Recommended resources
The freely available resources this lesson will direct you to.
Hello. In the previous lesson, you ranked evidence of product–market fit and identified gaps that could change your customer focus or buyer-facing claims. An interview can help explain those gaps, but only if it gets beyond “Do you like the product?” and reconstructs what the institution decided, what changed in practice, and what would have to happen next.
This lesson gives you a reusable guide for conversations with operations and collections leaders at lending institutions. It covers four things you need to test: adoption drivers, realized value, unmet needs, and expansion potential. Allow about 40 minutes to study the short resources and adapt the guide to your next customer conversation.
Start with an evidence question, not a list of product questions
Take the two most important gaps from your evidence ledger. Turn each into a question the interview could help answer. For example, “We do not know why one institution widened deployment while another remained in a pilot” is more useful than “We need feedback.” It tells you whom to speak with and which decisions to reconstruct.
The UK Government Service Manual’s Using in-depth interviews explains how to turn research questions into a discussion guide while leaving room for follow-ups.
Using in-depth interviews - Service Manual - GOV.UK
Read the UK Government Service Manual’s guidance for a practical way to structure a conversation without turning it into a rigid questionnaire.
In “Design the interview,” read the guide design steps, from the list beginning “review the research questions” through the explanation of what a discussion guide records. Then, in “Do the interview,” read the interviewing advice, stopping before “After the interview.” Notice the distinction between prepared starter questions and follow-ups prompted by what the participant actually says.
For an existing product, interview people whose decisions and experience differ: someone responsible for a durable deployment, someone involved in a stalled or limited rollout, and, where possible, the leader who considered wider use. Speak with both an operations and a collections perspective when they own different parts of the decision. Do not count two enthusiastic people at one institution as evidence from two institutions.
Before each conversation, write down the uncertainty you are testing and what could challenge your current belief. If you think staffing capacity drove adoption, for instance, an interview may reveal that the decisive factor was consistent handling of cases—or that capacity mattered to users but not to the budget owner. Both findings improve your message.
Y Combinator’s How To Talk To Users demonstrates how open questions about present workflows and follow-ups yield more than asking whether someone would use a proposed feature.
How To Talk To Users | Startup School
Watch “How To Talk To Users” from Y Combinator for its examples of questions that uncover work, motivations, and existing responses to a problem.
Watch the question examples. Notice the move from “What is difficult?” to “Why does it matter?” and “What do you do now?”, as well as the warning against leading or hypothetical feature questions. The speaker is discussing discovery with prospective users; with an existing customer, you should still be clear about which product you are researching, while avoiding a pitch during the interview.
A 35-minute guide you can use with an existing customer
The timings below protect space for all four topics. Treat the questions in bold as starters, not a script to recite. Ask one at a time, follow a concrete story when it emerges, and shorten a later section rather than interrupting an important answer.
Opening and context — about 5 minutes. Explain that you are trying to understand the institution’s experience, including what has not worked, rather than seeking an endorsement. Ask permission before recording or requesting internal figures; a note-taker or written notes are alternatives. Confirm the person’s role and the portfolio or workflow they can speak about.
“I’m reviewing how institutions decide to adopt and use this workflow, and where it does or doesn’t create value. Specific examples—including frustrations—are useful. Is it all right if I take notes?”
Then ask: “Which parts of the collections process are you responsible for today?” This establishes whether the person observed daily use, authorized a purchase, or both.
Adoption drivers — about 7 minutes. Reconstruct a past decision, starting before the product entered the process.
“Take me back to the period when your team decided to try a different approach. What was happening in collections then?”
Useful follow-ups are: “What were you doing instead?”, “What made the issue important at that point?”, “Who pushed for a change?”, and “What nearly prevented the rollout?” If this person joined later, ask who can describe the original decision rather than inviting a guess. For a stalled rollout, ask where the decision stopped and what the institution continued doing instead.
The answer should distinguish a trigger—perhaps rising queue volume or a change in operational priorities—from an approval condition, such as confidence in oversight. The person who felt the pain may not be the person who could authorize change.
Realized value — about 8 minutes. Move from the hoped-for benefit to an observed change, with a period and a scope.
“Since the workflow began handling eligible cases, what has changed in your team’s work?”
Follow with: “Can you describe a recent instance?”, “Which queue and period are you comparing?”, “How did you check that change?”, and “What else changed during that period?” If the leader cites saved time or improved results, ask where the underlying measure lives and who could verify it. Do not press for a number they do not have in front of them. Record “reported improvement; not yet verified” rather than silently turning an impression into an ROI claim.
