The AI growth gap: why company size matters more than location

Forty four per cent of UK businesses now use AI, but only 6% of those have reached the stage where it actually changes how the business runs, according to NatWest's AI Adoption Report, published 25 September 2026 from a survey of 1,400 UK SMEs and mid-market businesses. The gap between those two numbers is the real story, and it has almost nothing to do with where a business is based. Company size explains around 39% of the variation in AI use; region explains only about 8%. A small firm in Edinburgh and a small firm in a market town in England are starting from roughly the same place. A small firm and a 150-person firm are not.

Why does company size matter more than location?

The NatWest research found that 36% of smaller firms use AI, against 67% of businesses with more than 100 staff, almost double. That is not a Scotland problem or a regions-versus-London problem. It is a resourcing problem. Bigger firms have a person whose job includes "look into this," a training budget, and someone technical enough to get past the first awkward week. Smaller firms have an owner already doing five jobs, and AI becomes the sixth thing on a list that was already too long.

Robert Begbie, CEO of Commercial and Institutional Banking at NatWest, put it plainly in the same release: "There is a real risk of an AI adoption divide, with smaller businesses facing greater barriers, including limited specialist expertise, skills and investment."

Are Scottish businesses actually behind on AI?

Not in the way the headlines imply. According to Glasgow Chamber of Commerce's coverage of the same NatWest data, 82% of Scottish businesses already using AI report increased efficiency, 80% report greater innovation, and 66% report higher profitability. Scottish businesses that get going with AI see results in line with, or ahead of, the UK picture. The issue is not that Scotland lags on outcomes. It is that fewer small Scottish businesses have started, for the same size-driven reasons as small businesses everywhere else in the UK.

That distinction matters for how an owner should read this. If the gap were about location, the fix would be infrastructure or connectivity. Since it is about size, the fix is time, a specialist to ask, and a clear starting point, which is exactly what most small businesses are short of.

What's the difference between "using AI" and closing the gap?

Most of the 44% in the NatWest figures are using AI the way a person uses a calculator: open it, ask it something, close it again. That is real use and it has real value, but it is also exactly why only 6% reach the transformational stage. A chatbot that answers a question when someone remembers to open it is not the same as a system that pulls in a new enquiry, checks it, and starts the next step without anyone prompting it.

Think of it as a ladder. On the first rung, a person uses AI as a helper on demand. On the second, a specific job runs unprompted and a person checks the output before it goes anywhere. The jump from the first rung to the second is where the growth gap actually closes, and it is the first step from the 44% who use AI towards the 6% of users who are transforming. We have also written about why your team is probably already using AI informally, before anyone decided to build anything on purpose, which is usually the first thing worth finding out before building a system around it.

Is there help available for small Scottish businesses closing this gap?

There is a genuine, funded route worth checking. Scottish Enterprise has appointed Insight to deliver its AI Scaling Adoption Programme, working alongside Highlands and Islands Enterprise and South of Scotland Enterprise, as reported by ChannelWeb on 14 September 2026. It is part of the Scottish Government's wider AI Strategy and sits alongside the National AI Adoption Programme for SMEs. We cover the current grants and advice routes, what each one actually funds, and where to check live status, in AI grants for Scottish small businesses rather than repeating it here. Worth a look if advice and training cost, not just technology access, has been the thing holding you back.

What to do this month

You do not need a strategy document to start closing this gap. You need one finished example.

1. Pick one job that repeats every week: intake, quoting, chasing, reporting, or admin.
2. Write down exactly how it is done today, step by step, including who touches it.
3. Decide who checks the output before it goes anywhere else, and make that a rule.
4. Run it for a few weeks and measure the hours it actually saved.
5. Only then move on to the next job.

That sequence, repeated, is how a small business moves from occasional use towards the 6% of AI users who have actually changed how the work runs. If you want an outside read on where to start, the AI Ops Scorecard is a short set of questions that names the first thing worth fixing, and the AI Operations Day is built to turn that first job into a working system rather than another pilot that stalls.