The short version
ServiceTitan’s new 2026 State of AI in the Trades survey of 1,017 contractors has one number every headline leads with — 66% of contractors using AI save at least three hours a week — and two numbers most headlines bury: only 47% report better profitability and only 43% report more revenue. Time is being saved. For most shops, it isn’t turning into money.
- Trust is low: 11% highly trust their AI tools; 58% trust them little or not at all.
- Cost isn’t the barrier anymore (17%) — knowing how to use it is (44%).
- Why contractors try AI: they can’t hire enough people (37%).
- The fix isn’t more AI. It’s pointing saved hours at the three things that make money: answering calls, dispatching fast, and getting paid on the job.
Below: the full numbers, how to read a survey published by a company that sells AI, and a 30-day plan to turn saved hours into revenue.
On this page
The numbers at a glance
ServiceTitan released The Update: 2026 State of AI in the Trades on September 29, 2026. The research was run by Thrive Analytics on ServiceTitan’s behalf. Here’s what it found.
Time saved vs. money made
Trust in the AI tools contractors use
| Finding | Number |
|---|---|
| AI users saving 3+ hours per week | 66% (40% save 5+) |
| AI users reporting productivity gains | 64% |
| AI users reporting faster decision-making | 55% |
| AI users reporting improved profitability | 47% |
| AI users reporting increased revenue | 43% |
| High or complete trust in AI tools | 11% |
| Little or no trust | 58% |
| Top concern among AI users: accuracy and reliability | 35% |
| Top reason to try AI: difficulty hiring | 37% |
| Barrier: understanding how to use AI | 44% |
| Barrier: high purchase cost (down 12 points) | 17% |
| Actively engaged with AI (Dec 2025 → Sept 2026) | 46% → 52% |
| Prefer AI built into their core platform | 61% of users |
Full release: ServiceTitan via GlobeNewswire
The gap the headlines skip
Two out of three AI users save three or more hours a week. Fewer than half say their business is more profitable, and fewer still say it’s bringing in more money. ServiceTitan’s own framing calls this an “opportunity to translate efficiency” into results. For an owner, it’s simpler: the hours are real, but most of them aren’t landing anywhere that pays.
Three reasons that happens in real shops:
- The time comes back in crumbs. Fifteen minutes saved writing a job note, ten on an email, twenty on a schedule. Three hours spread across a week in 10-minute pieces doesn’t become one more job. It becomes a slightly shorter day.
- It’s saved in the office, but revenue is made in the field. Most AI tools help the owner or office manager. Revenue comes from calls answered, trucks rolling, and invoices collected. If the saved hour doesn’t change any of those, the P&L doesn’t move.
- Nobody decided where the time goes. Saved time with no plan fills up with more admin. That’s not a technology problem — it’s a management one.
What 3 hours a week is actually worth
Here’s the same three hours, pointed two different ways. The dollar figures are placeholders — swap in your own average ticket.
Absorbed into more admin: $0 in new revenue.
Pointed at one extra booked job a week: 50 jobs × a $350 average ticket = $17,500 a year.
Pointed at answering calls you’re missing today: depends entirely on how many you miss — which is exactly why you should count them. Illustrative only. The point isn’t the number — it’s that the same saved hours are worth nothing or five figures depending on where they go.
The survey tells you the hours exist. Whether they become revenue is decided in your shop, not by the tool.
The trust problem: 11% vs 58%
Only 11% of contractors place high or complete trust in the AI tools they use. More than half — 58% — report little or no trust. Among AI users, the top concern is accuracy and reliability, cited by 35%.
That’s not contractors being stubborn. It’s contractors being right. In a trade business, a confidently wrong answer about a price, a code requirement, or an arrival time costs you a customer. Low trust is a reasonable response to tools that are useful most of the time and wrong some of the time.
The practical rule: let AI draft; let a person decide. Anything involving money (prices, discounts, refunds), promises (arrival times, warranties), or safety (codes, diagnostics) gets a human sign-off before it reaches a customer.
Why contractors are trying AI in the first place
The top reasons contractors gave for experimenting with AI:
- Difficulty hiring / the workforce gap — 37%. AI can’t turn a wrench. But it can take work off the people you have, and good dispatch can squeeze more billable jobs out of the same crew.
- Off-hours support — 34%. This is where AI pays fastest: a 9 p.m. call that gets answered and booked is revenue a voicemail loses.
- Reducing operating costs — 33%. Fine — but cost savings are capped. Revenue isn’t.
Notice that the #1 reason is a people problem. The best return from any tool — AI or not — comes from getting more paid work out of the team you already have.
The barrier moved: from price to know-how
Last year, cost was a bigger worry. This year, high purchase cost fell 12 points to 17% and ranked last among the barriers measured. What replaced it:
| Barrier to using AI | Share of contractors |
|---|---|
| Understanding how to use AI in their business | 44% |
| Lack of clear use cases | 43% |
| Integration with existing software | 40% |
| High purchase cost | 17% |
And when asked what would encourage more adoption, contractors pointed to clearer ROI (40%) and AI trained on trade-specific data (38%). In plain English: show me what it’s for, and show me it pays. The rest of this post does exactly that.
How to read a survey published by a company that sells AI
This is useful research, and it was run by an independent firm. It’s still worth reading with the publisher in mind:
- ServiceTitan sells AI. Its Max platform is the center of its current strategy, and the survey found 61% of AI users prefer AI built directly into their core platform — which lines up neatly with what it sells. That doesn’t make the number wrong. It makes it a number to think about, not just repeat.
- It’s self-reported. “Saves three hours a week” is what contractors estimated, not what a stopwatch measured.
- Many findings are about AI users only, a subset of the 1,017. About half of contractors reported active engagement with AI.
