Short answer: steering without indicators does not cost you a bad decision, it costs you a late one. An organisation that discovers its utilisation rate at year end, its bench at the close of the quarter and a fixed-price overrun on delivery day does not have an analysis problem. It has a latency problem. Three simple numbers, held weekly and shared, beat fifty indicators consulted twice a year.
The short version
- The symptom: running on instinct, with no view and no strategy, and finding out about mistakes too late.
- The mechanism: by the time the figures land, the decision they should have informed has already been taken. Reacting after the fact is not steering.
- The three blind spots named: utilisation excluding leave, bench time, and fixed-price project profitability.
- The fix: not a thousand KPIs. A few simple numbers, weekly, shared, and a twenty-minute ritual.
Why a late decision is a lost decision
Most organisations are not short of data. They have plenty of it, spread across a capacity spreadsheet, a time-tracking tool, the accounts and the memory of two or three people. What they are short of is the date on which that data becomes readable.
That is the distinction the video draws. A figure that is accurate but late is not half a figure: it is a post-mortem. You cannot restaff an employee onto a month that has already closed, renegotiate a fixed price that has already shipped, or recover an invoice that should have gone out six weeks earlier. Steering happens inside the window where something can still be decided, and that window closes fast.
The three blind spots
| What is not tracked | What you do not know | When you find out |
|---|---|---|
| Utilisation, excluding leave | Whether teams are genuinely productive | At the close, once margin is already booked |
| Bench time | How many sellable days quietly disappear | At the quarter, once the days are gone |
| Fixed-price profitability | Whether the effort budget still holds | On delivery, when no lever is left |
Excluding leave from utilisation is not an accounting nicety. A rate computed on raw working days makes a team on holiday look like a team with nothing to do, and an indicator that cries wolf twice a year is an indicator nobody reads any more.
To put a number on the second line, our bench cost calculator works out what an unsold day actually costs, per day, per month and per year, entirely in your browser.
The twenty-minute ritual
The end of the video is the cheapest part to put in place and the part almost nobody does: a short weekly slot, with the same numbers, in front of the same people.
- Weekly, because that is the cadence at which a staffing decision can still change.
- Simple, because a forty-row dashboard is never read to the bottom.
- Shared, because a number only the founder sees triggers nothing in the managers who actually staff.
- Short, because twenty minutes fits into a week and a half-day steering committee does not.
What that asks of the tooling is unglamorous but not optional: time, staffing and invoicing have to sit on one set of records. That is the whole point of the data repository; without it, the twenty-minute meeting turns back into an arbitration between two spreadsheets that disagree.
Video transcript
Full transcript, translated from the French and lightly tidied for reading.
Too many founders navigate on instinct, with no indicators, in the dark. The result: no view, no strategy. And the problem is that they find out about mistakes too late. By the time the numbers finally arrive, the decision has already expired. Reacting after the fact means taking what comes.
For example, with no tracking of utilisation excluding leave, you have no idea whether your teams are genuinely productive. With no tracking of bench time, you lose margin in silence. And with no profitability indicator on fixed-price work, you discover the overrun on delivery.
You do not need a thousand KPIs, just a few simple numbers, weekly, shared. Not hours of meetings: twenty minutes to steer instead of being steered. With the right map and the right ritual, your company stops improvising. It anticipates, and it moves.
Read next
- Three key indicators to move beyond instinctive decision-making : the long version of this video, figure by figure.
- The business steering module : capacity plan, staffing and project margin on one repository.
- Bench cost calculator : what an unsold day really costs, in two minutes.
Frequently asked questions
How many indicators does an organisation need?
Three to five, provided they are weekly and shared. The video's point is that the problem is not the number of indicators but their freshness: a rich but monthly dashboard always arrives after the decision it was meant to inform.
Why exclude leave from the utilisation rate?
Because a rate computed on raw working days turns a holiday period into apparent under-utilisation. Excluding leave means the indicator only measures what a manager can act on: the days that were genuinely available, and how they were staffed.
How often should these numbers be reviewed?
Weekly, in a short slot of around twenty minutes. That is the cadence at which a staffing move, a reinforcement or a chase can still change the outcome; monthly, all that is left is to note what happened.
