Journal 020

Businesses May Be Less AI Ready Than They Think

Our first benchmark suggests many businesses may be harder for AI to understand than their owners realise. It's an early observation—not yet a conclusion.

5 min

Written by

Businesses May Be Less AI Ready Than They Think

Before we began this investigation, we had an assumption.

Surely established organisations would already be easy for AI to understand.

Many had been online for years.

Some had invested heavily in search engine optimisation.

Others were household names.

Surely AI would understand them too.

The evidence suggested otherwise.


The First Surprise

Across our initial benchmark we kept seeing the same pattern.

The problem wasn't that businesses lacked information.

The problem was that the information AI needed wasn't always easy to find.

Sometimes it was buried.

Sometimes it was implied rather than stated.

Sometimes it simply wasn't there.

The evidence existed.

The understanding didn't always follow.


Why That Matters

People are remarkably good at filling in the gaps.

AI isn't.

A visitor might browse several pages before understanding what a business does.

AI has to build that understanding from evidence.

If the evidence is weak...

...confidence falls.


Good Businesses Didn't Always Produce Good Understanding

This was perhaps the biggest surprise.

Some excellent organisations communicated surprisingly little about themselves.

Others clearly described their products but never explained who they served.

Some inspired trust with people...

...but provided very little evidence AI could confidently use to reach the same conclusion.

Being a good business doesn't automatically make you easy for AI to understand.


That Doesn't Mean They're Wrong

It's important not to overstate the evidence.

This doesn't mean those businesses are failing.

Nor does it mean AI is always right.

It simply suggests there is often a gap between what a business believes it communicates...

...and what AI can confidently understand.

That gap is exactly what we're trying to measure.


An Observation, Not A Conclusion

One hundred organisations is a useful engineering benchmark.

It is not a representative sample of every business.

These are observations.

Not conclusions.

We're publishing them because we believe understanding should grow alongside evidence.

If later evidence proves today's observations wrong...

...we'll say so.


Where This Leads

Our benchmark is already growing.

Every new assessment gives us another opportunity to test these observations.

Some patterns will become stronger.

Others may disappear.

That's exactly how we want this investigation to work.

Evidence should shape the conclusions—not the other way around.

The value of this project isn't that it confirms our assumptions.

It's that it's willing to challenge them.