I work with a lot of trades businesses on the Hibiscus Coast, and the AI conversation goes differently with them than it does with, say, a professional services firm. Sparkies and builders are refreshingly hard to sell abstractions to. The question is always some version of: what would this actually do on site, on a Tuesday, in the rain.
So here's my answer, as an order of operations. If I ran a trades business with three to ten people tomorrow, this is the sequence I'd work through. Note how far down the list the interesting technology appears.
First: answer the phone, or make sure something does
Unglamorous, and it beats everything below it.
In most trades businesses the single biggest source of lost work is an enquiry that arrived while everyone was on the tools and got answered two days later, or not at all. You can have the best website in the district and it won't survive that.
Before anything else, I'd make sure every enquiry — phone, form, Facebook message — lands somewhere that isn't a person's memory, and that someone has a defined job of clearing it daily. That's a decision, not a purchase.
Second: say what you do, where you do it, and roughly what it costs
The second biggest leak. People ring to ask whether you cover their suburb and roughly what a job runs to, and if your site doesn't say, a chunk of them ring someone whose site does.
Write the ten questions you get asked constantly and answer them plainly on the site. Areas covered. Whether you charge for a call-out. Typical price ranges for your three most common jobs, even as a range with caveats. Emergency availability. How quickly you usually respond.
This also happens to be exactly what makes you visible now that a lot of searching gets answered without a click. Specific, factual, well-organised answers are what those systems quote. Vague pages get skipped.
Third: make it easy to be paid and to book
Online booking, or at least a proper quote request that captures what you need to know without three follow-up texts. Invoicing that goes out the same week the work finished, not the following month.
Nothing AI about it. It's the difference between a business that runs and one that limps, and there's little point automating parts of a process this shaky.
Fourth: stop typing things twice
Here's where the interesting technology finally earns its place, and it's the first thing on this list I'd actually build.
In most trades businesses the same information gets written down three or four times. The customer's details on the phone. Again on the job sheet. Again into the accounting system. Measurements from a photo, re-typed into a quote. Someone in the office spends hours a week on this, and it's the kind of work that's both tedious and easy to get wrong.
A photographed job sheet or supplier invoice turning itself into a proper record is genuinely one of the things current technology does reliably well. It's a contained build, usually a couple of thousand dollars, and the sum is easy: five hours a week at $35 an hour is about $9,100 a year.
Fifth: make your own records answerable
Once jobs and quotes live somewhere consistent, being able to ask them questions in plain English becomes useful. What did we charge for a job like this last year. Which jobs are still unpaid past thirty days. When did we last do work at this address.
That one comes up constantly and is oddly satisfying — a customer rings, and instead of "I'll check and call you back", you know while they're still on the phone.
Sixth, and only now: the marketing stuff
Content, social posts, review responses. Fine to use tooling for the tedious shaping, as long as the substance is yours — your jobs, your photos, your actual opinions about how something should be done. The generic version reads like every other trades business in the country, which is a poor advertisement for a business that isn't like every other one.
What I'd skip entirely
A generic off-the-shelf chatbot, at least until the website answers the ordinary questions properly. A real assistant that knows your jobs, your pricing and your availability is worth having, but it belongs on top of a clear site, not in place of one — and it's a different purchase from the widget you'll be quoted for. Anything sold with a deadline. Any subscription where I can't say who looks at the output and what they do differently because of it. And an AI strategy document, which for a business of this size is a way of feeling busy.
The pattern
Four of the six steps above aren't AI at all, and they're the four I'd do first. That's not me talking myself out of work — it's that the things underneath have to hold before anything built on top of them is worth having. A tool sitting on a process nobody agrees on just makes the disagreement faster.
If you're wondering where your own business sits in that order, the honest answer is usually available in about twenty minutes of looking at your last ten lost jobs. Start there rather than with the technology. The technology will still be here.
