What is a qualitative survey?
A qualitative survey collects people’s descriptions, explanations or experiences, usually through open-ended questions. It lets respondents answer in their own words. Unlike an interview, it normally offers limited opportunity to clarify an answer or ask a follow-up question.
Use it when you want accounts from people who can respond independently and when written responses fit their circumstances. It can reveal issues you did not anticipate. It is less suitable when the topic requires extensive probing, respondents face writing barriers or the context is too complex for a short form.
Decide whether a survey fits the question
| You need to understand… | Consider… | Watch for… |
|---|---|---|
| Experiences across several locations | A short open-response survey | Differences in access and language. |
| A complex personal journey | An interview | Consent, privacy and interviewer skill. |
| How often an issue occurs | A suitable quantitative measure and sample | An open-response theme count is not automatically prevalence. |
| What people actually do | Observation or routine records | Self-report can differ from behavior. |
A survey can combine open and closed questions. The important point is how each answer contributes to the question you are investigating. See mixed method surveys.
Build a short qualitative survey
| Part | Example | Purpose |
|---|---|---|
| Introduction | Explain the purpose, expected use and whether responding is optional. | Support an informed decision to participate. |
| Context | Which service or event are you describing? | Make the answer interpretable. |
| Experience | What happened when you tried to use the service? | Invite an account without assuming success. |
| Explanation | What made the process easier or harder? | Understand conditions around the experience. |
| Improvement | What would you change? | Identify practical suggestions. |
| Closing | What important detail have we not asked about? | Allow unanticipated information. |
Do not ask every possible question. Pilot the form with people similar to the intended respondents and check whether the answers are specific enough to use.
Recruit the people whose experience matters
Define who should be included and why. A purposive sample may help explore different experiences; it should not be described as statistically representative unless the sampling supports that claim. Track invitations and responses where appropriate without compromising promised anonymity.
Consider who may be missing: people who left, could not access the service, use another language or cannot complete an online form. Offer an appropriate alternative when feasible. More responses from one convenient group do not necessarily solve a coverage problem.
Analyze responses systematically
- Read the responses before finalizing the coding categories.
- Define each theme and record examples and boundary cases.
- Allow an answer to contain several themes where appropriate.
- Review contradictory and unusual responses, not only the most frequent ones.
- Keep the original text connected to the interpretation.
- State whether a count refers to people, responses or mentions.
AI can assist with organizing text, but check the categories and source passages. Translation, sarcasm and context can affect interpretation. Do not replace a participant’s meaning with a more convenient label.
Worked example: do not confuse mentions with people
In an illustrative survey, 40 people respond. Twenty mention delays and 12 mention unclear instructions; five mention both. These counts cannot be added to claim that 32 different people had problems. Nor do they establish the percentage of all service users affected.
A useful finding states the observed pattern, gives the response context and identifies what to investigate next. For example: “Delay was mentioned in 20 of 40 responses; we will compare these accounts with waiting-time records before changing the process.”
Report what the evidence supports
Describe who responded, when, how questions were asked and how coding was done. Include a few permission-appropriate excerpts to explain themes, with enough context to avoid changing their meaning. Keep dissenting evidence visible.
The King Center’s published story illustrates the value of making open-ended feedback usable rather than leaving it unread. It is a practice example, not proof that any one survey design produces better outcomes. Read the story.
How Sopact reduces coding and reporting work
An open-ended survey earns its place when the team can use the whole response set, rather than selecting the few comments it has time to read.
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
Is a survey qualitative or quantitative?
It depends on the questions and analysis. A survey can collect either or both.
How many responses do I need?
It depends on the purpose, diversity of experience and analysis plan. There is no universal qualitative survey sample size.
Can I use percentages for themes?
Yes, if the denominator and coding rules are clear. Avoid generalizing beyond the observed sample without a suitable design.
Should I promise anonymity?
Only if the collection and analysis genuinely support it. Free text itself can contain identifying information.
