Do You Actually Know How Your Organisation Works?

Most leaders know their organisation.

They know their people, their customers, their products or services, their financial position, their strategy and the systems the organisation relies upon.

But there's another kind of organisational knowledge that can be surprisingly difficult to see from the top.

How does work actually get done?

Not how the process map says it gets done.

Not how the policy says it should happen.

Not how the system was designed to support it.

How does it actually happen on a Tuesday morning when somebody needs to get something done?

There are two versions of most organisations

There is the formal organisation.

It has systems, processes, policies, organisational charts, procedures and defined responsibilities.

Then there is the organisation people experience every day.

That's where someone exports information into Excel because it's easier to work with. Someone keeps their own document because they don't trust the version in the system. A team has developed an unofficial process because the official one takes too long. An employee knows who to call when something goes wrong, even though that knowledge has never been documented.

None of this necessarily means people are doing anything wrong.

Quite often, they're doing exactly what organisations need people to do.

They're solving problems.

But over time, those solutions become part of an invisible operating layer.

AI makes the invisible visible - or exposes the fact that we can't see it

This matters because AI increasingly depends upon organisational context.

If important knowledge lives primarily in people's heads, AI can't reliably access it.

If two teams use different sources for the same information, which should AI trust?

If the official workflow bears little resemblance to how work actually happens, which one are we teaching AI?

If documents exist across several locations, which ones represent the organisational record?

These aren't theoretical AI questions.

They're questions about how well an organisation understands itself.

The people doing the work know things leadership doesn't

One of the best ways to understand an organisation is simply to become curious about how people actually work.

Ask someone to show you.

Watch where they go for information.

Ask what happens next.

Ask what they do when the system doesn't work.

Ask which spreadsheet they couldn't live without.

Ask who they call when they don't know what to do.

Those conversations can reveal more about an organisation than a beautifully documented process map.

And importantly, they bring the people who actually perform the work into the AI conversation.

Because readiness can't be understood entirely from the boardroom or the IT department.

Organisational reality matters

AI is going to make our ability to understand how organisations really operate increasingly important.

Not because every workaround needs to disappear.

Not because every process needs to be documented perfectly.

But because we need to distinguish between the organisation we think we have and the organisation that actually exists.

The gap between those two can contain some of the most important information about where AI will work, where it won't, and where the greatest opportunities may be hiding.

Before AI can understand our organisation, perhaps we need to make sure we understand it ourselves.