Practical guide · September 2026 · about 4 minutes · part of 10 practical ways AI could help Irish trades businesses
“I need a plumber.” “Can you look at my roof?” Most enquiries start with almost nothing, and you spend the first five minutes of every callback finding out where, what and how urgent. There’s a way to have that ready before you ring — as long as it doesn’t get in the way of the customer reaching you.
The short version
- A short automated step can collect the basics so your callback starts with the job, not twenty questions.
- The goal is a better first conversation, not fewer conversations.
- It collects and organises. It never inspects, diagnoses or prices.
What’s worth asking
| Ask for | Why |
|---|---|
| Location | Is it in your area? |
| Type of work | What skills, tools and time it needs |
| Urgency | Routine, or needs a quick answer |
| Photos | Helps you prepare — never a substitute for seeing it |
| Access | Parking, keys, height, tight spaces |
| Timing | When suits them for a call or visit |
| What they’re after | A repair, a quote, a service, or advice |
Only ask what you’ll actually use. A long form puts people off and makes a local business feel like a call centre.
How it works
It kicks in when someone fills in a form, starts a web chat, or when you miss a call. The customer is told an automated assistant is taking the basics. It notes what’s missing and creates a callback reminder for you.
You get a short summary, not a long transcript: where, what, how urgent, and any photos. Then you take over — diagnosis, advice, price, booking.
The danger: sounding sure
AI turns a vague description into a tidy, confident summary — more confident than the information deserves. “The boiler’s making a strange noise” shouldn’t come back with a diagnosis. A photo of a roof isn’t an inspection. A message mentioning a smell of gas or a burning socket needs a person, fast.
That’s a real risk, not a theoretical one: in the US Federal Reserve’s 2026 survey of small employer firms, accuracy was the most common problem reported by businesses using AI (46%).
The rule: it can collect and organise information. It must never act as if a qualified person has looked at, diagnosed or priced the job.
Always leave a way to reach you
Nobody should have to argue with a chatbot to get a tradesperson. Keep your number visible and the callback option obvious. Outside working hours, say when they’ll hear back and point urgent problems to the right help — without offering a diagnosis. And tell people, in a sentence, what happens to their photos and details.
Try it for a month
One channel, one type of job. Say an electrician tries a short web form just for rewiring enquiries: town, type of property, what they want done, rough timing, and photos if they’re happy to send them.
Look at every enquiry yourself for the first month, and track:
- how many enquiries came in with the details you needed
- time from enquiry to your callback
- how many unsuitable jobs you spotted before ringing
- how many people started the form and gave up
- how many turned into site visits, quotes and booked jobs
- which questions confused people
Don’t aim for the most automation. Aim for a better first conversation — the one that wins the job.
What this research can’t tell you
I haven’t found any study measuring whether automated enquiry forms win more work for trades businesses, or how Irish customers feel about them. The accuracy figure is from US businesses across all industries. The only way to know if it helps you is to try it on a small scale and count.
How quickly can you get back to people?
I’m running a short survey asking local tradespeople which parts of the admin cause the most trouble — including missing calls while you’re working and getting back to messages quickly. It takes 6–8 minutes, and leaving your contact details is optional.
Keep reading
- How much could missed calls be costing your trades business?
- Can voice notes replace the tradesperson’s notebook?
- The human side of AI in the trades: how trust is built
Sources
- Federal Reserve Banks — 2026 Report on Employer Firms (Small Business Credit Survey, fielded September–November 2025)