Three names for the same service — how a company model became the single source of truth
A B2B services company had no written way of operating — services, prices, and processes were scattered across spreadsheets, boards, and unwritten knowledge. We built a company model: a structured description of the organization readable by both people and AI tools, from which offers and documents are now generated.
Knowledge about the company was scattered and unwritten
The company grew faster than it formalized how it worked. The service structure lived in the sales team's spreadsheet, the client-facing methodology on a virtual board, delivery processes in the task management system, and pricing rules remained the unwritten knowledge of a few people.
The effects were felt daily. When someone asked for the current version of the price list, scope, or document, the answer pointed to a specific person, not a place. The organization depended on the availability of individual employees to a degree no one had previously measured.
This same problem limited the value of AI tools. A system without information about which version of the price list applies doesn't flag uncertainty — it generates a plausible-sounding answer, often incorrect.
What we didHow we built the company model
- Defined entities and relationships — offer, service, package, pricing model, client segment, outcome, competence, process — creating a map of the organization: which service addresses which segment, who delivers it, and by what pricing rule.
- Resolved contradictions revealed during the description — the process forced decisions that had been deferred for years. Every service received one name, one scope, one pricing rule, and one owner.
- Separated layers — the task management system remained the team's working interface but stopped serving as the source of truth about the company's structure. That role moved to a separate model.
- Connected the model to AI tools via a connector, so the entire team works from the same context without repeating instructions for every task.
What changed
Offers, scopes of work, and orders are generated from a single source — changing a price in the model updates all documents instead of requiring parallel corrections across several files.
The client methodology, previously an individual competence, was written down as a framework accessible to the entire team.
The company now has an unambiguous answer to the question of what its own services cost and what they include.
Next stage: bidirectional synchronization between the model and the task management system.
“If the answer to what your price list looks like depends on who you ask, no AI tool can fix that — you'll just generate the same confusion faster.”
Let's write down how your company actually works.
- 45 minutes, and no preparation needed
- An honest read on where you sit today
- The first process worth automating — and why that one
