Recruit for decision proximity, not friendliness
Start with the role that owns the problem, then expand to the people who can block or fund a change. In mid-market U.S. companies, that often means an operations lead plus someone who reconciles spend. When a controller says “we already do this in QuickBooks,” treat it as a map of existing habits, not a competitor name-drop.
Screen for recent purchase involvement. Ask who joined the last vendor evaluation, who wrote the business case, and who signed. Friendly users who never sat in procurement meetings produce colorful stories with low decision value. nixbux recommends logging proximity to the last buying event as a recruitment filter, not a vanity demographic.
Question design that resists leading
Open with timeline reconstruction: “Walk me through the last time this problem interrupted a week.” Follow with artifact questions: screenshots of workarounds, invoice exceptions, escalation emails. Then probe consequences before exploring solutions. Save category awareness for the second half so you do not train respondents to perform as informed shoppers.
When finance workflows appear, ask how month-end changes if a new tool arrives. Mentions of QuickBooks, bank feeds, or approval queues are invitations to study integration anxiety and training cost — topics that rarely show up in NPS-style prompts. Cross-check those answers later with market observation at real contact points.
Listening for friction language
Buyers rarely say “friction.” They say “we waited on legal,” “finance needed a new class code,” or “the champion left mid-pilot.” Capture verbs and nouns exactly. Exact language feeds later purchase friction analysis and keeps synthesis honest.
Use silence. After a vague claim about “easy ROI,” wait. Many decision-makers fill the pause with the real constraint: risk of disrupting close, fear of dual entry, or skepticism after a failed migration. nixbux treats those admissions as higher signal than polished opinions about brand preference.
Ethics and capture
State purpose, recording consent, and anonymization rules before recording. Separate identifiable notes from synthesis memos. Do not promise outcomes you cannot control. If a respondent asks for a product roadmap commitment, document the request and redirect — interviews are for learning, not negotiation theater.
“When they mention QuickBooks, listen for the workflow underneath — chart of accounts, approvals, month-end — not for a feature checklist.”
After eight to fifteen interviews with decision proximity, move into affinity clustering described in pattern analysis, then draft restrained claims via positioning synthesis. For a method overview, see research methods.