AI for small business: how to train a team of 10 to 20

What does AI for small business actually mean?
For most small businesses, AI means the general assistants: ChatGPT, Microsoft Copilot, Google Gemini and Claude. Everyone uses them on the work they already do, such as emails, quotes, proposals, meeting notes, spreadsheets and customer replies. It rarely means a custom system or a new piece of software.
That is why so many guides on this topic are tool lists, and why they don't help much. A business of 15 people doesn't need a list of 40 apps. You probably already have access to one of the big assistants through Microsoft 365 or Google Workspace. What's missing is that most of your people haven't been shown how to use it well on their own work, and nobody has said what's allowed.
So this post is about the people. If you own or run a business of 10 to 20 staff, here is how to get all of them using AI properly in about a month.
Which AI tools does a small business need?
Fewer than the tool lists suggest. Most teams of 10 to 20 need one general assistant for everyone, plus whatever AI features are already built into software they use.
- One general assistant for the whole team. If you run on Microsoft 365, start with Copilot. If you run on Google Workspace, start with Gemini. ChatGPT and Claude are strong choices if you use neither, or if the work is mostly long documents and writing.
- The AI already inside your other software. Many accounting, customer service and design tools now include AI features. Switch on the ones that save a clear step, and teach them alongside the main assistant.
- A meeting notes tool, if your team spends a lot of time on calls. Check where it stores recordings before you turn it on.
Choose one main assistant and stick with it for the first few months. A team that learns one tool well gets far more from it than a team trying five. Use the business version rather than free personal accounts, because the two handle your data differently.
How many small businesses use AI?
Fewer than larger firms, and the gap is clear in the official figures. The ONS Business Insights and Conditions Survey for September 2026 found that 30% of all UK businesses used at least one AI technology. That was 29% for businesses with fewer than 10 employees, 38% for those with 10 or more, and 51% for those with 100 or more. Among firms with 10 or more staff, the most used kind was text generation with large language models, at 21%.
The government's AI Adoption Survey 2026, published by the Department for Science, Innovation and Technology on 29 October 2026, put overall adoption at 32%, up from 16% in 2025. It also asked what held businesses back. Limited AI skills, expertise or knowledge affected 67% of businesses already using AI and 78% of those planning to. Among businesses not using AI, 65% said they had no clear business case or proven return.
Read those two findings together. The usual barrier isn't cost or access. People don't know how to use the tools on their own work, so the owner can't see what it's worth. Hands-on training on real tasks deals with both, because people learn the skill and the owner sees the result on work they recognise.
Why training the team beats buying more tools
In a small business, one keen person usually starts using AI first. They get faster, the others don't, and the habit stays with one person. If that person leaves, so does the know-how.
Training the whole team at once changes that. Everyone learns the same way of working, uses the same approved tools and follows the same rules. A prompt that works for one person's quotes gets saved and shared, so the whole team benefits from it. In a business of 15, that shared habit counts for far more than any single tool.
It is also a safety issue. If staff have never been told which tools are approved, some will paste customer details into a free personal account. Our guide on how to use AI at work without creating a problem covers how common that is.
Set three rules before anyone trains
You don't need a long AI policy. A small business needs one page that answers three questions, agreed before the first session so the training can teach it.
- Which tools are approved. Usually the business version of whatever you already pay for, such as Copilot in Microsoft 365 or Gemini in Google Workspace, because the business versions handle your data differently from free personal accounts. Check the terms of your own plan.
- What must never go in. Customer personal data, staff records, passwords, and anything under a confidentiality agreement, unless your approved tool and your data protection position allow it.
- Who checks what goes out. Anything that leaves the business, such as a quote, a contract, advice or a figure in a report, gets read by a person who knows the subject before it is sent.
Write it down, share it and put it on the agenda for the first session. In a team of 15, everyone can know the rules by name within a week.
Pick everyone's first tasks
Training on generic examples is quickly forgotten. Training on a task someone does every week sticks, because they use it again the next morning. Before the sessions, ask each person for two or three recurring jobs that take time and are easy to check. In a business of 10 to 20 the first tasks tend to fall into a few groups:
| Who | Good first tasks | What a person still checks |
|---|---|---|
| Owner or manager | First drafts of proposals and tenders, summaries of long documents, thinking through a decision | The numbers, the commitments and anything sent to a client |
| Sales and account staff | Researching a prospect, follow-up emails, tidying call notes | Every claim about price, product or delivery |
| Office and admin | Turning rough notes into letters, minutes and procedures, sorting a busy inbox | Names, dates and anything personal |
| Finance or bookkeeping | Explaining a variance, drafting a cash-flow note, cleaning up a spreadsheet | Every figure against the source |
| Marketing | Drafts in your house style, social posts, rewriting web pages | Facts, claims and tone |
| Customer service | Draft replies to common questions, summarising a complaint | The reply before it goes out |
Start where a mistake is cheap and easy to spot. The point of the first month is to build the habit and learn where the tool goes wrong, not to automate your riskiest process. If someone asks for a bigger project, our guide on which workflow to automate first will help you choose.
