Three vantage points on the same climate. The distance between them is the report.
A sponsored engagement runs the instrument across leadership, managers and employees at once — then reads where their accounts diverge. Individual answers stay non-traceable throughout.
The smallest engagement is 5 people — one answering under named authority and four others. Below that, no employee group can be reported without identifying someone, so there is nothing to set the leadership answers against.
Each seat answers a different instrument
The evidence-based Leadership Diagnostic — counts, budgets, dates, named projects. Claims you can check, not impressions.
The layer that translates intent into permission. They report what they can actually authorise, fund and protect.
The lived climate: what happens when you raise a concern, ask for time, or turn out to be wrong.
A run has three states, and no clock
We don’t promise you a four-week programme, because the system doesn’t have dates in it. A run is a draft, then open, then closed — and it moves when you’re ready, not when a schedule says so.
Scope, slices and the three declarations: employee representation consulted, which is required, plus benchmark contribution and public mention, which are optional. Each is recorded with a version and a timestamp. Until the text is final the run does not open: the API refuses provisional text.
Invitations go out by seat. While it’s open you see participation counts — how many people have finished, not what they said, and not who.
Six dimension scores per vantage point, the gaps between them, and the patterns those gaps fire — each labelled with the rung it stands on. Then a readout with your team.
You get distributions. You never get people.
The limits below are not settings an administrator can turn off. A required check in our pipeline fails the build if they stop being true: no migration may add a per-respondent identity column to the results table, and no organizational code path may read the answers table outside the sanctioned aggregation boundary.
Dimension scores per vantage point, gap magnitudes, fired patterns with confidence, and the questions to ask next.
Individual responses, re-identifiable slices, free-text attributed to a person, or any slice under five respondents.
A slice that would identify a team is suppressed and shown as suppressed — not quietly folded into the average.
Employee representation consulted where required, declared with a version and a timestamp before a run may open. Two further declarations, on benchmarking and on being named publicly, are optional. The API refuses to open a run on provisional text.
Five steps from an enquiry to a reading with your leadership team.
There is no self-serve sign-up for a sponsored engagement, and that is deliberate: the declarations below have to be made by a named person before anyone is asked a question.
- 01Scoping call
Who you want to hear from, how many of them, and which question the engagement is meant to settle. We size the wave and quote it here — nothing is priced before we know the shape.
- 02Declarations
Before a run may open, a named sponsor records what will be measured and who will see it. The platform enforces this: a wave cannot be fielded without it.
- 03Invitations go out
Each seat gets its own instrument — leadership, managers, employees, and partners or customers if the wave includes them. Respondents answer from a link; they need no account.
- 04Fielding
You watch response counts as they arrive, never answers. Nothing is shown below five respondents in a group, so a small team cannot be read back to itself.
- 05The reading
We compute the wave and go through it with your leadership team: the six dimensions, where the accounts diverge, and which named patterns fired with what confidence.
A second wave later measures the same organisation again and is reported beside the first, so a change is a change rather than a new number.
Tell us who's asking, and why now
You'll hear back from a person, not a sequence. Zenshin Kaishi is small; we answer as soon as we can.
Prefer to see it first? Run the free opening questions yourself — it's the employee instrument, and it's one sitting.