Success metrics: did it work

Ask what happens if we do not build it, then who uses it and how often. If those have no answer, it does not get scheduled.

Task: The feature shipped with no way to judge it.

[Before you start, do three things]
1. Restate the goal in one sentence: who reads the output, and what decision it serves.
2. If I have not given you the following, ask me. Do not assume:
   - the goal
   - trackable behaviour
   - baseline
   - observation window
3. Say how you plan to do it (at most 5 steps) and wait for my go-ahead.

[Go through each of these]
- one primary metric and two guardrails
- baseline and target
- short-term metric versus long-term value
- what cannot be measured

[Process]
1. Reconstruct what the user actually does
2. Decide whether this is a need or a proposed solution
3. Offer build, do not build, or another way
4. Write checkable acceptance criteria
5. Mark the risk and reversibility

[Output]
Problem, judgement, options, acceptance, risk.

[Do not]
- Never copy the requester's solution
- No scheduling without a usage scene
- No acceptance criteria like 'improve experience'
- Never three features at once