A four-week plan for a team of 10 to 20
Here is the shape that works for a business this size. It fits around the day job, and nobody needs to be technical.
- Week 1: agree the one-page rules and the approved tool, and make sure everyone can log in to it. Collect each person's two or three first tasks.
- Week 2: hands-on sessions. Split the team into two groups of 5 to 10, grouped by the kind of work they do, so everyone gets time on the keyboard and the trainer can help each person. Each person works on their own tasks, using their own documents, under your rules.
- Week 3: practice on real work. Everyone uses AI on their chosen tasks every day and saves the prompts that worked to a shared document. Pick one or two people who take to it to answer colleagues' questions.
- Week 4: a short follow-up session. What is everyone using now? What stopped? Fix what stalled, share the best prompts, and choose the next tasks.
Two half-day sessions or one full day per group are both fine. What matters more is that the follow-up happens, because a single session without it fades fast. We explain why in why one-off AI training workshops don't stick.
How to tell whether it worked
Small businesses don't need a measurement framework. Before the sessions, note three things: how many people use AI on real work each week, how long two or three common tasks take, and a sample of a typical piece of work such as a quote or a customer reply.
Look again four to six weeks later. If more people use it every week, the common tasks are quicker and the work is as good or better, it worked. If not, you'll be able to see which group or which task stalled, and fix that rather than starting again. Our guide to measuring AI training success has more detail if you want it.
How much does AI training cost for a UK small business?
It depends mostly on how the training is priced. Open courses that individuals book onto are priced per person, so the bill grows with every member of staff you send. A private session for your own team is usually a flat fee for the day, so the cost per person falls as the group grows, and the content is built around your work rather than a standard course.
For a business of 10 to 20, that usually makes a private team day the better buy. Remember to count your team's time as well, because a day out of the business is a real cost for a small firm. Our guide to what AI training costs for a small team compares published UK prices and explains what moves a quote. If you want the simplest option, a hands-on AI workshop for your team is priced as a flat fee for the day.
Free courses, in-house or a trainer?
Start with what's free. Since January 2026, every adult in the UK can take free AI courses through the government's AI Skills Hub, some in under 20 minutes, checked against Skills England's AI foundation skills benchmark. They are a good way for each person to learn what the tools do.
They won't train your team to use AI on your work under your rules, and that is where the value is. If someone in the business already uses AI well every day and has time to teach, run the four weeks in-house. If nobody does, bring in a trainer for the hands-on sessions and the follow-up. Our checklist for choosing an AI training provider and our guide to what AI training costs will help you compare quotes.
Where Fautons fits
We run live, hands-on AI training for UK teams, and small businesses are a good fit. Sessions work best with 5 to 15 people, so a business of 10 to 20 usually trains in two groups. We start with a short scoping call about your team, your tools and the work you want AI to help with. Then your people practise on their own tasks, on-site or live online, and leave with prompts and playbooks they can reuse, followed by a check-in.
See AI training for small teams for how it works, or book a discovery call and we'll talk through your team.
Frequently asked questions
How can a small business use AI?
Mostly through general assistants such as ChatGPT, Microsoft Copilot, Google Gemini or Claude, used on everyday work: drafting emails and proposals, summarising documents, writing up notes, tidying spreadsheets and drafting customer replies. In ONS figures for September 2026, text generation with large language models was the most used kind of AI among UK businesses with 10 or more employees, at 21%.
Which AI tool should a small business start with?
Usually the one you already pay for. If you use Microsoft 365, look at Copilot. If you use Google Workspace, look at Gemini. ChatGPT and Claude are strong general choices too. Pick one approved tool for the whole team so everyone learns the same way of working, and use its business version rather than free personal accounts, because the two handle your data differently.
How long does it take to train a small team to use AI?
About four weeks for a team of 10 to 20: a week to agree rules and pick tasks, hands-on sessions in two small groups, a week or two of practice on real work, then a short follow-up. The sessions themselves can be one day or two half-days per group.
Does a small business need an AI policy?
Yes, but a short one. One page that says which tools are approved, what must never go into them, such as customer personal data, staff records and passwords, and who checks anything before it leaves the business. Agree it before training so the sessions can teach it.
Is AI training worth it for a small business?
It usually is, because the skills gap is the main barrier. The government's AI Adoption Survey 2026 found limited AI skills affected 67% of businesses already using AI and 78% of those planning to. The tools are cheap or already paid for, so the return depends almost entirely on whether people know how to use them on their own work.
How much does AI training cost for a small business?
It depends on group size, format and how tailored the sessions are. Some providers charge per person for open courses, and others charge a flat fee for a team day. Our guide to what corporate AI training costs compares published UK prices and explains what moves them.
Sources
- ONS: Business insights and impact on the UK economy, 8 October 2026 (BICS Wave 165, 21 September to 4 October 2026)
- GOV.UK (DSIT): AI adoption survey report 2026, executive summary (published 29 October 2026; 3,503 interviews, 7 January to 21 April 2026)
- GOV.UK: Free AI training for all, as government and industry programme expands to provide 10 million workers with key AI skills by 2030 (28 January 2026)


