What is critical incident technique?

Also called: CIT, Critical incident method

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.

Updated

Where it comes from

John Flanagan formalised the method in "The Critical Incident Technique" (Psychological Bulletin, 1954), drawing on studies of aircrew performance and training in the Aviation Psychology Program during the Second World War. He set out five steps: define the aim of the activity, set the plan for which incidents to collect, collect them, analyse them, then interpret and report.

Why it works in product research

Ask a customer "How do you find reporting?" and you get a vague average: "It's fine, a bit slow." Ask "Tell me about the last time a report didn't give you what you needed" and you get a story: the board meeting, the export to a spreadsheet, the colleague who rebuilt the numbers by hand. Stories have causes, workarounds and consequences you can design against.

Memorable incidents are also the ones that drive renewal, churn and word of mouth, so they point at what matters to customers.

How to collect incidents

  • Ask for one specific event: "Think of a time in the last month when..." A recent, bounded time window gives more accurate detail.
  • Ask for both kinds: a time it went well and a time it went badly. Positive incidents show what to protect.
  • Probe for the sequence: what were you trying to do, what happened, what did you do next, what was the outcome?
  • In surveys, use one open question per incident type and keep it optional. Expect shorter stories than in interviews.
  • In churn or exit interviews, ask about the moment they decided to look for another tool.

How to analyse them

  • Treat each incident as one unit, not each participant. One interview can produce three or four incidents.
  • Group incidents into categories bottom-up, by cause or by stage of the task, in the same way as an affinity map.
  • Count incidents per category to see which problems come up most, while remembering the counts reflect what people remember, not true frequency.

Go deeper

Questions

What counts as a critical incident?

Any specific event the participant remembers clearly that had a noticeable effect on what they were trying to do, good or bad. It doesn't need to be dramatic. A failed import before a deadline counts; a general feeling that imports are clunky doesn't.

How many incidents do you need?

Flanagan didn't set a fixed number. A practical rule is to keep collecting until new incidents stop adding new categories, the same idea as data saturation in interviews. Count incidents, not participants, when you judge this.

Is the critical incident technique biased toward negative events?

It can be, because bad experiences are easier to recall. Ask for a positive incident in every session, and don't treat the ratio of good to bad incidents as a satisfaction score.

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
  • Probing questions

    Probing questions are follow-ups an interviewer asks in response to what a participant just said, to get the detail, reasons, examples or clarification the first answer left out, without introducing a new topic or suggesting an answer.

  • Leading questions

    A leading question is worded in a way that suggests the answer the asker expects or wants, so participants drift toward that answer and you learn more about your own assumptions than about their experience.

  • 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.