What is a qualitative interview?
A qualitative interview is a conversation designed to explore someone's experience, perspective or understanding in depth. The interviewer uses questions and follow-ups to learn how the person describes a situation, rather than limiting every answer to a predefined choice.
It can help a service team understand a difficult handoff, a program team learn how participants use support, or a membership organization explore what local representatives need. The value comes from the quality of the conversation and its interpretation. A transcript does not automatically explain a survey score, and an interview does not need an accompanying score to be useful.
This guide covers interview analysis and the collection choices that make that analysis useful. For a broader introduction, use the qualitative-analysis walkthrough or choose an analytical method.
How to analyze qualitative data from interviews
Prepare the transcript or notes, read the full accounts, choose an analytical approach, organize the material, examine patterns and differences, and write a supported interpretation. These activities require judgment; assigning labels is only one possible part of the work.
1. Prepare and become familiar with the accounts
Keep the interview question, date and relevant context alongside the material. Check uncertain wording against the recording where one is available. Read each interview as an account before cutting it into extracts. Write notes about what you notice and what you may be assuming.
2. Decide what the analysis needs to answer
A review of recurring support issues, an interpretation of experience and an analysis of a person's story are different tasks. Select the method accordingly. Do not choose a codebook or a frequency table merely because it is the software's default output.
3. Organize the evidence without losing its context
In a structured applied review, you might label passages about finding help, contacting someone and receiving a response. Record what each label means and revise it if necessary. Keep an extract connected to the surrounding account. Other analytical approaches may develop interpretation differently.
4. Examine the finding against supporting and contrasting accounts
This fictional example shows an early review of coaching interviews. It is a teaching illustration, not a completed research finding.
Scroll horizontally to see all columns →
| Illustrative extract | Initial observation | What to examine next |
|---|---|---|
| “I had the office number, but I thought it was only for booking sessions.” | The participant had contact information but understood its purpose narrowly | How was the contact route explained? |
| “My coach told me to call between sessions if I needed help.” | A contrasting account describes an explicit invitation | What differed in the communication or context? |
| “I knew I could call, but I wanted to try it myself first.” | Not making contact does not necessarily mean the route was unclear | How did the participant make that choice? |
“Participants do not know whom to contact” would flatten these differences. A more useful provisional interpretation is that knowing a number, understanding its purpose and deciding to use it are distinct parts of getting help. Read the complete accounts and the wider material before adopting that interpretation.
5. Write the result and its limitations
Explain whose interviews were reviewed, how they were selected and how you reached the interpretation. Include evidence that complicates the finding. A practical next step could be reviewing how support options are explained, followed by checking participants' subsequent understanding. The interviews alone do not establish that a communication change will improve outcomes.
If you report category counts, state whether you counted people, interviews or passages. Several extracts from one person are not several participants. Selected interview accounts do not automatically estimate how common an experience is across the whole program.
AI can help organize material, but review any proposed label, extract and summary against its source. A fixed codebook does not guarantee identical interpretations or unbiased results. A human remains responsible for the method and conclusion.
Structured, semi-structured and unstructured interviews
Scroll horizontally to see all columns →
| Format | How it works | Useful when | Tradeoff |
|---|---|---|---|
| Structured | The interviewer follows a consistent set of questions and order | The review needs a closely standardized conversation | There is less room to pursue an unexpected account |
| Semi-structured | A topic guide gives direction, with follow-up questions and flexibility | The team needs to cover shared topics while learning from individual experience | Interviewers need judgment to explore without leading |
| Unstructured | The conversation follows a broad purpose with more participant-led development | The study calls for an open exploration of experience or a story | It requires skill and may produce material that is harder to compare on a fixed set of questions |
“Qualitative” does not simply mean “unstructured.” The format should fit the purpose and analytical approach. A semi-structured interview is often practical for an applied review because it leaves room for the unexpected while retaining a shared set of topics.
When an interview is the right choice
Use an interview when you need to explore how someone experienced something, clarify an unfamiliar process or understand the context behind an account. It is especially useful when a short form answer leaves important questions unresolved.
It may be the wrong primary method if your question is how many people completed a task, how frequently an event occurred across the whole population, or whether someone can use an interface. Those questions may need records, a suitable survey, observation or a task-based study. Interviews can complement that evidence, but they should not be expected to answer every question.
The GOV.UK in-depth interview guide recommends planning the research questions, topic flow and follow-ups, then testing the discussion guide before using it. It also addresses practical arrangements such as accessibility and recording.
