Guide

How to sell AI receptionists to local businesses (the playbook that works in 2026)

By Rey · Scale Client One · · 6 min read

Most guides on this topic are written by people who have never called a plumber. This one is the process I use, with the numbers I use. If you do nothing else, read the section on the demo link. It is the whole business.

Who actually buys an AI receptionist

Not "small businesses". A specific kind of small business: one where the owner is the person who answers the phone, and the owner is usually doing something else with their hands when it rings. Plumbers under a sink. HVAC techs in an attic. Roofers on a roof. Electricians in a panel. Auto shops with a wrench in hand. Martial-arts schools mid-class. Salons with a client in the chair.

For those businesses a missed call is not an inconvenience, it is a lost job, and the caller does not leave a voicemail. They ring the next name on Google. The owner knows this. Ask any plumber how many calls he misses a week and he will give you a number without thinking, because he has already done the maths on what it costs him.

That is your buyer: a trade or appointment business, one to ten people, owner answers the phone, phone rings while they work. Dentists with a front desk are a harder sale. Restaurants are a maybe. Start with plumbing and HVAC.

What you are actually selling

Not software. The owner never logs into anything. What they get is an email and a text within seconds of every call: caller's name, number, what they need, and whether it is urgent. That message is the product from their side. Everything else, the voice, the AI, the dashboard, is your problem, not theirs.

When you describe it, describe the text message. "Your phone buzzes, it says Dana Ruiz, 847-555-0142, water heater leaking, still flowing, 218 Maple. You call Dana back." That sentence sells more than any feature list.

Here is the part that matters. Do not pitch. Build the receptionist first, then let them hear it.

With a modern platform you paste the business's website and sixty seconds later you have a receptionist that knows their services, their hours and their service area, on a public link that works for 48 hours with no login. The prospect calls it, says what a customer would say, and hears their own business name come out of it. That is the moment people buy. Nobody buys from a description; they buy after hearing it handle a burst pipe at 9pm.

The sequence, over whatever channel the owner actually reads:

Message 1, day 0. "Hi Dana, Rey here. I built an AI receptionist for Northside Plumbing from your website. Call it here, no login, works for 48 hours: [link]. Say what a customer would say and see what shows up in your email after."

Message 2, day 1 or 2. "Did you get to call it? If you did, what did you think? If you didn't, try it on your drive home, it takes a minute."

Message 3, after a positive reply. The proposal: one page with what missed calls cost them, your price, the demo embedded, and an Accept button that takes payment.

Message 4, day 7. "The proposal's still open. If you're in, hit Accept and I'll have it live on your line within a day. If something's holding you back, tell me what it is and I'll answer straight."

Rules that make it work: build the demo before you send anything, never send without one. One call to action in every message, always the same one: call the link. Under 90 words. Stop after five touches. Any reply, including "no", ends the sequence, and you reply as a human within the hour.

The 60-second phone call

If you would rather call than message, this is the script. Time it; it really is a minute.

Opener, 10 seconds. "Hi, is this Dana? Rey here, I'll be quick. I built an AI receptionist for Northside Plumbing off your website. Do you have 30 seconds for me to tell you what it does, or is now a bad time?" If bad time: "No problem, I'll text you the link so you can call it when you're free. What's the best number?" Send it. Done.

Pitch, 20 seconds. "Every call you can't answer, it picks up. Gets the name, number, what they need, whether it's urgent. You get an email and a text within seconds. Works nights and weekends, and you never log into anything. Quick question: how many calls a week do you reckon you miss when you're under a sink?"

They will give you a number. Then: "Right, and each of those is a job. Here's the thing, I already built yours. Can I text you a link so you can call it and hear what your customers would hear?"

Close, 15 seconds. "I'll text it now. Call it, say you need a quote, then check your email. If it's good, it's $397 a month, no contract. If it's not, tell me and I'm gone. Sound fair?"

Gatekeeper: "I'm not selling anything on this call, I built something for the owner to listen to. Can I text or email a link so they can hear it?" Voicemail: fifteen seconds, say you built it, say you are texting the link, then text the link immediately.

The twelve objections, answered

"We have voicemail." How many people leave one? Most hang up and ring the next name. This picks up, gets the details, and texts you, so you are calling back a lead instead of a silent voicemail.

