Making organizations work
This page was automatically translated from the original French version.
Organizations rarely stop working because nobody knows a management method. More often, they accumulate friction: uncertainty is hidden behind precise-looking numbers, the same decisions are discussed again, urgent matters consume the time meant for important work, and strategy grows into a list of ambitions that never requires a choice.
Over more than twenty years building Obeo, I have gradually assembled ways of dealing with some of these problems. None of them arrived as a complete theory. Most began with something that repeatedly went wrong, an obvious solution that proved insufficient, or a belief I eventually had to reverse.
This is not a collection of universal management rules. It is a record of things I believe now, and what happened to make me believe them. Each idea is provisional: useful in a particular context, open to objections, and sometimes still difficult to apply consistently.
Decide without pretending uncertainty has disappeared
Organizations need commitments, but commitment is not the same as artificial precision. A single estimate, forecast, or target is easy to communicate and dangerously easy to mistake for certainty.
That is why I now prefer estimates that preserve their uncertainty. In Estimating Without Erasing Uncertainty, I describe how early software estimates led us from late-night deliveries to three-point estimates: probable, low, and high. The arithmetic matters, but the most valuable change is in the conversation. It makes what we do not know part of the decision instead of silently discarding it.
The same issue appears in financial communication. Making company figures available does not automatically make them meaningful. People also need familiarity with orders of magnitude, recurring costs, cash flow, and the scale of the organization before the same number can produce a shared understanding.
Strategy creates a related tension. Organizations are usually good at listing desirable ambitions. They find it much harder to state what they will not pursue. Yet a strategy that never rules out an attractive option has probably not made the choices needed to guide everyday decisions.
Reduce friction before it becomes a crisis
Some practices look expensive because their cost is visible while the problems they prevent are not. Regular management is a good example. Frequent conversations take time, but they allow weak signals, misunderstandings, and small disagreements to be addressed while they are still inexpensive. Less frequent management can feel efficient until the accumulated problem requires a large meeting and a difficult recovery.
The calendar works in much the same way. Time for concentrated work, email, preparation, and recurring responsibilities has to be protected before everything becomes urgent. I have increasingly come to see “important and urgent” not merely as a priority category, but often as evidence that the system failed to anticipate something important early enough.
These are not recipes for filling every hour or scheduling more meetings. They are attempts to make important work visible before urgency makes the choice for us.
Give the organization a memory
Long-running organizations repeatedly lose the context behind their own decisions. Teams change, discussions are separated by months, and a reasonable question — “Why did we decide this?” — can restart an entire debate.
A shared decision log, inspired by architecture decision records, can preserve the context, alternatives, and consequences of important organizational choices. The idea is simple. Making it a durable habit across teams is much harder, and I have not solved that part completely.
Documentation creates a similar trap. When an intranet contains years of obsolete material, the obvious response is a complete cleanup. But nobody knows what can safely be deleted, so the task becomes both risky and enormous. A more useful question is often not “What should we remove?” but “What deserves to be indexed as the current reference?” Curation can restore a reliable path through the information without first repairing the entire past.
How these notes are written
I do not want these articles to read like management lessons with the story added afterwards. The story is the evidence: the early belief, the problem that resisted an obvious solution, the disagreement that exposed a missing condition, or the sequence of events that changed my mind.
Each note therefore concentrates on one claim that should be interesting enough to discuss on its own. It also tries to preserve the counterexample, the limits of the practice, and what remains unresolved. The most honest conclusion will often be: this is where I am today.
The first note in this series is about replacing single-number estimates with ranges that retain uncertainty. The other questions on this page will become articles as the experiences and stories behind them are ready to be told.