Marcus Andrey Study
Ler em português Library

Framework · Research · 3 min read

Insight statement

A one- or two-sentence format for an insight: who does what, why, why it matters to a goal and where it points, plus the evidence, confidence and owner that make a team act on it.

1Who

A specific group from your synthesis, never just "users".

Supervisors planning night shifts
2What

What they do or struggle with, observed, not assumed.

Export the plan before publishing it
3Why

The reason, in their terms. The part a finding doesn't have.

Absences don't show in the plan
4So what

The effect on a user outcome, a business metric or a risk.

Late plans, understaffed shifts
5Points to

A direction, not a solution. A "How might we" works well.

"How might we show absences while planning?"
+Credibility

Evidence, limits, confidence, the decision it informs and an owner.

Confidence: medium

Who + what, because why. This matters because so what, so where it points.

An insight statement has five parts in one or two sentences, plus the rows that make it credible.

Use it when

  • You've finished synthesis and have a wall of findings that nobody will act on.
  • You're writing the three to five points a busy stakeholder will actually read.
  • You want to share early results without anyone mistaking one quote for proof.

How to use it

  1. Start from a pattern, not a quote. Pick a finding that shows up across participants, from your themes or affinity map. Merge findings that share a reason. Output: one finding with the number of participants behind it.
  2. Find the why. Go back to the extracts and write the reason in the participants' terms. If you can think of two different whys, note both: they point to different designs. Output: a reason you can back with quotes.
  3. Write the sentence. Who, what, because why. This matters because the effect on a goal, so the direction it points to. Name a specific group, and keep the direction a direction, not a solution. Output: one or two sentences.
  4. Add evidence, limits and confidence. List participant counts, key quotes and other sources, say who wasn't in the study, and give a level: high, medium or low. Output: the rows that make the sentence credible.
  5. Test it with three questions. Would anyone disagree? Does it explain something? Could a team act on it next sprint? If not, rewrite it, merge it or move it to the appendix. Output: three to five insights per study, not thirty.
  6. Attach the decision and an owner. Write the decision it informs, a recommendation proportionate to the confidence, and a team with a review date. Output: an insight someone owns.

Worked example

A study with 8 supervisors who use an internal shift-planning tool.

Part Text
Finding 5 of 8 supervisors exported the weekly plan to a spreadsheet before publishing it.
Insight Supervisors who plan night shifts rebuild the plan in a spreadsheet before publishing, because the plan view doesn't show who is on leave or sick. This matters because plans are published late and shifts start understaffed, so planning should show real availability where the plan is made.
Evidence Observed in 5 of 8 sessions (P1, P2, P4, P6, P7). "I don't trust it until I've checked the leave list." P4
Limits Night-shift supervisors at two sites; no day-shift planners.
Confidence Medium: a clear pattern, said and observed, from one method. Not yet checked against usage data.
How might we …show absences while someone is planning, not after?
Decision and owner Next quarter's planning roadmap. Planning team, review in six weeks.

Watch out for

  • A finding dressed as an insight. "62% skip onboarding" is a number without a why. Ask "because?" until you have a reason.
  • A solution dressed as an insight. "Supervisors need a dashboard" closes the design space. Write the need and let the direction stay open.
  • "Users". A specific group tells the team who to design for and who wasn't in the study.
  • No confidence level. Without it, a single quote reads like a law. Write the level next to every insight, and recommend anyway.

Template

Copy one block per insight into your report or research repository.

Insight (1-2 sentences)
[Who] [does or struggles with what], because [the reason, in their terms].
This matters because [effect on a user or business goal],
so [the direction it points to].

Evidence:      [n of N participants, quotes with participant IDs, other sources]
Limits:        [who wasn't in the study, what wasn't observed]
Confidence:    High / Medium / Low, because ...
How might we:  ...?
Decision:      [the decision it informs]
Recommendation:[specific, proportionate to the confidence]
Owner, review: [team], [date]

Test: Would anyone disagree? Does it explain? Could a team act next sprint?
Source
Data, findings and insights: NN/g (Sara Ramaswamy, 2023)

"See on Amazon" links are affiliate links. As an Amazon Associate I earn from qualifying purchases, at no extra cost to you.

Saved only on this device.

↑ ↓ to move · Enter to open · Esc to close