An AI readiness assessment for service businesses

Most AI readiness assessments are written for companies with a data team and a transformation budget. If you run a service business with five or fifty people, the questions that matter are smaller and more practical, and you can answer most of them yourself in an afternoon.

Here are those questions, what a useful AI readiness assessment should hand you at the end, and what the answers looked like in businesses we've built for. If you'd rather do it with someone, there's a free 30-minute version at the bottom.

What "ready" means for a small service business

You don't need clean data lakes or an AI strategy document. You're ready for AI and automation when three things are true. A piece of work happens the same way most of the time. The information it needs already lives in a tool rather than in someone's head. And getting it wrong costs you something you can measure, whether that's hours, missed leads or late invoices.

If one of those is missing, the honest answer is that you aren't ready for that piece of work yet. That's useful to know. It usually means the first job is writing the process down, before anyone buys software.

Eight questions to answer first

Work through these for one area of the business at a time. Lead handling and invoicing are good places to start, because they repeat and the cost of slipping is easy to see.

1. 1Can you describe the steps in order? If two people in your team would describe the same job differently, pick one way before you automate anything.
2. 2Where does each step's information live? A CRM, a spreadsheet, an inbox, a booking tool. Write the tool next to each step.
3. 3Does each of those tools let other software read and write to it? Most mainstream tools do. A few don't, and that decides what's possible.
4. 4How often does it happen? Something that happens twice a month rarely pays back. Something that happens forty times a week usually does.
5. 5What does a mistake cost? A lead that goes cold, an invoice that goes out late, a client report that's wrong on the call. Put a rough number on it.
6. 6Who notices when it goes wrong today? If the answer is nobody, an automated version needs an alert built in from day one.
7. 7Which steps need judgement? Keep those with a person. Automation is good at the steps either side of a decision, and the decision itself stays human.
8. 8Who will look after it once it's built? Someone has to own it, even if that's only checking a weekly summary.

What a good assessment should give you

At the end of an AI readiness assessment you should have one clear answer that fits on a page. That means the single gap that costs you most, a rough idea of what closing it is worth, and what fixing it would involve. If the assessment can't name a gap in money or hours, it hasn't finished.

It should also tell you what not to do. An engineering firm once asked us for ten AI agents across four frameworks. When we ran the reliability numbers, a fleet that size only got the job done about 60 percent of the time, so we recommended a much smaller system. We wrote that up, because more AI isn't always more ready.

What the answers looked like in real businesses

A crane hire business kept quotes, field tickets, invoicing and payment tracking in separate places: spreadsheets, paper and a lot of re-keying. The cost was easy to measure, two working days a week lost to invoicing and chasing payment, and the steps were the same on every job. We built Crane Ops, quote-to-cash software where crews log tickets from a phone and invoices go out in one click.

A solar installer was losing leads across four inboxes. The steps were simple and the leads were already in tools, so the gap was joining them up. The lead engine moved everything into one Pipedrive pipeline: 569 leads captured, zero duplicates.

A coaching business spent 2 to 3 days a month building client reports by hand from several tools. The same report was due for every client, every month, which made it a clean fit. XPC Coaching's reports now build themselves before each check-in call.

In each case the assessment found one gap worth closing first. None of them started with "use more AI".

Doing it yourself, or with us

You can run the eight questions on your own, and if you do, start with whichever area your team complains about most. That's usually where the hours are going.

If you'd like a second pair of eyes, we run a free 30-minute call. We trace how a lead becomes a paying client in your business, show you the biggest gap, and give you a rough idea of what closing it is worth. If AI isn't the fix, we'll say so on the call. Any build after that is a fixed fee, agreed in writing before work starts.

Book the free 30-minute AI readiness call

Common questions

How long does an AI readiness assessment take? The questions above take an afternoon for one area of the business. Our call is 30 minutes and focuses on the path from lead to paying client.

Do I need to have used AI before? No. What matters is that the work repeats and the information is in a tool.

Is this only about AI, or automation too? Both. A lot of what gets called AI is plain automation: moving a record, raising an invoice, sending a reminder. AI earns its place in the steps that need writing or reading, like drafting a follow-up or summarising a report.

What if the answer is that we're not ready? Then you've saved the cost of a build that wouldn't have worked. Write the process down, put the information in one tool, and come back to it.