A general contractor does not want more calls. He wants to stop driving to free estimates for people who were never going to buy. Here is how to sell, scope and price an AI system for remodelers and design-build shops, including who to walk away from.
Lumina··13 min read
Sell a general contractor the same AI receptionist you sold a plumber and you will lose the deal. A plumbing company wants every call answered because every call is a job. A remodeler who runs $40,000 kitchens does not want more calls. He wants to stop driving forty-five minutes to a free estimate for someone with an $8,000 budget. The system you build for a contractor has to qualify, schedule and follow up. Answering the phone is the least interesting part of it.
Get that distinction right and the contractor leans in, because you are describing his actual week. Get it wrong and you sound like the fourth marketing company that called him this month promising leads.
AI agency software for sales and fulfillment
Custom fulfillment from our engineering team, live AI demos, and a client portal, all 100% white-labeled.
This is a playbook for general contractors, remodelers and design-build shops. It assumes you are a one to ten person agency delivering the work, or having someone deliver it for you.
Why is a contractor a harder sale than a plumber?
Three things make it different.
The owner is the salesperson. In a plumbing or HVAC company of any size, a dispatcher books jobs and a tech runs them. In a remodeling company doing a few million a year, the owner still runs most estimates himself. His calendar is the constraint on the entire business. Every hour he spends at a kitchen table is an hour he is not managing the three jobs already in progress. So the thing you are protecting is not his phone. It is his Tuesday.
The sales cycle is long and the ticket is big. A drain clog closes in four minutes on the phone. A bathroom remodel takes two site visits, a design conversation, a proposal and three weeks of silence. The money in a contractor's business is not sitting in the unanswered calls. It is sitting in the twelve open proposals nobody has touched since they went out.
Lead quality is the complaint, not lead volume. Contractors buy from Angi and Thumbtack and Google Local Services and they are tired. Most of them have already paid for leads that turned out to be tire kickers, and they are allergic to the word. If your pitch is "we will get you more leads," you have been sorted into the same bucket as everyone else before you finish the sentence.
What does the remodeling market look like going into 2027?
Slow and flat, which matters for how you pitch.
Harvard's Joint Center for Housing Studies puts annual spending on home improvements and repairs at $509 billion in late 2025, rising to roughly $518 billion by the end of 2026, with year-over-year growth easing from 2.1% in mid-2026 to 1.6% by the fourth quarter, according to its Leading Indicator of Remodeling Activity. The market is not shrinking. It is just not growing fast enough to cover sloppiness.
Read that into your pitch. In a boom you sell growth. In a flat market you sell conversion rate. The contractor is not getting more inbound next year. He has to close more of what he already gets and spend fewer windshield hours doing it, which is a better pitch anyway because it is measurable on his existing numbers.
One more number worth carrying into the room. In a Moneypenny survey of 500 construction and building trade companies and 2,000 consumers, reported by Contracting Business in May 2026, 83% of US consumers said they go with the business that responds first, 41% said a response within minutes would meet their expectations, and response speed ranked as the top loyalty factor at 36%, above reliable quality and brand trust. For a homeowner choosing between three contractors for a $60,000 job, first contact is doing more work than any contractor wants to admit.
Have contractors already tried AI?
Yes, and you need to plan for it. This is the biggest change in this vertical over the past year.
The 2026 US Houzz State of AI in Construction and Design Report, which surveyed 601 US construction and design businesses in June and July 2026, found 52% of construction firms using AI for everyday business tasks, up 20 points in a year. The top function was sales and marketing at 64%, ahead of planning and design at 61%. Over half of pros said they save three or more hours a week.
So when you walk in, assume he has used something. Probably ChatGPT, to write a proposal or clean up an email. He may believe he has already "done AI."
He has not done what you are selling, and the gap is easy to name. Writing a proposal faster is a tool he operates. A receptionist that answers at 7pm and books a qualified estimate into his calendar is a system that runs without him. That difference is the entire pitch, and it lands better than any feature list. If you want the longer version of this conversation, we wrote it up in what to say when a prospect asks why they cannot just use ChatGPT.
Two more findings from that research matter. Concern about AI reliability and accuracy rose from 33% to 37% year over year, so skepticism is climbing alongside usage. And on the homeowner side, 22% have used AI on their projects while 80% still hire a professional to do the work. Nobody in this trade thinks AI is replacing the contractor, which makes your job easier. You are selling a better front desk, not a replacement.
