Originally written in French. Translated by AI — the meaning has been preserved, not the prose.
Main idea
A Jira ticket can say who opened it, which customer it belongs to, what its status is and who owns it. A Notion page can say what it contains and when it was modified. A database row can say that a fact occurred on a date. All these objects hold the event perfectly.
What none of them holds is what the event taught the team: why this ticket mattered more than the others, which product tension it connected to, which past decision it called into question, which weak signal it confirmed, which idea will be left of it in three months. That reasoning exists at the moment the ticket is handled, and it has nowhere to be deposited — the "comment" field receives it, but nothing makes it retrievable as reasoning.
The value is therefore not in the object, it is around it. An information system that records only the object loses everything that surrounded it, and never notices, since what it keeps stays accurate.
Why it matters
It shifts the diagnosis when an organization notices that it "forgets". The reflex is to look for a better hierarchy, a better naming scheme, a better search engine — that is, to treat a filing problem. But retrieving better what was never written yields nothing.
It also says what to look at to judge a documentation tool: not its ability to classify, but its ability to receive a piece of reasoning and connect it to the object that prompted it.
Nuances and limits
Some objects carry part of the reasoning by construction: a decision minute, an incident review, an arbitration note. They remain the exception, and they are written once, not updated.
That exception has a tooled version on the product side, and it confirms the gap more than it closes it. A cross-cutting decision rule written with its context, its rejected alternatives and its accepted consequences keeps exactly what no card keeps: why the company ruled that way, and what would have to change for it to rule otherwise. But the setup is only worth something while it stays rare — a few rules a year — whereas the information system records everything. The gap pointed at here is therefore not filled by one more object: it is displaced toward the question of knowing which pieces of reasoning, among all those that run through the organization, deserve to be written down.
Layer added by "Product Decision Record: tracing the product choices that shape the company" (2026-06-03).
And not every piece of reasoning deserves to be kept. Much of what is said around a ticket is disposable; the gap pointed at here only becomes costly for what would have served elsewhere.
Open questions
- At what point in handling a ticket can you capture what it teaches without weighing down the work of whoever is handling it?