Running interviews and focus groups: recruiting, questioning and listening for useful evidence (Research and Evidence for Better Decisions) | READOS

Some questions cannot be answered with numbers. Why did a long-standing customer stop ordering? How do maintenance planners actually decide which jobs to do first? What frustrates new staff in their first month? What would make a buyer trust a new supplier? Interviews and focus groups answer questions like these. They reveal reasons, processes, language and priorities that surveys cannot capture, and they often uncover issues nobody thought to ask about.

They are also easy to do badly. An interviewer who talks too much, asks leading questions or invites people to imagine their future behaviour collects polite opinions rather than evidence. Interviewing only friendly customers, or only the staff a manager nominates, produces a comfortable but distorted picture. A focus group dominated by one confident participant reflects that participant’s views, not the group’s. Notes written days later record what the interviewer remembers, which is usually what they expected to hear.

This article explains when to use interviews and focus groups, how to recruit participants, how to write a discussion guide, how to ask questions that produce reliable answers, how to listen and probe, how to run focus groups, how to record and handle information ethically and what to do immediately after each session. It is part of a series on research and evidence for business decisions.

When to use interviews and focus groups

Individual interviews suit questions about personal experience, decisions, processes and sensitive topics. They allow depth and follow-up, and they avoid the social pressures of a group. They are the default choice for most business research, including customer research, staff research and research with buyers in other organisations.

Focus groups bring six to eight people together for a guided discussion. They suit exploring reactions to ideas, concepts, messages or designs, and generating a range of views through interaction, as participants build on each other’s comments. They are less suitable for sensitive topics, for detailed individual processes or where participants hold different levels of power, such as managers and their staff in the same group.

Recruiting the right people

Who is interviewed shapes what is learned. Recruitment should be purposive, deliberately covering the range of perspectives relevant to the question: large and small customers, new and long-standing, satisfied and dissatisfied, recent defectors, people who considered the business and chose a competitor. In business-to-business research, it often means speaking to several roles within a customer organisation, such as the user, the buyer and the person who approves spending, who may see things very differently. The who can say no article explains how buying decisions involve several people.

Practical points include:

Screening participants with a few questions to confirm they fit the profile.
Avoiding only friendly contacts, who are easy to reach but unrepresentative.
Offering a modest thank-you, such as a gift card or donation, where appropriate and permitted by the participant’s employer.
Explaining purpose, time and confidentiality clearly when inviting people.
Scheduling flexibly, including short calls, for busy participants.

The discussion guide

A discussion guide lists the topics and key questions to cover, in a sensible order, with suggested probes. It ensures consistency across interviews while leaving room to follow unexpected leads. A typical structure is:

Introduction: purpose, confidentiality, recording consent and how long it will take.
Warm-up: easy questions about the person’s role and context.
Core topics: the main questions, moving from general to specific.
Reactions: where relevant, responses to ideas, concepts or prototypes.
Closing: anything missed, anything else they want to say, and thanks.

Semi-structured interviews, which follow a guide but adapt to the conversation, suit most business research. Fully structured interviews ask fixed questions in a fixed order and work more like spoken surveys. Unstructured interviews explore freely and suit very early exploration.

Asking questions that produce evidence

Ask about real, specific past behaviour

The most reliable information comes from what people have actually done, recently and specifically. “Could you walk through the last time a machine broke down? What happened next?” yields far better evidence than “How do you usually handle breakdowns?” or “Would you use a faster repair service?” People describe their real past behaviour with reasonable accuracy, but they generalise about their habits optimistically and predict their future behaviour poorly.

Use open, neutral questions

Open questions starting with “how”, “what” and “describe” invite description. Closed questions starting with “do”, “is” or “would” invite yes or no. Both have uses, as the open and closed questions article explains, but interviews rely mainly on open ones. Neutral wording matters: “How did you find the delivery?” rather than “Were you happy with our fast delivery?”

Probe

Probes draw out detail and reasons:

“Could you give an example?”
“What happened then?”
“Why was that important to you?”
“How did you decide?”
“What else did you consider?”

Repeating “why” gently, sometimes called laddering, moves from surface preferences to underlying needs. A buyer who wants “faster delivery” may, on probing, really need predictability so that production can be planned.

Use silence

After an answer, a short pause often prompts the participant to add the most important part. Interviewers who rush to the next question lose this.

Avoid common traps

Leading questions that suggest an answer.
Hypothetical questions treated as forecasts.
Pitching or defending the business’s products during the interview.
Asking two questions at once.
Explaining what other participants said, which influences answers.

Listening

Good interviewers talk little and listen a great deal. They follow the participant’s language and logic, notice what is emphasised or avoided, and resist correcting misunderstandings or defending the business until the interview is over. Interviewing one’s own customers or colleagues makes neutrality harder; it can help for someone not directly involved in the relationship to conduct the interview. The listening article covers the skills involved.

Running focus groups

A focus group needs a skilled moderator who keeps discussion on topic, draws in quieter participants, manages dominant voices and avoids expressing opinions. Group dynamics shape what is said: people may conform to early or confident views, avoid disagreeing in public or perform for the group. Techniques that reduce these effects include asking participants to write down their initial views before discussing them, using round-robin questions so everyone answers, and explicitly inviting different opinions. Groups are typically recorded, with an assistant taking notes on who said what.

Remote and in-person sessions

Video and phone interviews are cheaper, easier to schedule and allow participants from anywhere. In-person sessions build rapport more easily, allow observation of the participant’s environment and work better for showing physical products. Remote sessions benefit from tested technology, a backup phone number and shorter durations. For interviews about work processes, visiting the workplace often reveals details that would not come up in conversation.

