Expert Interviews for iGaming Decisions

Discussion with specialists can clarify industry knowledge that documents leave unresolved.

This guide explains how to plan and assess expert interviews; it does not report interviews conducted by this publication. For an online gaming team in New Zealand, an interview can help map how a safeguard or supplier process works. It cannot establish how common that practice is or whether it reduces harm. Check claims against records, regulator guidance or other sources.

The key is not simply speaking to someone with an impressive job title. A useful expert interview connects a defined research question with the right subject matter expert, a disciplined discussion and a transparent analysis of the evidence. Otherwise, a persuasive opinion can easily be mistaken for a market-wide fact.

What are expert interviews?

Expert interviews are a form of primary, qualitative research in which researchers speak with people who have specialised knowledge of a sector, process, technology, community or policy issue. The interviewee may be an industry expert, technical specialist, senior manager, regulatory professional, academic, former practitioner or other key informant.

Unlike a general consumer interview, the emphasis is often on professional knowledge: how decisions are made, how a value chain operates, what has changed in market dynamics and which operational constraints are not visible in secondary research. This includes tacit knowledge gained through experience as well as explicit facts that can be checked against other sources.

Expert interviews may be structured, unstructured or, most commonly for exploratory work, semi-structured . A semi-structured interview uses a discussion guide with consistent core questions while allowing probing questions and follow-ups when unexpected insights emerge. Expert consultations and expert calls are related terms, although a paid consultation does not automatically constitute a rigorous research study.

When should you use expert interviews?

Industry specialist explaining a complex topic during an interview.

Use expert interviews when the central uncertainty concerns mechanisms, context or judgement , rather than the prevalence of an opinion across a population. They can inform strategic decision-making, but the correct method depends on what must be established.

Do not use interviews as a substitute for representative measurement. If the decision requires knowing what percentage of New Zealand customers prefer an option, a properly designed survey or other quantitative data is more appropriate. Interviews can help formulate the survey questions first.

Expert interviews, surveys, focus groups and document review

These methods answer different questions and are often strongest in a mixed-methods research design.

• Expert interviews: uncover explanations, industry insights, tacit knowledge and decision pathways. Their strength is depth; their weakness is limited statistical generalisability.
• Customer interviews and voice-of-customer research: reveal experiences and unmet needs from the user's perspective. A technical expert cannot speak for every customer.
• Surveys: estimate patterns across a defined population when sampling and questionnaire design are appropriate. They rarely explain complex causes without supporting qualitative evidence.
• Focus groups: show how views develop in a group and how participants react to each other; group dynamics can suppress dissenting opinions.
• Desk research: uses existing reports, statistics, policy documents and other secondary sources. It is efficient for established facts but may miss recent or undocumented changes.

A practical sequence is desk research to map existing evidence, expert interviews to investigate gaps, then surveys, administrative data or field observation to test whether the emerging explanation holds more broadly. This is triangulation , not an assumption that agreement between experts proves a claim.

How to conduct expert interviews

1. Turn the business problem into focused questions

Start with a research brief that states the intended decision, hypotheses, target population and evidence gaps. “Understand the New Zealand construction market” is too broad. “Identify which approval and procurement steps delay adoption of a specified building technology” identifies a mechanism that interviewees can discuss.

Separate questions requiring firsthand experience from those better answered by published statistics. Record what would change the decision, what would count as disconfirming evidence and which findings require further validation.

2. Recruit experts for knowledge, not seniority alone

Use purposive sampling to identify participants with experience relevant to the precise research question. An executive may understand strategy while an operations manager, distributor, buyer or frontline specialist better understands day-to-day constraints. Snowball sampling, where participants refer other informed people, can help reach a hard-to-find expert group. It can also reproduce the same professional network and its biases.

Screen candidates against a short, explicit set of eligibility criteria:

• Direct involvement in the process, market or decision under investigation, with a relevant timeframe.
• Ability to distinguish personal observation from second-hand knowledge or speculation.
• Coverage of a needed perspective: supplier, buyer, regulator, practitioner, community stakeholder or independent analyst.
• Potential conflicts of interest, confidentiality restrictions or incentives that may colour the account.

