Back to blog Growth

Customer Discovery Interview Questions for B2B

PublishedNovember 20, 2026
Read9 min

Useful customer discovery interview questions reconstruct a problem someone has already experienced. They help you learn what happened, how the person handled it and what would have to change inside their business before a different approach could be considered.

For a B2B founder, the important distinction is between an operator who feels the friction and a buyer who can authorize a change. Both perspectives matter. A painful task does not automatically create an approved project, and enthusiasm for your idea does not establish demand.

This guide stays with pre-offer problem research. Copy the interview-and-evidence sheet, adapt the prompts and leave with a better question to test. If you are investigating a completed purchase or lost deal, use the separate win-loss analysis template.

1. Define the assumption you want to challenge

Begin with one assumption about a particular role, task and consequence. Write it so an interview could weaken it. A broad statement such as businesses need better operations gives you no useful boundary. You could interpret almost any complaint as confirmation.

Hypothetical research assumption: operations leads at small B2B service firms repeatedly rebuild client onboarding information because sales handoffs omit delivery requirements. This creates enough disruption for someone to consider changing the handoff process. This is an invented research scenario, not a client finding.

Split that assumption into questions: does rebuilding happen, what causes it, what consequence follows and who owns the decision to change it? Evidence for the first question cannot answer all the others. A person may redo a task without seeing it as a priority.

Set a disconfirming condition before the conversation. Perhaps the information already arrives reliably, the workaround is acceptable or the problem occurs only in unusual projects outside your intended segment. Keep your ideal customer profile provisional while researching rather than excluding anyone whose experience challenges it.

2. Choose people with relevant recent experience

Screen for involvement in the work, not interest in your proposed solution. Ask whether the person recently performed, supervised or approved the task you are studying. Record when that happened and their role. A senior title does not prove firsthand knowledge.

Include different perspectives deliberately: the person doing the task, the manager accountable for the result and someone familiar with approval constraints. They may be the same person in a small firm. Ask rather than assuming that every company has separate departments or a formal procurement process.

Track how you found participants. Friendly contacts and people already interested in your idea may provide a narrow view. Seek relevant people who use a satisfactory alternative or have chosen to tolerate the problem. Keep those conversations visible alongside enthusiastic responses.

Explain the research purpose, how notes will be used and whether you want to record. Obtain permission before recording and check applicable requirements. Offer a notes-only conversation. Do not request customer identities, credentials, confidential contracts or files the participant is not authorized to share.

3. Ask about the last real occurrence

The NSF I-Corps description of customer discovery connects interviews with potential customers and industry stakeholders to evaluating potential product-market fit and informing a business model. That purpose is broader than collecting compliments about a concept.

Use these original prompts as a sequence, following the participant's account rather than reading every question mechanically:

  1. What is your responsibility in this task, and when did you last do it?
  2. What started that particular instance of the work?
  3. Can you walk me through what you did, from the first input to the finished result?
  4. Where did you have to stop, check something or repeat a step?
  5. What information was missing, and how did you get it?
  6. Who else became involved, and what did each person need?
  7. What happened to the work waiting behind it?
  8. How did you know the task was complete?

Keep the participant's sequence intact. If they describe an approval wait, ask what was being approved and what the approver needed. Avoid supplying your preferred explanation, such as assuming the software caused the wait.

The GOV.UK guidance on in-depth interviews recommends open, neutral questions and specific stories. Applied here, that means asking for the actual incident before asking what an ideal process might look like.

4. Explore workarounds and approval constraints

Customer discovery questions should reveal the current alternative, including doing nothing. Ask what the person tried before, what they kept using and why. A spreadsheet can be an adequate solution. Do not assume an improvised process is something its users want replaced.

  • What do you currently use to get through this step?
  • When did that workaround last fail, if it has?
  • What other work did you postpone to handle the incident?
  • Which consequence mattered to someone outside your role?
  • When your team last changed a comparable process, who had to agree?
  • What information, security checks or internal resources did that change require?
  • What would prevent the team from changing this process even if the problem continued?

Separate a remembered estimate from an observed measurement. If the participant cannot quantify time lost, record UNKNOWN. You can seek a permitted, redacted workflow example later. Do not multiply a vague estimate into a financial impact and present it as established evidence.

Ask the buyer about their own recent approval experience. An operator's account of what a director would approve is secondhand evidence. Conversely, a director's confidence that a task is easy does not erase the operator's account of having to redo it.

5. Copy this customer discovery interview template

Use one sheet per participant and incident, linked to a separate assumption register. The blank fields below are a working template. Keep identifying details and original notes in a restricted location, and use a reference ID in the synthesis sheet.