Worth tempering all that with how little money is actually changing hands. Intuit's construction AI adoption report, drawing on anonymized payment records from more than 524,000 US construction businesses, found only 7% currently paying for AI tools, with bookkeeping the top paid use case at 46%. Privacy and security was the leading barrier at 30%, with cost third at 23%. Plenty of free experimenting, very little paid deployment. And separately, Dodge Construction Network's survey of 235 general and trade contractors, covered by Construction Dive, found 40% already allocating a dedicated AI budget and 51% actively evaluating changes.
Put those together and you get the real state of the market: high interest, real budget forming, almost no paid systems in place, and a bias toward back-office uses rather than the front desk. The front desk is wide open.
What actually breaks in a contractor's front office
Ask these five questions on a discovery call and you will find the money.
How many estimates did you run last month, and how many turned into jobs? If the answer is twelve and three, his problem is qualification, not volume. He gave away nine unpaid afternoons.
What happens when someone calls at 6:30pm? Usually voicemail, or his cell while he is at dinner. Contractors answer their own phones far more than they admit, and they hate it.
How long does a web form sit before someone replies? This is where it gets ugly. Many contractors check the inbox at the end of the day. Against that 83% number above, a six-hour reply is a lost job.
How many proposals are open right now that you have not followed up on? Every contractor has a number here and every contractor is embarrassed by it. This is the single most profitable thing you can automate for them, because the leads are already qualified and already interested.
Who answers the "can you just give me a rough price" call? The answer is him, badly, from the truck, and it either scares the customer off or anchors a number he regrets.
Notice that none of those five are "I am not getting enough calls."
What should you build for a general contractor?
Three tiers, and I would push almost every contractor to the middle one.
Entry. A voice agent that answers the phone and the website chat. The script does three jobs: find out the project type, the rough scope and the timeline, confirm the property is in the service area, and book an estimate slot only if those clear a threshold the contractor sets. Everything else goes to a message. The point is not answering calls. The point is that his calendar only fills with people worth driving to.
Middle, and the one that fits most shops. Add outbound follow-up on open proposals. A text, then an email, then a call attempt on a schedule he picks. Add after-hours chat on the website with the same qualification logic. Add a review request once a job closes. This is where a contractor feels it, because the follow-up sequence works on deals he has already paid to create.
Top. Add a CRM with the estimate pipeline in it, so the owner can see every open proposal, every booked site visit and every lead that failed qualification and why. The "why" matters. After two months the contractor can see he is turning away every bathroom under $15,000, which is a business decision he has been making by accident for years.
One rule across all three. Do not let the agent quote prices. Not a range, not a ballpark, not "most kitchens run between." A custom remodel price depends on the house, and a number from a phone agent becomes an anchor the contractor has to argue down in person. Set the script to collect scope and hand it off. Say this out loud on the sales call, because it is the moment a skeptical contractor decides you understand his business.
The vertical pack behind the demo already carries trade-specific services, prices and the objections you will hear, and the library covers 133 trades, so there is a general contractor pack plus adjacent ones for remodeling and design-build. If his niche is narrow, you can build a pack from his own website. We go deeper on how the packs change demo outcomes in what's inside a vertical pack.
How do you run the demo call?
Send the demo before the call, not during it. Pick the contractor pack, type his company name, publish, send the link. He plays with it on his phone at 9pm and comes to the call already past the "does it sound human" question. That is the single biggest change in close rate available to you.
On the call, do three things.
Run the website audit against his site live. Most contractor sites have no chat, a form that goes to an unmonitored inbox and no after-hours path. You are not insulting his site. You are showing him three specific holes with his own domain on the screen.
Call the demo together and ask it a hostile question. "Can you give me a price on a kitchen?" Let him hear it decline cleanly and move to scheduling. That one exchange sells the no-quoting rule better than any explanation.
Then put his number on it. Twelve estimates, three closes, nine wasted afternoons. If qualification cuts the wasted visits in half, what is four and a half afternoons worth to him in November. Let him do the math out loud. Our full demo call playbook covers the sequencing in more detail.
What should you charge a contractor?
