You are halfway up a ladder with both hands inside a ceiling cavity when the phone goes off, which is where AI for electricians starts to earn its keep. You let it ring. By the time you are back on the ground with your gloves off, the caller has already rung the next sparky on the list and booked them in.
That is the real reason most electrical businesses stop growing, and it is exactly the gap this technology is built to close. Not the wiring. The phone, the quotes, the follow-up, and the two hours of admin that eat your evening after the last job.
This post covers what to hand over first, what to leave well alone, what an ordinary Tuesday looks like once it is running, and how to get the first piece live inside a fortnight.
Key Takeaways
- AI for electricians is not about the wiring. It is about the phone, the quote follow-up, and the job list that eats your evenings.
- Most sparkies lose work to response time, not price. The first business to respond wins the job roughly 78% of the time.
- An AI receptionist answers every call, sorts urgent from quotable, and books the job while you are still on the tools.
- Quotes go cold after 48 hours of silence. Automated follow-up recovers work you already did the site visit for.
- One dental practice lifted booked appointments 44% by fixing missed calls alone, with no extra marketing spend.
- Start with answering the phone. Trying to automate quoting judgment first is why most trade AI projects stall in month two.
- Track booked jobs per 100 enquiries, not calls answered. That number tells you whether the system is actually earning.
Why Electricians Lose Jobs to the Phone, Not the Price
Ask an electrician why a quote did not land, and you will hear “they went cheaper” nine times out of ten. Sometimes that is true. More often, the customer never got far enough to compare prices because nobody picked up.
Think about how the work actually arrives. Someone’s RCD keeps tripping at seven at night. A property manager needs a tenant’s oven wired in tomorrow. A builder’s sparky has gone quiet, and the inspection is Thursday. In every one of those cases, the customer is not shopping. They are solving a problem, right now, and they will hand the job to whoever answers.
Research on response time is consistent and unforgiving. A Harvard Business Review study of 2,241 companies found that firms responding to an enquiry within an hour were nearly seven times more likely to have a real conversation with a decision maker than those that waited just one hour longer. The figure quoted most often in sales circles is that the first business to respond wins about 78% of the deals. In the trades, the window is tighter still, because your customer has three names open in the Google map pack and a thumb.
The maths nobody runs
Here is the exercise I put in front of every trade owner I sit down with. Pull your mobile call log for the last month. Count the calls you did not answer. Then count how many of those numbers you ever rang back.
The second number is always smaller than people expect. Not because anyone is lazy, but because a missed call at 11:20 am gets buried under four texts, a materials run, and a job that took longer than quoted. By six o’clock it is gone.
Now put a job value on it. If a fifth of those missed calls were switchboard upgrades, EV charger installs, or a full rewire, you are not looking at a marketing problem. You are looking at a leak in the bucket you already filled. Most electrical businesses I look at are paying for leads at the top and losing a chunk of them to a phone that rang out.
Why hiring does not fix it
The default answer is to put someone on the phone. It works, up to a point, and then it does not. One person covers roughly forty hours out of the 168 in a week. They take annual leave. They get sick. They are on another call when the third one comes in, and they are definitely not there at 7 pm when the RCD trips.
Worse, everything they learn about your business lives in their head, in the same way everything about your business currently lives in yours. That is the trap. You are the only person who knows which jobs are worth chasing, what you charge for a callout, and which builder always pays late. Adding another head does not fix a business with no memory of its own. It just spreads the dependency around.
What AI for Electricians Actually Means, and What It Does Not
Let me kill the obvious objection first. AI is not going to pull cable, terminate a board, or sign off a Certificate of Compliance. There is no version of this where a machine climbs into the roof space at a Grey Lynn villa and sorts out the old TPS. The regulated, hands-on, licensed part of your work is not the target and will not be for a very long time.
What AI is genuinely good at is everything that happens either side of the job. The phone call, the qualifying questions, the booking, the confirmation text, the quote that needs chasing on day three, the invoice reminder, the customer who had a switchboard quote eighteen months ago and never got back to you.
