Per-minute pricing has a design flaw nobody mentions in the sales call: it charges you the most in exactly the months you can least think about it. Your busiest month, the storm week, the promotion that finally works, that's when the meter spins hardest. A pricing model that punishes your best months isn't a detail. It shapes how you use the service, and eventually it shapes how you grow.
How does answering service pricing actually work?
Most human answering services and some AI products price one of three ways:
- Per minute. A base fee buys a bundle of minutes; overage bills at a per-minute rate. The industry's default.
- Per call. Same idea, counted in calls. Long calls make it look cheap; a flood of short calls makes it expensive.
- Flat rate. One monthly price, use it like you own it. Rarer, because meters are profitable.
Meters aren't a scam; they map cost to usage, which is fair from the vendor's side. The problem is what they do to your side.
What's wrong with paying by the minute?
Your bill is unpredictable in the wrong direction. Quiet months you overpay for a bundle you didn't use. Busy months you blow through it. You can't budget a number that moves against you.
It punishes success. Ads finally working, a storm bringing demand, a viral review: your reward is an overage invoice. The service gets a raise every time your marketing works.
It quietly rations quality. Per-minute pricing rewards short calls, and everyone responds to incentives, including your answering service. But short calls and good calls are opposites: qualifying, answering questions, and booking take minutes. A pricing model that taxes minutes taxes thoroughness.
It bends your own behavior. Owners on metered plans start doing the math mid-month: routing fewer calls, trimming the service's scope, telling it to "just take a name and number." You bought coverage, then talked yourself into using less of it.
What does flat pricing change?
One number, every month, no matter how good the month is. ZilchWork plans are flat per business: Solo $997, Growth $1,997, Multi $4,997, Enterprise by consult (pricing). No per-minute meters, no per-call charges, no overage invoices after your best week.
The reason we can price this way is the reason to prefer it: our AI, Anna, doesn't cost more when she works more. Three simultaneous calls at 8 a.m. cost her nothing extra, so they cost you nothing extra. The incentive alignment flips: we want her taking long, thorough calls that qualify, book, and close, because that's what keeps a customer, and nothing about the bill argues otherwise.
One honest note so this stays fair: flat plans are sized to business scale. A solo operator and a three-location company don't pay the same because they don't use the same. What never happens is a surprise line item because a Tuesday got busy; if a business outgrows its tier, the conversation is an upgrade, never a shock invoice.
What does the math look like side by side?
A labeled example with stated assumptions, so you can swap in your own numbers. Say a service business takes 400 calls a month averaging 4 minutes: 1,600 minutes.
| Per-minute service (example assumption: $1.20/min blended) | ZilchWork Growth | |
|---|---|---|
| Normal month (1,600 min) | ~$1,920 | $1,997 |
| Slow month (900 min) | ~$1,080, paying for silence either way | $1,997 |
| Best month ever (3,000 min) | ~$3,600, plus the anxiety | $1,997 |
| What happens on the call | Message taken, forwarded to you | Qualified, booked, closed where appropriate, logged |
The per-minute column varies by vendor; published rates change often, so check current pricing before comparing (for reference, Smith.ai's published plans in July 2026 run $97.50 a month for AI answering and $292.50 for the human-backed hybrid, with plan limits). The structural point survives any specific rate: the meter's best case is a quiet business, and a quiet business isn't the goal.
The deeper difference is in that last row. A message-taker on a meter and a closer on a flat rate aren't two prices for the same product; they're two products. The full comparison of the options covers that side by side.
How should you compare offers?
Five questions for any vendor, metered or flat:
- What did your average customer my size pay last month, actual invoices, not the base rate?
- What happens to my bill in my busiest month of the year?
- Does anything in your pricing reward keeping my calls short?
- What does the service DO with a call beyond taking a message?
- Can I test it before committing? (Ours: $99 three-day trial credited to month one, $999 setup waived until August 31, live in 7 days or your first month is free. Call 518-629-1415 or try the browser demo first.)
Then run your call volume against both models with your real numbers, and check what the leak costs you today with the lost-lead calculator.
FAQ
How much does an answering service cost per minute?
Rates vary widely by vendor and plan, typically with a monthly bundle plus overage. Published prices change often; get the vendor's current rate card and, more usefully, an actual sample invoice for a business your size.
Why do most answering services charge per minute?
Their cost is human labor time, so they meter it. Fair on their side; the side effects (unpredictable bills, short-call incentives) land on yours.
Is flat-rate answering always better?
For low, stable call volume, a metered plan can price out cheaper. Flat pricing wins when volume is meaningful, spiky, or growing, and when you want thorough calls instead of fast ones.
Does ZilchWork ever charge overages?
No surprise charges, ever. Plans are sized to the business; outgrowing a tier is an upgrade conversation, not a bill.
What's included in the flat rate?
The whole front office: 24/7 answering, lead callbacks in seconds, qualifying, booking, closing on unregulated sales, and a CRM that fills itself in. What done-for-you means has the full breakdown.