AI doesn’t create commercial inconsistency. It usually reveals how much was already there.

When the organisation still relies on people to hold things together

For a long time, most companies could absorb quite a lot of inconsistency without it becoming fully visible.

Not because the inconsistencies didn’t exist.

Mostly because experienced people were quietly compensating for them every day.

You would see it in customer conversations.

A sales lead reframes something instinctively because they can feel the discussion drifting away from the actual buying problem.

A founder steps into a meeting and reconnects the conversation back to what customers really care about after the messaging starts becoming too feature-heavy.

Someone in Customer Success explains the product differently again, but in a way that still somehow lands because they’ve spent years close to the operational reality underneath the customer environment.

Very little of this tends to exist formally.

The organisation just develops people who know how to carry the reasoning properly because they’ve spent enough time inside the patterns. They know which explanations hold up under pressure and which ones only sound convincing internally.

And for a while, that’s usually enough.

The gaps stay manageable because humans are continuously repairing coherence in real time.

Most organisations probably rely on this far more than they realise.

Where things start becoming heavier

What starts happening after a certain stage of growth is that the organisation gradually loses the ability to rely on shared memory.

The people joining later inherit fragments instead of origins.

They inherit a positioning statement, but not necessarily the customer conversations that shaped it. They inherit a sales deck, but not always the trade-offs that sat underneath it when it was first created. Messaging gets adapted by different teams over time, each trying to make it more useful for their own operational context, and with every adaptation it becomes a little harder to see where the original reasoning began.

None of this usually feels dangerous while it’s happening.

In fact, most of the people involved are trying to make the organisation more effective. Marketing is trying to improve engagement. Sales is trying to move conversations forward. Product is trying to explain value through the lens of what is actually being built. Viewed locally, most of the decisions make sense.

That is probably why the drift is so difficult to recognise early.

The organisation experiences adaptation. What it doesn’t necessarily see is the growing amount of interpretation sitting underneath that adaptation.

Over time, more explanation becomes necessary to reconnect things that previously held together naturally. Meetings begin to appear that exist largely to clarify what somebody originally meant. Enablement gets rewritten because different teams are solving for slightly different assumptions while still using the same terminology. None of these activities seem significant on their own, but collectively they create a surprising amount of invisible coordination work.

This is usually the point where organisations become increasingly dependent on a small number of experienced people.

Not because they are better communicators.

More often because they still remember where the reasoning came from before it spread across the organisation. They can reconnect customer problems to positioning decisions, positioning decisions to sales conversations, and sales conversations back to the operational realities customers are actually trying to navigate.

The organisation starts relying on them as a kind of continuity layer, even if nobody describes it that way.

What AI seems to expose very quickly

Before AI, organisations could often carry a surprising amount of inconsistency without it becoming fully visible because people compensated instinctively.

Someone would hear an explanation that wasn’t quite right and quietly correct it. An experienced salesperson would adjust the positioning in real time because they could feel the conversation drifting away from how customers actually buy. A founder might step into a discussion and reconnect everyone to the underlying business problem before confusion spread any further.

These corrections rarely appeared in systems, documents, or processes.

They happened in conversations.

They happened because people had accumulated context over time.

What stands out with AI is that it changes those dynamics quite significantly.

AI does not know which interpretation inside the organisation is the one that consistently survives contact with customers. It doesn’t know which explanation emerged from years of learning and which one was introduced three months ago because it happened to perform well in a particular campaign.

It simply works with the reasoning structures it is given.

That may not be the whole issue, but it is probably one of the stronger signals emerging from organisations experimenting heavily with AI-assisted commercial workflows.

When the underlying commercial logic is coherent, AI can be remarkably useful. Teams move faster, content production becomes lighter, and certain operational bottlenecks start disappearing.

But when fragmented interpretations already exist underneath the surface, AI can reproduce and distribute those inconsistencies very efficiently.

What makes this difficult to spot is that many of the visible signals improve at the same time.

The language becomes clearer. Outputs become more structured. Confidence increases. On the surface, everything appears to be moving in the right direction.

Yet those improvements can sometimes conceal a different question altogether.

Is the organisation becoming more coherent?

Or is it simply becoming better at expressing multiple interpretations of the same thing?

That distinction started to matter more and more the deeper we looked into it.

The difficult part is that alignment can still feel real

What makes this awkward operationally is that companies often still believe they are aligned while this is happening.

And to some extent, they are.

At a high level, everyone usually still feels connected to the same story. The terminology still matches. The positioning still sounds familiar and, from a distance, the strategic language still appears coherent.

Nothing obviously looks broken.

It is usually only after spending time inside the day-to-day operating conditions that the tensions start becoming visible.

Different teams begin carrying different assumptions about what customers value most.

Economic framing shifts depending on who is speaking.

Sales conversations evolve faster than enablement.

Marketing starts optimising around signals that product teams don’t fully recognise.

Customer Success reframes outcomes around operational adoption because that’s what customers are actually struggling with after implementation.

None of this is irrational individually.

But collectively, the organisation starts carrying more and more invisible coordination work just to preserve coherence.

That operational weight builds slowly.

More clarification.

More interpretation.

More translation layers between teams.

More dependency on institutional memory.

More people spending time reconnecting explanations that used to hold together naturally.

What we kept circling back to

Over time we found ourselves becoming less interested in content generation itself and more interested in the conditions underneath it.

How does an organisation preserve commercial reasoning once interpretation starts spreading across teams, workflows, tooling, AI systems, and operational scale?

How do you allow adaptation without gradually losing coherence underneath it?

How do organisations stop generating new variations of the same explanation every time another layer of growth gets added?

That probably sounds like a messaging problem on the surface.

I don’t think it really is.

The stronger signal is probably somewhere underneath that.

More to do with whether organisations can preserve a shared understanding of why customers buy once the business no longer fits inside the same conversational space.