Strategyzer’s Customer Gains image is a useful prompt when a leader initially defines value too narrowly. Its questions cover savings, quality, ease of work, and other desired outcomes. A desired gain is not evidence that it has occurred.
Unmet needs — about 6 minutes. Ask about ordinary exceptions, not just a wish list of features.
“Tell me about a recent case or situation where the current workflow did not meet your team’s needs.”
Probe for what happened, how often similar situations arise, who handled the workaround, and the consequence if nothing changes. If someone requests a feature, ask, “What happened that made you need it?” This separates the underlying operational problem from a proposed design. Ask which issue they would address first and why; “all of these matter” does not tell you what constrains use.
Expansion potential — about 5 minutes. For an active customer, investigate an actual internal decision or discussion rather than asking for a speculative forecast.
“What is the most recent discussion you had about using this workflow in another queue or portfolio?”
Ask what prompted it, which coverage was considered, who would need to support it, and what evidence or change they requested. If no such discussion has occurred, record that plainly. You can ask what currently determines whether a new queue receives attention, but do not interpret a courteous answer to “Would you expand?” as an expansion signal. For a customer whose rollout has stalled, use this time instead to understand what would have to be resolved before existing use could continue or deepen.
Close — about 4 minutes. Summarize the account in neutral terms: “I heard that the rollout addressed [problem], but [constraint] still affects [part of the workflow]. Is that accurate?” Invite correction. Ask which operational or finance owner could help check a disputed measure, and agree on any follow-up without turning the research conversation into an unannounced sales pitch.
Lean B2B’s B2B Customer Discovery Interview Questions offers additional prompts on business priorities, purchase decisions, and the inputs to an ROI story. Use it to refine your follow-ups, not to add every question to a 35-minute call.
B2B Customer Discovery Interview Questions for B2B Entrepreneurs
Read the Lean B2B question table for prompts that connect an operational problem to priorities, decision-making, and measurable effort.
In the “TOPIC / SAMPLE QUESTIONS / WHAT IT TELLS YOU” table immediately before “How to Use the B2B Customer Discovery Questions,” scan the question table. Focus on the rows “Business drivers,” “Decision-making power,” “Buying process,” and “Calculation of Return on Investment.” Select only follow-ups relevant to your evidence gaps; notice that estimating effort is a starting input, not proof of realized savings.
Make the answers comparable without flattening the conversation
Use the same four topics across institutions, but adjust the follow-up to the person’s vantage point. A head of collections can often describe queue handling, exceptions, and changes in day-to-day outcomes. A head of operations may be better placed to describe staffing, cross-team handoffs, rollout decisions, and what must be true for wider deployment. Neither should be asked to certify a measure they do not own.
After the call, separate four kinds of notes:
| Record | What to capture |
|---|---|
| Observed action | A rollout, continued use, workaround, funded extension, or decision to stop—with its date or period |
| Reported outcome | The person’s description of value, including the queue, comparison period, and any other changes they mention |
| Verification needed | A metric, document, or owner that could confirm or challenge the account |
| Your interpretation | The adoption driver or gap you infer; keep it separate from the person’s words |
This separation matters when you later explain the company to a buyer. “A collections leader reported fewer manual interventions in one queue” and “the workflow reduced operating cost” are different claims. The first may accurately report an interview; the second needs a defined baseline and supporting data.
A short review immediately after each conversation will make the next one better. Note which question produced a specific event, which produced only a general opinion, and whether you heard something that contradicts your starting hypothesis. Keep a common core across calls so you can look for patterns by institution, but revise unclear wording when a question repeatedly fails.
For a quick hands-on pass, adapt the guide before your next meeting: mark one high-priority gap from your evidence ledger, choose the starter and two follow-ups most likely to test it, and read them aloud once. If a question contains your hoped-for answer—“How much time did we save?”—make it neutral: “What happened to the time spent on this work?”
What to carry forward
A strong customer interview guide has a clear evidence objective, a logical flow, neutral starters, and probes for specific decisions and recent events. It tests why the institution adopted the workflow, what value it can substantiate, where workarounds remain, and whether broader use has moved beyond interest into an internal decision. Keep reported benefits, verified results, and your own interpretations distinct.
The next lesson will use what you learn about the ideal customer and its alternatives to write a positioning statement. Interview language can make that statement sharper—but only claims supported by customer behavior and evidence should make it stronger.