Almost every week, a business owner asks us what they should be doing with AI. Usually they've tried a chatbot, written a few emails with it, and come away unsure. It's clever, but it hasn't changed how the business runs.
That's normal. On its own, AI is a tool looking for a job. The value shows up when you point it at a specific, boring task your team does every day. Here's where we'd start.
Start with the boring work
Good first uses of AI have three things in common. The task happens often, it follows a pattern, and a person can check the result in seconds. Four places usually pay off first:
- 1Reading documents. Invoices, purchase orders and forms are read and entered for you, then checked by your team.
- 2Drafting routine writing. Quotes, follow up emails and replies to common questions are drafted from your own data, ready to review.
- 3Answering questions about your data. Ask "which customers haven't ordered this month?" in plain words, without waiting for someone to build a report.
- 4Flagging what needs attention. Late payments, low stock and hot leads surface on their own, instead of being found by accident.
None of these replace anyone. They take away the typing, searching and checking that fills your team's day, so people spend more time on customers and decisions.
What to skip for now
Some ideas sound great but rarely work as a first step:
- A chatbot that talks to customers with nobody checking it. One wrong answer about a price or a delivery date costs more than it saves.
- "AI for everything" projects. If you can't name the task, you can't tell whether it worked.
- Buying an AI tool before your data is in order. If your orders live in five spreadsheets, AI will read five versions of the truth.
That last one matters most. AI is only as useful as the information it can see, which is why AI projects often start with getting sales, stock or orders into one system.
Keep a person in the loop
The rule we follow on every project is simple: AI suggests, people approve. It drafts the quote, and someone presses send. It reads the invoice, and someone confirms the total. Every action is logged, so you can see what happened and why.
This keeps mistakes small while you learn where AI is reliable in your business. Over time, you'll trust it with more.
How to pick your first AI task
- 1List the tasks your team repeats every day.
- 2Circle the ones that follow a pattern.
- 3Pick the one where a mistake is easy to spot and cheap to fix.
- 4Time how long it takes today, then try AI on it for a month.
If it saves real time, keep it and pick the next one. If it doesn't, you've lost a month, not a year.
| Task | Good first AI project? | Why |
|---|---|---|
| Entering supplier invoices | Yes | Frequent, patterned and quick to check |
| Drafting follow up emails | Yes | Saves time and a person reviews before sending |
| Answering customers alone | Not yet | Mistakes reach customers before anyone sees them |
| Setting prices automatically | Not yet | High impact errors that are hard to spot |
One thing to try this week
Ask your team: "What do you type, copy or check every single day?" Write the answers down. Somewhere on that list is your first AI project, and it's probably less exciting than you expected. That's a good sign.