What Does Agile and Lean Mean for Customers? An Analysis of Mobile App Crashes

Abstract

This research provides a modeling framework to quantify the economic loss from app crashes in terms of users and usage, and to suggest and evaluate policies to minimize that loss. We focus on the number of page-views as a key metric of engagement, given its close tie with advertising – the main revenue source of most apps. We collected consumer usage data of a mobile app and applied a non-homogeneous Poisson Process Copula that accounts for page-view loss due to consumer attrition, prior crashes, and truncated browsing sessions when the app fails. The results do not reveal “digital stockpiling” behavior despite the supply shortfall, but rather, a downward adjustment of consumer engagement after experiencing an app crash. A relatively low clumpiness of prior crashes leads to greater consumer attrition and a smaller number of pages viewed in subsequent sessions. Greater prior usage experience, longer time passed, and updates of the operating system attenuate the negative impact of recent app malfunctions. App publishers can apply our framework to estimate the loss in page views and advertising revenue due to app crashes, and more importantly, to identify the optimal scope and timing of app-update release to mitigate those losses in various situations, which is illustrated via several policy simulations.