From local workflow to shared ontology
One reason organisational modelling struggles is that it often feels too large to begin.
The enterprise is too complex. The terminology is unstable. The systems do not line up. The teams use different language. The documentation is incomplete. The real work contains too many exceptions.
So the organisation postpones the harder structural work. Or it launches a giant modelling exercise that becomes abstract before it becomes useful.
Both responses are understandable. Neither is very good.
A better path is to grow shared structure from local workflow outward.
Organisations do not need to model everything up front
A lot of modelling efforts fail because they aim for totality too early.
They try to define the whole organisation in one pass. Every capability. Every process. Every relationship. Every object. Every term.
That creates a heavy burden before value is visible. It also creates pressure to settle concepts that the organisation has not yet learned how to use consistently.
In practice, this often leads to two outcomes:
- a large abstract model with weak operational adoption
- a stalled effort that never becomes part of real work
Neither outcome helps much.
The more useful path is gradual structural learning
An organisation can build a stronger model by starting where knowledge is already being exercised.
That usually means local workflow. The places where work is already happening, decisions are already being made, handoffs are already occurring, and ambiguity is already producing friction.
Those local points are not a distraction from the enterprise model. They are one of the best starting points for building it.
They show:
- which concepts actually matter
- which relationships are operationally important
- where terminology breaks down
- where dependency knowledge is missing
- where the same patterns recur across teams
This is valuable because it turns modelling into learning instead of speculative completeness.
Local workflow is where organisational truth shows itself
Official documents tell part of the story. Local workflow often reveals the rest.
It shows what people actually need to know to get work done. It shows where formal structure is sufficient and where it is not. It shows which categories are real enough to support coordination and which ones collapse under use.
If an organisation pays attention to that layer, it can start turning local working knowledge into shared structure.
That is how ontology becomes grounded instead of ornamental.
Shared ontology does not have to begin as a grand taxonomy
The word ontology can sound heavier than it needs to.
At a practical level, it means building shared understanding of the things the organisation needs to reason about, and the relationships between them.
That can start simply.
For example:
- what counts as a service
- how a capability differs from a team
- what a dependency really is
- what states matter in a workflow
- how work, decision, risk, evidence, and ownership connect
These are not just modelling questions. They are coordination questions.
When the organisation gets them clearer, local work becomes easier to connect. When it does not, every team ends up reinventing its own private ontology.
The bridge is contribution, not central declaration
A lot of organisations assume shared structure must arrive through central definition first. Sometimes that is necessary. But if it becomes the only path, progress slows down and adoption weakens.
A stronger pattern is contribution.
Local workflow surfaces contribute useful structural learning back into a shared model. The shared model then improves later local experiences. That creates a feedback loop.
In that loop:
- local work informs shared structure
- shared structure improves local work
- repeated use sharpens concepts over time
This is more realistic than expecting a central group to define the whole ontology in isolation.
Gradual growth also makes disagreement more usable
Another advantage of growing a shared ontology gradually is that disagreement becomes informative instead of fatal.
When teams use different language or categories, that does not have to block progress completely. It can help reveal:
- where concepts are unstable
- where the organisation has hidden variation
- where shared terms are overloaded
- where local differences are real and where they are accidental
That gives the organisation a better basis for deciding what truly needs to be common and what can remain contextual.
This approach supports coherence without demanding perfection first
Many organisations delay structural work because they think the model has to be clean before it can be useful.
That is backwards.
A useful model often becomes cleaner because it is used. Because gaps are exposed. Because conflicts become visible. Because recurring workflow pressure reveals what the organisation actually needs to stabilise.
That is one reason I think of this as a path from local workflow to shared ontology. The point is not to avoid structure. The point is to let structure emerge through meaningful organisational use.
Why this matters for knowledge and governance
When local workflow can feed a shared structure, the organisation gets more than a better model. It gets a better foundation for governance and knowledge reuse.
It becomes easier to:
- trace where a concept is used
- compare patterns across teams
- detect recurring gaps or contradictions
- preserve useful reasoning near the point of work
- reduce duplicated interpretation effort
- build experiences that stay locally relevant while remaining structurally connected
This is part of how an organisation becomes more legible to itself.
The point
An organisation does not need to choose between local pragmatism and enterprise structure.
It can use local workflow as the path into enterprise structure.
That is often the more honest route anyway. Because local workflow is where the organisation reveals what its concepts, relationships, and gaps actually are.
If that learning is captured, connected, and reused, the organisation can gradually grow from scattered local knowledge toward a more shared ontology.
Not all at once. Not as a giant abstract exercise. But as a living structural accumulation of what the organisation learns through work.