This month, AI got cheap. Google's newest model dropped to around a dollar to read a million words of text — a fraction of what the same power cost a year ago — and a wave of new tools launched alongside it: automatic bookkeeping that flags your dodgy transactions, marketing tools that build a whole campaign from one sentence, and "agents" that run tasks on their own while you sleep.
If you run a trades or home-service business, you've probably felt the pressure to finally "do something about AI." And the usual reason to wait — it's too expensive — just disappeared.
Here's the uncomfortable part: cheaper AI is about to cost a lot of trades businesses money. Not because the tools are bad, but because the barrier was never really price. It was knowing which tool fixes a problem you actually have — and cheap, plentiful tools just make it faster to buy the wrong ones.
What actually changed this month
Three things worth knowing, in plain terms:
- AI got much cheaper. Google released Gemini 3.7 Flash at an introductory price of about $0.75 per million tokens in and $3.75 out (a "token" is roughly three-quarters of a word). The tools built on top of models like this get cheaper too.
- The tools multiplied. In a single week: Canva added one-sentence campaign creation, Xero launched AI that flags unusual transactions in your books, and no-code tools like Bardeen now watch how you do a task and then do it for you.
- "Agents" went mainstream. Several big platforms now offer AI that runs unattended — doing multi-step jobs without you clicking through each one.
All real, all genuinely useful to someone. The question is whether that someone is you, and for which job.
Why cheaper AI is a trap for trades
When something is expensive, you think hard before buying. When it's cheap and everywhere, you buy on impulse — a subscription here, a trial there — and three months later you're paying for five tools, using none of them, and no better off.
For a trades business, the risk is sharper than for most, for three reasons:
- Your time is the scarce resource, not money. A $20 tool that needs six hours of setup you don't have is more expensive than it looks. Most tools that "sat untouched" didn't fail — they were never set up.
- Tools that don't talk to each other create work. A quoting app, a separate booking app, a separate follow-up app, none connected, means you're the glue — copying details between them by hand.
- Automating a broken process just breaks it faster. If your quote process is a mess in your head, automating it doesn't fix it — it scales the mess.
The pattern we see constantly: a business's real problem isn't the thing they think it is. "I spend too long on admin" is usually downstream of a deeper issue — there's no single place a job lives, so everything gets done twice. Buy a tool for the symptom and you'll still have the disease.
Where AI genuinely pays off for trades
Strip away the noise and the tools that reliably make trades businesses money do one of two things: they stop you losing work you've already earned. Both are revenue leaks you can't see, because you only notice the jobs you win — never the ones that quietly slipped away.
Notice what these have in common: each targets a named, specific leak, and each earns its cost back from a single recovered job. That's the test. Not "is this clever?" but "which job I'm losing does this stop?"
Not sure which leak is costing you most? That's the actual question worth answering before you spend a dollar — and it's exactly what a Custom AI Roadmap is built to figure out for your specific business.
The one question that decides everything
Before you buy any AI tool, answer this: what is the single biggest way my business loses money that I'm not already fixing?
Not "what can AI do." Start from your leak, then find the tool — never the other way around. If the honest answer is missed calls, start there and fully solve it before touching anything else. If it's quotes going cold, start there. If it's that jobs live in your head and three notebooks, the fix is a system to hold them, not another app.
The businesses that win with AI don't use the most tools. They fix one leak completely, bank the return, and then use that to fund the next fix. Cheap AI makes that easier — but only if you're pointing it at the right problem in the right order.
What to do before you buy anything
- Write down your top three revenue leaks. The jobs, calls, or quotes you know you're losing. Rank them by money lost.
- Check what you already pay for. Your booking or invoicing software may already include follow-up or reminders you've never switched on.
- Fix one, fully. One tool, set up properly and actually running, beats five half-configured trials.
- Don't automate a process you haven't written down. If you can't describe how a job flows from enquiry to paid, a tool can't either.
AI getting cheaper is good news for trades businesses — but only if you resist buying on impulse. Price is no longer the constraint; clarity is. Pick your single biggest revenue leak, fix it with one proven tool, and ignore everything else until that one is paying off. That discipline is worth more than any tool on its own.
Common questions
Is AI worth it for a small trades business?
Yes — but only for a specific problem, not as a general upgrade. The tools that pay for themselves in trades are the ones that plug a known revenue leak: missed calls while you're on a job, and quotes that go cold because nobody follows up. AI that automates those tends to earn back its cost within the first month. AI bought because it's cheap or trendy, with no clear job to do, is where trades businesses waste money.
Which AI tool should a tradesperson start with?
Start with the tool that fixes your single biggest revenue leak, not the most advanced one. For most trades businesses that means missed-call text-back (so a missed call becomes a booked job instead of a call to your competitor) or automated quote follow-up. A trades-specific platform like Jobber or Housecall Pro bundles several of these together. Pick one problem, fix it fully, then move to the next — don't buy five tools at once.
How much should a trades business spend on AI?
Less than most people now assume. The core tools that move the needle for trades run roughly $25 to $65 per month each, and you usually only need one or two to start. The cost of AI has fallen sharply in 2026, so price is no longer the real constraint — knowing which tool fits your business, and setting it up so it actually runs, is. Spending more does not mean getting more.
Want this worked out for your specific business instead of in general? Get a free report →