Start with a decision you need to understand
Define the research question before recruiting. “Why do operations leaders replace their accounting system?” is useful; “Do buyers like our dashboard?” is already steering toward a feature. Recruit people who recently participated in the relevant decision, including customers, lost prospects, and buyers who stayed with the status quo. Recent memories produce concrete sequences: the trigger, first search, internal conversation, shortlist, objection, and final compromise.
A finance manager who compared an unfamiliar platform with QuickBooks may describe very different risks from the CFO who signed the contract. That difference is not noise. It is evidence that roles shape the buying story. Pair this work with decision-maker research to identify the economic buyer, champion, technical evaluator, and blocker.
Use jobs-to-be-done prompts
Open with context: “Take me back to when you first realized the current process was no longer workable.” Then move chronologically. Ask what happened next, who became involved, what alternatives were considered, and what nearly stopped the purchase. Jobs-to-be-done interviewing treats a purchase as progress a buyer is trying to make under constraints. Functional needs matter, but so do social concerns, personal credibility, and fear of disruption.
Request artifacts when appropriate: a comparison sheet, an approval memo, or the language used in an internal message. Never request confidential data. The goal is to ground the account in observable details and map those details against actual journey touchpoints.
Avoid leading and hypothetical questions
“Would automated reconciliation save you time?” invites agreement. Ask instead, “How did reconciliation work before the purchase?” Follow with “What made that difficult?” Avoid naming benefits before participants do. Silence is useful; give people room to remember. When they use broad words such as “easy” or “powerful,” ask for an example. Do not correct their category vocabulary, even if it differs from your internal terminology.
Hypothetical willingness to pay, future feature use, and recommendation intent are weak substitutes for behavior. Ask what budget was approved, which option they rejected, and what tradeoff they accepted. Record with permission, take sparse notes, and preserve exact phrases separately from interpretation.
Synthesize without overstating
After each conversation, write a short decision timeline and tag evidence by trigger, desired progress, alternatives, friction, stakeholders, and outcome. Compare cases only after individual summaries are complete. Patterns across a small qualitative sample can guide messaging and further investigation, but they do not represent every U.S. buyer.
Mark contradictions instead of averaging them away. A controller may value familiar workflows while an operations lead prioritizes integration. Those tensions can explain stalled decisions better than a neat persona. Use market observation to check whether stated behavior appears in reviews, sales calls, and onboarding, then document the limits of the evidence.