What is generative research?

Also called: Exploratory research, Discovery research, Formative research

Generative research is research done before there is a solution to test, to understand people's needs, behaviours and problems well enough to decide what to build. It produces opportunities and questions, not a verdict on a design.

Updated

What it's for

Generative research answers questions like: how do finance teams close the month today, where does it break, and what have they tried? The answer shapes the roadmap. Evaluative research comes later, once there's a design or product to check.

Product teams usually run generative research when entering a new market or segment, when a metric moves and nobody knows why, before a major redesign, or when a team keeps building features that don't get used.

Generative research methods

  • User interviews: open conversations about past behaviour and context. The default method for most teams.
  • Contextual inquiry and field studies: watching people work in their own environment and asking about it as they go.
  • Diary studies: participants log activities over days or weeks, which suits habits and tasks that happen rarely.
  • Jobs-to-be-done interviews: reconstructing the moment someone switched to a new product, to find what they were trying to get done.
  • Critical incident questions: asking about specific moments that went well or badly, inside any of the methods above.

How to scope a generative study

  • Frame the question around people, not your product: "How do agencies plan client reporting?" instead of "What do agencies want from our reports feature?"
  • Recruit for the behaviour you're studying, including people who don't use your product, if the question is about a market.
  • Keep the discussion guide loose, with topics and probes instead of fixed questions.
  • Plan synthesis time. Generative studies produce more raw material than evaluative ones, and the value is in the themes.

Common mistakes

  • Showing a prototype halfway through. Once people see a solution, the conversation turns into feedback on it.
  • Asking people what they want built. Ask what they do and why; the solution is the team's job.
  • Stopping at a list of quotes. Generative research needs a synthesis step that names opportunities the team can act on.

Go deeper

Questions

What's the difference between generative and evaluative research?

Generative research happens before there's a solution and helps you decide what to build. Evaluative research checks whether a design or product works, through methods such as usability tests, concept tests and A/B tests. Most teams need both, generative early and evaluative later.

Is generative research always qualitative?

Mostly, because open questions need room for answers you didn't expect. A survey can be generative if it explores behaviour you don't yet understand, but it works better after interviews have shown what to ask about.

How many interviews does a generative study need?

Plan per segment and stop when new interviews stop adding themes. For a narrow question with a similar group, that can come within 10 to 12 interviews; broad questions across different roles take more.

Next step

Run the study, not just the method

Great Question handles recruiting from your own customers, scheduling, incentives, interviews, tests and analysis in one place.

Book a demo

No credit card required.

Related terms

Full glossary
  • Evaluative research

    Evaluative research tests an existing design, prototype or product against a goal, to find out whether people can use it and whether it does what the team intended. It checks a solution, where generative research looks for the problem.

  • Data saturation

    Data saturation is the point in qualitative research where new interviews or observations stop producing new themes or insights, which tells you that adding more sessions would mostly repeat what you already know.

  • Critical incident technique

    The critical incident technique is a research method where participants describe specific past events that went especially well or badly, what led up to them and what happened next, so you learn from concrete behaviour instead of general opinions.