A Method for Mobile Download Conversion Rate Measurement based on Device Fingerprinting

Users download mobile applications after being drawn to the application stores, including referrals from advertising campaigns on websites. To determine if a mobile application install originates from a specific campaign, mobile devices need to be uniquely identified before referral to the mobile application store and after a successful install. However, the mobile sandbox environment makes it impossible to exchange device identifiers between a mobile web browser and other mobile applications on the same device. This paper introduces an alternative approach that makes it possible to create identifiers based on measurable device characteristics. The proposed use of device fingerprinting allows to uniquely identify devices even across multiple mobile applications and regardless of the mobile operating system. The presented comparison algorithm is capable of finding two independently created identifiers that were measured on the same mobile device, which makes it possible to determine successful mobile application installs that originated from a specific website.