Crashing Engagement: Savannah Wei Shi’s Research on How App Failures Impact User Retention and Advertising Revenue

We’ve all been there; you’re about to reach a new level of a video game, Venmo your landlord for rent, or find out if the guy in the rom-com gets the girl, and… your app crashes.  

App crashes are extremely prevalent, even with mature and popular apps like Instagram, DoorDash, and Netflix. Apps are constantly updating with new features, fixing bugs, and enhancing the overall user experience to stay competitive in the app market, all of which cause app crashes. Sometimes a crash is so minor that a user may not even register it as an app crash; however, other times, with more popular apps, it becomes national news. 

Having experienced many app crashes themselves, Savannah Wei Shi, Associate Professor of Marketing, her co-author and colleague, Kirthi Kalyanam, as well as Seoungwoo Lee from Yonsei School of Business, and Michel Wedel from the University of Maryland, began thinking about how these crashes affect users’ engagement with apps.

“Previous research provides insights into the impact of a system-wide app crash on consumer purchase. However, the impact of app crashes on consumers’ engagement with apps, which is the main driver of advertising revenue in mobile commerce and widely acknowledged to be crucial for a company’s long-term growth and success, remains unclear,” Shi says. 

Shi and Kalyanam’s research, The Impact of App Crashes on Consumer Engagement, offers a theoretical framework that predicts consumers increase app engagement after encountering a single crash, but reduce their engagement level after experiencing repeated, concentrated crashes. The reasoning behind these findings lie in human psychology.

“Consumers will immediately go back to and even browse more on the app that just crashed because humans have a psychological need for closure, known as the Zeigarnik effect, and they are curious about what was left unseen in their last browsing session,” says Shi. “App crashes are so common for consumers that it’s almost habitual for them to re-open the app when it crashes. However, when consumers are subject to repeated interruptions, especially within a short time frame, they become fatigued with the faulty app and will eventually stop using it.”

Because user engagement drives advertising revenue in mobile commerce, any loss in interaction translates to a direct decrease in ad impressions and, consequently, lower earnings for the company. So, when app crashes are compounded over time, it makes it harder for apps to maintain a steady revenue stream. Without strategically addressing app crashes and managing user experiences apps risk not only losing users but also a significant portion of their advertising budget and potential income.

Shi explains how this pioneering research has real-world applications to help address these concerns:

First, it provides managers with a framework to quantify the economic impact of crashes. For example, when an app developer is looking to roll out a major system update that they believe will cause 70% of the app to crash, using this framework, they are able to estimate how it is going to affect consumer engagement and consequently, advertising revenue.

Second, because the framework allows managers to estimate how each customer will react to an app crash, they can adopt a targeted app release strategy. Knowing that some users are more tolerant of app crashes than others, app developers can target those users with new app features and updates first. By waiting to roll out updates to less tolerant users, after the major kinks are worked out, apps can minimize disruptions to consumer engagement and avoid a potential decrease in advertising revenue, while continuing to release new app features quickly and frequently. 

Third, the experiment shows how app managers can address the loss of engagement due to repeated crashes by strategically designing in-app messages when crashes occur. The research finds that deploying a post-crash message that communicates to customers the cause of an app crash, gives consumers peace of mind that the app developers are aware of and resolving the issue, relieving some of the pressure of the app crash. 

“There are a few companies who are already implementing these practices really well. Namely, Apple with the targeted rollout of Apple Intelligence and the Google Play Store, which now allows app developers to release pop up messages when apps crash,” Shi states. “Yet, many companies aren’t doing this because they don’t see the benefit.” 

Shi hopes that this study will change that and will provide managers with a framework to quantify the economic impact of crashes and offer insights into the benefits of targeted feature releases and strategic design of post-crash messages, all of which have direct impact on a firm’s bottom line. She and Kalyanam also hope to expand on this new dimension of research on app failures to focus on consumers’ reactions to app crashes on different types of apps, such as banking or educational apps, and to continue to elicit greater interest in the business problems posed by app crashes.