The fastest way to build the wrong thing is to start from the solution. BUC Populi started from the pain: professors spending hours marking essays, and the slow-burn worry of students slipping into an attendance-compliance problem nobody noticed until it was late.
Find the pain that actually costs them
Grading was the obvious time sink - hours per stack of essays, every week. The less obvious pain was risk: a missed attendance flag isn't an inconvenience, it's a compliance problem. Both were costing people real time and real stress, and neither showed up in a feature request. You only find them by listening to how the work actually feels.
Users ask for features. Listen for the pain underneath - that's where the product actually lives.
Build to the pain, in order
Once the pains were clear, the product almost designed itself:
- Drafting grades with Claude attacks the time sink, with a human still deciding.
- A weekly attendance audit attacks the risk - flagging compliance exposure before it's a crisis.
- Each feature traces back to a pain someone actually feels, not a capability someone wanted to show off.