Plan the conversation before recruiting
Define the learning goal
Replace a broad goal such as “get feedback” with a question the interview can address. For example: “How do participants describe getting help between scheduled coaching sessions?” That question gives the conversation direction without assuming that the current support is either effective or inadequate.
Write down the decision this learning will inform. A team considering a new contact route needs different evidence from a researcher exploring how trust develops over several years.
Choose whose experiences you need
Recruit around the question. Include relevant variation, such as different stages of participation, locations or experiences of the process. Recruiting only the most available or enthusiastic people can leave important experiences unheard.
There is no universal interview count that guarantees a good study. Plan around the scope, diversity, depth and analytical approach. Review what the material covers as work proceeds. Record who was invited, who participated and any important gaps without implying that a purposive sample is statistically representative.
Make participation understandable and workable
Explain the purpose, what participation involves, how the material will be used and who will see it. Decide how recording, quotation and any later follow-up will be handled. Arrange a setting and format that allow the participant to speak comfortably.
Consider the relationship between interviewer and participant. Someone receiving a service or working for an organization may be cautious about criticizing it. Where possible, choose an interviewer and process that reduce pressure and make the participation choices clear.
A sample semi-structured interview guide
This fictional example explores how a participant gets help between scheduled coaching sessions. Adapt the topics to your setting and test the wording. The questions are interview questions, not instructions for an AI assistant.
Scroll horizontally to see all columns →
| Part of the conversation | Starter question | Possible follow-up |
|---|---|---|
| Context | “Tell me a little about how you take part in the program.” | “What does a typical week look like?” |
| A recent experience | “Can you tell me about the last time you wanted help between sessions?” | “What happened first? What happened next?” |
| Finding support | “How did you decide whom to contact?” | “What information did you have at that point?” |
| What followed | “What happened after you made contact?” | “How did that fit with what you expected?” |
| Variation | “Was there another occasion that went differently?” | “What seemed different about that occasion?” |
| Closing | “What have I not asked that would help us understand this?” | “Is there anything you would like to clarify?” |
Follow the person's account rather than asking every follow-up mechanically. If the participant has never needed help between sessions, do not pressure them to invent an example. Explore that experience on its own terms.
Questions about a concrete recent event often produce more detail than a general request for satisfaction. Ask about the sequence, setting and expectations, while avoiding assumptions about what caused the experience.
Improve questions that lead the answer
Scroll horizontally to see all columns →
| Question to reconsider | Why it can distort the conversation | More neutral alternative |
|---|---|---|
| “How frustrating was the slow response?” | Assumes both delay and frustration | “What happened after you sent the request?” |
| “Was the coach helpful and easy to reach?” | Combines two different issues | Ask separately about reaching the coach and the help received |
| “Would a new app solve this?” | Invites speculation about the team's proposed solution | “How do you manage this now? What is difficult about that?” |
| “Why didn't you follow the instructions?” | Can sound accusatory and assumes the instructions were understood | “What information did you have when you decided what to do?” |
Neutral wording does not remove every source of influence. Tone, timing, the interviewer's role and earlier questions all shape the conversation. Note those conditions when interpreting the account.
During the interview: listen, clarify and leave room
Begin with a brief explanation of the session and confirm the agreed arrangements. Use the guide to maintain direction, but give the participant time to finish a thought. Avoid filling every pause with another question.
Clarify unfamiliar terms in the participant's language. A question such as “When you say the process felt complicated, what happened?” asks for detail without supplying an explanation. Reflect back your understanding tentatively so the participant can correct it.
Distinguish an interview from a service intervention. If someone needs immediate practical help, follow the appropriate process rather than relying on a future analysis report. Do not promise that every suggestion will lead to a change.
At the close, explain what happens next and how any agreed follow-up will work. A clear ending helps avoid the impression that the team has collected a personal account without a purpose.
Keep records that preserve meaning
If a session is recorded, check that the recording worked and that its use matches the arrangement with the participant. If you rely on notes, label them as notes rather than a verbatim transcript. Keep direct quotations separate from the note-taker's interpretation.
Review a transcript against the recording where wording matters or appears uncertain. Automatic transcription can mishear names, technical language or speech. Translation can introduce another layer of interpretation. Preserve the original language where appropriate and have important translated passages checked by someone qualified to assess them.
Store the relevant interview date, topic-guide version and context. Use a participant reference when it serves the design and keep identifying information appropriately restricted. Not every reviewer needs access to the person's identity.
Repeat interviews across locations or over time
Different sites may need different follow-ups. Agree on the limited shared topics and context required for the overall question, then allow local depth. A common topic guide can support coverage without forcing every conversation to be identical.
