Measurement and attribution

AppLovin’s attribution system connects ad engagement to conversion events like purchases. AppLovin uses this system for its reporting and to optimize its models.

Attribution reporting: Clicks vs. Clicks and views

Ads Manager includes an Attribution reporting setting that controls which conversions appear in your reporting. You'll find it on the reporting, campaign, and creative set pages. There are two modes:

• Clicks and views: counts transactions that happened after a click, or within one day of viewing an ad. This gives you a more complete picture of your campaigns' impact, including users who saw your ad but didn't click through, and is consistent with industry-standard default reporting on other major ad platforms
• Clicks: includes conversions attributed to clicks on AppLovin's ads that result in a visit to the promoted website

Switching modes recalculates your reporting view, including historical data. This setting only affects how data is reported. It doesn't change how your ads are optimized, targeted, or billed. After switching, you can save it as your default view so it stays selected every time you come back to the page.

Reporting for web promotion campaigns

AppLovin reports conversions using the following windows:

• Day 0 (D0): Purchases that occur in the 24 hours following the attributed ad engagement (click vs. click and view)
• Day 7 (D7): Purchases that occur in the 192 hours following an AppLovin ad click

In the campaign or reporting tab, you can select columns for most metrics on a D0 or D7 basis. For many metrics longer windows are available, such as D14 and D28, with which you can see the longer term impact of ad clicks.

Optimization for web promotion campaigns

AppLovin offers models that optimize for either D0 or D7 metrics. This setting tells the model which conversion window to prioritize; it doesn't restrict which windows you can view in your reporting.

Why your numbers may not match other tools

In-platform attribution is based on clicks (and, if enabled, same-day views), so it won't capture every conversion AppLovin's ads may have influenced. A few things to keep in mind if your numbers differ from another measurement tool:

• Click-based attribution misses non-click conversions. A user who sees an ad without clicking, and later converts through a search, a different channel, or a different retailer, won't be attributed to AppLovin in-platform, even if the ad played a role.
• Different tools measure different things. MTA (multi-touch attribution) and MMM (media mix modeling) tools are built to capture parts of this picture that click-based, in-platform reporting isn't designed to see, so it's expected for their numbers to diverge from AppLovin's dashboard.
• The gap between in-platform and controlled-test results is growing. AppLovin's internal incrementality-test data shows the gap between in-platform reported ROAS and controlled-experiment (geo holdout) iROAS has more than doubled over the past year.
• Triangulate rather than relying on one number. Using in-platform reporting, MTA, MMM, and incrementality testing together gives a more complete read than any single source alone.

If your reporting doesn't match a third-party tool, this is expected and doesn't necessarily indicate a tracking issue. For more on interpreting differences across measurement approaches and on running incrementality tests, see Making sense of AppLovin through a measurement lens.