Getting your team on board with AI in a small business
By Keith Hamilton · · 4 min read
Getting your team on board with AI is harder than choosing the tool, and it decides whether the tool gets used. If your team sees AI as something being done to them, they will work around it or quietly ignore it. If they see it as something that takes the tedious parts of their job away, they will help you make it work. This article covers the practical steps for introducing AI to employees in a small business.
What are the common staff concerns about AI?
The common staff concerns about AI are about their own jobs first. People ask:
- Will this replace me, or reduce my hours?
- Will I be judged on how well I use a tool I have never been trained on?
- Will I be blamed when it gets something wrong?
- Is this just more work on top of my normal job?
These are reasonable questions, not resistance to change. Ask people what concerns them, listen properly, and answer as honestly as you can.
How do you explain why you are introducing AI?
Explain what you are trying to achieve in plain terms, tied to a real problem in the business. For example:
- the team is stretched and you want to stop hiring temporary staff for busy periods
- customers wait too long for replies
- you want people spending more time on the work they are good at and less on copying details between systems
Be honest about what you know and what you do not. If you cannot promise that no roles will change, do not promise it. People cope better with honest uncertainty than with reassurance that later turns out to be wrong.
Why involve the people who do the work?
Involve the people who do the work because they know where it goes wrong, which cases are unusual and which shortcuts are safe. That knowledge is exactly what an AI project needs.
Ask them directly: which parts of your week follow the same steps every time, and which would you happily never do again? You will get a better list than any manager produces alone. Our guide on where to start with AI shows how to turn those answers into a first project.
Involve the same people in testing. They spot problems quickly, and they trust a tool more when they helped shape it.
Which task should you start with?
Start with a task people dislike. Your first project sets the tone.
If it takes away a job people find dull or frustrating, such as re-keying invoices or sorting a crowded shared inbox, support follows easily. If it starts with the part of someone's job they enjoy and take pride in, you are fighting uphill.
How much time do people need to learn AI tools?
People need protected time to learn, even if it is only an hour a week for the first month. New tools take practice, and training works best when it uses the team's own tasks rather than generic examples, so people see the point straight away.
Encourage people to share prompts and approaches that work, and keep them in a simple shared document so the whole team benefits.
Expect early results to be slower than the old way. That is normal while people adjust.
What rules should staff follow when using AI?
Staff need clear answers to four questions:
- Which tools are approved?
- What information must never be put into a public AI tool?
- Who checks outputs before they reach a customer?
- Who do they ask when they are unsure?
People are more confident when they know what is allowed. A short written policy answers these questions, and our guide to writing an AI policy shows what to include.
Who stays in charge of judgement?
People stay in charge of judgement, and the AI supports them rather than overrules them. Staff must be able to correct or override what it produces, and should know that reporting a problem is welcome rather than a nuisance.
For anything sensitive, such as a complaint, a vulnerable customer or a decision that significantly affects someone, a person stays responsible. Saying this openly reassures both staff and customers.
Should you appoint an AI champion?
Yes. In AI change management for a small business, a champion works better than an enforcer. Most teams have someone who is curious about new tools and patient with colleagues.
Give that person a little time and recognition to help others, answer questions and pass feedback back to you. Peer support works better than instructions from above.
How should you share results with the team?
Share results honestly, using real measures from your own business, such as hours spent on a task before and after, or how quickly customers now get a reply.
Here is how the arithmetic works. A five-person team each spends 2 hours a week on a task that AI now handles in part, saving half that time. That is 5 hours a week across the team, or 260 hours a year at 52 weeks.
Saying where those hours now go, such as calling customers back or clearing a backlog, shows people the change is about better work rather than fewer people.
Be just as open about what did not work. Admitting that a tool was dropped because it was not good enough builds more trust than pretending everything succeeded.
Where Forwardcycle fits
Our Team Enablement workshops, run as half or full days, help your team use AI well on their own real tasks and leave you with a library of prompts and workflows. Our AI Governance Pack provides a policy and staff guidance. To talk through where your team is now, arrange a free call.