Answering Service Contracts: The Billing Terms That Decide What You Actually Pay

Answering Service Contracts: The Billing Terms That Decide What You Actually Pay

Nobody chooses an answering service by reading the agreement. You compare a per-minute rate against a per-minute rate, pick the one that seems reasonable, and sign. Then the first invoice arrives and it is forty percent higher than the plan you selected, and the explanation is entirely legitimate — every dollar of it is in the terms you agreed to.

As of August 2026, the price of an answering service is set less by the headline rate than by five mechanics buried in the agreement: the billing increment and how it rounds, what the provider counts as billable, the monthly minimum, the overage rate outside your block, and what happens at the end. None of these are hidden in any dishonest sense. They are simply in the part nobody reads, and they compound quietly across hundreds of short calls.

This is a plain-English reading of those terms. It applies whether you are evaluating a traditional human answering service, a virtual receptionist, or an AI system — the billing structures differ, but the questions do not. No competitor's numbers are quoted here, because rate cards move and per-account terms vary; the goal is to teach you what to ask and how to check the answer against your own call log.

The single most important number: the billing increment

The billing increment is the smallest unit of time the provider will bill. Common increments are per-second, six-second, thirty-second, and full-minute. This one setting can change your bill by a third without any change in the rate.

Why it matters so much for service trades. Your call mix is not what the rate card assumes. A locksmith, plumber or towing operator gets a large volume of short calls: price shoppers, wrong numbers, "are you open," "do you work on Hondas," vendors, and a meaningful share of outright spam. Those calls are frequently under a minute. Under full-minute rounding, every one of them bills as a full minute.

The arithmetic, worked plainly. Take a hypothetical shop with 300 answered calls in a month, where 180 of them run about 40 seconds and 120 run about 2 minutes 20 seconds.

• Per-second billing: 180 x 40s = 7,200s, plus 120 x 140s = 16,800s. Total 24,000 seconds, which is 400 minutes.
• Full-minute rounding, per call: 180 x 1 minute = 180, plus 120 x 3 minutes = 360. Total 540 minutes.

Same conversations. Same rate. A 35 percent difference in billed minutes, entirely from the rounding rule. At a hypothetical $1.50 a minute that is $600 versus $810 on identical work. This example is illustrative, not a measurement of any specific provider — the point is the mechanism, and the mechanism is real.

Where it gets worse. Rounding is usually applied per call, not per month. That is the compounding effect: the more calls you get, the more rounding events you pay for. A business with lots of short calls is penalized precisely for having the call profile it actually has. If you are a trade where half the phone volume is qualification and screening, ask about the increment before you ask about the rate.

Questions to ask:

1. What is the billing increment, in seconds?
2. Is rounding applied per call or aggregated across the month?
3. Is there a minimum billable duration per call, and does it apply to calls that were answered but not connected to anything?

What actually counts as a billable minute

The second big variable is where the clock starts and stops. Providers differ, legitimately, and the differences are material.

Time before a human. If calls route through an automated greeting, a menu, or a queue before a person picks up, is that time billed? Some providers bill from the moment the call hits their platform, others from when the agent connects. On a service with any hold time at peak, this is not a rounding error.

Hold time. If the agent puts the caller on hold to look something up or to reach you, that is usually billable time. It is also time you have no control over.

Transfer time. When a call is warm-transferred to you, the period during which the agent is bridging, announcing, or waiting for you to pick up is frequently billable — sometimes to both legs. Ask specifically.

Wrap-up and after-call work. Some agreements include the agent's post-call time typing the message. This is defensible from the provider's side, since it is real labor, but it can add a meaningful chunk to every single call and it is invisible on a call recording. If a contract mentions "after-call work" or "wrap time," ask for the typical duration.

Spam, robocalls and wrong numbers. This is the one operators are most surprised by. If a robocall reaches the service and an agent answers it, that is usually billable. In a trade with a heavy spam load, you can pay a nontrivial share of your monthly bill for calls you would never have wanted. Ask whether there is spam screening, whether screened calls are billed, and whether there is any credit process for obvious junk.

Outbound calls made on your behalf. Callbacks, confirmations, and dispatch notifications may bill at the same rate, a different rate, or not at all.

Messages and notifications. Some plans bill per SMS or per email notification. Small individually, not small at volume.

