What is biased attribution?
Biased attribution happens when an attribution platform has a conflict of interest. This usually happens when the attribution platform is owned by the same company that provides traffic for ads. Such situations exist with Facebook or Google: they provide analytics into the installs their ads are driving but also have a clear interest to show a prettier picture than what really exists.
For example, Facebook uses attribution models that attribute installs to an ad even if a user just viewed it and didn’t tap on it. So, in practice, a user could tap on an ad from a different network, install an app, and Facebook would still attribute that install to themselves (if it happened during the Facebook attribution window).
Why biased attribution is important?
Understanding biased attribution is the first step in getting better visibility into ad performance. Most importantly, it shows that depending too much on a platform that is likely to have a biased attribution problem may skew your analysis results and lead you to the wrong decisions. This can be solved by using a third-party attribution platform such as AppsFlyer, Adjust, or Kochava platform.
Biased attribution and App Store growth
The biased attribution problem unveils a big issue in the mobile marketing analytics space where it’s impossible to see the full picture by analyzing performance in silos.
In the example above, we saw how a simple analysis would point towards Facebook being the top-performing channel for acquiring new installs although it might not be the case.