To train your team on AI without losing a week, skip the all-day workshop and teach in short sessions on real work. Set up business accounts and a one-page rule sheet first, give each person one task from their own week, run a hands-on session of about an hour, then hold brief check-ins while the habit forms. People learn AI fastest on work they already know.
Here is what to prepare, what a first session looks like, what to teach first, how to bring hesitant people along and how to keep the habit going.
Why does AI training on real work stick?
A slideshow about AI shows what the tool did for someone else. A session on this week's emails shows what it does for you, and that is what people remember on Monday morning. When the task is familiar, people can judge the result right away. They know a good quote, a good reply or a good summary when they see one, and that judgment is the skill that matters most.
Real work also answers the question many people carry and few ask out loud: will this make my job harder? The answer is easier to see when the tool takes a task they dislike off their plate.
What should you set up before the first AI session?
A little preparation keeps the session itself short:
- Business accounts in the company's name for everyone taking part, with logins tested the day before.
- A one-page rule sheet: which tools, what can go in, what stays out, who reads AI work before it goes out and who to ask.
- One task per person, chosen with them: something they write, copy or look up every week.
- Two or three real examples for each task, with names and private details removed.
- A shared document, owned by the company, where good instructions get saved.
If you have not chosen a tool yet, do that first. Our guide on choosing between ChatGPT and Claude shows how to test both on your own work in a week.
What does a first AI training session look like?
Keep it to about an hour, with laptops open, on the accounts people will use the next day. A simple plan:
- Ten minutes on what AI does well, where it makes mistakes and what the rule sheet says.
- Ten minutes on one live example, start to finish, including a draft that needs fixing.
- Thirty minutes where each person works on their own task, with help nearby.
- Ten minutes where each person shares one result, and the best instructions go into the shared document.
End with something done. When people leave with a draft they would actually send, they come back to the tool the next day.
What should you teach first?
Four habits cover most everyday use. Teach them in this order:
- Give context. Say who the message is for, what it needs to do and what the tool should know about the business.
- Show an example. Paste a reply or a quote you are proud of and ask for the same style.
- Ask for changes. Shorter, warmer, simpler, in Spanish: the first draft is a starting point, and asking again is normal.
- Check before it goes out. Read every result for facts, names, numbers and tone. A person sends the final version.
Then add the rules that keep the business safe: what stays out of AI tools, which accounts to use and what to do when unsure. Repeat them at every session until nobody needs the sheet.
How do you train people who are hesitant about AI?
A team usually holds a mix: someone excited, someone quietly worried and someone who expects it to pass. Each learns best a little differently:
- For the worried, start with a task they dislike, and say plainly that the decisions stay with them.
- For the doubtful, use their own hardest example and let them judge the result.
- For the excited, hand them the shared document and make them the person others ask.
- For anyone more at home in Spanish, train in Spanish. People learn faster in the language they think in.
Pairing helps too. Two people at one screen, one typing and one asking, catch what one person alone might miss, and the quiet questions get asked.
How do you keep AI training from eating the week?
Spread it out. One hour, followed by short check-ins, fits around a normal week better than a full day away from the work:
- A fifteen-minute check-in each week for the first month, where people share one thing that worked and one that did not.
- One new task per person once the first one becomes routine.
- A go-to person on the team for quick questions, so the owner does not become the help desk.
- A second session after a few weeks, built on the questions people asked.
Keep every session tied to real work, and nobody has to catch up afterward, because the training was the work.
How do you know the AI training worked?
Watch for small, visible changes:
- People open the tool without being reminded.
- The shared document of instructions keeps growing.
- Questions change from "how do I" to "could it also".
- Drafts reach the owner already checked.
If a few weeks pass and someone still is not using it, sit with them on a task and find out why. The task may be the wrong fit, and a different one may click.
How Banana Blueprint trains teams on AI
Banana Blueprint is a business design and engineering firm in Northeast Ohio. Our AI training includes talks for owners, teams and groups, and hands-on sessions where everyone works on tasks from their own week, with the ChatGPT or Claude accounts the business uses. Sessions run in person in Kent, Akron, Cleveland and across Northeast Ohio, or online with teams anywhere in the U.S., in English or Español.
Training often goes hand in hand with setup. Through our AI consulting, we find where the hours go, set up the business accounts and assistants and stay on as the tools change, so what the team learns keeps matching what they use. To shape a session around your team, start the conversation by phone or video.
Filed under AI & Technology
Written by Adolfo Polanco for owners who want their business to run with less chasing and more clarity. We help owners put AI where it saves the most time, with people in charge of every decision.