On 25 September, NatWest published research on AI adoption across UK businesses, based on a survey of 1,400 small and mid-sized firms. The headline is that a gap is opening. 44% of businesses now use AI, but only 6% have reached what the report calls transformational adoption. And the split by size is sharp: 67% of firms with 100 or more employees use AI, against 36% of smaller businesses.

For the owner of a smaller agency, that can read like a verdict. It is worth reading the detail before accepting it.

The most interesting finding is what the gap does not follow. NatWest reports that headcount explains 39% of the variation in AI use, turnover 24%, and region just 8%. In other words, where a business is based matters far less than how many people it has. The bank's chief economist, Seb Burnside, describes the opportunity as not confined to the UK's major technology centres. A studio in Leeds or Bristol is not held back by its postcode.

So why would size matter? The report does not settle that, and the survey covers UK SMEs and mid-market firms generally rather than agencies, so any answer here is Go Wisely's reading rather than NatWest's finding. The likeliest explanation is capacity. A larger firm can give someone the time to test tools, write guidance and train colleagues. In an agency of thirty or forty people, everyone is on billable work, and the person best placed to work out how AI fits is also the person the clients need on Thursday. The problem is one of spare hours and structure, and neither is a matter of talent or ambition.

That matters because it changes what a smaller agency should do about it. If the obstacle were ability, the answer would be a long learning curve. If the obstacle is capacity, the answer is to make the work of adoption smaller and to have someone else carry part of it.

There are a few practical ways to do that. Start with an honest picture of where the team is now, because most agencies overestimate how much AI use is happening and underestimate how much of it is hidden in individual habits. Choose one or two recurring jobs rather than a general ambition to "use AI more". Protect a small, regular slot for practice, an hour a week is enough to begin with, rather than hoping it happens between projects. And agree who checks the output, so the tools do not become a source of quiet risk.

One events and film agency of around 36 people, a Go Wisely client, is a fair example of a smaller firm doing this. Go Wisely began with a full AI readiness diagnostic across six dimensions, so the work started from where the team actually was. It then ran a six-session bespoke AI Accelerator programme, built around the agency's real work and the tools already in its environment, which was Copilot. Across the programme, satisfaction averaged 4.6 out of 5 from 131 survey responses, and not one response was dissatisfied. The agency did not need to become bigger to get there. It needed the work to be structured and the time to be protected.

A note on the source. NatWest is a bank, and the same release announces its own upskilling and partnership offers, so read the report as useful evidence from a party with an interest rather than as neutral research. The pattern is still worth taking seriously, because it matches what smaller businesses tend to report about time and support.

The reassuring part is the 6%. Only a small fraction of UK businesses of any size have reached transformational adoption, so no agency, large or small, is far behind a finish line that almost nobody has crossed. Being in the 36% is not a failing. It is a starting position, and the evidence suggests it is shaped more by spare capacity than by anything about the people in the team.

Smaller agencies do not need to match the pace of larger ones. They need a sensible first job, a little protected time, and someone to help them keep it simple.

Go wisely.