What is the interview method of data collection?
The interview method collects information through a planned conversation in which an interviewer asks questions and, where the design allows, follows up for clarification and detail. Interviews can produce accounts of experience, explanations of decisions and descriptions of events. Structured interviews can also collect standardized numeric or categorical responses.
Use an interview when you need to understand how or why something happened. A survey may show that customers are dissatisfied; an interview can explore the sequence of events behind a particular experience. Neither method automatically makes a sample representative.
Structured, semi-structured and unstructured interviews
| Approach | How it works | Useful when | Main trade-off |
|---|---|---|---|
| Structured | The same questions are asked in a specified form and order. | Consistency across respondents is a priority. | Limited room to explore unexpected explanations. |
| Semi-structured | A topic guide is combined with neutral follow-up questions. | You need comparable topics and room for detail. | Interviewers need skill to probe without leading. |
| Unstructured or in-depth | A broad purpose guides a flexible conversation. | The experience or topic is not yet well understood. | Comparison and analysis can require more time. |
These are approaches to questioning. In-person, telephone and video interviews are modes of delivery. A video interview can be structured or semi-structured; the mode does not determine the research design.
Interview, survey or focus group?
An individual interview gives one person space to explain their experience. A focus group uses interaction between participants as part of the data. A survey can reach more people with standardized questions, although open responses usually allow less probing.
| Question | Possible method | Reason |
|---|---|---|
| Why did this customer stop using the service? | Individual interview | Explore a sequence and ask for clarification. |
| How do team members discuss a shared practice? | Focus group, where safe and appropriate | Observe agreement, disagreement and group language. |
| How common is a predefined experience in the target population? | Suitable survey design | Estimate prevalence using a defensible sample and measure. |
| Do people understand a question as intended? | Cognitive interview | Explore how they interpret and answer it. |
CDC describes cognitive interviewing as a method for examining the thought processes behind survey answers. It is different from an interview about the service experience itself. Read the CDC method description.
How to collect data through interviews
- Define the question: decide what an interview can clarify and what needs another source.
- Choose participants deliberately: explain whose experience matters and why. Include variation relevant to the question.
- Prepare a guide: start with concrete experience, then explore detail and interpretation.
- Plan consent and records: explain participation, recording, use, access and the option to decline.
- Pilot the conversation: check clarity, timing and whether questions produce useful detail.
- Interview and write notes: separate what the participant said from your interpretation.
- Review while collection continues: identify gaps and document any change to the guide.
Do not recruit only the people easiest to reach or most likely to praise the service. If people who left are central to the question, make a feasible plan to hear from them and report who could not be reached.
An example semi-structured interview guide
The guide below explores a service experience. Adapt the language and sequence to the setting; it is not a validated questionnaire.
| Purpose | Neutral question | Possible probe |
|---|---|---|
| Establish the sequence | Tell me about the last time you used the service. | What happened next? |
| Understand the need | What were you trying to do? | What would a useful result have looked like? |
| Find obstacles | Was anything difficult or unclear? | Can you describe a specific moment? |
| Explore responses | What did you do when that happened? | Did someone help, or did you try another route? |
| Understand the result | How did the situation end? | What remained unresolved? |
| Identify improvement | What would you change about that experience? | Why would that change matter to you? |
Avoid questions such as “How helpful was our excellent support?” They introduce the answer you hope to hear. Ask about concrete events before requesting an overall judgment. Leave space for an account that challenges your initial explanation.
Advantages and limitations of interviews
Interviews allow clarification, detail and unexpected topics. They can help explain the meaning of a survey result and reveal terms that should inform a later questionnaire. They also require time for recruitment, interviewing, transcription or notes and analysis.
What people recall may be incomplete. Their answers can be influenced by the interviewer, the setting and perceived consequences. An employee speaking with their manager may describe a problem differently from an employee speaking with an independent interviewer.
These limitations do not make interviews unusable. Reduce avoidable bias through neutral wording, appropriate interviewer roles, clear participation information and comparison with other evidence. Report what the method can and cannot establish.
How many interviews and questions do you need?
There is no universal number. The breadth of the question, diversity of participants, depth of interviews, analysis approach and decision all matter. A narrow operational question may require a different design from a study comparing several locations.
Plan an initial sample and review what it covers. Record whether additional interviews reveal new relevant patterns or gaps. Do not declare “saturation” simply because several people repeated a theme; explain what was examined and how the stopping decision was made.
Similarly, choose the number of questions by the time and depth required. A short guide with good follow-up can produce more useful evidence than a long list rushed through.
From recordings and transcripts to analysis
A transcript is a source, not a finding. Review the recording or notes where meaning is uncertain, remove information that should not enter the analysis copy, and keep source access appropriately restricted.
- Use a codebook or another explicit analysis approach suited to the question.
- Retain links between themes and the passages supporting them.
- Examine contradictory cases and variation, not only frequently repeated themes.
- Separate participant accounts from the team’s interpretation.
- Document changes in codes, definitions and review decisions.
AI can propose themes, classify passages or summarize a conversation. Check its output against the source and review omissions, ambiguity and unsupported inferences. Do not infer a person’s intent or diagnosis from a short passage.
Connect interviews to the right context
When the research question concerns change over time, a permission-aware identifier can connect interviews with earlier check-ins, service records or assessments. Keep the interview date, guide version, participant group and relevant stage with the source.
An interview collected after a service failure should not be compared uncritically with a routine onboarding interview. Context makes the comparison interpretable. Existing document and research tools may support this work; evaluate the actual workflow rather than assuming files are unusable outside a particular platform.
For selecting complementary methods, read qualitative data collection methods.
Continue learning in the Academy
Use these existing Academy guides for the practical next step. They are suggested companion readings; follow each guide’s course navigation for the full sequence.
Use the practical guides to connect the model to a collection plan, review new evidence and decide what to improve.
Plan the handoff from interview to reviewed finding
For each interview, retain a source identifier, date, guide version and the context needed for interpretation. Store the participation and access information separately where that better protects the participant. Give the analyst enough context to understand the account without automatically sharing every identifying detail.
Suppose a customer describes a problem during onboarding and later says it was resolved. Keep the two accounts dated and distinguish the customer's report from a staff note saying the case was closed. If the accounts differ, record that uncertainty and decide whether a permitted follow-up can clarify it.
Before sharing a theme, name the reviewer who checks the source passages and the person responsible for the next decision. This makes collection, analysis and follow-up part of one planned workflow. The customer feedback course develops that operational plan; it does not replace the study's methodological requirements.
How Sopact reduces coding and reporting work
Interview design and interpretation still need human judgment. When a codebook-based approach fits, plan how your team will reapply revised definitions across transcripts and connect permitted interview context with other evidence.
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.
Frequently asked questions
Are interviews qualitative or quantitative?
They can support either. Open, probing interviews usually generate qualitative evidence; structured interviews can collect standardized categorical or numeric responses.
Is a Zoom transcript an interview dataset?
It can be a source within a dataset, but it still needs appropriate consent, context, quality review and an analysis method.
Should every interview be recorded?
No. Choose an appropriate recording or note-taking approach, explain it to participants and obtain required permission. Respect a refusal to be recorded.
Can interview findings be generalized?
The strength and type of generalization depend on design and sampling. A small purposive sample should not be presented as a population prevalence estimate.
Can AI code interviews?
AI can assist coding and synthesis, but researchers need to review source fidelity, confidentiality, the code definitions and the conclusions.
