Is an AI receptionist right for your industry?

Rarely, and that is the useful answer. Your trade decides less about whether an AI receptionist fits than almost anyone expects, because what actually decides it is the mix of calls your line takes — and two businesses on the same street in the same industry can take completely different mixes. A plumbing company running on scheduled maintenance and one running on 2am burst pipes are buying different things. The question worth asking is not "does this work for dental" but "what share of my calls are answerable from written-down facts, and what share need somebody's judgment in the first sentence?" BotsDontSleep builds for fifteen industries and takes custom builds for anything with an inbound line, 24/7, at $0.60 CAD per answered minute with $0 setup. The sorting below is the part worth your time.

What actually decides whether an AI receptionist fits your industry?

Three questions about your calls, and none of them mention your trade.

What share of your calls are answerable from written-down facts? Hours, location, parking, price ranges, whether you take a given insurer, how long a service takes, what to bring, whether you are open on the Monday of a long weekend. If that is most of your volume, the fit is strong whatever your sign says.

What share need a decision nobody has written down? Quoting a job that has to be seen first, reading a medical situation, deciding whether to waive a fee for a customer of nine years. These are escalations rather than failures — but if they are the majority of your calls, what you are buying is a switchboard rather than a receptionist, and that is a smaller purchase than it sounds.

What happens to a call you miss? A caller who leaves a voicemail and waits is a different economic problem from one who hangs up and dials the next name on the list. The second case is where the arithmetic turns lopsided quickly, and that is a property of your market rather than your industry category.

Answer those three honestly and you have your answer before you have read a single industry page.

Which calls does an AI receptionist handle well in any industry?

Four kinds, and they look much the same in a dental office and a roofing company.

• Repeat information. The same eleven questions, arriving in a different order every time, answered from what you wrote down. It gives the fortieth caller of the day the same answer as the first.
• Booking, rescheduling and cancellations. Into your calendar, against your real availability, following your own rules about notice periods, deposits and who may be double-booked.
• Intake and screening. Name, callback number, address or file reference, the matter type, and the three qualifying questions you would have asked anyway — collected in the order you specified, with a readback on anything that has to be exactly right.
• Routing and escalation. Who gets this call put through now, who gets an SMS, and who finds a summary, transcript and recording waiting in the morning.

What those four have in common is that every one is a reliability job rather than a clever one. The argument for software on a business line is not that it is sharper than whoever would otherwise have picked up. It is that at 7pm on a Friday it does exactly what it did at 10am on a Tuesday.

Which calls does it hand straight to a person?

Anything where the next sentence needs judgment rather than information.

A quote on work nobody has looked at. A clinical question — it is not a triage line and does not behave like one, which is the whole design of how we handle a medical practice line. An argument about a bill. A negotiation over price. A caller distressed enough that what they need is a human voice rather than a correct answer.

Handled properly, none of those is the agent failing. Each is the agent recognizing a boundary you drew and moving the call across it with the context attached, so the person who picks up already knows who is calling and why. It never invents an answer. When something sits outside what it has been taught, it says so and escalates — and a provider who tells you their agent handles everything is describing a risk rather than a feature.

How do fifteen industries actually sort?

By call pattern rather than by trade. Five patterns cover nearly all of it.

Two things fall out of that table. First, the pattern predicts the fit, not the industry — a driving school and a hair salon are effectively the same purchase, and a roofing company has more in common with a plumber than with the law firm two doors down. Second, no row is fully automated. Every one of them carries a column of calls that stay with a person, and any provider showing you a version of this table with that column empty is selling you something.

What changes between industries, and what doesn't?

Less than the marketing in this category implies.

What does not change: the price, at $0.60 CAD per answered minute billed by the second, with $0 setup, no contract and no monthly minimum. The one-week launch. The 90+ languages, including a caller who switches language mid-sentence. Answering under a second on the first ring, with no limit on how many calls arrive at once.

What does change is all configuration. The vocabulary — procedure codes, unit numbers, plate and VIN references, matter types. The triage rule, which is a sentence you write and we follow rather than something the agent decides for itself. The booking object, which is the part that quietly does the most work: a salon books a stylist and a chair, a driving school books an instructor and a car, a clinic books a room and a practitioner, and a property manager books a viewing against somebody else's availability. And the escalation ladder, which is simply a list of names and the order they are tried in.

That distinction is worth carrying into any shopping you do. A provider charging an industry premium is charging you for configuration work, so ask what specifically costs more and what it buys.

When is an AI receptionist the wrong tool for your industry?

Four cases, plainly, and none of them are about which industry you are in.

Your volume is genuinely low. If you take fifteen calls a month and you answer nearly all of them yourself, the phone in your pocket is already doing the job. Nothing here beats that, and the honest recommendation is to spend the money elsewhere.

Nearly every call needs judgment in the first sentence. A sole practitioner whose callers all want the practitioner, specifically, and will not be screened or booked by anyone else, is buying a very expensive message pad. The calls still get answered and logged, which has some value, but the main benefit is not there.

Your intake is regulated in a way you cannot write down. If you cannot state your own rule in a sentence someone else could follow, nobody can configure it — not software, and not a new hire either. That is a process problem to fix before it is a phone problem.

The phone call is the relationship. Some businesses are bought because a particular person answers. If your callers would notice and mind, a staffed service is the better fit, and a human voice on every single call is exactly the case made on our Smith.ai comparison. We are voice-only and we would rather say so here than have you find out in month two.

What if your industry isn't one of the fifteen?

It is still a configuration question rather than a capability one.

Accountants, electricians, movers, gyms, tow operators, and one-person consultancies all sit on one of the five patterns above — usually booking-led or dispatch-led — and the build follows the same week and the same rate. The any-industry page covers what a custom deployment asks of you. The practical test is the one in the next section, and it works regardless of whether a page exists with your trade in the heading.

How do you test the fit in a week?

Five steps, in order, and the first four cost nothing.

1. Count one week of calls on a single sheet of paper. Time, who called, what they wanted, and whether it was answered. A week is enough and a month is procrastination.
2. Sort them into the four buckets. Repeat information, booking, intake and screening, judgment. Add up the first three as a share of the total — that number is your fit, and it is more reliable than any case study.
3. Write your escalation rule in one sentence. "Anything involving water, gas or heat loss wakes the on-call number; everything else is a message." If you cannot write it, that is the finding.
4. Call the demo line with your two hardest calls. Maryann answers at +1 437-494-9110 at any hour. Do not test it with an easy booking; use the awkward one and listen for what it does at the boundary.
5. Point one slice of your line at it for a month. After-hours only, or overflow only. Keep everything else exactly as it is, then read the invoice against your own call log. If the calls that mattered got handled, widen it. If they did not, you have spent a month of overflow minutes and learned something specific about your business.

National Auto Repair in Brampton, Ontario runs on this system, and so does Aion Appraisals, our founder's own appraisal firm, which we put on the same service we sell.

What does it cost to find out?

The same in every industry: $0.60 CAD per answered minute, billed by the second, with $0 setup and nothing to cancel.

Three hundred minutes a month is about $180. A twenty-second wrong number is about twenty cents rather than a whole call. An after-hours-only test on a quiet line can land under $30 for the month, which is the real reason to start narrow — the experiment is cheap enough that the answer costs less than the meeting about it.

For comparison, live answering services in this category typically advertise somewhere between $1.50 and $3.00 a minute. Treat that as the category's advertised range rather than any one provider's rate, and ask anyone you shortlist for a current rate card with the setup, minimum and rounding lines on it. Our price index sets out twenty providers' published figures with the sources and the dates, and our own rate sits on one screen on the pricing page.