Why Interview Experts?
An expert interview provides a focused opportunity to learn how a person with relevant experience understands uncertainty affecting an objective. It can bring out concerns that are difficult to express in a group, details that are not apparent from written material, or perspectives that differ from those already voiced in a workshop. The interviewer’s task is not to confirm a preselected risk. It is to create conditions in which the expert can describe what may happen, why it may happen, and what the consequences could be.
A useful interview is structured but not rigid. The interviewer prepares a consistent set of topics and core questions, while leaving room to follow relevant answers. Structure supports coverage and comparability; flexibility allows the expert to identify matters the interviewer did not anticipate. This is different from asking an informal sequence of questions and recording only the points that seem immediately important.
Plan the Interview Before Asking Questions
Begin with the assessment objective and scope. State what the assessment is intended to understand and what is outside its boundary. Then identify why this expert is being interviewed: for example, the person may have knowledge of a decision, a control, a customer-facing activity, or an area of uncertainty. Do not assume that a job title alone makes someone knowledgeable about every part of the scope.
A short plan should specify the interview’s purpose, expected duration, participants, topics, recording method, and intended use of the information. Tell the expert how notes will be handled and whether comments will be attributed. Follow applicable organizational policies on confidentiality and records. Do not promise anonymity if the assessment process cannot guarantee it.
Prepare a sequence rather than a script to be read without variation. Start with context and experience, move to open questions about uncertainty, then use probes to clarify the answers. Finish by checking the interviewer’s understanding and inviting anything important that has not been discussed. A consistent core sequence helps compare interviews, while follow-up questions adapt to each person’s knowledge.
Name the objective, boundary, and relevant time horizon in plain language. Avoid inserting an expected failure into the scope statement.
Connect the person’s experience to the topic and explain how the information will be used, recorded, and followed up.
Use questions that invite description rather than a yes-or-no confirmation. Order them from broad experience to more specific clarification.
Ask neutral follow-ups when an answer is unclear or consequential. Summarize the interpretation and let the expert correct it before closing.
Ask Open Questions, Then Probe
An open question allows the expert to choose which information to provide. “What uncertainties could affect the objective?” is more open than “Could a supplier delay disrupt the objective?” The second question names a particular cause and consequence; it may be useful later to test that specific possibility, but it can steer the initial discussion toward the interviewer’s hypothesis.
Good core questions are concise, use familiar language, and avoid assuming that an event has occurred or that a control has failed. Examples include: “What has changed in this area?” “Where do outcomes depend on an assumption?” and “What could make it difficult to achieve the objective?” A question should generally address one topic at a time. Compound questions can cause the expert to answer only the easiest part or leave the interviewer unsure which part was answered.
Probes are short follow-ups that clarify an answer without supplying the answer. They are not a checklist that must be exhausted after every response. Select the probe that addresses the actual gap in the expert’s statement.
| Purpose | Neutral probe | What it helps clarify |
|---|---|---|
| Clarify meaning | “What do you mean by ‘unstable’ in this context?” | The expert’s intended description |
| Explore cause | “What conditions could lead to that?” | Possible contributing conditions |
| Clarify event | “What might happen?” | The event or change being described |
| Explore consequence | “If that happened, what could follow for the objective?” | Potential effects and their connection to the objective |
| Understand basis | “What are you basing that view on?” | Experience, records, assumptions, or other evidence |
| Explore uncertainty | “What would make that more or less likely?” | Conditions and limits on the expert’s assessment |
| Check for omissions | “What else should we consider?” | Information not yet raised |
Neutral language matters throughout the interview. Avoid praise or surprise that rewards one answer over another, such as “That’s exactly the problem we expected.” Avoid minimizing a concern with “That probably won’t happen.” If the expert uses an evaluative term such as “rare,” ask what experience or information supports it rather than silently converting it into a likelihood category.
From Answer to Clear Risk Information
An expert may describe a concern in shorthand: “The renewal will be difficult,” “we have a staffing problem,” or “the new platform is fragile.” These statements are useful starting points, but they do not yet explain what could happen or how the objective could be affected. Ask for clarification in the expert’s terms. Where helpful, organize the answer into cause, event, and effect, using the risk-statement approach already introduced in the earlier material on writing cause-event-effect risk statements. Do not force an answer into that format if the expert does not have enough information.