I split it into two halves, because they do different work and they fail in different ways.
The brain: the part that knows your business
The brain is the layer that holds everything currently stored in your head. Your callout hours and what you charge to attend after them. Your service area, and the suburbs you will only travel to for jobs over a certain size. Which work you take (residential maintenance, commercial fitout, EV chargers) and which you refer to the mate who does data cabling. Your team, their tickets, who is on the tools on Saturday. How you talk to customers.
Without that, an AI is a stranger answering your phone. With it, the AI answers the way you would answer, because it knows the same things you know. This is the piece most trade owners skip, and it is why their first attempt with ChatGPT felt clever for a week and then went nowhere. A tool with no context gives generic answers. A tool with context gives your answers.
The workforce: the part that does the repetitive jobs
The workforce is the doing. A voice agent that picks up on the second ring, any hour. A text and email follow-up that runs on a schedule instead of on whether you remembered. A booking that lands in the calendar with the address, the fault description, and the property manager’s name already attached. A note written into the CRM so the next person to touch that customer knows the history.
None of these are impressive on their own. Stacked together, they change what your business is capable of holding. That is the difference between buying AI agents for a small business one at a time and building something that compounds.
The Five Jobs to Hand Over First
Order matters more than ambition. Every stalled AI project I have looked at in a trade business started with the hardest thing on the list. Start at the top of this list instead and work down.
1. Answering every call, including the ones at 9 pm
This is the first one, every time, because it is the one costing you money today. An AI receptionist answers on the second ring, at any hour, on any number of simultaneous calls. It works out whether the caller has an emergency or wants a quote, takes the address and the fault, books the job into the diary or flags it to you, and texts the customer a confirmation.
The proof I keep coming back to is not from the trades, but the shape is identical. Dr Claire ran a dental practice with two receptionists on staff and still had 47% of calls going unanswered. Not a staffing problem, a systems problem. Once the AI receptionist went in, missed calls dropped to zero and booked appointments rose 44%. Same marketing spend. Same team. The only change was that the phone stopped ringing out.
For an electrical business, the after-hours side is worth more, not less, because your emergency work is the highest-margin work you do. I have written about the after-hours call answering side of this in more detail if you want the mechanics.
2. Replying to web and Google enquiries within 90 seconds
The form fill on your website, the message through Google Business Profile, the Facebook enquiry. These sit unread for hours because they land somewhere you do not check while you are on a site.
An automated first response closes that window. Ninety seconds after someone submits an enquiry, they get a text and an email, both with a real answer to what they asked, and a link to book. Most of them are still holding the phone. That single change usually lifts conversion more than any amount of extra ad spend, because you are converting leads you already paid for.
3. Chasing the quotes you already wrote
You did the site visit. You spent forty minutes measuring, working out the run, and pricing it. You emailed the quote. Then nothing.
Most electricians follow up once, maybe, when they remember. The customer is not saying no; they are waiting on their partner, or a builder, or payday. Silence from you reads as “he is probably too busy anyway.”
A follow-up sequence over text and email at day two, day five, day twelve and day thirty runs whether or not you remember. It answers the common objection at each step. It offers to book without asking the customer to ring anyone. The jobs it recovers are the cheapest work you will ever win, because the site visit is already paid for.
4. Confirming jobs and killing no-shows
Nothing wastes a day like arriving at an empty house. Automated confirmation the day before, a reminder the morning of, and a text when you are on your way. The customer can reply to reschedule, and the diary updates itself. This is the least exciting item on the list and one of the fastest to pay for itself. The same AI appointment booking mechanics apply whether you are running two vans or ten.
5. Waking up the customer list you already own
Every electrician has a database, and almost nobody works it. Everyone you have ever invoiced. The switchboard quotes that never converted. The customers whose smoke alarms are now nine years old. The commercial clients whose test and tag is overdue.
A multi-touch reactivation over text and email, written to sound like a person rather than a mailout, re-opens those conversations. The clearest number I have on this comes from a finance broker rather than a sparky: 319 dormant contacts his team had written off, worked systematically, recovered $49,000. No ad spend, because these were people who already knew the business.
