The Win-Interview Question That Doesn’t Plant the Answer

A buyer says an AI assistant helped build the shortlist. The statement is useful, and its value depends on how the buyer came to say it. If the interviewer asked, “Did AI help you choose us?”, the buyer was handed the idea along with the question. People tend to answer the question they are asked, and a question that names a channel invites the channel back. The account is not worthless, but it is a different kind of evidence than one the buyer volunteered.

A better opening is open and anchored in time: “Before we first spoke, what helped you decide which vendors to consider?” It mentions no channel. It points at the period before the vendor was involved, which is when a shortlist forms. And it leaves room for any answer, such as a colleague’s recommendation, an analyst report, a search, an event or an assistant. Ask it before AI comes up anywhere else in the conversation, including in the interviewer’s small talk.

Follow-ups come next, in order. Start with “What else?” until the buyer runs out. Then try “What did you do first?” and “Where did you look?” Only if the buyer mentions an assistant, whether ChatGPT, Claude, Gemini, Perplexity or another, ask which one, what they asked it, and what it helped with: building a list, comparing options, checking a claim. Ask last whether they would be willing to share the conversation, and accept a no without pressure. A saved conversation shows what the buyer actually saw. Without it, the account is a recollection of an answer.

If the buyer never mentions AI, record that, and consider holding any direct question until the end. A direct question such as “Did you use an AI assistant at any point?” is legitimate, but it produces a different kind of evidence. Label it as prompted and keep its answers apart from the unprompted ones. Mixing the two makes a count look stronger than it is.

Record what was said in the buyer’s words, not a summary. A short template helps: the date, the deal, who conducted the interview, the exact wording of the question, the buyer’s answer, any assistant named, what it helped with, and whether the buyer shared the conversation. The interviewer matters. A buyer may answer one way to the salesperson who closed the deal and another way to someone neutral, so note who asked. Timing matters too, since recollection of how a shortlist formed changes with distance from the decision, and the record should say how long after the decision the interview took place.

Use the same wording every time. If the question changes from one interview to the next, the answers are no longer comparable, and changed wording amounts to a different test, not a repeat. Store the exact sentence with the template so that anyone running the next interview uses it. If the wording must change, note the date it changed and treat the earlier interviews as a separate group.

An illustrative row, with no real buyer behind it, shows how little the template needs. Date, deal and interviewer filled in. Wording: the exact sentence used. Buyer’s answer in their own words: “A colleague mentioned two names, and I looked at a couple of review sites.” Assistant named: none. Conversation shared: not applicable. That row is as valuable as one that names an assistant, because it shows what the shortlist rested on, and a set of such rows is what makes a later count meaningful.

Buyers who chose another vendor are a different group. Asking them the same question is possible but harder, since they have less reason to talk. If you do, keep their answers separate, because a count among buyers who chose you and a count among buyers who did not answer different questions.

What can the answers support? A single unprompted mention supports a narrow statement: this buyer recalls an assistant as part of how the shortlist formed. A set of interviews with the same wording supports a count, reported with its denominator: how many interviews took place, how many buyers were reachable, and how many mentioned an assistant unprompted. A count of mentions does not show that the assistant caused any sale, and an interview in which AI never came up does not show that the buyer never used it. Absence of a mention is not evidence of absence.

Interviews also have their own limits. They reach buyers who agreed to talk, they capture what people remember, and a small team may have few of them. None of that is a reason to skip the question. It is a reason to state, alongside any number, how it was collected.

The question costs a minute, and the record costs a few more. In return, the team stops arguing about whether a buyer was led, and starts working with evidence whose origin it can describe.

FAQ

What is a good first question for a win interview about AI? “Before we first spoke, what helped you decide which vendors to consider?” It mentions no channel and points at the period when a shortlist forms. Ask it before AI comes up anywhere in the conversation.

Why not ask the buyer directly whether AI helped? A question that names a channel invites the channel back, so the answer partly reflects the question. A direct question is acceptable at the end, if it is labeled as prompted and kept apart from unprompted answers.

What should I ask if the buyer names ChatGPT, Claude, Gemini, or Perplexity? Ask which assistant it was, what they asked it, and what it helped with. Then ask whether they would be willing to share the conversation, and accept a no.

Why use the same wording in every interview? Changed wording is a different test, not a repeat. Using the same sentence keeps the answers comparable across interviews.

What does it mean if the buyer never mentions AI? Record it as such. It does not show that the buyer did not use an assistant, only that it did not come up unprompted.

Is this article reporting a result from Axis Suite? No. It describes a way to ask and record, and reports no result for any brand.

About Axis Suite Axis Suite is the independent intelligence layer that tracks what AI says about your brand and the observable evidence behind it, retrieval, narrative, recommendation, and corroboration signals, across engines including ChatGPT, Claude, Gemini, and Perplexity. Built by TrendAxis, Axis Suite works from a framework that separates Mentioned, Cited, Recommended, and Chosen. Axis Suite is measurement infrastructure, not a marketing quick fix. Explore the Proof Center.