Playbook5 min
The discovery questions that actually change the deal
Most discovery calls collect facts nobody uses. Twelve questions that change what happens next, why each one works, and the four habits that quietly ruin the call.
Product managers and researchers
Leave the interviewwith something you did not know.
A user interview goes wrong quietly. You ask the user to agree with you, they oblige, and you leave with a validated belief you walked in holding. That is visible in the transcript and almost never visible in the room.
Apple Silicon · free forever for individuals · no account · v0.1.9
User interviews, coached while you run them
What it saysin the moment
Cues arrive during the call, and a cue that cannot point at the line in the transcript it is reacting to is discarded in the backend before it reaches you. These are the shape of the ones that survive.
That was a leading question. Ask what they did last time instead of whether they would use it
They asked for a dashboard. Nothing yet on what they do with the number once they have it
You have confirmed the belief you arrived with twice and tested it zero times
A workaround was described in passing and never priced
Live signals
The scores that move as the call moves
- Pain intensity
- How much this actually costs them, not how much they say it annoys them
- Requirement clarity
- Whether what they want is now specific enough to build
- User urgency
- Would they change anything about their week over it
- Workflow understanding
- How much of the surrounding process you now know
- Feature relevance
- Whether the thing they asked for addresses the thing they described
The scorecard afterwards
Five dimensions, each with its quote
- Problem probing
- Did you dig into the workflow behind the request, or take the feature ask at face value?
- Listening
- Did you let the user describe the problem in their own words?
- Testing assumptions
- Did you check a belief you walked in with, or only confirm it?
- Non-leading questions
- Did you ask what happened, or ask the user to agree with you?
- Next step
- Is there a concrete follow-up, a doc or a prototype or another session, with a date?
A dimension the call never tested is left unscored rather than quietly given a zero, because a zero you did not earn teaches you nothing.
About 40 seconds after you hang up
The review, on one screen
A scored rubric, a verdict and the one next step that matters. Everything else is a tile you open, and the page underneath never moves.
Post-meeting brief
Tell it what you are trying to achieve
Name what you are trying to find out and it will notice when you stop trying to find it out.
You pick one of these when you set the meeting up. It becomes the thing the call is measured against, and the reason a cue can say you have not got there yet:
- Extract user pain points
- Discover requirements
- Concept or prototype feedback
- Churn or cancellation interview
- Stakeholder alignment on scope
What it is not
It does not synthesise across interviews or build an insight repository. It works one conversation at a time.
Worth reading before the next one
All posts
Free, and it takes two minutes
No account, no card, no bot in your next call. Download it and run one meeting through it.
Apple Silicon · free forever for individuals · no account · v0.1.9