Why Do Ad Platforms and App Analytics Show Different Mobile Install Numbers?
Airbridge explains install discrepancies through attribution windows and one winning touchpoint, with SKAN on iOS and install referrer matching on Android.
Key Takeaway
- Airbridge uses an attribution window to define how long after a target event any follow-up actions are still considered connected to it, measuring only the events that take place during that period.
- When a target event happens, the model ranks touchpoints from the lookback window by priority and assigns credit to the single winning touchpoint that occurred closest to the target event.
- For OS-specific measurement, the platform supports iOS privacy-centric attribution alongside other methods and uses install referrer matching specifically for Android.
If you are comparing your ad partner dashboards against your app analytics and wondering why your attributed installs and post-install numbers never quite match up, the answer comes down to how each system defines attribution windows, prioritizes ad touchpoints, and processes iOS and Android data. When you are running a lean team, spending hours reconciling spreadsheet numbers is the last thing you need. Understanding the exact rules behind how an install gets credited—and how long that credit lasts—lets you read your reports with confidence and see what is actually driving growth.
Understanding the Attribution Window for App Installs
When you are trying to figure out why two reports show different post-install results, the first setting to check is the attribution window. In the Airbridge Attribution Model, the attribution window sets the exact length of time after a target event during which later actions can still be treated as connected to that target event. The model only measures events that take place inside that active window. If an event happens after the clock runs out, it will not be counted under that target event's window.
To see how this works in practice, consider an app install that happens at 2 PM on January 4 with a 3-day attribution window applied. Under this condition, the model measures the events that occur between 2 PM on January 4 and 2 PM on January 7. Because the cutoff is tied to the exact hour of the install rather than calendar days alone, an action taken at 3 PM on January 7 falls outside that 3-day window.
If your ad channels and your analytics dashboard are set to different time frames, their totals will naturally diverge. You should review your current lookback and attribution window configurations on the Management > Attribution Rules page to make sure your reporting rules match how your users actually behave.
How a Winning Touchpoint Gets Credited for an Install
Ad networks often claim credit for any user who interacted with one of their ads prior to downloading your app, even if the user clicked another channel right before installing. To keep your data clean and avoid double-counting attributed installs, Airbridge follows a step-by-step sequence to select a single winning touchpoint.
- Log user ad interactions as touchpoints: Every time a user interacts with an ad, the system records that event as a touchpoint. In day-to-day campaigns, these touchpoints are typically impressions and clicks.
- Filter eligible interactions using the lookback window: Before any touchpoint can be considered for credit, it has to fall within the lookback window. This window defines how far back in time before a target event an ad interaction can still be treated as related to that event. If any ad interaction happens outside that defined time frame, it is excluded from receiving credit.
- Determine or assign the touchpoint timestamp: To know which interaction happened most recently before the install, the system checks when each touchpoint occurred. However, under Attribution Methods that use install referrer matching, the touchpoint timestamp may be unavailable. Only when the timestamp is missing does the model, across all ad channels, assume the touchpoint occurred one hour before the install event. Knowing this one-hour assumption is crucial when you are troubleshooting why an install referrer touchpoint beat—or lost to—another touchpoint that occurred shortly before the download.
- Rank touchpoints by priority and select the closest one: Once a target event like an install takes place, all eligible touchpoints inside the lookback window are ranked by priority. From those priority-ranked touchpoints, the model credits the single winning touchpoint that is closest in time to the target event. Because only one winning touchpoint receives the credit, your attribution dashboard will often show lower install numbers for an individual ad network than that network's own self-reported dashboard.
How Subsequent App Events Keep or Shift Install Credit
For a subscription app, an install is only the starting line; you need to know which ad touchpoint actually drove subsequent events like trial starts or purchases. In the Airbridge Attribution Model, a subsequent event is any event that takes place within the attribution window after a target event occurs. By default, the performance of that subsequent event follows the exact same touchpoint to which the original target event was attributed.
However, user journeys are rarely a straight line, and a user might click a retargeting link or promo link soon after installing your app. When two target events happen close together, their attribution windows can overlap. As explained in the Attribution Scenarios - Airbridge Help Center, when app attribution windows overlap, a deeplink open takes priority over an app install. Specifically, when a deeplink open happens while the app install attribution window is still active, the deeplink open attribution window applies first.
This priority rule directly changes which touchpoint gets credit for your downstream conversions. For instance, when those app attribution windows overlap and an add to cart event occurs inside the deeplink open attribution window, that add to cart action is credited to the deeplink open rather than the initial app install. If you notice that a post-install event did not get credited to the campaign that drove the original install, checking whether the user triggered a deeplink open in between will clarify why the credit shifted.
Measuring Attributed Installs Across iOS and Android
When you break down your install reports by operating system, you will quickly notice that iOS and Android rely on very different underlying technologies. Airbridge supports both platforms in the Core Plan, which includes Mobile SDKs as well as SKAN 4.0 reporting, a conversion value manager, and SKAN overlap features.
On the iOS side, privacy rules limit how user-level data can be shared across apps and websites. To address this, Airbridge supports SKAdNetwork—Apple’s privacy-centric attribution framework for iOS—running in parallel with other attribution methods. Because SKAN enables app install measurement without relying on user or device identifiers, you can continue measuring iOS campaign results alongside your other attribution models.
On Android, measurement can take advantage of store-level referrer data. Airbridge’s install referrer matching attributes installs using the information carried by the install referrer, a method that is available only for Android. In the standard app store flow, a user clicks an ad to download your app from the store, and Airbridge sends the store a referrer that contains the user’s web ID. When the user opens the app for the first time after installation, that referrer is passed back to Airbridge along with the user’s web ID and app ID, which enables user matching for attribution.
Because Android can pass a web ID and app ID through the store referrer on the first app open while iOS runs SKAN in parallel without user or device identifiers, reports separated by OS will naturally reflect these structural differences.
FAQS
FAQ
How can I verify that an organic iOS app install is being logged properly on a test device?
Under the Testing - iOS SDK guidance, if the app is already installed on your test device, delete it before testing, then download and launch the app and check whether the install event details appear in App Real-time Logs. Keep in mind that event details may take up to 10 minutes to appear in the real-time logs.
Why does an individual ad network dashboard often report more attributed installs than my cross-channel analytics?
An individual ad network typically counts any install that followed an interaction with its own ads, without visibility into whether the user interacted with another channel afterward. Cross-channel attribution evaluates all eligible touchpoints together and assigns credit to a single winning touchpoint, which prevents the same install from being counted twice across multiple networks.
What should I check first when a post-install subscription event is credited to a different campaign than the original install?
Start by checking your user journey logs to see if the user interacted with a retargeting campaign or opened the app via a promotional link after installing. If a new target event occurs while the initial install window is still open, downstream actions can be reassigned to the campaign that drove the newer interaction rather than the original acquisition campaign.
Read clean install data
Measure attributed installs across iOS and Android with the Core Plan.
