Before scaling demand, learn who recently bought, what event made them look, what they compared, what evidence settled the decision and why suitable prospects said no. Use interviews and deal records together. The goal is a testable buyer hypothesis, not a lengthy persona document.
Reconstruct recent decisions
Choose a few recent customers who represent the market you want more of, plus suitable prospects who did not buy. Ask what changed before they began looking, what they tried first, who was involved and what almost stopped the purchase. Compare answers with CRM notes and observed product behavior. The interview questions guide can help keep the conversation factual.
Separate fit from trigger
Firmographics describe who might be able to buy; a trigger explains why they act now. Write both. A company may fit by size and industry yet have no urgent problem, while a smaller account with an active migration may be ready to evaluate. Demand generation aimed only at job titles can miss this difference.
Turn research into a test
Draft one statement naming the buyer, trigger, problem and credible proof. Show it to sales and to someone who recently bought. Ask what sounds wrong and what question remains. Then test it in one channel with a defined success signal. The research is useful when it changes a message, targeting rule or offer—not when it fills a deck.
A bounded example
A hypothetical workflow tool assumes operations leaders buy because they want “automation”. Interviews reveal that recent buyers acted after a compliance deadline made manual review risky. That finding changes the page headline, proof needed and distribution list. It does not prove every operations team has the same trigger; the next campaign tests whether the pattern holds beyond the interview sample.
A short research plan tied to one campaign decision
Choose the exact decision the campaign requires, such as which use case to lead with or which buying trigger to target. Select recent buyers and suitable non-buyers who can speak to that question. In each conversation, ask for a concrete timeline: what happened before the search, what alternatives were considered, who joined and what nearly stopped the decision. Avoid asking respondents to predict whether they would click a future ad; recollection of actual behavior is more useful.
After each call, write the evidence in the respondent's language and distinguish observation from your inference. Compare cases for a repeated trigger or objection, then test the tentative message with a small campaign or sales conversation. State what finding would make you revise the hypothesis. If the interviews disagree, segment by context rather than averaging away the difference. Research is complete enough for the next test when you can explain why a specific buyer would act now, what proof they need and what you still do not know.
Frequently asked questions
How many customer interviews are enough before a campaign?
There is no universal number. Start with a small set of recent, relevant wins and losses, then look for repeated patterns and contradictions before spending more.
Can analytics replace buyer interviews?
Analytics shows observed behavior in your systems but rarely explains the buyer's internal alternatives, trigger or risk. Combine it with conversations and deal notes.
Are campaigns being planned from an ICP slide nobody can defend?
Bring recent wins and losses. We can identify the buying evidence that should shape your next demand test.
Talk through the problem ↗See how we could work together
