Customer interviews that stay with the buyer’s words

What the interview is for

The point is to learn how a purchase is justified inside an account. You are listening for the job the buyer is trying to get done, the people who must agree, and the moment the conversation usually dies. You are not listening for a number you can paste into a board slide. A dozen interviews do not become a representative sample because the team wishes they did. Say what you heard. Name how many conversations you held, in which roles, and over what kind of deal. Then stop.

Marketing research of this kind is closer to a careful debrief than to a tracker study. The value is specificity. “Walk me through the renewal you almost cancelled in March” produces a scene. “How do you think about vendors?” produces a speech the speaker has given before. klulkler favors the scene. Speeches are useful only as evidence of the story a company tells itself, and they should be labeled that way in the claude notes.

Who is worth an hour

Users, champions, economic buyers, procurement, security, and the skeptical operator who will live with the tool after the contract is signed. If you only interview fans, you will write positioning for people who already agree. Decision-maker research is the companion page: it explains why a champion’s vocabulary and an economic buyer’s vocabulary are different documents, even when they work on the same evaluation.

Recruit from a real evaluation, not from a panel that opted into “IT decision-makers” for a gift card. A title on LinkedIn is not proof that the person was in the room when legal sent the contract back. Ask for the last purchase, loss, or renewal in the category. If they cannot name the month, the other vendors, and one colleague who disagreed, thank them and end the call. Polite interviews with spectators feel productive and teach almost nothing about purchase friction.

A sequence that does not lead the witness

1. The last time they bought, renewed, or rejected something in this category.
2. Who else was in the decision, and who could have stopped it.
3. What they compared, including the option of doing nothing.
4. Where the process stalled, and what the stall was waiting on.
5. The words they used internally to defend or kill the choice.
6. What they still do not trust, even if they signed.

Stay with the past or with a deal that is already in motion. Hypothetical questions (“Would you pay more for…”) invite courtesy. Courtesy is not evidence. When a buyer offers a percentage, write it down as something they said, not as a finding. The finding is that they reached for a number. The number itself is not a market result.

What claude notes capture during the hour

While they talk, write the sentence, not your interpretation. Mark hedges: “sort of,” “if legal is comfortable,” “we would need to see it in our tenant.” Mark lines they repeat. Mark refusals, including questions they will not answer about price, risk, or internal politics. Those marks are claude notes. Interpretation comes later, in pattern synthesis, and only if a mechanism shows up in more than one account.

Do not tidy the grammar in the moment. A broken sentence often holds the criterion. “It’s not the price, it’s that I can’t walk this into the Friday staff meeting” is a positioning clue. A cleaned-up version — “the buyer seeks internal alignment” — has already thrown the clue away. Leave the Friday staff meeting in the note.