"Customers hate robots." They hate hold music and voicemail more. Call the demo and judge it yourself. If you would hang up on it, don't buy it.

"Too expensive." What is one job worth to you? If it captures one call a month you would otherwise miss, it is paid for.

"What if it says something wrong?" It only knows what is on your website and what you tell me. It does not quote prices or make promises unless you tell it to. If something is off, I fix it the same day.

"I'll think about it." Fair. Do one thing first: call the demo tonight when you would normally not be answering. Then decide. I'll check in Thursday.

"We already have a receptionist." Good. This covers lunch, after hours, weekends, and when both lines are busy. It does not replace her; it stops calls falling through when she cannot get to them.

"What happens after hours?" Same thing that happens at noon. It answers, tags it urgent if it is, and texts you. You decide whether to call back now or in the morning.

"Is my data safe?" Caller details go to your email and your phone. Nobody sells them. Do not invent certifications you do not have.

"Can it book appointments?" Yes, with a calendar it can connect to. Otherwise it takes the request and you confirm the time on the callback.

"I don't want to be locked in." Month to month. Cancel any time. No contract, no cancellation fee.

"Does it sound robotic?" You tell me. Call the demo. That is the exact voice your customers would hear.

"What does it cost you to run?" Less than I charge you, same as any service. You are paying for a phone that gets answered every time and a text in your pocket seconds later.

Pricing and the margin math

Charge $300 to $500 a month per business. $397 is a good number: under $400, obviously more than a gadget, obviously less than a person. No setup fee for the first few clients; add one ($250 to $500) once you have references.

Your cost on a platform built for this is a flat subscription plus a phone number per client and a per-minute voice cost. On Scale Client One that is $29 to $119 a month depending on how many clients you run, $5 a month per number, and roughly a quarter a minute beyond the included minutes. Five clients at $350 is $1,750 a month against a cost under $150 before minutes. Your margin is not the interesting number to the buyer, though. What sells is the $400 next to the $29: they can see the business.

Do not discount to close. Discounting tells them the price was made up. If they hesitate, offer the first month at cost and a cancel-anytime promise instead.

From yes to live, same day

When they say yes, the demo becomes the real receptionist in one click; same voice, same knowledge, call cap lifted, alerts switched on. You add a local number (or forward their existing line to it), you put the owner's email and mobile in, they reply YES to one text to confirm alerts, and it is live. Send them a one-line message: "You're live. Call your own number and watch your phone." Then leave them alone for a week and let the alerts do the selling.

Every Monday they get a short email with the week's calls and the estimated value of the leads it took. That email is your retention plan. Nobody cancels the thing that texts them jobs.

Your first week

Day 1: pick one trade and one town. Find twenty businesses on Google Maps with a website and a four-plus rating. Build twenty demos. Days 2 to 4: send Message 1 to all twenty. Call the ones you can. Reply to anyone who answers within the hour. Day 5: send Message 2 to everyone who has not replied. Send proposals to anyone who called the demo. Day 7: send Message 4 to open proposals. Count: demos sent, demos called, replies, proposals, accepts. Fix the biggest drop-off first.

Twenty demos a week, honestly done, gets most people their first paying client inside two weeks. The second is easier, because now you have someone to name.

Questions people ask

Do I need technical skills to sell AI receptionists?

No. With a platform that builds the receptionist from the business's website, your job is finding businesses, sending the demo link, and following up. Nobody writes prompts or touches servers.

How much should I charge for an AI receptionist?

$300 to $500 a month for a local service business, no contract. $397 is a sensible starting point. Add a setup fee once you have references.

What kind of businesses buy AI receptionists?

Trades and appointment businesses where the owner answers the phone while working: plumbers, HVAC, electricians, roofers, auto repair, martial-arts schools, salons. Start with plumbing and HVAC.

How long does it take to get the first client?

Doing twenty demos a week with the follow-up sequence above, most people close their first client within two weeks. It is a numbers game with a very short cycle, because the demo does the convincing.

Does the business owner have to use software?

No. They never log in. They get an email and a text for every call, and a weekly summary. That is the entire product from their side.