Separate what the expert observed from what the expert inferred. “Three experienced reviewers left during the quarter” is a reported observation; “the review process will fail” is a prediction. Record the source and status of each point. An interview note should not turn an opinion into a fact or a possibility into a certainty.
Worked Example: Interviewing a Distribution Specialist
A hypothetical insurer is assessing whether it can meet its objective of maintaining timely service for policyholders. A distribution specialist has experience with broker submissions and renewal work. Design the opening questions and follow-up probes for an interview. Avoid assuming that brokers are already submitting incomplete information.
Method. State the scope neutrally, begin with the expert’s experience, and use open questions before probing any specific answer. This avoids the common mistake of treating the interviewer’s suspected cause as established fact.
Step 1: Frame the purpose. The interviewer says: “We are looking at uncertainties that could affect timely service for policyholders during renewal work. I would like to understand what you have observed in your role and what conditions could affect that objective. We will record notes for the assessment and distinguish observations from views about what may happen.”
Step 2: Invite the expert’s perspective. Ask: “What parts of renewal work have the greatest effect on timely service?” This is open: it does not assert that one part is already failing or identify a preferred cause.
Step 3: Follow the answer without supplying a conclusion. Suppose the specialist says, “Information arrives in different forms, and some files take longer to resolve.” Ask: “What differences in the information matter?” and “What tends to happen when a file takes longer to resolve?” The first clarifies the shorthand; the second explores possible consequences.
Step 4: Establish the basis and boundaries. Ask: “What are you basing that on?” and “Does this occur in all renewal work or only under particular conditions?” Record whether the answer refers to personal experience, a specific record, or a general impression. Do not infer that all brokers submit incomplete files.
Step 5: Check the summary. Summarize: “I heard that variation in the form of incoming information can extend resolution time for some files. Is that accurate, and what would you change?” The specialist can correct the wording or add a condition. The method produces a clearer candidate risk while preserving uncertainty and avoiding an unsupported generalization.
Keep the Interview Neutral and Useful
A neutral interviewer listens carefully, allows pauses, and does not rush to fill silence with a suggested answer. The interviewer can ask direct questions, but should distinguish a neutral probe from a leading one. A leading question embeds a preferred response, assumes a disputed condition, or frames one outcome as expected.
For example, “Why are the new controls not working?” assumes both that the controls are new and that they are not working. A more neutral version is: “How are the controls operating in practice?” If the expert reports a problem, the interviewer can then ask what happened, under what conditions, and what evidence supports the account. This sequencing allows the expert to introduce the concern rather than merely agree with the interviewer.
Take notes that retain the meaning and limits of the answer. Record key phrases, relevant examples, the source of information, and any uncertainty or disagreement. If exact wording matters, ask permission before recording or quote only what the notes reliably capture. Avoid interpreting silence as agreement. If the expert cannot answer, note that limitation rather than filling the gap with an assumption.
At the close, summarize the main points and ask whether the summary is accurate. Invite corrections and ask what important issue has not been covered. Confirm whether follow-up evidence or clarification is available and agree on a practical next step. A respectful close also helps the expert understand that the interview is about learning, not testing the individual’s performance.
Worked Interview Decisions
Worked Example: Rewriting a Leading Question
A hypothetical claims operation is reviewing uncertainty affecting accurate and timely claim decisions. The interviewer plans to ask a senior adjuster, “The new review rule is slowing claims and causing complaints, isn’t it?” Rewrite the question and specify a sequence of follow-ups if the adjuster raises delay concerns.
Method. Remove the assumed cause, effect, and expected agreement. Then move from open description to clarification and basis. This avoids the common mistake of treating a confirmation prompted by the question as independent evidence.
Step 1: Identify the embedded claims. The original wording assumes the review rule is new, that it slows claims, that complaints result, and that the adjuster agrees. None of those points has yet been established by the interview.
Step 2: Replace it with an open question. Ask: “What, if anything, has changed in how claims are reviewed, and how has that affected the work?” The phrase “if anything” permits the answer that no material change has occurred.
Step 3: Probe only if delay is raised. Ask: “Which part of the review takes longer?” followed by “Under what conditions does that occur?” and “What effect, if any, does it have on claimants or the operation?” Each question addresses one clarification, rather than presuming the next conclusion.
Step 4: Check the basis. Ask: “What observations or records support that view?” If the adjuster describes a few cases, record them as examples rather than generalizing to all claims. The revised sequence elicits a possible risk while preserving the distinction between experience and inference.
