Every service business eventually hits the same wall: the phone rings more than the people available to answer it. You're on a roof, under a sink, or driving between jobs, and the calls keep coming. Each one is maybe a customer, and voicemail is where maybe-customers go to call your competitor.
At that point you have three realistic options: hire someone, contract a human answering service, or use an AI receptionist. All three can work. They have genuinely different cost structures, and the right answer depends on your call volume and what you need the phone-answerer to actually do.
We sell one of these three options, so you know our bias going in. But the math below is just math: pricing models you can verify with any vendor, and a framework for running your own numbers. No invented industry statistics, no scary made-up percentages. Bring your own call volume and the answer falls out.
First, get one number: your monthly call volume
Everything below keys off how many calls you actually get. Most owners guess wrong, usually low, because you only remember the calls you answered.
Pull your phone bill or carrier app and count inbound calls for the last full month. Include the ones that went to voicemail. If you've claimed your Google Business Profile, its performance reports also show calls placed from your listing, which is a useful cross-check on how many of those calls come straight from search. That number, calls per month, is the input for every calculation in this post. For the worked examples we'll use 200 calls a month, roughly 7 a day, which is in the normal range for a small trades business, but swap in your own number.
Option 1: Hiring a receptionist or office person
The pricing model: payroll. A wage, plus the costs that ride along with a wage.
The math most owners do is "wage times hours." The real math is bigger. For a full-time hire you're paying:
- The wage itself, 40 hours a week, 52 weeks a year, whether the phone rings or not
- Payroll taxes on top of the wage
- Some mix of benefits, paid time off, and workers comp depending on your setup
- Recruiting, training, and the productivity dip every time the seat turns over
The SBA's guidance on the cost of hiring is a good grounding here: the rule of thumb that an employee costs meaningfully more than their wage exists because the add-ons are real. Whatever hourly wage is normal in your market, the loaded cost is the wage plus payroll taxes plus benefits plus overhead, and you can compute it precisely from your own books.
The coverage problem: one full-time person covers about 40 of the week's 168 hours, minus lunch, breaks, vacation, and sick days. Evenings, weekends, and the moments they're already on the other line are uncovered. True 24/7 human coverage isn't one hire; it's a rotation of several, which is why almost no small business does it.
Where hiring wins anyway: a real employee does far more than answer phones. They schedule, invoice, order parts, chase paperwork, calm down upset customers, and hold context an outside service never will. If you have 30+ hours a week of office work beyond the phone, you're not really comparing receptionist options anymore. You need the employee, and the phone is one of their duties.
The honest summary: hiring is the most expensive way to answer phones and the cheapest way to get everything else an office person does. Judge it on the whole job, not the phone alone.
Option 2: Human answering service
The pricing model: this is where you have to read carefully, because the industry bills several different ways:
- Per minute. You pay for operator talk time, often with rounding rules (billed in 30- or 60-second increments, which inflates short calls).
- Per call. A flat rate each time an operator picks up, regardless of length.
- Bucket plans. A monthly fee includes some number of minutes or calls; overage rates kick in above the bucket, and the overage rate is usually worse than the base rate.
The structural thing to understand: you're renting slices of a human's time, so the cost scales linearly with your volume. Double your calls, roughly double your bill. There's no point at which it gets cheaper per call, and growth in your business shows up as growth in this line item forever.
The fee fine print to check on any quote: setup fees, monthly minimums, holiday surcharges, after-hours premiums, charges for patching a call through to your cell, charges for sending you the message by text or email, and the rounding increment on per-minute plans. Ask for the effective cost of a typical 3-minute call at your volume, all fees included. That one number makes plans comparable.
The capability ceiling: a traditional answering service operator is working from a script across dozens of client businesses. They take a message, maybe ask three intake questions, maybe patch urgent calls through. Most can't quote your prices, can't see your schedule, and can't book a job. For many businesses the output is a politely-taken message you still have to return, which means you paid to move the callback from voicemail to a notepad.
Where answering services win: low call volume with high stakes per call, where a human voice matters and the conversations are unpredictable. Also businesses with compliance or sensitivity requirements where a trained human is non-negotiable.
