 per cent in quiet months.
For a wider framing of how reactivation spend shows up in measurement, our attribution for re-engagement campaigns guide covers the channel cuts (push, retargeting, email, WhatsApp) where reattributions accumulate.
Auditing Your Re-Engagement Spend Against the Benchmark
The most actionable use of the benchmark is a sanity check on the proportion between re-engagement spend and reattributed installs. Two patterns to look for:
- Re-engagement budget overstates its impact. If your retargeting line items claim 20 per cent of total installs but the reattribution flag on those installs sits at 3 to 5 per cent, your channel is taking credit for users who would have returned anyway. This is the cousin of paid-organic cannibalisation, where paid campaigns capture credit for organic users; the paid-organic overlap guide covers how to measure the overlap.
- Reattribution rate is suspiciously high. If your reattribution rate sits well above the vertical band, audit the reattribution window setting in your MMP. The default 90 days or longer often credits routine app reopens to ad touches that had no causal role. A practitioner pattern worth flagging: across attribution audits on apps running active retention budgets, reattribution rates above 5 per cent often indicate the reattribution window has been left at a vendor default rather than tuned to the app's true reactivation cycle.
A quick decision rule: tighten the reattribution window until your reattribution rate sits inside your vertical band, then validate that retargeting campaigns still show measurable lift on user behaviour and revenue. If they do not, the spend was buying reattribution credit rather than genuine reactivation.
Adjusting Your CAC and ROAS Math
Once you can see the reattribution flag, the CAC and ROAS math should change in two specific ways.
- CAC denominator: use net new installs only, not blended install totals. Reattributed users were already paid for once. Including them lowers reported CAC but does not lower true CAC.
- ROAS numerator: revenue from a reattributed user attached to a fresh campaign often overstates that campaign's first-touch performance. Split ROAS by install type before making reallocation decisions.
Reporting cut to set up this week:
- Install volume, split into new and reattributed columns
- CAC, computed on new installs only, with reattribution-CAC reported separately
- ROAS, computed on new installs and reattributed installs separately, then a blended figure for context
The how attribution data powers retention marketing guide shows how this split feeds into churn prediction and lifecycle marketing once the data is exposed.
How to Validate This in Your Own MMP
The minimum-viable cut works in any MMP that exposes a reattribution flag on install records.
- Export installs from the last 30 days with project id, install date, campaign, and is_reattribution flag
- Compute reattributions divided by total installs as a percentage
- Compare against the 3 per cent population average and the band that matches your vertical
- Drill into reattribution rate per campaign to find outliers
Red flags worth investigating:
- Blended reattribution rate above your vertical band: reattribution window is likely too long
- A single retargeting campaign accounting for the bulk of reattributions: the campaign is probably running on a recently-installed audience that would have returned anyway
- Reattribution rate trending up week over week without a matching spend increase: the window setting may have changed inadvertently
Tools like Linkrunner make this cut easy to maintain daily because the flag sits inline with the install table rather than behind a separate retargeting report.
FAQ
What counts as a reattribution in mobile attribution?
A reattribution is an install credited to a new campaign for a user who already had the app previously. The previous attribution window has closed, and a fresh ad touch within the reattribution window earns credit for the reopen.
Is a reattribution the same as a re-install?
No. A re-install is a user who uninstalled the app and downloaded it again. A reattribution can happen even if the app was never removed; the user simply returned after the original attribution window closed.
What is a healthy reattribution rate?
Across the Linkrunner customer base in the 30 days to 29 June 2026, the population average is around 3 per cent. Subscription apps typically sit higher (5 to 8 per cent), high-frequency utility apps sit lower (under 2 per cent). Compare against your vertical band rather than the headline.
Should reattributions be counted in CAC?
No. The user was already paid for the first time around. Compute CAC on net new installs. Track reattribution-CAC separately so retention spend has its own efficiency metric.
How do I configure the reattribution window in my MMP?
Most MMPs default to 90 days or longer. Tune the window down until your reattribution rate sits in the band that matches your vertical, then validate that retargeting campaigns still show downstream behavioural lift. If they do not, the window is buying credit rather than causing reactivation.
What to Do Next
The highest-value action this week: pull your reattribution rate from the install report, compare against the 3 per cent population average and your vertical band, and audit any campaign attributing more than 10 per cent of its installs to reattributions. That usually surfaces either a window misconfiguration or a channel taking credit for organic returnees.
If you would like to see the new and reattributed split alongside your CAC and ROAS in one view, request a demo from Linkrunner and we will walk you through how the reattribution flag flows into the dashboard for your project.