What we know
What we know is one organisation's register of insights: short statements of what has been found, each with the evidence behind it and a plain note of how sure we are. It sits alongside the survey results rather than replacing them. A result says what people answered; an insight says what you have concluded from it, and keeps that conclusion honest over time.
Open an organisation and choose What we know. The list shows its insights, grouped by the lens you choose.
What an insight is
An insight states what was found, never what caused it. "Fewer people in Operations say they can raise a concern" is an insight. "Operations managers are discouraging people from speaking up" is a theory about a cause, and belongs in a hypothesis in the listening purpose until something tests it.
Each insight is one of four kinds:
- Fact: something found in the results, for one group or for everyone who answered.
- Difference: two groups compared on the same field, such as two departments or two locations. The groups must have no value in common, so nobody can be counted on both sides.
- Pattern: a regularity you have seen in the results.
- Trial result: what happened when an action was tried. A trial result needs an action review, so you will not be able to record one until that part arrives later.
How sure we are
Every insight carries a How sure rating: sparse, emerging, moderate or strong, with a sentence saying why. Choose the rating the evidence supports today. An insight starts as a draft or current, and can later be marked contested or retired.
Nothing is deleted. Every change to the wording, the rating or the status needs a reason, and each change adds a revision rather than overwriting the last one, so you can always see what was said before and why it changed. Marking an insight Contested or Retired also needs a reason, and a contested insight cannot be rated strong.
Evidence
Attach findings from a closed survey to show what an insight rests on. Choose the survey, the finding, and whether it supports, challenges or mixes with the insight. What you attach is a snapshot of that finding as it stood, so a later change to the survey cannot quietly alter an insight you have already recorded. If the survey's results have changed since you looked, you are asked to choose the finding again.
You can withdraw evidence, with a reason. It stays visible in the history; it just stops counting.
The small-group rule
A finding about a group is only shown while the group is at or above the survey's size limit. If it falls below, the group is hidden, its result is not offered as evidence, and the insight says that the group is no longer shown. This holds for both sides of a difference, and the two groups are checked so that taking one away from the other cannot reveal a small group by subtraction.
Listening gaps
Below the list, Gaps shows the themes in the chosen lens that have no current insight, or only sparse ones. A gap is not a failure. It is a candidate learning question for the next survey.
Who can use it
Practitioners with the right plan can add and change insights. Client users and anyone viewing as another person cannot, and a record can only be seen by the organisation that owns it.