Free AI Checklist for Businesses

Where AI can actually help a business, what to leave alone, and seven questions to ask any AI consultant. Answer a few quick questions and it's yours straight away.

What kind of business do you run?

How many people work in the business?

Why are you looking at AI?

What are you trying to solve? (optional, but it helps)

Your name

Your email

Business name (optional)

Yes, Chris can email me about what I've told him. Leave this unticked if you just want the checklist.

We use your answers to understand what you're after and, only if you tick the box above, to get in touch. We don't add you to any mailing list. Details are in our privacy policy.

Your checklist

Here's your checklist.

DOWNLOAD THE PDF

Most businesses don't need an AI strategy. They need to know which two or three jobs are worth handing to AI, and how to tell a good idea from an expensive distraction. Use this checklist to work that out. It takes about 20 minutes.

1. Start with the problem, not the tool

Before you look at any tool, answer these three questions in writing. If you can't, nothing you buy will help.

Where does your week actually go? List the five jobs that take the most time, and roughly how many hours each.
What do you do over and over? Look for tasks you repeat every week in nearly the same way.
What gets dropped or done late? Unanswered enquiries, forgotten follow-ups and late invoices are often the most expensive gaps.

2. Ten jobs AI often helps with

For each one: what it looks like, the sign it might suit you, and when to be careful.

1. Answering enquiries

What it looks like: First replies to common questions, with you checking anything sensitive before it goes out.

Look for: The same handful of questions arriving by email, DM or phone.

Be careful if: A reply could commit you to a price, a date or a promise without your review.

2. Quotes and proposals

What it looks like: Drafting quotes from your past work and price lists.

Look for: You rewrite near-identical quotes several times a month.

Be careful if: Your prices vary a lot from job to job. AI will happily invent a figure.

3. Bookings, scheduling and reminders

What it looks like: Taking bookings, rearranging them and sending reminders.

Look for: No-shows, or lots of back-and-forth messages to agree a time.

Be careful if: Your rules are complicated. Write them down first, then automate.

4. Admin and paperwork

What it looks like: Filing, data entry, and pulling details from emails, invoices and forms into one place.

Look for: You copy the same information between systems.

Be careful if: The documents contain personal or financial data. Check where it is processed.

5. Following up with customers

What it looks like: Thank-you messages, review requests and chasing late invoices.

Look for: Follow-ups that only happen when you remember.

Be careful if: Messages go out automatically with nobody glancing at the tone.

6. Marketing content

What it looks like: First drafts of social posts, emails, product descriptions and website copy.

Look for: The blank page is your main obstacle.

Be careful if: You publish without editing. Generic copy makes you sound like everyone else.

7. Summarising calls, meetings and documents

What it looks like: Notes, action lists and plain-English summaries of long documents.

Look for: You spend time writing up or reading things nobody else will.

Be careful if: The content is confidential. Check the tool's data terms before uploading.

8. Answering staff questions

What it looks like: An internal assistant that answers 'how do we do X?' from your own procedures.

Look for: You or a senior colleague are interrupted with the same questions.

Be careful if: Your procedures aren't written down or are out of date. It will repeat what is wrong.

9. Spreadsheets and reporting

What it looks like: Cleaning data, spotting patterns and producing regular reports.

Look for: A monthly report takes half a day to assemble.

Be careful if: The numbers drive decisions. Always check the totals yourself.

10. Research and preparation

What it looks like: Summarising competitors, suppliers, tenders or regulations before a meeting.

Look for: Preparation eats hours before every pitch or meeting.

Be careful if: Facts matter. AI can state wrong things confidently, so verify anything you rely on.

3. Is this job a good candidate? Five signs

Score each job on your list. Four or five ticks means it's worth a closer look. Two or fewer, leave it for now.

It happens at least once a week.
It follows a pattern a new starter could learn in a day.
It involves a lot of text, numbers or messages.
A mistake is cheap to catch, because someone checks the output.
It costs you at least an hour or two a week.

4. When to leave it alone

These are the signs that AI will cost you more than it saves.

The process is broken or unclear. Fix the process first. AI speeds up whatever you already do, good or bad.
The task needs judgement, empathy or accountability that has to be yours, such as a hard conversation or a sensitive complaint.
A wrong answer would be costly and nobody checks it: legal, medical, financial or safety advice.
It happens rarely. The set-up time will never pay back.
You would have to put customers' personal data into a free consumer tool.

5. Protect your data

UK basics. This is general guidance, not legal advice.

Don't paste customer personal details into free or personal AI accounts. Business plans usually give you more control over how data is used, but check the provider's current terms.
Find out where your data is processed and stored, and whether it is used to train the provider's models.
Keep a person in the loop for anything that directly affects a customer.
If AI helps make decisions about people, such as hiring, credit or pricing, take extra care and take advice.
If in doubt, the ICO publishes free guidance for small businesses at ico.org.uk.

6. Seven questions to ask any AI consultant or tool provider

A good answer is specific and in plain English.

What exactly will it do, and what won't it do?
What happens to my data? Where is it stored, who can see it, and is it used to train models?
What happens when it gets something wrong? Who checks, and who is responsible?
What will it cost in total, including ongoing fees and my own time?
Can I leave? Can I take my data and my set-up with me?
How will we know it worked? Agree a measure before you start, such as hours saved per week.
Who will I actually deal with day to day?

Be wary of guaranteed results, vague promises of 'transformation', pressure to decide quickly, and anyone who recommends AI for everything.

7. Measure it with a simple pilot

One job, a few weeks, a clear before and after.

Pick one job from your list.
Time it for a normal week. Write down the hours and any mistakes.
Try the AI-assisted way for two to four weeks.
Compare the hours saved, the quality, and how much checking it needed.
Decide: keep it, change it, or stop.

8. Your first 30 days

A realistic plan for a busy owner.

Week 1: write your problem list (section 1) and score each job against the five signs (section 3).
Week 2: pick one job and time it as it is today.
Weeks 3 and 4: run a small pilot with one tool or one person, then review (section 7).

Want to talk it through?

Hutch Media Group helps Gloucestershire businesses work out where AI is worth using, and builds bespoke tools where that is the answer. Email hello@hutchmedia.group or visit hutchmedia.group/contact.