How should you run B2B demand generation customer interviews?

Decision brief: The business choice the interviews are intended to inform.

Outcome cohort: Interviewees grouped by whether a purchase was won, lost or left undecided.

Purchase episode: A specific past buying process recalled by a participant.

Neutral prompt: An open question that lets someone describe events without being led.

Recording boundary: The agreed arrangements for recording, storing and using interview data.

Personal data: Information in a recording that identifies a living person.

Pipeline label: A sales-stage or qualification term used to describe progress and blockers.

Name the decision that research must settle

A decision brief identifies the business choice these conversations need to inform. State the product line, purchase context and buyer roles that can describe it. This gives recruitment a boundary and tells the interviewer which evidence matters, rather than inviting a general discussion of demand generation.

Keep the remit narrow enough for the conversations to inform a real choice, such as changing pricing, revising onboarding copy or choosing sectors to pursue. The audience research guide can help organise wider evidence around the defined scope, without turning these interviews into a review of unrelated offers.

Map outcomes before contacting people

Define the outcome cohorts before sending invitations, so every account can be compared by how its purchase concluded. Separate closed-won, closed-lost and stalled or no-decision outcomes; each group reveals a different part of the buying process and helps show which account stories remain missing.

Invite signed customers, buyers who selected another supplier, and accounts where a decision stopped or ended without action. The useful participant is the person who remembers evaluation obstacles, not necessarily the most senior contact. Keep the product line under study consistent across these groups so their accounts can be compared.

For one product line, use six to ten interviews for won accounts and six to ten for lost accounts, with four to eight for stalled or no-decision accounts. Treat these as planning ranges rather than quotas. Deal scale and the number of buyer roles can affect whether further conversations are likely to add useful evidence.

Compare the cohorts side by side: won accounts can reveal triggers and proof points, lost accounts can expose exclusion, and stalled accounts can surface internal blockers. Look for explanations that could change the decision, rather than filling every planned slot. Repeated objections may mean further recruitment adds little.

Use the ranges as starting points, not fixed targets. Deal scale and the number of buyer roles can change how many conversations are useful. Review what each interview adds as recruitment proceeds, then favour accounts that may explain something not yet captured over participants likely to repeat an existing point.

Reconstruct a purchase episode

Let the purchase episode guide the conversation through a real, recent buying process rather than a prediction about future behaviour. Ask who took part, what the participant consulted first and what nearly stopped progress. A remembered sequence gives the interviewer events to examine instead of opinions about an imagined purchase.

Use a neutral prompt to ask what the participant saw, in what order, and what happened next. Avoid asking for a rating or confirmation of your theory. The buyer personas resource can help name roles in the process, while a pause after a question gives the participant room to supply detail.

Before the call, assign one person to ask questions and another to capture notes. Where practical, use an interviewer who did not manage the account, especially for a lost deal. A founder or marketer can listen without defending an earlier sales interaction, keeping attention on the participant's account of events.

Set the recording boundary in advance

Agree the recording boundary before recruitment, rather than improvising the arrangements at the start of a call. Settling the approach early lets each invitation explain what participation involves and gives the interviewer a consistent process to follow across the conversations.

Explain what will be captured, where files will be kept, the retention period, who may listen, and the intended uses. Obtain written consent in advance, and decide the lawful basis and subject access process before scheduling, so each invite can describe the same arrangement.

A voice or recording that identifies someone is personal data. For a formal review, consult statutory wording on personal data and the provision on processing lawfulness. Make the arrangements clear, and consider where the interviewee is based as well as where the business is established. The data protection guidance can help review the handling plan.

Check who can access recordings, where they remain stored and when files will be deleted, rather than leaving those choices to each interviewer. Requirements can depend on the circumstances, so one policy may not suit every programme. Set the process before inviting participants and apply it consistently across the interviews.

Carry buyer language into pipeline labels

Use pipeline labels to transfer a buyer account into shared sales and marketing definitions. Tag each interview by funnel stage, then compare the participant's language with the organisation's stage and qualification criteria. The purpose is a vocabulary teams can use, not a transcript collection disconnected from the decision.

If several lost buyers describe an unexpected security review, consider whether qualification criteria should capture it. Tie any revised terms to the operating workflow so teams can apply them in daily work. The measurement and reporting guide offers a separate frame for connecting decisions with reporting.

Keep the record beside the choice that prompted the calls. Note which outcome cohorts were represented, which objections recurred and which buyer phrases changed a stage definition. This comparison helps sales and marketing interpret progress and blockers together, while avoiding the assumption that one interview describes the whole market. For questions better answered at larger scale, the survey tools comparison describes a separate research route.