Know Exactly Where Your Team Stands on AI
Why I Do This
Most businesses I meet start their AI journey with the same question:
“Which AI tool should we buy?”
It feels like the logical first step.
It usually isn’t.
Before choosing anything, you need to know where your team actually stands.
Who already uses AI, and for what? Which tools are people using on their own accounts? Where does the time go each day? Which tasks are repeated again and again, and which ones genuinely need a person?
Most owners can’t answer those questions with confidence. That isn’t a criticism. Busywork is easy to miss from the top, because it’s spread across many small tasks done by many people.
That’s what an AI audit is for.
It doesn’t start with technology. It starts with the work.
What an AI audit looks at
When I audit a business, I look at four things, person by person and team by team:
The work
- What people do each day
- What they repeat
- Which tasks need judgment
The tools
- What you already pay for
- What people use on their own
- Where information is stored
The friction
- Where time is lost
- Where mistakes happen
- What gets forgotten
The fit
- What to automate
- What needs AI
- What should stay human
How I Actually Work
Here’s the process I follow. It’s simple on purpose, so it fits a small business without disrupting the week.
Talk to the People Doing the Work
Short, relaxed conversations with the people who do the work, not just the managers.
I use ten plain questions about their normal day. No AI jargon, no technical knowledge needed.
Map the Tools You Already Have
Email, CRM, accounting, Google Workspace or Microsoft 365, WhatsApp, spreadsheets, your website.
Plus any AI tools people already use, including personal accounts nobody approved.
Sort Every Task
Every task gets one honest label: automate it, give it AI help, connect two systems, organise the information, fix the process, or leave it with a person.
Show Where You Stand Today
I place the business on my 7 Levels of AI Maturity framework.
Not to give you a grade, but to show the next level that makes sense for you.
Rank the Quick Wins
A short list of quick wins, ranked by time saved and effort needed.
Plus the bigger opportunities for later, so nothing gets lost.
Start With One Pilot
We prove one improvement with one person or team for 30 days.
Measure the time saved, then decide what to roll out next.
10 Questions That Reveal Where AI Can Help
You don’t need to ask people complicated questions about artificial intelligence.
Start with simple questions about their normal working day. These ten are the ones I use most, and they work just as well if you ask them yourself.
What do you do every day that you repeat again and again?
The best place to start. Similar emails, reports, data entry, scheduling, posts. The more often a task repeats, the more it’s worth improving.
What takes up the most time in your working day?
The answer often surprises managers. People may spend hours preparing, searching and formatting rather than on their main job.
What information do you copy from one application to another?
My favourite question. Email to Excel, website to CRM, invoice to spreadsheet. Manual transfer usually means an automation opportunity.
Which emails or messages do you write repeatedly?
Ask to see examples. If only the name and details change, an assistant can draft it in their tone. They still review and send it.
Which documents or reports do you create every week or month?
Proposals, management reports, quotations, meeting summaries. If the structure repeats, AI can prepare the first draft.
What information do you spend time searching for?
Old proposals, the latest price list, a client’s history. Sometimes the fix isn’t AI at all, just better organised files.
Which tasks do you forget or need reminders for?
Follow-ups, quotation chasers, callbacks. Simple tasks, but forgetting them costs money. A workflow can create the reminder automatically.
Which tasks need you to read, understand and then decide?
This separates AI from plain automation. Reading an email to work out who should handle it needs AI. Copying an invoice number doesn’t.
What task would you happily never do again?
People know exactly what frustrates them. The weekly report, entering leads, minutes. Frustration points to the best opportunities.
If an AI assistant could help with one thing tomorrow, what would it be?
Don’t show them a list of tools. Ask what they actually want help with. The answer often reveals something management never noticed.
From Answers to the Right Fix
The point of these questions isn’t to finish the interview and say, “You need ChatGPT.”
That would be far too simple.
Each answer gets assessed, and different answers lead to very different fixes:
“I copy customer details from our website into the CRM.”
Same steps every time. Connect the two systems.
“I read customer emails and decide what reply they need.”
Needs reading and judgment. AI drafts, the person decides.
“I keep looking for the latest company documents.”
Organise the files first, then make them searchable.
“I move information between two apps by hand.”
A plugin or connector may already exist.
“I build the same report every Friday.”
The data and structure repeat. AI prepares the first draft.
“Nobody follows the same steps, so it’s confusing.”
No tool fixes a broken process. Agree the steps, then improve them.
Notice that only some of these answers lead to AI.
Others lead to an automation, an integration or simply a better-organised shared drive. And one of the most valuable findings is often the last one: the process itself is the problem, and no tool will fix it.
That distinction saves money, avoids unnecessary complexity and makes the AI you do adopt far more likely to stick.
One Company, Many Different Answers
So the outcome of an audit isn’t one AI tool for the whole company.
A finance officer, a salesperson and a marketer do completely different work. Their recommendations should be different too.
A typical roadmap reads more like this:
Sometimes ChatGPT is the answer. Sometimes Claude. Sometimes Gemini or Copilot makes more sense, simply because it already sits inside the tools people use all day.
I explain how each one fits in Cut the Busywork Out of Your Day, and how automation fits in Stop Copying Data Between Tools.
What I Bring
You don’t get a thick report that sits in a drawer.
You get three practical things you can act on straight away:
Your Team’s AI Snapshot
Who uses AI today, for what, and on which accounts.
Plus where you sit on the 7 levels.
Your Opportunity Map
Every repetitive task we found, with the right fix for each.
Ranked by time saved and effort.
A 30-Day Starting Plan
Your first pilot, the quick wins and simple rules for using AI safely.
Ready to start the following week.
Frequently Asked Questions
Do my staff need to know anything about AI?
No. The questions are about their normal working day, not about technology. People who have never used AI often give the most useful answers.
We already use ChatGPT. Do we still need an audit?
Yes, and it’s often more useful. It shows who is using it, for what, and whether company information is going into personal accounts. It also shows where you’re getting value and where you aren’t.
Will the audit tell us to buy more software?
Not necessarily. Many quick wins use tools you already pay for. Some of the best recommendations involve no new tool at all.
How long does it take?
It depends on the size of the team. Each conversation is short, and most small teams can be covered without disrupting normal work.
Can I ask the ten questions myself?
Absolutely. They’re a good starting point on their own. Where an outside view helps is in sorting the answers and choosing the right fix for each one.
Want to know where your team really stands?
Book a 20-minute call. We’ll go through a few of the ten questions together and you’ll leave with at least one quick win.