You don't need an AI strategy. You need a first win.

Most of the business owners I speak to in Devon and beyond aren't sceptical about AI anymore. That argument is over. What they actually feel is stuck. They know they should be “doing something with AI”, but the gap between a headline about ChatGPT and a real change in how their business runs feels enormous. So they wait. They read another article. They book a webinar they never watch.

If that's you, here's the good news: you don't need a grand AI strategy to get going. You need one practical win that saves real time on a real task. Get that, and the rest follows. Here's how I'd start.

1. Pick a boring, repetitive task — not an exciting one

The instinct is to point AI at your biggest, most glamorous problem. Resist it. The best place to start is the dull stuff you do every week and quietly resent: drafting the same kind of email, summarising meeting notes, writing first drafts of proposals, turning a messy spreadsheet into something readable, pulling together a weekly update.

These tasks are perfect because they're low-risk, you already know what “good” looks like, and you do them often enough to feel the time saving immediately. Save two hours a week on admin and you've got two hours back every week for the rest of the year. That's the win that makes the case for everything else.

2. Treat it like a new team member, not a magic button

The single biggest mistake I see is people typing one vague line into a chatbot, getting a generic answer, and concluding “AI's not very good”. You wouldn't hand a new starter a one-line brief and expect brilliance — and it's the same here.

Give it context. Tell it who you are, who the output is for, what tone you want, and what a good result looks like. Paste in an example of how you'd normally write something. The quality of what you get out is almost entirely down to the quality of what you put in. When people in my workshops learn this one habit, the difference in their results is night and day.

3. Keep a human in the loop — always

AI is fast and confident, and it will occasionally be confidently wrong. It can invent a statistic, misread a nuance, or get a fact subtly off. That's not a reason to avoid it — it's a reason to use it properly.

Use it for the heavy lifting: the first draft, the summary, the rough version. Then you check, edit, and approve. Never send anything client-facing or commercially important straight out of the machine without reading it. Think of AI as a brilliant, tireless assistant who needs a sign-off from someone who actually knows the business. That someone is you.

4. Get your team involved early

If you crack AI yourself but your team carries on as before, you've changed almost nothing. The real value shows up when adoption is shared — when your whole team has a few tools they trust and use daily.

You don't need to send everyone on a computer science course. You need to show them, in plain terms, how it applies to their actual jobs, and give them permission to experiment. A short, practical session where people work on their own real tasks beats any amount of theory. People adopt what they've seen work with their own hands.

Where this leaves you

You don't need to transform your business overnight. You need to pick one task this week, give the tool a proper brief, check the output, and notice how much time you got back. Do that a few times and you'll stop asking whether AI is worth it and start asking what to point it at next.