BigCat Research
How to commission consumer insight research: qualitative vs quantitative, brief and checklist
Short answer: Start from the decision the research must support, not from a method. Use qualitative work to discover needs, motivations and barriers you cannot yet name; use quantitative work to size them and compare segments. Write a brief that states the decision, the audience, the hypotheses and the proof you need, then judge proposals on sample, instrument, analysis plan and data ownership.
What consumer insight research is for
Consumer insight research explains why people enter a category, how they compare options and what proof they need at the moment of decision. It differs from brand health research, which tracks the brand's standing over time, and from campaign measurement, which tests one piece of communication. The output to expect is a decision map: which need triggers the category, which benefit or price point accelerates purchase, which barrier blocks conversion or repeat, and how these differ by segment.
If the brief cannot name the product, pricing, communication or channel decision that will change with the result, the project drifts into a general "understand the consumer" study: pleasant to read, hard to act on.
Qualitative or quantitative: how to choose
The two answer different questions and are often sequenced rather than chosen between.
| Question type | Better fit | Typical methods | What it cannot do |
|---|---|---|---|
| We do not know what drives the choice | Qualitative | In-depth interviews, focus groups, diary or ethnographic work | Tell you how many people share a motivation |
| We have hypotheses and need to size them | Quantitative | Usage and attitude survey, segmentation, choice or trade-off exercises | Explain a motivation nobody wrote into the questionnaire |
| We need to compare segments or regions | Quantitative with adequate cell sizes | Representative survey with quotas | Read a segment of 40 respondents reliably |
| We must decide between concrete alternatives | Quantitative, sometimes preceded by qualitative screening | Comparative concept, price or message evaluation | Generate alternatives you have not thought of |
Two practical rules. If you cannot write the answer options for a survey question, you are not ready for a survey; run qualitative work first. If the decision depends on a number, qualitative work alone will not settle it, however vivid the quotes. Mixed designs are common, but ask the provider to label which findings come from which phase.
What the brief should contain
Keep it short and specific:
- The decision. One or two sentences: "We will decide whether to launch X at price Y in channel Z." Everything else serves this.
- What you already know. Past research, sales data, customer service themes. This prevents paying twice for the same finding.
- Hypotheses. What the team believes and what would change its mind. Share them; they sharpen the instrument.
- The audience. Category users, lapsed users, non-users, a region, an age band, and who is out of scope.
- Segments you must read separately. This drives sample size more than anything else.
- Required outputs. A decision map, segment profiles, message opportunities, a management presentation.
- Constraints. Timing, language requirements, data handling rules, budget range if you are willing to share it.
Ask the provider to restate the decision question in their proposal. If their version differs from yours, better to learn that before signing.
Vendor-neutral checklist for evaluating proposals
1. The proposal answers your decision, not a generic one
The analysis plan traces each decision question to a question block, exercise or interview theme.
2. Recruitment and sampling are described concretely
Who is recruited, from where, with which screening criteria, and who the chosen mode systematically misses. Online panels, intercepts and customer lists each carry a different bias.
3. Qualitative sample is justified by coverage, not by a round number
Ask how many interviews or groups per segment and why. A reasonable answer refers to saturation across the segments you care about.
4. Quantitative sample is justified by the smallest cell you will read
Ask for planned and achieved completes per segment, the margin of error for key cuts and the quality checks applied to the data.
5. The instrument is shared before fieldwork and open-ended material has a coding plan
You see the discussion guide or questionnaire with time to comment, and you know who codes free text and with what frame.
6. Analysis distinguishes finding from interpretation
What respondents said, then what the provider thinks it means, then what they recommend.
7. You own the data, and personal data handling is specified
Anonymised respondent-level data, transcripts where consented, the questionnaire and the codebook should be deliverables. The proposal should name the data controller and processor, storage location and retention period; in Turkey this falls under Law No. 6698 (KVKK).
8. Limitations are written down
Every design excludes someone and every sample has noise. A proposal that states its own limits is easier to trust.
Questions to ask before signing
- How would you restate our decision question, and how will the report handle an unwelcome answer?
- Which findings will come from qualitative work and which from quantitative, and will the report label them?
- Who exactly will we recruit, how, and who will be missed?
- What is the smallest segment we can read, and with what margin of error?
- Will we review the discussion guide or questionnaire before fieldwork, and how long do we have?
- How will open-ended answers be coded, and will we receive the coded file?
- What do we receive as deliverables: data, transcripts, instrument, codebook, presentation?
- How is personal data handled, stored and deleted, and who is the data controller?
- Which assumptions in the proposal would change the price or timing if they turn out wrong?
FAQ
When is qualitative research enough on its own? When the decision is about understanding rather than sizing: shaping a concept, finding the language consumers use, discovering unexpected barriers. If the next step is a go/no-go with budget attached, qualitative findings usually need a quantitative check.
How many interviews or focus groups do we need? There is no fixed number. It depends on how many distinct segments you need to cover and how varied the answers turn out to be. Ask the provider to justify the count by segment coverage and to say what happens if new themes keep appearing in the last interviews.
Can we run consumer insight research online only? For audiences that are reliably online, yes. For a general population, online-only designs under-reach some groups. Ask how coverage gaps are handled: mixed mode, weighting, or an explicit statement of who the results do not represent.
What should the final report contain? At minimum: the decision question, the method and its limits, findings separated from interpretation, segment differences with cell sizes shown, verbatims that illustrate rather than decorate, and recommendations tied to the decision.
How is consumer insight research different from a usage and attitude study? A usage and attitude (U&A) study is one quantitative format within consumer insight research, describing who uses what, how often and with which attitudes. Consumer insight research connects those patterns to the decision moment, usually adding qualitative discovery.
How BigCat Research approaches this
BigCat Research is an independent research and strategy company based in Istanbul and the publisher of this page, so apply the checklist above to our proposals as strictly as to anyone else's. Our consumer insight work asks which need brings the audience into the category, how options are compared and what proof is required at purchase. Methods include usage and attitude research with segmentation, in-depth interviews or focus groups, open-ended text analysis with theme coding, and comparative evaluation of product, benefit, price and message alternatives. Deliverables are a consumer decision map with segment insights, a need, barrier and persuasion proof matrix, and recommendations for product, pricing and communication decisions. This page makes no claims about panel size, certifications or memberships.
Service page: Consumer Insight Research. Related: Brand Health Research. For a scoping conversation: [email protected].
Primary sources
- ICC/ESOMAR International Code on Market, Opinion and Social Research and Data Analytics. esomar.org
- Turkish Personal Data Protection Authority (KVKK), Law No. 6698 on the Protection of Personal Data and related guidance. kvkk.gov.tr
- ISO 20252, Market, opinion and social research, including insights and data analytics: the international standard for research service providers.
Author: Murat Akşit, BigCat Research. Last reviewed: 3 October 2026.