If you want a rough figure for your own list before you talk to anyone, the Revenue Recovery Calculator at tools.octavius.ai will give you one in about two minutes. There are more examples of what this looks like across different trades in these AI automation case studies.
What a Tuesday Looks Like When It Runs
Here is an ordinary Tuesday for a three-van operation with the first four pieces live. It is a composite rather than one specific client, but every step in it is something the system does today.
6:40 am. You read a short brief on your phone before you get out of bed. Two jobs booked overnight, one of them an urgent oven circuit for a property manager in Papatoetoe. Three quotes still unanswered, one of which has now had two follow-ups and opened both. One invoice went past 30 days.
7:15 am. You are driving to the first job. The phone rings and the AI picks up before you can. A homeowner in Massey with lights flickering through half the house. The system asks the questions you would ask, works out it is not an emergency, and books it for Thursday morning with the fault description already in the job.
9:30 am. You are in a roof space. Three calls come in during the forty minutes you are up there. All three answered, two booked, one flagged to you because the caller wanted a commercial fitout price, and the system knows that is a conversation you take yourself.
12:10 pm. Lunch in the van. A quote you sent last Thursday comes back accepted. You did not chase it. The follow-up sequence did, twice, and the second one answered the question the customer actually had, which was whether the work needed an inspection.
4:45 pm. Last job done, and you are not doing admin tonight. The confirmations for tomorrow went out at two. The invoice for today’s work was generated when you marked the job complete. The customer who cancelled has already been offered the slot that opened up.
8:20 pm. An RCD trips in Titirangi. The phone is answered on the second ring. The system takes the details, tells the customer honestly what the after-hours callout involves, and books it for 7 am. You find out about it in the morning brief.
Nothing in that day required you to be at a desk. That is the whole point. The gap between a business that runs on your attention and one that runs on a system is not effort; it is whether the business can hold information without you.
Where AI for Electricians Goes Wrong, and How to Start in a Fortnight
Plenty of trade businesses have tried this and got nowhere. The reasons repeat.
Mistake one: buying tools instead of building a system
An MIT report on enterprise AI that did the rounds last year found roughly 95% of pilots delivered no measurable return. The failures were not technical. They were businesses buying a tool, plugging it in beside eleven other tools that do not talk to each other, and expecting the whole to be worth more than the parts.
You already know this pattern. The job management app you paid for and half use. The CRM with three years of contacts nobody has touched. Adding an AI chatbot to that pile gives you a twelfth login, not a better business. The system has to sit across the tools, not inside another one.
Mistake two: automating judgment before admin
The instinct is to start with the interesting problem. “Can AI price my jobs?” Sort of, badly, and not without a lot of setup. Meanwhile, the phone is still ringing out.
Pricing an electrical job is judgment work. It depends on access, on the age of the building, on whether the customer is going to be difficult, on how busy you are that month. That is exactly the sort of thing to leave with a human for now. Answering the phone at 9 pm is not judgment work. Start where the work is repetitive, and the cost of not doing it is obvious.
Mistake three: no context, so it sounds like a stranger
If you set up a voice agent that does not know your callout hours, your service area, or what you actually do, it will book you jobs you do not want in suburbs you do not travel to. Then you will turn it off and tell people AI does not work for trades.
Getting the context in is a couple of hours of structured questions, once. It is the same briefing you would give a new employee on day one, except this one does not forget it and does not leave. That work is what separates a system that books you good jobs from one that generates noise. It is also the piece that quietly solves the admin overload problem, because the business stops needing you to answer basic questions.
Do not hide it, and do not fake it
Tell people. A short line at the start of the call that the caller is speaking to your assistant costs you nothing and saves you the one bad review from someone who felt tricked. In practice, most customers do not care at all, provided the thing is quick, understands them, and gets them booked. What they hate is a phone that rings out and a voicemail nobody returns.
The fortnight version
You do not need to down tools for a week to start.