The question that covers most of this at once: ask the provider to describe, second by second, what a billed 90-second call actually consists of. A straight answer takes thirty seconds to give. Hesitation is itself information.

Minimum commitments, blocks and rollover

Most answering services sell a block of minutes for a monthly fee, with an overage rate beyond it. Three things determine whether that structure works for you.

The minimum monthly commitment. You pay it whether or not you use it. That is not unreasonable — the provider staffs for your account — but it means a seasonal business pays full freight in its slow months. If your volume swings materially between summer and winter, price the annual total, not the monthly one.

Rollover of unused minutes. Usually there is none. Minutes are use-it-or-lose-it within the month, which means buying a large block to secure a lower per-minute rate is a bet on your own volume. Buy the block that matches your realistic floor, not your best month.

The overage rate. The rate outside your included block is frequently higher than the effective rate inside it, sometimes substantially. This is where a busy month becomes an expensive month. Two things to establish: what is the overage rate, and is there any tiering — does going 20 percent over cost the same per minute as going 200 percent over?

The trap this creates. You are being asked to forecast your own call volume before you have any data on it, and both errors cost money. Under-buy and you pay overage; over-buy and you pay for minutes you burn at month end. This is structurally different from a flat-plus-metered model where the included allowance is generous relative to typical use and the metered rate is disclosed up front. We laid out that comparison in detail in cost per minute versus flat rate answering.

Per-seat versus flat pricing

A separate axis that catches growing shops.

Per-seat pricing charges by the number of users on the account — dispatchers, office staff, techs with app logins. It is common in software and it appears in some answering and receptionist packages when the service includes a portal or CRM.

Why it matters. Per-seat pricing turns hiring into a line-item decision. Adding a dispatcher raises your software bill. Giving three techs read access raises it again. The amounts are small at first and they scale with exactly the thing you want to be doing, which is growing.

What to ask: Is pricing per seat, per user, per location, or flat? Are there read-only seats, and are they free? What happens if I add a fourth truck — does anything in this agreement change?

For reference, our own plans have no per-seat fees at any tier; add as many users as you want. That is a policy choice, and you should ask any provider to state theirs in writing rather than assuming.

Setup, onboarding and change fees

These are one-time or occasional charges that do not show on the comparison page.

Setup and onboarding. A traditional answering service needs to build your call script, load your service area, learn your job types and configure routing. Some bundle this, some charge for it, and the charge can be a meaningful fraction of a first-year total.

Script change fees. This is the one that quietly shapes behavior. If changing your script costs money or takes a week of turnaround, you will stop changing it — and a script that never changes drifts out of alignment with your business. Prices change, service areas change, you drop a job type, you add a trade. If revisions are billable, ask for the rate and the turnaround, and factor in about how often you realistically change things.

Training or re-training charges when you add a service line or a second location.

Number setup or porting fees at the start, and — more importantly — at the end.

What to ask: what is the total first-year cost, including onboarding, and how many script revisions are included per year?

Auto-renew and the termination notice window

This is the term that determines whether "we can leave if it does not work" is true.

Auto-renewal is standard. The agreement renews for another term unless you give notice. Fine in principle. The mechanics are what matter:

• The notice window. Thirty, sixty or ninety days before term end is typical. Miss it by a day and you are in for another full term.
• How notice must be delivered. Some contracts require written notice by a specific method. A phone call to your account manager may not legally count.
• Early termination. Is there a fee, or a liquidated-damages clause equal to the remaining term? On a multi-year agreement that can be a large number.
• Month-to-month availability. Is it offered, and what is the price difference? Sometimes the month-to-month premium is smaller than it looks and worth paying for the first six months while you find out whether the service works.

Set a calendar reminder the day you sign, dated to the start of your notice window. This costs nothing and is the single highest-return thing in this article.

Number ownership and porting on exit

The most consequential clause in the whole agreement, and the one least likely to be discussed on a sales call.

The question is simple: who owns the phone number? If the service assigns you a tracking or forwarding number and you advertise it — on the van, on the website, on business cards, in your Google listing — then that number is your business identity. If the provider owns it and you leave, you may lose it.

Losing a published number is not a billing inconvenience. It is every sign, vehicle wrap, directory listing and printed card going stale at once, plus years of accumulated inbound from old customers dialing a dead line.