For repeated interviews with the same person, preserve the date and relevant circumstances of each conversation. Do not replace earlier context with their current profile. Record changes to the guide so a later interpretation can distinguish a change in experience from a change in questioning.
When interviews involve different people at each wave, describe the result as evidence from those groups rather than an individual trajectory. Keep changing participation visible and avoid treating an absent interview as evidence that an experience improved.
Choose a manageable collection and review workflow
A small interview study can be managed well with suitable recording, document and analysis tools. Recurring work across many locations adds questions about access, context, version history and how reviewed findings reach the people who can act.
Evaluate Sopact around that end-to-end process: how interview material is collected or brought in, how it retains relevant record context, how analysis is reviewed and how an approved finding is shared. Verify transcript formats, language support, access controls and review behavior using representative material.
The goal is to reduce avoidable reconciliation while preserving analytical judgment. A platform should not be assumed to make interviews comparable, eliminate interpretation differences or detect every important concern merely because it can process text.
Report the account without overstating it
Explain the interview purpose, selection, format, analysis and limitations. Use extracts to make the finding understandable, with enough context to represent the account fairly. Consider whether a person could be recognized from the details even when their name is removed.
State what the team will investigate or change, and how it will learn whether that action helped. For a broader report, use How to Write an Impact Report and browse report examples.
How Sopact reduces coding and reporting work
Long interviews still require careful interpretation. Where a structured coding approach fits, automation can reduce repeated application work while reviewers keep responsibility for meaning and context.
A workflow with repeated manual work
- Define from an initial sampleRead material and agree on the codebook.
- Apply it across the datasetCode responses and check the result.
- Revise a definitionReturn to affected material and recode it.
- Reconnect the numbersReconcile coded results with ratings and context, then rebuild the view.
The Sopact workflow
- Your team owns the definitionsDecide what each code means and improve it as you learn.
- Apply coding across the eligible dataAutomate application; people review quality and exceptions.
- Reprocess after a definition changesReapply the revised definition across the configured scope instead of recoding each response by hand.
- Ask across coded text and numbersKeep the response, rating and relevant record context connected; inspect the evidence behind the result.
This compares workflow patterns, not a claim that every research tool requires manual coding or separate files. Some already automate parts of this work; compare the complete cycle.
For this codebook-based workflow, the main saving is repeated application and reconnection—not the removal of human judgment. A changed definition can be reapplied across the configured data while reviewers concentrate on quality, exceptions and interpretation. Coded text stays connected to the relevant ratings and context.
Count the recurring work in ownership cost. Include setup, coding, recoding after revisions, source reconciliation, review and reporting, plus your actual platform and processing expenses. A worked scenario of four cycles of 4,000 responses illustrates 272 fewer annual staff hours; it is an assumption-based example, not a customer benchmark. Existing automation, review needs and implementation effort can substantially change the result.
Adjust the workload assumptions and compare total effort →
A reliable assistant should calculate from the selected records and let a reviewer open the supporting evidence. Check the data scope, definition, denominator and access permissions. Reproducible arithmetic does not make every AI interpretation correct.
Watch: Why Qualitative Analysis Stays Small — And How to Scale It
See why revising a codebook creates repeat work, and how connected coding and quantitative analysis change that workload.
Watch: connecting qualitative evidence to a review
This companion video introduces a qualitative-evidence workflow. It complements the interview planning and analysis guidance; it is not an interviewing tutorial.
Unified Qualitative Analysis | What Changes Everything
Frequently asked questions
What makes an interview qualitative?
Its purpose and approach explore experience, meaning or perspective in depth, usually through open questions and follow-ups. It is not defined solely by whether a recording is made.
How many qualitative interviews do I need?
The number depends on the question, scope, variation, depth and analytical approach. There is no universal count that guarantees adequate evidence.
Can I use the same questions at every location?
You can agree on shared topics where comparison requires them, while allowing local follow-ups. Preserve differences in context and guide versions when interpreting the material.
Should every interview be linked to a survey score?
No. Link sources when it is useful, appropriate and supported by the design. Interviews can provide valuable evidence independently of numerical measures.
Can I analyze notes instead of recordings?
Sometimes, if notes provide the material the question and method need. Be clear that they are notes, distinguish quotation from summary and acknowledge what detail may be missing.
Can AI analyze an interview without review?
AI can assist with transcription or preliminary organization, but errors and unsupported interpretations remain possible. Check source material and retain responsibility for the findings.
Choose your next step
Girls Inc. of Metropolitan Dallas
Explore how participant evidence stays connected across programs and years. The interview guide above is a fictional teaching example.
Read the customer story →