Coaching

Checks on your questions, your calls and your price, and the questions Mind asks you

Updated 1 Nov 2026 · 5 min read

On this page
  1. Before a call: check your questions
  2. After a call: the call check
  3. Blind spots in your beliefs
  4. Price guardrails
  5. Mind asks you one question at a time
  6. In Ask Mind
  7. Costs
  8. Limits

Coaching is a set of small checks that catch the common mistakes founders make in customer research. It is spread through the places you already work in. Rules do most of it, at once and for free; two reads by Mind cost 1 AI credit each. It is for people who can edit the product. Coaching never changes a belief, a quote, a tag or a price by itself.

Before a call: check your questions

Learn › Questions › a belief › Prepare. Write the questions you plan to ask, one per line. Each line is checked as you type against six rules, free and with no AI:

  • asking about the future ("Would you…", "Would you pay…");
  • naming your product;
  • words that tell them how to feel ("Isn't it frustrating…");
  • a "…right?" on the end;
  • two questions in one;
  • telling them what others said first.

Each problem comes with the reason and a better question about the past ("Tell me about the last time…"). Use it takes the suggested fix, or Rewrite with Mind asks for Mind's own rewrite (1 credit). Below, What your questions cover shows whether you ask how they handle it today, what it costs them, who else is involved and what they have tried, and whether you ask the two score questions word for word. Keep these as the script saves them.

After a call: the call check

Learn › Conversations › a conversation. Once a transcript is captured and who spoke is confirmed, the Call check counts, by rule and free:

  • who talked more, you or them;
  • when you first named the product, described it or named a price;
  • which of your questions led;
  • the opportunity score, when both score questions were asked and answered with a number. The scale they repeat back ("on a scale of 1 to 10…") is never read as their score, and Mind never guesses a number.

A conversation made from notes has no talk time or pitch to count. Capture the transcript, with their consent, for a call check.

Let Mind read it (1 credit) adds what you missed (a complaint nothing covers, a commitment they offered, a moment they changed their mind) and whether the quotes you logged say what you tagged them as. Every quote is their exact words, linked to the transcript, and Mind reads only their turns. If the transcript changes afterwards (a speaker swap, a redaction) the read is hidden until you press Read it again, which is free when their turns did not change.

Blind spots in your beliefs

On Learn › Questions and on a belief's page, by rule and free, each with a question to ask yourself: a belief written about yourself, an exact number nobody has said back to you, only safe beliefs, no belief about price, how people find you or whether you can build it, a test line that is too easy to reach, and a "what it blocks" that only repeats the belief.

Price guardrails

Market › Positioning › Pricing. Enter your monthly cash costs, your hours, what an hour of yours is worth and the customers you expect. You get a floor from your costs, a ceiling from what people said about money in your conversations (each quote with its conversation) and a test price between them. No AI, and never an outside benchmark.

  • Add your own guess at a ceiling until a buyer names one. It is labelled Provisional: your guess.
  • When one sentence names both prices, only the one called too much sets the ceiling. A tool they name ("We use FreshBooks") counts as what they use today even when no price is said. Money counts only when a line says it with a money word ("$40", "forty euros").
  • Make it a pre-sale test drafts a test at the test price as a plan you review. You set its line before you start it.

Mind asks you one question at a time

Mind keeps a list of what it does not know about your product: who it is for, your stop rule, what people use instead, your price, who owns an area, the line for an important belief, an onboarding question you skipped. It asks the most useful one: one on Today a day, up to three in the check-in, one line in your digest, and in Ask Mind when your question touches it.

Answer in your words, or press Not now, I don't know or Not relevant. An answer is shown as a change to review, saved only when you apply it, with Undo for 24 hours. A price is kept with the period you say ("$120 a year" stays yearly); a price with no period is kept without one, and the number field means a month. Not now twice puts a question away for a week, I don't know brings it back in a month, and Not relevant never.

The quarterly check-in adds a clarity check: three honest lines on what is true now, what you have stopped believing and what you are avoiding. They are kept with the check-in and are not scored.

In Ask Mind

Type / for three moves: Check my questions, Audit this call and Price it. The last two quote your conversations, so they are for editors; a viewer's Ask refuses them. Mind brings coaching in only for coaching questions ("what should I charge?", "are my interview questions leading?"), not for any question that happens to say "price" or "audit".

Costs

Rule checks, the call check's counts, blind spots, price guardrails and the questions Mind asks are free. Mind's rewrite of a question costs 1 AI credit and so does Mind's read of a call (Growth and Scale: Included). A credit comes back if Mind cannot answer.

Limits

  • The question rules catch common patterns, not every one. A flag is a prompt to think, not a verdict.
  • Coaching does not listen to calls live; it works on transcripts you capture.
  • Your own lines are never counted as evidence.

See also Questions and beliefs, Conversations and People and Positioning.