AI for tradies: the bullshit and the real.
You've been hearing for two years that AI is going to replace your accountant, your SEO guy, your receptionist, and possibly your business. You've also seen mates get burned by half-baked tools. So which is it? Here's the honest version, from people building it.
Most of what tradies hear about AI is either doom-scroll fear-bait or breathless marketing. We're an AI company and we're going to tell you the truth, because if we don't, you'll either pay too much for the wrong tools or avoid the right ones for the wrong reasons. Either way, you lose.
The problem with the AI conversation right now is that it's being run by two camps. The first is the breathless camp, agencies and consultants telling you AI is going to "automate everything" and "10x your business overnight." The second is the doom camp, telling you AI is going to flatten the trades and you should fear-buy whatever they're selling to "future-proof" yourself.
Both are wrong. AI in 2026 is genuinely useful for some things, genuinely terrible at others, and somewhere in between for most of what a tradie business actually needs. The trick is knowing which is which before you spend a cent.
Part one: the bullshitWhat you can safely ignore.
These are the AI claims you'll see in your inbox, on LinkedIn, and at trade shows over the next 12 months. They sound impressive. Most of them are either flat-out wrong, wildly oversold, or technically true but useless for any actual tradie business.
"AI will run your entire business for you"
The "set and forget" pitch. Sign up, walk away, watch the leads roll in while you're at the beach. It's nonsense.
What's actually true: AI can run repetitive, rule-based work with strong oversight, answering calls, drafting quotes, scheduling jobs, sending invoices. What it cannot do is run your business strategy, decide your pricing, manage your client relationships, or work out what kind of customer you actually want. Those are judgement calls. They need a human, specifically you, and they always will.
Anyone selling you "set and forget" is either lying or doesn't understand their own product. Both are red flags.
"AI is just better than humans now"
You'll see versions of this everywhere, AI passes the bar exam, AI beats doctors at diagnosis, AI writes better than copywriters. Most of these claims are technically true and practically misleading.
What's actually true: AI is good at narrow tasks with lots of training data and clear right-or-wrong answers. It's bad at anything that needs context, judgement under uncertainty, or accountability. A doctor is doing more than diagnosis, they're reading the patient, taking responsibility, and making a call about an unknowable future. AI can support those calls. It can't make them.
For your business, this means: AI can draft a quote. It cannot decide what your business is worth. It can write a social post. It cannot decide what your brand stands for. The question isn't "is AI better than humans", it's "what's the right division of labour."
"AI will replace tradies"
This one's the doom version of the same fantasy. AI can't change a tap, climb a roof, or wire a panel. It can't physically be on a job site. It cannot do the work you actually do.
What it can do is replace some of the people who support you, the receptionist who took your calls, the bookkeeper who reconciled your invoices, the SEO agency who managed your rankings. Some of those jobs are genuinely changing. Yours isn't.
If anything, AI makes tradies more valuable, not less. The supply of physical-skill labour stays the same; the supply of admin labour drops; the leverage on your hourly rate goes up.
"This AI tool gets smarter the more you use it"
Marketing-speak. Most consumer AI tools don't actually learn from your specific data, they use a fixed model and the "personalisation" is shallow context-stuffing.
What's actually true: there are real personalisation mechanisms, fine-tuning, retrieval-augmented generation, prompt engineering on your business context, but they require deliberate engineering and ongoing investment. They don't happen automatically because you used the tool a lot.
If a vendor tells you their AI "learns your business" without explaining how, ask them to show you. The honest answer is usually some combination of "we feed your data into the prompt" and "we adjust based on your corrections", both are legitimate, but neither is magic.
What's genuinely worth caring about.
The flip side. These are the things AI in 2026 is actually good at, not "kind of good," not "with caveats", genuinely good enough that ignoring them is now leaving money on the table. If you're not using AI for at least one of these, your competitors who are will start to pull away.
Answering the phone when you can't
This is the obvious one. AI voice quality crossed the "actually fine" line in 2024 and is now better than most virtual receptionist services for routine call handling. It captures details accurately, asks reasonable follow-up questions, and books tentative slots without needing a script you have to maintain.
What it won't do: handle a customer who's genuinely upset, pick up the unspoken context of a long-term client, or make a judgement call about whether to bend a rule. Those still escalate to you. The point is everything else, the bulk of inbound calls that are routine, gets picked up and handled in your voice while you're on the tools. The missed-call-equals-lost-job spiral stops being your default setting.
Drafting work in your voice
Quotes, emails, social posts, follow-up messages, review responses. Anything where the format is structured, the input is clear, and the output goes through your approval before it ships. AI is genuinely fast and genuinely good at this in 2026, better than most humans, including (sorry) the bookkeeper who used to draft your stuff in Word.
The trick is the approval layer. AI drafts; you review. Once you trust it for a specific task, you can promote it to auto-send with override. The drafts get sharper as you correct them, because the system learns your style and your standards. After a few months, your quotes go out faster and read better than they ever did when you were doing them yourself at 10pm on Sunday.
Watching everything at once
Humans are bad at monitoring. We get bored, we miss things, we forget to check. AI doesn't. If you set up an agent to watch your reviews across Google and Facebook, your local rankings across the suburbs you service, your unanswered enquiries, your overdue invoices, and your compliance expiry dates, it will watch all of those things, every day, and surface the things that need your attention.
This is unglamorous and underrated. It's also one of the biggest reasons small tradie businesses leak money, things fall through the cracks because no one is paid to watch them. AI watches them for free.
Doing structured analysis fast
Per-job profitability calculations. Pricing reviews. Quiet-period forecasting. Margin checks across services. Tax-time data prep. Anything that's mostly maths and pattern-matching across your existing data, AI does in seconds, work that used to take you two hours in a spreadsheet on a Sunday.
It still needs you to ask the right questions. It still needs you to act on the answers. But the gap between "I should look at the numbers" and "I have the answer" went from days to seconds, and that changes how often you actually look.
One rule that won't change.
If you take one thing away from this article, take this:
Anywhere a question has a clear answer your business already knows the rule for, pricing for a standard job, format of a quote, timing of a review request, sequence of a follow-up, that's execution. Hand it off. Let the system do it.
Anywhere a question requires judgement under uncertainty, what to charge for an unusual job, whether to take on a difficult client, how to handle a complaint, what direction to grow the business, that's a decision. Keep it. Don't ever hand it off, no matter what the marketing says.
The line between the two is what makes AI useful instead of dangerous. Tools that respect that line will help you. Tools that ignore it will eventually cost you a customer, a reputation, or worse.
This is also, incidentally, the reason F*IT exists. Every other tradie SaaS we looked at either treated AI as a gimmick (a chatbot bolted onto a CRM) or treated it as a replacement for the owner (autonomous everything, decisions included). Both miss the point. The interesting product is the one that takes execution off your plate cleanly while leaving you genuinely in charge of the decisions that matter.
Which is what we're building. But that's not the point of this article. The point is: now you know what to look for.
If a tool you're considering crosses the line, promises decision-making, "set and forget," or "AI runs your business", walk away. There are better ones. There always will be. The good versions of this technology respect what you're for and what they're for, and they explain the difference clearly.
Don't let the hype cycle pick your tools for you. Pick them yourself, with eyes open, knowing what's bullshit and what's real.
Honest tools, honest pricing.
Want to see what AI is actually doing for your business, and what we'd never let it touch? Audit your stack and see what we replace.