Build a sampling matrix across relevant roles and segments before scheduling. If every interviewee comes from one supplier network, the research cannot support conclusions about the entire value chain.

3. Prepare a semi-structured interview guide

Organise the discussion guide around decision-relevant topics rather than a long questionnaire. Begin with the interviewee's responsibilities and recent experience; then move to process, specific examples, counterexamples and likely changes. Avoid leading formulations such as “Why is the new system better?”

Useful open-ended interview questions include:

1. “Walk me through the last time this decision was made. Who was involved?”
2. “Where does the process usually slow down, and what causes that delay?”
3. “What differs between larger organisations and smaller operators?”
4. “Can you describe a case that contradicts the usual industry explanation?”
5. “What evidence would convince you that this market assumption is wrong?”

Probing questions include:?What happened next??,?How often have you observed that?? and?Was that your decision or someone else's?? These help distinguish a specific example from a general claim. Pilot the interview protocol with one suitable participant and refine unclear questions without removing the core topics needed for comparison.

4. Conduct the interview and maintain an evidence trail

Remote video calls, telephone interviews and in-person interviews can all work. Select the format based on participant access, sensitivity and the need to observe materials or workflows. An interview of roughly 20-60 minutes may be suitable for many studies, but the appropriate duration depends on the topic and participant.

Explain the research purpose, intended use of findings, recording arrangements, participation conditions and confidentiality limits. Obtain informed consent and separate permission for any audio or video recording. During moderation, distinguish direct observation, interpretation and prediction in your notes. Capture contextual details without collecting unnecessary personal information.

After each session, write a short fieldwork memo: new evidence, contradictions, unanswered questions and any revisions required for the next interview. Use accurate transcription where detailed analysis is necessary, with secure access to recordings and interview transcripts.

5. Analyse themes rather than collecting memorable quotes

Begin with a codebook tied to the research objectives, then add new codes as genuinely new patterns appear. Transcript coding can group evidence by barriers, decision criteria, market trends, operational risks and stakeholder differences. Thematic analysis identifies recurring explanations; content analysis may be useful where systematic categorisation is needed.

Look for consensus and divergence . Three experts agreeing may reflect a shared professional background rather than independent confirmation. Compare accounts across roles, organisations and locations. Preserve contradictory cases and record whether a claim reflects direct experience, expert interpretation or a forecast.

For every consequential finding, maintain a compact evidence trail: the claim, participant role, relevant excerpt or observation, alternative explanation, supporting secondary source and remaining uncertainty. This supports qualitative synthesis, a defensible research report and actionable recommendations.

6. Test the finding before acting on it

Translate each insight into a decision rule. For example, if multiple procurement specialists describe integration costs as a barrier to adoption, test that proposition against supplier documentation, customer interviews or a targeted survey. Specify what would falsify it. A finding becomes useful for business strategy when it changes a choice, not simply when it sounds plausible.

How many expert interviews are enough?

There is no universal minimum. Sample size depends on the complexity of the research question, participant diversity, level of disagreement and whether further interviews reveal new material. Data saturation refers to a point where additional interviews provide little new relevant information; theoretical saturation concerns whether the emerging explanation is sufficiently developed. Neither should be treated as an automatic stopping rule.

A systematic review by Hennink and Kaiser reported saturation in 9-17 interviews in many studies with relatively homogeneous populations and narrowly defined objectives. That range is not a universal target for B2B market intelligence, multi-stakeholder policy research or culturally diverse settings. Segment coverage and contradictory evidence may matter more than reaching a round number.

Plan an initial range, review findings after each batch and stop only when the remaining uncertainties are unlikely to change the decision. If perspectives differ materially between Auckland-based corporate buyers and regional operators, for instance, continue recruitment in the underrepresented segment rather than declaring saturation across the whole market.

