Latest
Task number 19
There is no quotable improvement percentage: the tool's limit is not deduced from difficulty, it is tested task by task.
Articles
Working notes
What we learn from directing projects, putting processes in order, and getting technology to actually work. No textbook recipes: concrete problems, how they are diagnosed, and what changes once they are solved.
Series
1 partAI without the hype
Task number 19
There is no quotable improvement percentage: the tool's limit is not deduced from difficulty, it is tested task by task.
2 partsThe CEO’s time
The 4%
Emergencies take up 4 percent of the day: you chose that Tuesday, and the sense of firefighting has another engine.
The shape of the calendar
Neither shape of calendar is better: what costs you is that yours is not the one your company needs today.
3 partsCompeting in turbulence
Is everything really speeding up?
The slide confuses churn with death: acceleration is an aggregation myth, and the only measurable climate is your local regime.
The right to change your mind
On new ground, the plan is not a forecast to be met but a device for converting assumptions into knowledge at the lowest possible cost.
The organization built for the storm
In the thick of the crisis, firms that delegate decisions to the ground fall less: local information is worth more than control.
3 partsDeciding under uncertainty
The machine that decides
Bias is not cured by experience or intelligence: what fails in the head is corrected by designing the process.
The terrain sets the rules
The costliest error isn't a bias: it's applying the method of the wrong terrain and confusing the complicated with the complex.
Gut, rule or experiment?
Expert intuition feels the same whether it is expertise or illusion: only the environment where it was formed tells them apart.
3 partsProblems that don’t come back
The chronic firefighter
The incentive system makes visible whoever puts out the fire and invisible whoever prevented it —and the trap reinforces itself.
From symptom to cause
A root cause that doesn't explain all the available data isn't a cause: it's a conjecture in the shape of a conclusion.
The problem that doesn't come back
Finding the root cause once doesn't prevent recurrence: without a station that measures whether the problem came back, there's no capability.
3 partsHigh-performing teams
The myth of the united team
Collective performance is designed before the first meeting: chemistry and pooled talent do not make a team.
What actually moves the needle
A team's intelligence comes not from talent but from how its members interact, and a misplaced individual incentive destroys it.
From conditions to capabilities
What pays off is the structured and somewhat boring: cohesion is the consequence of working well together, not its cause.
Standalone notes
leadershipYou don't lead everyone the same way
The little grid's recipe almost never passed the tests: it is worth exactly what the boss's judgment is worth — and that is where bias lives.
management and dashboardsWhat gets measured gets managed
A green dashboard guarantees nothing: once a metric becomes a target, it stops telling the truth.
change managementWhy change initiatives fail
Resistance almost never explains a failed change: the problem is the design, and the early dip is the sign that it has begun.
professionalizationSigns an organization has outgrown its structure
The signals look like people problems, but they are a structure that became too small and collects its cost in silence.
operating financeThe profitable company that ran out of cash
Every extra dollar of sales locks up cash: growing above the sustainable rate breaks companies that make money.
commercial advisoryPrice tells a story
The lever that moves profit the most is the only one without an owner: without rules, every discount is decided by the urgency of a Friday.
capital projectsWhy capital projects go off track
An investment's fate is not decided on site: it is set by the quality of the engineering done before the funds are approved.
Contact
Does any of this sound like your own problem?
If something you read describes your situation, let's talk. The first meeting is for understanding the problem, not for selling a method.
Schedule an initial meeting