As a product manager, your data-driven approach is key to understanding why users aren’t completing each step of your funnel. Your app’s heartbeat is the ongoing measurement of three data sets: app performance, user behavior, and business results. Today, we are pleased to announce that Apteligent now automatically collects two performance events that can affect user behaviors but are difficult to track: UI Hangs and Out Of Memory (OoM) faults. The functionality is available for all plans in the latest iOS SDK.

UI Hangs occur when your application freezes up and becomes unresponsive. UI Hangs are not crashes, the app is still running but the user is blocked from achieving their intent within your app. Almost invariably, the user is faced with a decision to abandon the session either to force kill the app or reboot the phone itself. In a way, UI hangs are some of the most disruptive issues for your users to experience. Making matters worse, UI Hangs are not easy to detect and many companies today lack visibility into the extent to which they are an issue.

When your app receives a memory warning from the OS, your app likely reacts by freeing up memory. (If it doesn’t, the OS will kill your app). Unexpected behavior can occur when your app frees memory and the user is still interacting with it. As your codebase becomes more and more complex, it is increasingly easy for a developer to forget to release memory allocated for variables that are no longer being used. Garbage collection systems do not cover all situations. While there is no problem with the functionality of the app so most Quality Assurance tests will pass, the end result on the user is still dire.

In the hierarchy of user experience issues, UI hangs and OoM faults cause failure and frustration, so it is important to understand their impact.

This change requires no new instrumentation on your part and is available for iOS – you can download the new version of the SDK here. Sessions that end with an OoM fault or a UI hang will appear on the Crash Summary page. The events themselves will be logged as automatic breadcrumbs so that your development team can quickly reproduce the issue.

Keep your eyes peeled for business-relevant insights that leverage these new performance data from our data science team. In the meantime, we look forward to your feedback.