Worked Example: Handling a Vague Concern
A hypothetical product team is assessing uncertainty that could affect its objective of releasing a planned policy administration update. An engineer says, “The data handoff is brittle.” The interviewer must clarify the statement without turning the interview into a technical inspection.
Method. Treat the phrase as a cue to ask for meaning, possible conditions, and objective-level effects. This avoids the common mistake of writing “brittle data handoff” as if it were already a complete risk statement.
Step 1: Clarify the expert’s term. Ask: “What does ‘brittle’ mean in this situation?” Suppose the engineer explains that a change in one source field may prevent a downstream system from accepting a record.
Step 2: Explore the condition and event. Ask: “What kinds of changes could cause that?” and “What would happen if a record were not accepted?” Suppose the engineer says that a change to a field format could result in records being rejected until the interface is corrected.
Step 3: Connect the effect to the objective. Ask: “How could that affect the planned release?” Suppose the answer is that unresolved rejected records could delay readiness testing. Record that as a possible effect, not a certainty.
Step 4: Establish basis and limits. Ask: “Has this occurred before, or is this a potential failure mode?” and “What checks or other conditions could affect the outcome?” Record the engineer’s response and any supporting information separately from the possibility described.
The clarified account is that a change in a source field may cause downstream record rejection, which could delay readiness testing if not resolved. The sequence uses the expert’s explanation and does not presume the event will occur. It also stays at the level of eliciting risk; a detailed examination of each operational step is outside the purpose of this interview.
Why the Question Sequence Works
The logic of a neutral sequence is simple: an answer is more informative when the expert supplies its substance rather than merely accepts wording supplied by the interviewer. Begin broadly, allow the expert to select what is relevant, and then narrow through probes that request meaning, conditions, effects, and basis. Each step reduces a different ambiguity.
For example, “The vendor will miss the deadline, correct?” gives the expert a proposed event and invites agreement or disagreement. “What could affect the deadline?” leaves the possible causes open. If the expert mentions vendor capacity, “What makes capacity uncertain?” clarifies the basis; “What could follow if capacity is insufficient?” explores the consequence. This sequence does not guarantee that the account is correct, but it makes the source and reasoning more visible and reduces the interviewer’s influence on the content.
Structured interviews are especially useful when experts have different roles or may be reluctant to raise concerns in a group. They do not eliminate bias: experts can have incomplete information, incentives, different interpretations, or limited experience. The interviewer should therefore record the basis and limitations of an answer, and should not treat one interview as a complete assessment of the topic.
Common Interviewing Mistakes
- Seeking confirmation instead of eliciting. Replace “This is a major exposure, isn’t it?” with an open question about what could affect the objective.
- Asking two or three questions at once. Ask one question, listen, and follow up on the part that needs clarification.
- Suggesting a cause too early. Start broadly; mention a specific possibility only when testing or clarifying it, and make clear that it is a possibility rather than an established fact.
- Turning an opinion into a fact. Record whether a statement is an observed event, a reported belief, an inference, or a future possibility.
- Ending without checking the summary. A brief read-back gives the expert a chance to correct interpretation and identify missing context.
Check Your Understanding
For each situation, choose an interview approach that elicits information without embedding an unsupported conclusion.
- A hypothetical underwriting manager says, “The referral queue is unpredictable.” Give one neutral clarification question and one probe about the possible effect on the underwriting objective. (Answer: “What do you mean by unpredictable in this context?” followed, after clarification, by “What effect could that have on the underwriting objective?”)
- An interviewer wants to ask whether a software supplier’s weak support is delaying a release. Rewrite the question so it does not assume weak support or delay. (Answer: “What factors could affect the timing of the release?”)
- A hypothetical finance specialist says that a reporting change “could create reconciliation problems,” but gives no basis. Give one follow-up on basis and one on conditions. (Answer: “What are you basing that possibility on?” and “Under what conditions could the reconciliation problem occur?”)
- An expert reports that a control “failed twice.” State two points the interviewer should clarify before generalizing from this statement. (Answer: Clarify what “failed” means and the circumstances or scope of the two reported instances; record them as reported examples, not proof that the control generally fails.)
- At the end of an interview, the interviewer has notes stating, “Late information will cause customer harm.” What should be checked before recording this as a supported risk account? (Answer: Confirm with the expert whether late information is an observed condition or a possibility, clarify how it could affect customers and the objective, and establish the basis and limits of the statement.)