The headline rate on an answering service quote is rarely what you end up paying. The pricing model matters far more, because it decides what happens in your busiest month.
The three models
Per call. A fixed amount each time they pick up. Simple to understand. Check whether spam, wrong numbers and hang-ups are billable.
Per minute. You buy a bundle of minutes. Cheaper if your calls are short and transactional, more expensive if callers ramble or the script is long.
Flat monthly with an allowance. A bundle of calls or minutes, then overage. The overage rate is the number to interrogate, because that is where the surprise lives.
The questions that actually change the bill
- Is there a minimum billed duration? A 20-second call billed as a full minute changes the maths considerably.
- Are spam and wrong numbers charged? Get it in writing.
- What is the overage rate, and is it tiered? A sharp cliff means your best month costs the most.
- Is out of hours priced differently? Nights and weekends are often the whole reason you are buying.
- Is there a setup or scripting fee?
- What is the contract length? Twelve months is common. Monthly is worth asking for.
Why the model matters more than the rate
Consider a roofing business. For eleven months the phone is manageable. Then a storm comes through and three days generate a month of calls.
On per-call or per-minute pricing, that is exactly when the invoice spikes. You pay most for the week that was already chaotic. On a flat plan with a small bundle, you blow through the allowance into overage. Either way, demand and cost move together.
That is not a scam, it is a fair reflection of the work. But it is worth knowing before you sign, because it is the opposite of how most small businesses would like their costs to behave.
The comparison people forget
An answering service prices your phone line. It does nothing for the visitor reading your pricing page at 10pm who has a question and no intention of calling.
Website AI prices differently, typically a flat monthly plan with a set allowance of replies rather than a charge per conversation, because the marginal cost of another conversation is small. SpideyChat's plans work this way, and the website receptionist page sets out what the chat does and does not cover. For the out-of-hours side of the same problem, see 24/7 support without hiring a night shift.
SpideyChat is this kind of tool: a chat assistant on your website that answers written questions from your own pages and captures the visitor's details. It does not answer phone calls, so it sits alongside phone cover rather than replacing it.
Ask the same cliff question of any AI tool, though. A plan with a fixed monthly reply allowance has a ceiling, so check what happens when you reach it and size the plan to your busiest month rather than your average one. Do that and the bill stays predictable even in the week the storm comes through. If a meaningful share of your enquiries start online, that side is usually far cheaper to cover.
The honest framing is not "which is cheaper" but "which channel am I losing work on". Answer that first.
Working it out for your business
Pull your last three months of call logs. Count the calls that came outside hours, and how many left a voicemail. Then look at website analytics for the same period.
If the website number is larger, and it usually is, start there. You can build a website chat assistant from your own site in about a minute and see what it would have said, before spending anything.
Related: AI receptionist vs answering service.