what the document was never holding
A research report can be matched by any competitor who hires the same firm. The thing that actually reads a market can’t be.
The free edition argued that AI training absorbs everything that was ever made legible — every recording, every transcription, every documented output — and nothing else. The Pythagorean ear, the Dadaist sensibility, the microtonal instinct that lives in the hands rather than in theory: these never became data because the people who held them never needed to extract them. This edition goes into what that means structurally for organisations whose position depends on holding territory the training data cannot reach.
Angine de Poitrine’s modified guitar produces notes that don’t exist in Western music because two people spent years building an instrument around an ear that already heard them. None of that ever became a document. It didn’t need to.
Most organisations have a version of this, in a less interesting key. The country lead who knows, before the research comes back, which line in a campaign won’t survive contact with the market she has run for twenty years. The person who was in the room when the brand decision actually got made, not the one who wrote the case study explaining it two years later. This is formation. It was never a deliverable because the people who held it were too busy using it to write it down.
what happens when formation is asked to become a deliverable
The instinct inside most organisations is to ask the person who holds this kind of knowledge to translate it — into a framework, a deck, a research report — so it can be passed on, scaled, taught to whoever comes next. The translation feels like good practice. It is also the precise act that makes the knowledge absorbable.
A deck can be copied by a competitor who hires the same research firm. A framework can be learned by anyone trained on enough frameworks, including a model. The deliverable outlives the person who built it. The formation that produced the deliverable does not travel with it — it stays with the person. When the person leaves, the organisation is left holding a very well-written description of something it no longer has.
This is not an argument against documentation. Every organisation needs shared language, case studies, and a record of what was learned. The error is in what organisations think they’re buying when they pay for it. They commission the translation and then treat the translation as the asset. What was actually expensive to build — and is genuinely difficult to replace — was never the document. It was the person who could write a different one tomorrow, and a truer one the day after, because what they hold is a way of reading a market, not a result they once produced from reading it.
where the formation has to sit
The timing question is not abstract. It determines whether the formation does any work at all.
