Journal
Build an honest application inventory before you analyze usage
Usage analytics without a current inventory produces elegant answers to the wrong questions. Here is a lightweight inventory method we use with Korean mid-market teams.
Many organizations keep an application list that mixes retired tools, shadow licenses, and aspirational purchases that never rolled out. Analyzing usage against that list creates noise: phantom apps with zero activity, and real apps missing from the register entirely.
Our inventory pass is deliberately plain. For each application we record owner, business purpose, environment (production versus sandbox), license model, and the source of usage evidence. We mark confidence for each field. Incomplete rows stay visible rather than being silently dropped.
Shadow tools surface during stakeholder interviews. Someone mentions a desktop utility used to clean extracts before they enter the official catalog. That utility belongs on the inventory even if it has no enterprise license record — otherwise usage stories stay incomplete.
Only after the inventory is agreed do we attach usage measures. The sequence feels slower at the start and saves weeks of rework when findings collide with reality during the readout.