Ask what an answering service costs and you will get a number. The number is close to useless on its own, because two services quoting the same headline rate can produce invoices that differ by a factor of three on the same call volume — and the difference is never in the headline, it is in the structure underneath it.
So this page is about structure. No prices, deliberately: rates vary by country, by industry, by whether calls are handled onshore, and by what is bundled, and a figure quoted here would be stale within a quarter and treated as a benchmark it was never fit to be. What does not go stale is how the models behave when your week goes wrong.
For transparency: Reign AI sells one of the options described below, and the comparison is written so that a reader who costs it out and picks a human service has still got what they came for.
The four pricing models
Per minute. You are billed for the time spent on your calls, usually with a monthly minimum. Simple, and the one most exposed to a bad week — a cold snap, a storm, a listing that suddenly gets attention — because the invoice scales with exactly the event you bought the service for. Also the model where the billing increment does the most damage, for the reason in the FAQ.
Per call. A fixed price per handled call regardless of length. More predictable than per minute, and it inverts the incentive: long calls become good value and very short ones become expensive. Businesses with lots of brief logistical calls — confirmations, directions, “are you open” — tend to do badly here.
Included minutes with an overage rate. A monthly bundle plus a rate for anything beyond it. This is the model most likely to be mis-sold, because the bundle is sized against your average month and the overage rate is where the money is. Ask what the overage rate is before you ask what the bundle costs, and ask whether unused minutes roll over. They usually do not.
Flat fee. One price, volume-independent. Predictable, and it is the only structure where a genuinely terrible week costs you nothing extra — which for seasonal and emergency trades is the entire point rather than a nicety. The trade-off is real: below a certain volume you are paying for capacity you never use, and at that point a usage model is honestly cheaper.
Which of these suits you is decided by the shape of your call volume, not its size. A business with steady low volume and a business with the same monthly total arriving in two brutal weeks should not buy the same structure. The seasonal case is worked through for a specific trade in answering for HVAC contractors — a spike-driven pattern that punishes usage pricing exactly when the service is most needed.
How is an AI receptionist billed?
Usually as a monthly plan rather than per call — either a flat fee or a bundle of included minutes with an overage rate — and often with a one-time setup fee for configuring it to your business. That puts it in the last two models above, so everything said about them applies, plus three questions specific to software.
What happens past the included minutes? A bundle sized for your average month is the same trap it is with a human service, so get the overage rate and price your worst month against it.
Do simultaneous calls cost more? Software can answer several calls at once; check whether the plan charges for that or caps it, because the night five calls arrive together is the night you are buying it for.
What is in the base plan? Booking into your calendar, putting a caller through to a person and texting a caller a booking link are often set up per business rather than switched on by default, so confirm which your quote includes and whether changing the setup later costs anything. The setup fee is worth reading too: it pays for turning your services, hours and emergency rules into instructions, and that work decides whether the agent answers correctly.
The lines that are not on the quote
Every one of these has changed a real comparison between two providers. Ask each in writing:
- The billing increment. Thirty seconds, sixty, or actual time. See the FAQ — on a short-call business this is not a detail.
- Whether hold time bills. If the receptionist puts a caller on hold to try your mobile, who pays for those ninety seconds?
- Whether wrong numbers and spam bill. Someone has to answer to find out.
- Setup and onboarding fees, and whether scripting changes later cost anything. A service you cannot afford to adjust is a service that slowly stops matching your business.
- The minimum term. A twelve-month commitment on a service you have not tested is the expensive version of a cheap rate.
- After-hours, weekend and holiday premiums. Overnight coverage is frequently the reason people buy, and frequently the rate that is not on the front page.
- Whether it can book into your calendar, or only take a message. This is a capability question that behaves like a pricing question, for the reason in the next section.
- What happens to call recordings and transcripts, how long they are kept, and who can read them. Relevant for every business and non-optional for a professional practice.
The cost that does not appear on any invoice
The largest cost in this decision is usually not on the bill at all: it is what the service produces at the end of a call.
A message is not a booking. If a caller wanted an appointment and got a promise that somebody will ring back, you have paid for the call and still have to make one — and on emergency or same-day work the callback frequently arrives after the customer has booked elsewhere. That converts a service you are paying for into an expensive way of finding out who you lost.
So a cheaper service that only takes messages can easily be the more expensive choice, and the gap does not show up anywhere you would look for it. The way to compare honestly is to price the outcome — bookings landing in your calendar — rather than the input.
The structural differences between the options, independent of price, are set out in the four-way comparison. If the option you are currently comparing against is your own mobile, the honest version of that comparison is in AI receptionist versus forwarding to your cell.
How to compare two quotes properly
- Get your own call data first. Volume, average duration, the split between business hours and out of hours, and what share are genuinely short. Without it you are buying against a guess and every model looks similar.
- Model your worst month, not your average one. Take the busiest month in the last two years and price every quote against it. The average month is not the month you are buying insurance for.
- Ask the eight questions above in writing. A provider who will not answer them in writing has answered them.
- Test the actual experience. Call each service as a customer would, with your own awkward scenario, and read what arrives on your end. A cheap quote attached to a call that embarrasses you is not cheap.
- Compare against doing nothing. Work out what the unanswered calls are actually costing first — how much a missed call is worth depends far more on your callers’ alternatives than on your industry, and for some businesses the honest answer is that the cheapest option is the one you already have.
That last step is the one most often skipped and the only one that can tell you not to buy anything at all.
Common questions
Why does this page not list prices?
Because a price list would be wrong within a quarter and misleading immediately. Rates vary by country, by whether calls are handled onshore, by industry, and by what is bundled, and a figure quoted out of context gets used as a benchmark it was never fit to be. What does not go stale is the structure: the four models below, and the terms that change what a given rate actually costs you. Get quotes with those questions attached and the comparison becomes straightforward.
What is a billing increment and why does it matter so much?
It is the block of time a provider rounds each call up to — commonly thirty seconds or a full minute. On a per-minute plan it is one of the largest hidden multipliers there is, because a service business gets a lot of very short calls. Twenty calls that genuinely last forty seconds each bill as twenty minutes on a thirty-second increment and as twenty minutes on a per-minute one, but forty calls of fifteen seconds bill as twenty minutes rather than ten. Ask for the increment in writing; it is rarely on the pricing page.
Do wrong numbers and spam calls get billed?
Frequently yes, and it is worth asking directly rather than assuming. Someone has to answer a robocall to discover it is one, and many services bill that time. For a business whose published number attracts a lot of noise this can be a material share of the invoice, and policies differ enough that it is a legitimate comparison point between two otherwise similar quotes.
Is a flat monthly fee always cheaper?
No — it is more predictable, which is a different property. Flat pricing wins when volume is high or spiky, because the cost of a bad week is zero extra. Usage pricing wins when volume is genuinely low and steady, because you are not paying for capacity you never touch. The mistake is comparing a flat fee against a usage estimate built from an average month, when the whole reason people buy coverage is the months that are not average.
What is the real cost of a cheap plan that does the wrong thing?
Higher than the difference in price, and it is the cost nobody quotes. A service that takes a message when the caller wanted an appointment has converted a booking into a callback, and on emergency or same-day work a callback is frequently a lost job. Judge a quote by what it produces at the end of a call — a booking, or a note saying somebody rang — before judging it on the rate.