Option 3: AI receptionist
The pricing model: typically a flat monthly fee that includes a bundle of calls, with a small per-call price above the bundle. Using our own pricing as the concrete example since we publish it: our Max tier includes a 24/7 AI receptionist with 200 calls a month included, and extra calls are $1.50 each. Other vendors structure similarly; the key is that the marginal cost per call is low and known.
What changed to make this real: until a couple of years ago, "AI answering" meant a phone tree that made people mash zero. Current AI receptionists hold an actual conversation: they answer questions about your services and hours, capture the caller's name, number, address, and problem, and depending on the setup, book the appointment directly onto your calendar. They answer on the first ring at 2 p.m. and 2 a.m., never put anyone on hold, and handle ten simultaneous calls the same as one.
The honest limitations: an AI receptionist is not an employee and shouldn't be sold as one. The realistic failure modes:
- Complex, emotional, or unusual calls need a human, so you want a setup that texts you transcripts immediately and escalates or routes the calls that need you.
- A badly configured AI is worse than voicemail, because it wastes the caller's time before failing. Setup quality matters more than the model behind it.
- Some callers simply want a human, and a small share will hang up on any automated answer, just as some hang up on hold music or voicemail. The relevant comparison isn't AI versus a perfect human; it's AI versus whatever currently happens when you can't pick up.
Where AI wins: after-hours and overflow coverage, high call volume relative to budget, and businesses whose calls are mostly the same five questions plus booking, which describes a large share of HVAC, plumbing, and landscaping call traffic.
The framework: run your own three-way comparison
Here's the whole exercise on one page. Use your real numbers.
- Step 1, volume: count last month's inbound calls, including missed and after-hours. Note what share came in outside business hours; that share is structurally unservable by a single 9-to-5 human.
- Step 2, hiring math: loaded monthly cost equals wage times hours times roughly 1.25 to 1.4 for taxes and overhead (use your accountant's real number). Divide by monthly calls for cost per call, but credit the hire for all non-phone work they'd do.
- Step 3, answering service math: get the all-in quote at your volume, with the typical-call effective rate including every fee, then model it again at 1.5 times your volume to see how it scales as you grow.
- Step 4, AI math: flat fee, plus per-call overage times any calls beyond the included bundle. At 200 calls on our Max tier, that's the flat fee, full stop; at 260 calls it's the flat fee plus 60 times $1.50.
- Step 5, coverage: for each option, write down which of the week's 168 hours are actually covered. This is where the comparison usually stops being close, because options one and two cover business hours well and everything else poorly, unless you pay substantially more.
- Step 6, output quality: for each option, what exists after the call? A voicemail, a message slip, or a booked job on your calendar? A booked job is worth more than a message you have to chase, and you can price that difference using your own close rate.
Notice there's no universal winner in that framework. A two-truck operation getting 60 calls a month has different math than a 200-call shop, which has different math than a business that needs a full-time office manager anyway.
The mistake to avoid: comparing against zero
The most common error isn't picking the wrong option. It's picking none of them, because each option's monthly cost feels like a new expense while the current "free" arrangement, voicemail, feels like zero.
Voicemail isn't zero. Every missed call has an expected value: the chance it was a real customer, times your close rate, times your average job. We walk that exact calculation in our companion post on what missed calls cost, and once you compute your own number, all three options above tend to look cheap against it.
Where we land, openly
Omnyra is a veteran-owned shop in Wilmington, NC. We build done-with-you websites live on a call with you: first draft in 24 hours, live in 7 days, guaranteed, with 1,500+ small business sites built in the last 90 days. Tiers run from $500 Minimal, to $2,000 plus $200 a month Standard with ongoing SEO and AI-search optimization, to $3,500 plus $400 a month Max, which is the tier with the 24/7 AI receptionist discussed above (200 calls a month included, extra calls $1.50). Custom back-office builds start at $6,000 with Super Max, and pay-in-4 or Klarna financing is available.
Full numbers are on the pricing page. If you want to run your call volume through the framework with a person instead of a blog post, book a call and bring last month's phone bill. The math takes about ten minutes.