Interviewing staff

Interviews with employees about processes, problems or workplace experience need particular care. Staff may worry about consequences for themselves or colleagues, especially if a manager conducts the interview or if findings could identify them. Explaining the purpose, agreeing how comments will be reported, interviewing away from supervisors and focusing questions on systems and processes rather than individuals’ performance all encourage candour. Interviewing people across roles and shifts, not just those who volunteer, avoids hearing only the most engaged or the most frustrated.

Language and culture

Participants vary in language, culture and communication style. Some may be reluctant to criticise a business directly, some may answer more fully in their first language, and some may defer to a perceived authority. Plain language, avoiding jargon, allowing extra time and, where needed, using an interpreter or an interviewer from a similar background help participants express their views accurately. Interviewers should be aware of how their own manner and assumptions may shape what participants feel comfortable saying.

Watching as well as asking

What people say they do and what they actually do often differ, not through dishonesty but because routine actions are hard to describe. Combining interviews with observation, such as asking a buyer to show how they place an order or a technician to demonstrate a task, reveals workarounds, shortcuts and frustrations that would not come up in conversation. The next article in this series covers observation and field research in more detail.

Recording interviews, with permission, allows accurate quotes and analysis. Participants should be told how recordings and notes will be used, stored and protected, whether their identity will be kept confidential and how they can withdraw. Recordings and transcripts containing personal information must be stored securely and deleted when no longer needed. Recording a conversation without consent may also breach the law in some circumstances, so explicit consent should always be obtained.

Immediately after each session

Within a day of each interview:

write a short summary of key points, surprises and quotes;
note the interviewer’s own impressions separately from what was said;
record any changes needed to the discussion guide; and
store recordings and notes securely with consistent file names.

After several interviews, a short debrief with colleagues helps identify emerging themes and gaps. Analysis then proceeds systematically, by coding the transcripts or notes into themes, rather than relying on the most memorable stories.

How many sessions

Qualitative samples are small and purposive. For a focused question with a fairly similar group, new insights often slow after about eight to fifteen interviews. Varied groups, or several distinct segments, need more. Focus groups are usually run in at least two or three groups per segment, because a single group can be dominated by its particular mix of people.

A worked example

This is an illustrative example. The business and findings are hypothetical.

A commercial cleaning supplies distributor notices that several mid-sized customers have reduced their orders over the past year. The sales team believes the cause is price. The general manager commissions interviews before cutting prices.

An external interviewer conducts twelve phone interviews: six with customers whose orders fell, four with stable customers and two with former customers who had moved to a competitor. Each interview follows a semi-structured guide that begins with the customer’s role, then asks them to describe their most recent order from start to finish and their most recent problem with any supplier.

Price comes up in most interviews but is rarely decisive. The strongest theme is the difficulty of ordering: customers describe phoning orders in, waiting for callbacks and receiving incomplete deliveries without warning. Two customers who switched say the competitor’s online ordering with stock visibility saved their office staff hours each month. The distributor invests in an online ordering portal with live stock levels and proactive shortage notices, rather than cutting prices, and follows up with a short survey to measure how widespread the ordering problem is.

How this applies to a small Australian business

Choose interviews by default for reasons and processes, and groups for reactions to ideas.
Recruit deliberately across the range of relevant perspectives.
Use a discussion guide but follow important leads.
Ask about specific recent behaviour, not hypotheticals.
Probe for reasons and use silence.
Use a neutral interviewer where relationships could bias answers.
Record with consent and summarise within a day.
Combine interviews with observation where routines are hard to describe.
Analyse systematically across all sessions.

Common mistakes

Interviewing only friendly customers.
Asking people to predict their future purchases.
Leading questions and pitching during interviews.
Letting one voice dominate a focus group.
Relying on memory instead of recordings or prompt notes.
Skipping the warm-up, so participants never relax enough to speak candidly.
Interviewing staff in front of their managers.
Drawing conclusions from the most vivid story.

Frequently asked questions

How long should an interview take? Thirty to sixty minutes suits most business interviews. Busy participants may agree to twenty minutes if the topic is focused.

Should the business owner conduct the interviews? Owners learn a great deal from talking to customers, but participants may be less candid with them. A neutral interviewer is better where criticism is likely.

Are focus groups cheaper than interviews? Per participant, sometimes, but they produce less depth per person and are vulnerable to group effects. They are not a cheap substitute for interviews.

Can interview findings be quantified? Themes can be counted across interviews to show how widespread they were in the sample, but small purposive samples cannot estimate prevalence in the population; surveys are needed for that.

How can staff be encouraged to speak honestly in interviews? Explain the purpose, agree how comments will be reported, use an interviewer outside their reporting line, focus on processes rather than people and, afterwards, show what changed as a result.

Questions to ask

Who have we not heard from, and how could we reach them?
Do our questions ask about real past behaviour?
Who is best placed to interview without biasing the answers?
How are recordings and notes stored and protected?
Are participants from different backgrounds able to speak comfortably, in their preferred language if needed?
What would we see if we watched the process as well as asking about it?
How will we analyse all interviews systematically rather than recalling highlights?

Bringing it together

Interviews and focus groups reveal reasons, processes and priorities that numbers miss. Recruit purposively across relevant perspectives, prepare a discussion guide, ask open, neutral questions about specific past behaviour, probe for reasons and listen more than you speak. Manage group dynamics carefully, record with consent, summarise promptly and analyse systematically. Above all, seek out the voices that are easiest to miss: the quiet, the dissatisfied and the people who chose someone else.

Source: KEVOS editorial notes. Examples and findings in this article are illustrations. This article is general information and does not constitute legal advice on recording or privacy.