How to avoid the problem entirely: keep your own number and forward it to the service. Then leaving is a forwarding change, not a migration. If you do end up needing to move a number, porting is a defined process with real rules — the Federal Communications Commission publishes consumer guidance on number portability, and it is worth knowing your rights before you negotiate. We walk through the practical steps in keeping your existing number when you switch.

Questions to ask, in writing:

1. Do I own the number, or do you?
2. If I cancel, will you release it for porting, and is there a fee?
3. How long does release take, and what happens to calls during the transition?
4. If I bring my own number and forward it, does anything about the pricing change?

Recordings, transcripts and data export

Two reasons this section matters: disputes and leverage.

Disputes. When a customer says they were told a price you would never quote, the recording or transcript settles it. When you are checking a bill, call-level detail is the only way to verify billed duration against actual duration. A provider that gives you a monthly total and no call detail has made your bill unauditable.

Leverage. Your call history is a record of your own demand. If you cannot export it, you cannot evaluate a competing provider properly, and you cannot analyze your own patterns.

What to ask: Are calls recorded, transcribed, or both? How long are recordings retained, and does retention change if I cancel? Can I export the full call log with per-call duration, timestamp, caller number and disposition? In what format? Is there a fee?

Also worth confirming: call recording consent rules vary by state, and some are two-party consent jurisdictions. The provider should be able to tell you how they handle disclosure. The Federal Trade Commission is a reasonable starting point for general consumer-protection context, but your state law is the operative rule.

How to reconstruct your true cost per answered call

Comparing providers on the rate card is not comparing anything. Here is a method that produces a number you can actually use.

Step 1 — get your real call profile. Pull 30 days from your phone carrier or current provider. You want, per call: duration in seconds, time of day, and whether it turned into a job. Most carriers can export this. If you have nothing, log a week by hand — it is tedious and it is worth it.

Step 2 — bucket by duration. How many calls under 30 seconds, 30 to 60, 1 to 2 minutes, 2 to 5, over 5. The shape of this distribution is what determines whether increments hurt you.

Step 3 — apply each provider's actual billing rules to your real profile. Round each call according to their increment. Add whatever they told you about wrap-up, hold and transfer time. Include spam calls if they bill for them. Now you have billed minutes, not talk minutes.

Step 4 — add the fixed costs. Monthly minimum, onboarding amortized over twelve months, expected script changes, per-message fees.

Step 5 — divide by answered calls. That is your true cost per answered call. Then divide by jobs booked for the number that actually matters, which is your cost per acquired job.

Step 6 — run the second scenario. Recompute at 150 percent of your current volume, using the overage rate. That is what a good month costs, and it is often the number that reorders the shortlist.

An honest comparison of the two pricing shapes

Both models are legitimate. They fail in different places, and which one suits you depends almost entirely on your call profile.

To be fair to the human side: a live answering service gives you a person who can improvise, handle an upset caller with genuine empathy, and deal with the situation nobody scripted. That is real and it has real value. The trade-offs between the two approaches for this trade specifically are laid out in AI receptionist versus human answering service and on our answering service comparison page. If you are evaluating specific national providers, we keep straightforward write-ups on AnswerConnect and Ruby Receptionists — feature and structure comparisons, not invented numbers about their performance.

Our own numbers, stated plainly

We publish pricing rather than quoting it, so you can apply the method above to us the same way you would to anyone else.

KeyBot Lite is $149 a month. That includes 100 calls. Beyond that it is 50 cents a minute. Lite takes structured messages only — it does not quote, book or dispatch — and messages are delivered by Telegram. There is a 7-day trial with the first 5 answered calls free, and setup takes about ten minutes.

The full platform adds quoting from a price sheet you upload and confirm on screen, plus booking on the call: Core $500/mo for 500 AI minutes with 45 cents a minute after, Pro $750/mo for 1,000 minutes at 40 cents after, Elite $1,200/mo for 2,500 minutes at 35 cents after. Each includes a 14-day free trial. No per-seat fees on any tier. Everything is on pricing.

Two things we will not claim: these are not unlimited plans, and there is a per-minute rate once you pass the included allowance. That is a metered model, disclosed up front, and you should apply the same six-step arithmetic to it that you apply to anybody else. If you want to hear the thing before running any numbers at all, the live demo line is +1 (716) 350-6391, or have it call you as your own business at try it free.

The checklist to run before you sign

Print this and get written answers. A provider that answers all fifteen quickly is telling you something good about how they operate.

