Depending on the issue and available access, I’ll examine the relevant parts of the system — such as frontend behaviour, backend execution, database access, logs, profiling data, architecture, and code — to understand where time is actually being spent.