Before you start
A compact reasoning spine
- 1Confirm the signal and bound what changed, when, and for whom.
- 2Generate a small set of plausible hypotheses grounded in product behavior.
- 3Prioritize the investigation by decision value, speed, impact, and reversibility.
- 4Choose a next action and state the evidence that would change it.
This is starting structure, not a model answer. The scenario and follow-ups stay unseen until you begin.
What feedback actually checks
Mode-specific practice rubric
Situation framing
Clarifies what changed, when, for whom, and which outcome is actually at risk.
Hypothesis quality
Generates plausible causes grounded in product behavior rather than random brainstorming.
Investigation priority
Chooses what to investigate first based on likelihood, impact, speed, and reversibility.
User & product context
Segments the issue when averages can hide different experiences or root causes.
Decision metrics
Uses the right primary metric and guardrails to decide whether to rollback, fix, or continue.
Communication
Explains a diagnostic sequence rather than narrating every possible thought.
Each dimension receives a visible 0–3 practice signal after the round. It is not a hiring prediction or a company interview scorecard.