54% of UK small businesses are now using AI. Here's why that number is almost meaningless.

I couldn’t help but notice this stat from The British Chamber of Commerce recently: 54% of UK small businesses are now using AI. This is up from 35% last year and 25% the year before. 

What at first might seem like cause for celebration is actually almost meaningless. I’m all for celebrating AI adoption in UK small business, but we’re measuring the wrong thing here, folks.

“Using AI” and “getting value from AI” are two very different things, and it’s in the gap between them where businesses are quietly getting left behind.

What does “using AI” actually mean?

This stat doesn’t distinguish between a small business using AI to restructure their lead gen process, or offhand work from multiple roles to an AI agent cluster, and a company with a Copilot licence occasionally drafting an email.

Consider it alongside other research from Deloitte that shows 37% of companies adopting AI do so at a surface level, with no measurable benefit. Or, more specific to AI in UK SMEs where only 11% of businesses were found to be using any sort of technology to a great extent to automate or streamline operations.

It’s obvious that AI adoption alone isn’t helping. If we count any level of interaction with an AI solution in work as “adoption”, we’re measuring it wrong and not giving ourselves the opportunity to find out what might actually work.

Surface adoption vs structural adoption

So, we need to distinguish between different types of AI adoption. There are at least 2. Let’s call them surface adoption and structural adoption. 

Surface adoption is that colleague who copy and pastes their emails into ChatGPT so it can draft responses quicker. Structural adoption operates at process level. It’s aligned with how the business works and embedded into their systems. It works unprompted and around the clock.

Why is this distinction important? It comes down to how the value compounds. Surface level adopters (individuals using AI to save themselves time) get value, but if they leave the business, the value goes with them. Structural AI (automations, agents, integrated workflows) compound over time, scale with the business, and work regardless of who is working around them.

The fact is most UK businesses are at surface level adoption. That’s not a failure on their part - it’s a starting point. The question is where do they go from here?

How and when to move

The obvious answer is they move towards structural adoption. The problem with doing this is most jump into advanced solutions before fixing the foundations, or invest in off-the-shelf tools that go unused. This leads to high failure rates due to unrealised ROI.

Successful AI adoptions in small businesses are almost always preceded by some sort of technical audit. Mapping what they do, finding out where time is really being lost, and identifying where exactly an automation or agent would have the biggest impact. This might be just 1-2 processes, but they’re the right processes.

It’s not sexy, it’s not particularly groundbreaking, but that’s exactly why it works.

What now?

So where does that leave us? That 54% will keep climbing of course, but that doesn’t mean much for the businesses that make up the number if they’re just ticking the AI box and getting nothing from it.

The real question isn’t “are you using AI?” it’s “how are you using AI? Does it keep working when you’re not watching?”.

If you’re a UK small business interested in doing structural adoption right, it’s worth a quick chat with us. Contact us to find out more.