Curious Labs
Instruments that measure how leadership actually works.
We build measurement instruments for the people who run organisations. Each one asks the leader, asks the people around them, and reports the distance between the two answers.
The Time and Authority Audit
Made for one senior leader at a time, and for the organisation that needs to know what is actually holding them back. It names one constraint, not four, and it states how sure it is about every claim it makes.
What it measures
Four dimensions, read as a loop.
Unclear expectations push decisions upward. Decisions held too high eat the calendar. A calendar with no room is where standards go unsaid. And that feeds the first one again. The audit names the one link holding the other three.
Expectation clarity
Do people know what good looks like without asking.
Decision location
Are decisions held at the level that can carry them.
Time integrity
Does where the time goes match what is said to matter.
Consequence follow-through
Does anything happen when a standard slips.

Not one questionnaire
Nine questions decide which audit you get.
A leader three months into a bigger job is not carrying the same problem as one drowning in detail. Each version pulls different evidence, and only one of the four opens a calendar at all.
The Overloaded Operator
Time consumed by decisions that should have settled below them.
The New Scope
Recently promoted. The calendar is not full, it is undefined.
Influence Without Authority
Delivers through people who do not report to them.
The Standards Gap
Has the time. The team does not deliver.
How it runs
About ninety minutes of the leader’s time.
- Nine questionsRoute the leader to the version that fits their situation.
- Twenty-four statementsScored by the leader about themselves.
- Three to five peopleScore the same twenty-four statements independently. No answer is ever stored against a name.
- The readOne constraint named, every claim carrying how sure it is, and three commitments tied to real moments in the leader’s week.
- At thirty daysThe person named against each commitment confirms whether it happened. Not the leader.
- At ninety daysThe same people score the same statements again, against the baseline they set. Improved, held, gone backwards, or not separable from noise.
Where the data lives
Frankfurt. No passwords. Deleted on a schedule.
Every response is stored in a Postgres database in Frankfurt, Germany, run by Neon on AWS eu-central-1. The application code that reads it also executes in Frankfurt, so a European response is not routed through a server outside Europe to be scored. There are no passwords anywhere in the system, so there are none to steal. Respondent answers are never stored against a name, and below four respondents no individual row is shown at all. Everything is deleted on a schedule a job enforces daily, 365 days from the start or 90 days after the read is delivered, whichever comes first. No soft delete, no recovery bin.
The full security statement · What participants are told · A complete worked report
Enquire
Run it on one leader.
Tell us who, and what you are trying to find out. If the instrument is wrong for the situation we will say so.