Main idea

Choosing a direction isn't enough to make an arbitration. A committee that leaves a session with a formulation acceptable to marketing, product, sales and finance may only have found the sentence in which each one recognizes its own priority. The minutes will record a decision; the initial contradictions will be intact.

What distinguishes an arbitration is that it states the counterpart: accelerating a launch means accepting a narrower promise; preserving quality means accepting a longer timeline; protecting the margin means giving up part of the volume; holding a brand's coherence means turning down an appealing local opportunity. As long as that second half isn't written, the decision keeps in reserve the possibility that everything still holds together — and that is precisely what execution will find to be false.

The named loss is therefore the only observable sign that a choice took place. It can be read in the minutes, it stands against the requests that weren't kept, and it names the functions that will have to live with it.

Product decision documents draw a formal consequence from this: they reserve a section for negative consequences, next to the positive ones. A rule setting French and English as the only languages supported by default records there that some commercial opportunities will be turned down or deferred. This isn't an editorial precaution: the section is what keeps the rule from circulating as a pure improvement, and it makes the loss consultable months after the session, when nobody remembers what had been weighed against what.

Layer added by "Product Decision Record: tracing the product choices that shape the company" (2026-06-03).

The existence of such a document guarantees nothing, however: a decision log can record choice after choice without ever naming what was sacrificed, and it then becomes the record of a series of formulations that settled nothing. The criterion doesn't move to the artifact — the question remains, on each entry, what it forced you to write on the side of what was given up.

Layer added by "The Tools of Organizational Coherence" (2026-09-15).

Why it matters

This gives a simple test, applicable as you leave a meeting: reread the minutes and look for what the organization is giving up. If nothing is there, the subject wasn't settled, whatever sense of progress the participants have.

It also keeps you from confusing the production of an action plan with a decision: a plan can perfectly well roll out a promise nobody agreed to reduce.

Nuances and limits

Not every situation calls for an arbitration. When priorities really are compatible, looking for a loss to name would manufacture a pointless sacrifice — synthesis is then the right move.

And a loss can be named without being held to: minutes that write down what is given up but which each function then reinterprets its own way are worth no more than a soft formulation.

Open questions

• To what degree of precision must a loss be written down to withstand reinterpretation by the teams that execute it?