Hi everyone,
I’m exploring ways to apply process mining concepts to analyze performance and user interaction flows in a small digital product, and I’m curious how others here approach similar use cases outside of traditional enterprise processes.
In my case, the product is an interactive application where user actions (inputs, retries, session restarts, drop-offs) generate a high volume of short-lived events. While this isn’t a classic order-to-cash or procurement process, the challenge feels similar: identifying where friction, delays, or bottlenecks occur in real usage flows.
Some of the questions I’m currently working through:
-
How to best model non-linear user journeys where loops and retries are expected
-
Whether it makes sense to normalize event logs when actions happen in milliseconds rather than seconds
-
Techniques for identifying performance-related drop-offs using event sequence analysis
-
How others distinguish between “expected variation” vs actual inefficiencies in highly interactive processes
For context, one of the test projects I’m analyzing is a lightweight, performance-sensitive application where timing and responsiveness directly affect user behavior (example environment: https://cuphead-apk.com/). The goal isn’t the content itself, but understanding how interaction data can be structured to reveal meaningful insights.