1. What is the billing increment, in seconds?
2. Is rounding per call or aggregated monthly?
3. Is there a minimum billable duration per call?
4. Does billing start when the call reaches your platform or when a person or bot engages?
5. Is hold time billable?
6. Is transfer time billable, and on which leg?
7. Is agent wrap-up or after-call work billed, and what is the typical duration?
8. Are spam calls, robocalls and wrong numbers billable? Is there a credit process?
9. Are outbound calls and SMS or email notifications billed separately?
10. What is the monthly minimum, and do unused minutes roll over?
11. What is the overage rate, and is it tiered?
12. Is pricing per seat or flat? What does adding a user cost?
13. What are the setup, onboarding and script-change fees, and how many revisions are included?
14. What is the term, the auto-renew behavior, the notice window and the notice method? Is there an early termination fee?
15. Who owns the number, will you release it for porting, and can I export my full call log with per-call durations?

The bottom line

The rate on the quote is the least informative number in an answering service agreement. What decides your bill is the billing increment and how it rounds — a per-call full-minute round on a short-call trade can move billed minutes by a third with no change in the rate at all. Then it is what counts as billable: hold, transfer, wrap-up, IVR time before a human, and the robocalls somebody answered on your behalf. Then it is the monthly minimum you pay in slow months, the overage rate that makes a busy month expensive, and any per-seat charge that taxes you for growing. And at the end it is the notice window and the number: miss the window and you owe another term, lose the number and every van, sign and listing you own goes stale at once. Do not compare rate cards. Pull thirty days of your own calls, apply each provider's real rules to your real durations, add the fixed costs, and divide by jobs booked. Whichever model wins that arithmetic is the right one — and any provider unwilling to give you the inputs has already answered the important question.

Frequently asked questions

What is a billing increment on an answering service contract?

A billing increment is the smallest unit of time the provider bills, and common increments are per-second, six-second, thirty-second and full-minute. It matters more than the headline rate because rounding is usually applied per call rather than across the month, so every short call is rounded up individually. For a trade with a lot of screening calls and price shoppers, moving from per-second to full-minute rounding can raise billed minutes by roughly a third on exactly the same conversations.

Do answering services bill for hold time, transfers and wrap-up?

Frequently yes, and this is where quoted rates and real invoices diverge. Depending on the provider, billing may start when the call hits the platform rather than when a person engages, hold time is normally billable, transfer time can be billed on one or both legs, and some agreements include the agent's after-call typing time. Ask the provider to describe second by second what a billed 90-second call actually contains — a straightforward answer takes about thirty seconds to give.

Am I billed for spam calls and wrong numbers?

Usually yes, if the service answered them. That surprises most operators, because in trades like locksmithing, towing and plumbing a meaningful share of inbound volume is robocalls, telemarketers and misdials, and paying an answering rate for those is paying for nothing. Ask specifically whether spam screening exists, whether screened calls are billed, and whether there is any credit process for obvious junk.

What should I check about auto-renewal and cancellation?

Check four things: the length of the notice window before term end, the method notice must be delivered by, whether an early termination fee or remaining-term clause applies, and whether a month-to-month option exists. Missing a sixty or ninety day notice window by a single day typically commits you to another full term, so set a calendar reminder for the start of that window on the day you sign. That reminder costs nothing and it is the highest-return step in the whole process.

Who owns my phone number if I cancel?

It depends entirely on the agreement, and this is the clause worth resolving before anything else. If the provider assigned the number and you advertised it on vans, signs, cards and your Google listing, losing it on exit means all of that goes stale at once plus years of old customers dialing a dead line. The clean way to avoid the risk is to keep your own number and simply forward it to the service, so leaving is a forwarding change rather than a migration.

How much does TheKeyBot cost compared to a per-minute answering service?

KeyBot Lite is $149 per month with 100 calls included and 50 cents per minute after, with a 7-day trial where the first 5 answered calls are free; Lite takes structured messages only and does not quote, book or dispatch. The full platform is Core at $500 per month for 500 AI minutes with 45 cents per minute after, Pro at $750 for 1,000 minutes at 40 cents after, and Elite at $1,200 for 2,500 minutes at 35 cents after, each with a 14-day free trial and no per-seat fees. These are metered plans with a published overage rate rather than unlimited plans, and all tiers are listed at https://www.thekeybot.com/pricing.