Start from what the build costs you. Lumina's published wholesale pricing runs from $100 to $200 setup for one service on its own at $25 to $45 a month, through Tier One at $650 to $775 setup and $170 to $215 a month, Tier Two at $800 to $1,075 setup and $270 to $415 a month, up to Top Tier at $1,150 to $1,400 setup and $315 to $460 a month plus $0.35 a minute for voice. Every build is priced on a thirty-minute call before work starts. What you charge on top is yours.
Contractors are a good vertical for retail pricing because they already think in setup plus monthly. That is how they sell. They quote a job and then a maintenance plan, or a project and then a service agreement. You are speaking their structure.
Two specifics for this trade. Price the setup against one job, not against a software budget. A contractor whose average job is $35,000 is not comparing your $2,000 setup to a $99 app. He is comparing it to one additional closed bathroom. Say that. Second, the monthly is easier to hold than in cheaper trades, because the value scales with his ticket size. The margin math across any tier is laid out in what you actually keep on a white-label AI client.
Where does this fall apart, and who should you walk away from?
Some contractors should not buy this, and saying so early will win you more deals than it costs you.
The owner-operator with one crew. If he runs five estimates a month and answers 90% of his own calls, the qualification argument does not hold up. There is not enough volume for triage to matter. Sell him the after-hours chat and the proposal follow-up as a single service, or tell him to come back when he has two crews. Pitching him a top tier build is how you get a cancellation in month three.
The contractor whose problem is production, not sales. A lot of remodelers are booked eight weeks out and losing money on change orders and scheduling. AI at the front desk does nothing for that. If he tells you he has plenty of work and cannot deliver it, you are the wrong vendor. Be honest and ask for a referral instead.
Anything touching permits, code or insurance claims. Keep the agent away from all of it. A wrong answer about a permit requirement or a restoration claim is not an embarrassing moment, it is liability. Scope those to a message and a callback.
The design-heavy firm where the first conversation is the product. Some architecture-led design-build shops sell on the quality of the first twenty minutes with the principal. For them, an agent in front of that conversation is a downgrade. The receptionist still helps after hours, but you should not promise it will qualify. It will reject people who would have been good clients.
And the thing nobody mentions: voice reliability on a job site. Contractors call you back from a truck with a compressor running. Background noise degrades any speech system. Test it. Call the demo from a loud environment during the sales process rather than discovering it in week two.
FAQ
Is a general contractor a better client than a plumber or an HVAC company?
Usually yes on ticket size and no on sales cycle. A contractor's bigger average job makes your retail price easier to justify and the monthly easier to hold. But he takes longer to decide, because he is the one person in his company who makes every decision and he is on a roof. Expect two to four weeks from demo to signature where a plumber might take four days.
What if he says he already uses AI?
Ask what for. Almost always it is writing, in line with the Houzz finding that sales and marketing is the top AI function in construction firms at 64%. Then draw the line between a tool he operates and a system that runs when he is asleep. Do not argue that his ChatGPT use does not count. Agree that it is useful and point at the different job.
Should the AI give price ranges on common jobs?
No. For repeatable work with fixed pricing, a service plumber's drain clear, a range is fine. For custom remodeling it anchors the customer against a number that depends on things nobody has seen yet. Collect scope, book the visit, let the contractor price it.
How long does a contractor build take to deliver?
The demo you send takes about ten minutes to make. The client build is scoped and priced on a thirty-minute call, and the timeline comes out of that call based on what he needs. Tell him the demo timing and the call, and do not invent a delivery date you cannot stand behind.
What about seasonality?
Sell into the slow months. A remodeler in a cold climate signs more easily in January than in June, and the system is live and tuned before his season starts. In peak season, lead with the proposal follow-up, because that is the piece that pays off inside thirty days.
The next step
Pick two contractors in your area whose websites have no chat widget and a contact form with no visible response promise. Those two things take ninety seconds to check. Build a demo for each one with the general contractor pack, put their company name on it, and send the link with one sentence: "I built this for your company, call it and ask it about a kitchen." Then run the website audit on their site before the call so you have three specific holes ready.
That is two demos and about twenty minutes of work. If you do not have a workspace yet, the plan is $49 a month and includes ten live demos at once, 100 voice minutes and the trade library, which is enough to run this play on a dozen contractors before you owe anyone anything else.