- The fine print is honest about limits: a margin of error of ±3.7 points for the full sample and wider for subgroups, question wording lightly revised between survey waves, and a disclaimer that ServiceTitan gives no assurances about the data’s accuracy.
Our read: the direction of the findings is believable, because it matches what owners say every day — AI helps with paperwork, and most shops haven’t connected it to revenue yet.
The AI revenue test: 5 questions before you buy anything
Run any AI tool — or any software — through these before you pay for it:
- Does it answer or book more calls? Missed calls are the most expensive leak in a service business.
- Does it get a tech moving sooner? Minutes from call to dispatch decide who gets the job.
- Does it get you paid faster? Money collected on the job beats invoices chased for 30 days.
- Does it cut unbilled time? Drive time, wrong-part trips, and callbacks are hours nobody pays for.
- Does it work with the system you already dispatch from? 40% of contractors named integration as a barrier. A tool your team has to copy data into will be abandoned.
If a tool can’t say “yes” to at least one of the first four, it’s a time-saver at best. That’s fine — just don’t pay revenue-tool prices for it.
Where AI actually pays for a 1–10 tech shop
| Use | Why it pays | Measure it by |
|---|---|---|
| After-hours call answering and booking | Catches jobs that go to voicemail | Missed calls, after-hours bookings |
| Following up on unsold estimates | The cheapest revenue is a quote you already gave | Estimate close rate |
| Drafting estimates and job notes (human reviews) | Faster quotes, cleaner records | Time to send a quote |
| Replying to reviews | Keeps your rating and response rate up | Review response time, rating |
| Helpful website content | Gets you found — by people and by AI answers | Calls and form fills from search |
Where to keep a human in charge: pricing and discounts, diagnoses and code questions, warranty promises, anything said to an upset customer, and anything that sends money.
On the website point: we tested it on our own site. See how we got cited by Google’s AI — the short version is that helpful, accurate content beats any AI trick. For reviews, our guide to handling bad Google reviews covers what to automate and what not to.
A 30-day plan to turn saved hours into revenue
Week 1 — Measure your leaks
Count three things: missed or unanswered calls, average minutes from call to tech dispatched, and average days from job complete to paid. You can’t prove AI paid off without a before number.
Week 2 — Pick one leak, one tool
Choose the biggest leak from week 1 and one tool aimed at it. One. Stacking tools is how the 44% who “don’t know how to use it” get there.
Week 3 — Decide where the saved time goes
Write it down: “The hours we save go to ___.” Returning calls, one more job per day, same-day estimate follow-ups. If you don’t assign it, admin will.
Week 4 — Compare and decide
Re-measure the same three numbers. If one moved, keep the tool and assign the next leak. If nothing moved, cancel it — and be glad you didn’t sign a multi-year contract to find that out.
The Unglamorous Part That Makes the Money
VortechPro isn’t an AI product. It’s dispatch built for the three leaks above: jobs ping the nearest qualified tech, the office sees every truck on live GPS, techs attach photos, and customers pay by card before the tech leaves. $99/mo for 5 techs, +$20 per extra tech. Month-to-month.
The bottom line
ServiceTitan’s survey is good news with a catch. Contractors are using AI, and most who do are saving real time. But fewer than half are seeing it in profits or revenue, trust is thin, and the new barrier isn’t price — it’s knowing what to point AI at.
So point it at money. Measure your leaks, fix one at a time, and assign every saved hour a job before it disappears into admin. And whatever you buy — AI or not — make sure you can leave if it doesn’t pay. If you’re on a long contract now, here’s how to get out of a ServiceTitan contract without paying for months you won’t use.
FAQ
What did ServiceTitan’s 2026 State of AI in the Trades report find?
Among contractors using AI, 66% save at least three hours a week, but only 47% report improved profitability and 43% increased revenue. Only 11% highly trust their AI tools, while 58% report little or no trust. Top reasons to try AI: difficulty hiring (37%), off-hours support (34%), and reducing costs (33%).
Who conducted ServiceTitan’s AI survey?
Thrive Analytics, on ServiceTitan’s behalf, polling 1,017 U.S. residential and commercial contractors across HVAC, plumbing, electrical, roofing, garage doors, pest control, landscaping, and fire and life safety. Fielded May 26 to June 15, 2026; margin of error ±3.7 points, wider for subgroups.
Does AI increase revenue for contractors?
Sometimes. In the survey, 43% of AI users reported increased revenue and 47% better profitability, while 66% saved three or more hours a week. Saved time becomes revenue only when it goes into answering more calls, dispatching faster, running more jobs, or collecting payment sooner.
Why don’t contractors trust AI tools?
Accuracy and reliability were the top concern among AI users, cited by 35%, and only 11% reported high or complete trust. Keep a human approving anything involving money, pricing, or promises to customers.
What is the biggest barrier to AI adoption in the trades?
Know-how, not price. 44% cited understanding how to use AI, 43% a lack of clear use cases, and 40% integration with existing software. High purchase cost fell 12 points to 17%, ranking last.
Where does AI pay off for a small trades business?
In tasks tied to revenue: answering and booking after-hours calls, following up on unsold estimates, drafting estimates and notes for human review, replying to reviews, and helpful website content. Measure each against a baseline like missed calls, close rate, or days to get paid.
Should I trust a survey published by a software company?
Read it as useful but not neutral. ServiceTitan sells an AI platform, and the survey found 61% of AI users prefer AI built into their core platform. The research was run by an independent firm, but the findings are self-reported and the publisher has a stake in the conclusions.
Is VortechPro an AI product?
No. VortechPro is field service dispatch software. It targets the same gap the survey shows — turning time into revenue — through faster dispatch to the nearest qualified tech, live GPS, job photos, and card payments in the field, at $99/month for 5 techs plus $20 per additional tech.
