AppsFlyer uses the Google Play referrer parameter with fingerprinting and ID-based attribution technology backup.
AppsFlyer runs NativeTrack™ 2.0, a proprietary, patent-pending set of heuristic algorithms combined with statistical analysis and machine learning. These are combined to analyze multiple data points from the device, fingerprints, traffic, profilers, carrier, application and others, to statistically determine - on an anonymous basis - the likely unique user of the device, app download and conversion.
AppsFlyer assigns an internal unique user identifier to the user profile developed for that individual. This user identifier is not based on a UDID or MAC address, and is completely anonymized.
AppsFlyer also supports ID-based attribution when available, together with fingerprinting to provide the most accurate mobile attribution without compromising user experience or accuracy.
While PC browser and Android device fingerprints are quite unique, the situation with iOS devices is different. AppsFlyer was the first to apply what the industry refers to as device fingerprinting. According to the data we collected over time, we developed a system that goes beyond device fingerprinting. NativeTrack™ is proved to provide the most accurate conversion and installation tracking for native applications.