Week one: count. Pull the call log, count the missed calls, and count the quotes sitting unanswered past seven days. Write both numbers down. Do not fix anything yet. Almost every owner I do this with is surprised by at least one of the two, and the surprise is what makes the rest stick.
Week two: one channel. Pick the phone, because it is the biggest hole. Get the context written down, get an agent answering, and let it run on overflow first so you keep the ones you can answer yourself. Watch a week of transcripts. Fix what it got wrong. Then let it take the after-hours calls too.
Everything else on the list can wait until that first piece is boring. Quote follow-up next, because it is the cheapest revenue in the business. Then the dormant list. Then the invoicing and confirmations.
The number to watch
Not calls answered. Not hours saved, which nobody ever measures honestly. Booked jobs per 100 enquiries.
That single number tells you whether the system is earning. If you were converting 40 enquiries out of 100 and you are now converting 55, you know exactly what changed and what it is worth in dollars. It also stops you from congratulating yourself for automating things that were never costing you anything.
The Real Problem Underneath
Missed calls are a symptom. The condition is that your business has no memory and no ability to act without you in the room. Every decision, every price, every bit of history about a customer lives in one head, and that head is also the one that has to be up a ladder for seven hours a day.
Answering the phone with AI for electricians fixes one expensive version of that. It does not fix the whole thing. What fixes the whole thing is giving the business a brain that holds what you know, and a workforce that does the repetitive work on top of it. The phone is just the first and most obvious place the gap shows up, because it is the place where the cost has a dollar figure attached.
Start there anyway. One job, handed over properly, changes how you think about the next ten. Most electricians I talk to do not need a strategy. They need the phone answered on Saturday, and then a reason to keep going.
Where to Start
If you want to work out which piece is costing you the most, book a 30-minute Discovery Call. It is a straight conversation about where the work is leaking out of your business and what is worth fixing first. No deck, no pitch, and if the answer is that you should just get better at ringing people back, I will tell you that.
Bring your call log for the last month. That is usually all the diagnosis needs.
Frequently Asked Questions
Can AI really answer the phone for an electrical business?
Yes, and it is the most common starting point. A voice agent answers on the second ring at any hour, takes the address and fault description, works out whether the caller needs an emergency callout or a quote, and books the job straight into the diary. It handles several calls at once, which a person cannot, and it sends you a summary of every call it took.
Will my customers know they are talking to AI?
Some will, some will not, and you should tell them either way. A short line at the start of the call is enough. In practice, customers care far more about being answered quickly and getting booked than about who answered. The complaints come from phones that ring out and voicemails nobody returns, not from an assistant that sorted the problem in ninety seconds.
What should an electrician automate first?
Answering the phone. It is where the money is leaking most obviously, and the value is easy to measure. Second is quote follow-up, because you have already paid for the site visit and those jobs are the cheapest revenue in the business. Leave pricing and anything requiring judgment about a job until the simple things are running properly.
Can AI write my quotes?
Not the pricing judgment, and I would not try. Working out what a job is worth depends on access, building age, how busy you are, and how the customer is likely to behave. What AI does well is everything around the quote: sending it, formatting it, following up on days two, five and twelve, answering the common questions, and booking the job when the customer says yes.
Will it work with my existing job management software?
Usually, yes. Most trade software has an integration path, and the ones that do not can still be connected with a bit of work. The system sits across your existing tools rather than replacing them, so your job management app, accounting software and calendar stay where they are. If you have no CRM at all, Nexus covers that layer as part of the build.
Is this worth it for a one-van electrician?
It can be, though the case is different. A solo operator has no one to answer the phone at all, so the missed-call leak is usually worse in percentage terms than it is for a three-van outfit. The running cost is far below a part-time receptionist. The honest test is how many calls you missed last month and what those jobs would have been worth.
About Octavius
Titus Mulquiney is the founder of Octavius AI, where he builds AI brains and AI workforces for founder-led businesses stuck running everything out of their own head. Twenty years in marketing, ex-Sony product manager, ex-GM Zeal NZ. Based in Auckland, working with operators across NZ, Australia, and the US. Connect on LinkedIn.