Common limitations and ways to reduce bias

Expert interviews can produce high-value contextual knowledge, but they carry distinctive risks:

• Selection bias: accessible experts may not represent the relevant knowledge landscape. Use a sampling matrix and seek dissenting perspectives.
• Confirmation bias: researchers may pursue only claims that support the initial hypothesis. Include deliberate counter-questions and document contrary evidence.
• Recall and reputational bias: interviewees may misremember events or frame their decisions favourably. Ask for recent examples and corroborating records where appropriate.
• False precision: an expert's market-size estimate is still an estimate. Validate quantitative claims independently.
• Confidentiality and compliance risk: participants may be asked about protected client data, trade secrets or other restricted information. Do not solicit material they are not permitted to share.

Researcher reflexivity also matters. Record how the interviewer's background, assumptions, incentives and relationship to the participant may affect the discussion or interpretation. Strong qualitative research is transparent about limitations rather than presenting every quote as equal evidence.

Expert interviews in Aotearoa New Zealand

New Zealand research can involve businesses, central and local government, public-sector agencies, non-governmental organisations, sector associations, iwi, hapū and other community stakeholders. The relevant participant mix must follow the actual decision and affected groups; it should not be reduced to an Auckland-only corporate sample by default.

When research concerns Māori people, communities or data, consider appropriate engagement from project design onward, including who defines the questions, who interprets the results and who controls access and reuse. Māori data sovereignty is not satisfied simply by adding a Māori participant to a pre-set interview schedule. Te Mana Raraunga's principles address authority, relationships, obligations, collective benefit, reciprocity and guardianship. Their practical application should be determined with relevant Māori partners and project context, rather than assumed to be identical for every study.

Where applicable, Te Tiriti o Waitangi obligations and institutional research ethics requirements also need to be considered. Community engagement is not interchangeable with interviewing a technical expert, particularly when the research could affect people whose interests or knowledge are being represented.

Research teams handling identifiable interview information should assess their duties under the Privacy Act 2020 , including collection purpose, transparency, security, retention, disclosure and any relevant indirect-collection notification requirements. The Privacy Commissioner's guidance notes that Information Privacy Principle 3A took effect on 1 May 2026. Recording consent and privacy-law compliance are related but separate matters; research teams should check the specific obligations relevant to their data and project.

What belongs in an interview report?

A useful final deliverable is more than a transcript bundle or presentation of quotations. It should make clear what the evidence supports, where uncertainty remains and what decision follows.

• Research design: objectives, hypotheses, interview formats, recruitment criteria, role coverage and exclusions.
• Evidence map: key themes, participant segments, supporting observations and divergent views, with appropriate anonymisation.
• Triangulation: checks against desk research, surveys, operational data or independent sources.
• Decision implications: what to proceed with, what to test, what to reject and which risks need mitigation.
• Limitations: sampling gaps, potential conflicts, unresolved questions and boundaries of generalisability.

Before commissioning or approving a study, ask whether the deliverable will distinguish observed facts , expert interpretations and research hypotheses . If those categories are blurred, stakeholders may act on a confident narrative that has never been validated.

Frequently asked questions

Do expert interviews provide primary or secondary evidence?

They are primary research because researchers gather information directly from participants. Reports and publications used to corroborate interview claims are secondary sources.

Can expert interviews represent a population?

No. Purposive expert samples are designed to access specialised knowledge, not to estimate population percentages. Quantitative surveys or other suitable datasets are required for population-level measurement.

Should interviewees be paid?

An incentive may be appropriate to compensate for specialist time, especially when recruitment is difficult. Disclose the arrangement, apply consistent eligibility standards and assess whether payment creates a conflict of interest or influences responses.

How do expert networks differ from interviews?

No. An expert network may help recruit and schedule knowledgeable participants, while the research team remains responsible for the question design, ethics, interview moderation, analysis, validation and interpretation.

How do you know whether the insights are trustworthy?

Check whether the expert had direct relevant experience, how the account was documented, whether contrary views were sought and how the claim was triangulated. A transparent evidence trail is more convincing than the number of quotations included.

From expert knowledge to a defensible decision

A professional interview should produce a decision trail that can be checked against other evidence.

Start with one decision that cannot yet be made confidently. Identify which uncertainty requires specialist experience, recruit for complementary knowledge, ask for concrete examples and check the explanation against independent evidence. The result should be a small number of well-supported insights, clear limits and specific next steps-not simply more interviews.