What Double Counting Is
Double counting is what happens when two or more sources each take credit for the same install or purchase. Each ad network reports conversions in its own dashboard, using its own data and its own rules. When you add those dashboards together, the total can exceed the number of conversions that actually occurred.
It is most visible with self-attributing networks such as Meta and Google, which each measure their own results and do not see each other's ads.
How It Happens
- Overlapping audiences. A user sees a Meta ad, then clicks a Google ad and installs. Both networks have a legitimate claim under their own rules.
- Different windows. One network credits clicks for 7 days, another for 30, so the same install falls inside both. See attribution windows.
- View-through credit. One network counts an ad view as a conversion trigger, while another counts only clicks. See view-through attribution.
- Repeated events. The same purchase is sent twice, for example from the app and from the server, and is recorded as two revenue events.
What It Does to Your Numbers
Double counting inflates reported installs, revenue and ROAS for each network. The sum of network-reported conversions is higher than the number your backend records, and every channel looks more efficient than it is. Budget moves toward whichever dashboard claims the most credit, not the one that drove the most users.
How to Remove It
- Measure every channel with one independent source of truth that applies the same rules to all of them.
- Let each install be credited to one source, with clicks taking priority over impressions.
- Record each payment once, using a unique payment ID, so a repeated event does not add revenue twice.
- Compare networks on that deduplicated view, and treat network dashboards as in-platform optimisation signals.
Linkrunner credits each install under a single rule set and records payments once per payment ID, so the channels in its reports add up to the installs and revenue actually recorded.