Copyable interview-and-evidence sheet
FieldWhat to enter
Hypothesis and contrary condition[Role, problem, consequence and evidence that would weaken the assumption]
Participant reference and role[Research ID; operator, manager, approver or other; recruitment source]
Screening criterion[Relevant task performed or approved, incident date and eligibility reason]
Consent and recording status[Agreed note use, recording permission or no recording, access restrictions]
Original prompts[Questions actually asked, including follow-ups that affected the answer]
Past incident and evidence[Participant-reported sequence, observed artifact if permitted, source and date]
Current workaround[What they use, why they keep it and known limitations]
Buyer constraint[Approval owner, required evidence, competing priority or UNKNOWN]
Contradiction or missing evidence[Conflicting accounts, acceptable workaround, unanswered question or UNKNOWN]
Interpretation[Your explanation, author and alternatives; separate from the participant account]
Next test and owner[Specific evidence to seek, responsible person, review date and decision]

During the interview, summarize the incident and invite corrections. Label paraphrases as summaries. Use quotation marks only for words faithfully captured with appropriate permission. Never reconstruct a polished quote from memory.

End by asking what you misunderstood and which role could clarify an unresolved step. Do not turn that closing into a product demonstration. If a separate offer discussion is appropriate, distinguish it clearly from the research session and let the participant choose whether to continue.

6. Separate observation, interpretation and contradiction

Hypothetical B2B role-conflict example: an operator describes rebuilding an onboarding checklist after a sales handoff. A permitted, redacted checklist shows fields added during delivery. Record the account as participant-reported and the visible document changes as observed evidence. Neither proves the cause by itself.

Your interpretation might be that inconsistent intake creates avoidable rework. A buyer interview then suggests that accepting unusual client requirements is a deliberate commercial choice, and no one has assessed a standardized intake policy. The disagreement changes the research question from replacing a checklist to understanding who can constrain exceptions.

A contrary case matters too: another eligible operator uses a simple handoff note successfully and reports no meaningful disruption. Keep that result. Investigate differences in project complexity and responsibilities rather than discarding it because it weakens your original idea.

This example contains no real interview findings or quotes. It illustrates why a problem validation interview should preserve multiple explanations. Avoid turning agreement among a few selected participants into a claim about the entire market.

7. Decide which evidence to seek next

Review assumptions one at a time. Mark each as supported within the observed cases, challenged or unresolved. Attach the underlying incidents and contrary evidence. These are research judgments, not statistical confidence levels or permission to claim a validated market.

Choose the next action from the gap. A missing approval perspective calls for a relevant buyer interview. An uncertain cause calls for observing the workflow with permission. An acceptable workaround may justify changing the problem or pausing the idea entirely.

Use what you learn to revisit your positioning assumptions. When the evidence supports a direction, the go-to-market from zero guide covers the next operating steps. Interview completion alone does not establish demand, willingness to pay or a repeatable acquisition channel.

FAQ

What are customer discovery interview questions?

They are prompts used to investigate a potential customer's actual problems, recent experiences, current workarounds and decision constraints before developing or pitching an offer. Ask about specific incidents and keep participant evidence separate from your interpretation.

What should I ask in a B2B customer discovery interview?

Ask when the task last happened, what triggered it, which steps were difficult, how the person handled them and what consequence followed. Explore who approves a process change and what could prevent it, including an acceptable current workaround.

How many interviews validate a market?

No fixed interview count establishes market validation in this framework. Interviews can reveal patterns and challenge assumptions within the cases studied. Disclose recruitment limits, missing perspectives and contrary evidence, then seek behavioral evidence for the next decision.

Should I pitch my idea during customer discovery?

Keep the problem research session focused on the participant's experience. Presenting your solution can steer later answers toward that idea. If a separate offer conversation is useful, explain the change in purpose and let the participant decide whether to have it.

What belongs in a customer discovery interview template?

Include the hypothesis, participant role, screening criterion, consent status, prompts, past incident, workaround, buyer constraint, contradictions, interpretation and next test. Record sources and dates, and leave unanswered questions as UNKNOWN rather than filling them with assumptions.

Need a clearer problem before building the offer?

Bring the assumption you are testing and a de-identified outline of the evidence. We can discuss how it could inform your next growth decision through business growth consulting.

Book a Call
Nikhil Rai
Written by

Nikhil Rai

I work across strategic partnerships, business development, digital marketing, lead generation and automation, helping teams find opportunities, build relationships and scale.