You Asked Until You Heard Yes

You Asked Until You Heard Yes — Volume I, Understanding Humans Before Products · Seeing Reality · Investigation · Foundations. Maya books six interviews to confirm that teams want a shared inbox, asks every one of them whether it would help, gets six yeses, and learns that a question nobody can answer no to has told her nothing

1. 01 · Opening

   You Asked Until You Heard Yes

   Maya books six interviews to confirm that teams want a shared inbox, asks every one of them whether it would help, gets six yeses, and learns that a question nobody can answer no to has told her nothing

2. 02 · The setup

   Maya has a hunch: teams want a shared inbox. She books six interviews to confirm it.

   Maya thinks: Six calls and I'll have my answer.

3. 03 · The setup

   Priya: What are you actually asking them?

   Maya: Whether a shared inbox would help their team.

4. 04 · The evidence

   Six interviews. Six yeses. Maya starts writing the summary before the last call has finished.

   Interview tally — Interviews run: 6; Said it would help: 6; Said it would not: 0; Asked what they'd give up for it: 0

5. 05 · The evidence

   Maya: Six out of six. Nobody said no.

   Priya: That's the part that should worry you.

6. 06 · The evidence

   Dev: You asked if it would help. Everything helps, if it's free.

7. 07 · The evidence

   A question that can only be answered yes has told you nothing

   "Would this help?" costs the person answering it precisely nothing. They are not choosing it over anything, not paying for it, and not living with it. Agreement is the cheapest thing a user can give you, so it is the least informative.

   Maya's sample was fine. All six were the right people. The failure was upstream of the sample: the question had one available answer, so six interviews produced one data point repeated six times.

8. 08 · The evidence

   She rereads her notes. Every question named the feature first. Not one asked what it would replace.

   Maya thinks: I wasn't testing the idea. I was introducing it.

9. 09 · The evidence

   The Leading Question Test (1.0)

   Could they have said no? — If no is not a realistic answer to your question, you are not measuring demand. You are measuring politeness.

   What does yes cost them? — Money, time, a workflow they already have, a tool they already pay for. A yes that costs nothing predicts nothing.

   Did you name the solution first? — Ask about the last time the problem happened. If you say the feature's name, you have handed them the answer.

10. 10 · The evidence

    Maya: Let me ask it the other way round.

11. 11 · The evidence

    Same six teams. New question: what would you stop paying for to get this?

    Interview tally — second pass — Interviews run: 6; Would drop a paid tool for it: 1; Already solved it in a spreadsheet: 4; Wanted it, would not swap anything: 5

12. 12 · You make the call

    Six of six said the shared inbox would help. What did that actually establish?

    The story does not tell you first.

    A. Strong demand — six out of six is unanimous

    B. That the idea is not objectionable

    C. Nothing, because six interviews is too small a sample

13. 13 · What happened

    Maya: Four of them built it in a spreadsheet already.

    Priya: That's worth more than six yeses.

14. 14 · Complete

    You Asked Until You Heard Yes

    Agreement is free, so it is worthless. Ask what a yes would cost them. Next: what to do when the number moves and you cannot tell what moved it.