On this page
An AI receptionist is software that answers your business line with a natural-sounding voice, asks the questions your office would ask, books visits on your calendar, flags emergencies to an on-call person, and sends you a written summary. It works well for after-hours, overflow and routine intake. It shouldn't diagnose problems, quote complex jobs or replace judgment, and it should be set up with call-recording and disclosure rules in mind.
Key takeaways
- An AI receptionist handles routine intake: who, what, where, how urgent, and when.
- It works best behind your own routing, answering only the calls your team can't.
- Give it firm limits on pricing, diagnosis and promises, and an easy path to a human.
- An AI receptionist transcribes calls and many keep recordings, which brings state recording-consent laws into play.
- The FCC treats AI-generated voices as artificial voices under the TCPA, so an AI that places calls usually needs the called person's consent first.
What is an AI receptionist?
An AI receptionist is a phone answering system built on a language model and a synthetic voice. When a call comes in, it greets the caller in your company's name, listens to what they need, asks follow-up questions, and takes an action: books a visit, sends a text with a link, alerts your on-call tech, or takes a detailed message. Then it writes a summary of the call to your phone or CRM.
It differs from the phone trees most people hate. There's no "press 2 for service". The caller just talks, and the system responds to what they actually said. It also differs from an answering service staffed by people: it doesn't take breaks, some products can take more than one call at a time (Jobber's Receptionist page says it handles multiple calls and texts simultaneously), and it knows only what you've taught it.
For a contractor, the appeal is simple. Calls arrive while you're on a ladder, driving, in a crawlspace or asleep. A caller who reaches voicemail is free to try the next company on the list. Even Thumbtack's refund policy page tells its pros that most customers want to hear back right away and that many hire the quickest and most helpful pro.
How does it work on a real call?
A typical setup sits inside your existing phone routing:
- A customer calls your business number.
- Your routing rings the office or your crew lead first, during the hours you choose.
- If nobody picks up within a set number of rings, or it's after hours, the call goes to the AI receptionist.
- It answers in your company's name and asks your intake questions: name, address or town, what's wrong or what they want, how urgent, best callback number.
- Based on your rules, it books a slot on your calendar, transfers the call, alerts the on-call phone, or takes a message.
- The caller gets a confirmation text from your business number. You get a summary with the details.
The quality depends heavily on setup. A receptionist trained on a one-paragraph description of your company will sound generic and miss things. One trained on your actual services, service area, hours, policies and common questions will sound like it works for you.
What can it do well?
- Answer every call, immediately. Nights, weekends, lunch, the busy week after a storm.
- Qualify the job. Ask the questions you'd ask, in order, and politely decline work you don't do or areas you don't serve.
- Book on a real calendar. Offer open slots and confirm one, if it's connected to your scheduling system.
- Spot urgency by your definition. You decide what counts: no heat, active leak, sewage backup, a tree on the house. Those calls ring your on-call person right away.
- Answer routine questions. Hours, service area, whether you handle a certain brand, what to expect at the estimate.
- Write things down. Every call becomes a short written record, so nothing lives only in someone's memory.
What can't it do, or shouldn't?
- Diagnose. It can ask what the caller sees and hears. It shouldn't tell a homeowner what's wrong with their furnace or whether a wall is load-bearing.
- Quote complex work. If you price from site visits, it should book the visit, not guess a number. If you have fixed prices for simple services, you can let it share those.
- Make promises you haven't approved. Arrival times, warranties, discounts and refunds should come only from rules you've written.
- Handle angry or sensitive calls as well as a person. A billing dispute, a damage complaint or an upset longtime customer should go to a human fast.
- Be right every time. Speech recognition makes mistakes with addresses, names and accents. Confirmation texts and summaries let you catch errors before a crew drives to the wrong house.
Before launch, write down what the receptionist may say about prices, scheduling, warranties and emergencies, and what it must never say. Let any caller ask for a human and get one. Review a sample of call summaries every week for the first month.
Disclosure, recording and the law
None of this is legal advice. State rules differ, so have a lawyer look at your setup.
Recording. An AI receptionist has to transcribe the call to understand it and write the summary, and many products also keep a recording. Federal law (18 U.S.C. § 2511(2)(d)) permits recording with one participant's consent, but several states require every party's consent. Many businesses play a short notice at the start of the call, such as "This call may be recorded." Whether a notice is enough in an all-party state is a question for your lawyer. Our guide on recording business calls covers the details.
Saying it's AI. We recommend the receptionist identify itself as your company's virtual or automated assistant. There's a legal reason in some places too. California's bot disclosure law makes it unlawful to use a bot to communicate with a person in California online with the intent to mislead them about its artificial identity for the purpose of knowingly deceiving them about the content of the communication to encourage a purchase or sale, and it says a person using a bot isn't liable under that section if they disclose that it is a bot. Whether that law reaches phone calls is a question for a lawyer, but being upfront avoids the question entirely, and saying so costs you nothing.
Outbound calls. If you want the AI to call people rather than only answer, different rules apply. In February 2024 the FCC issued a declaratory ruling (FCC 24-17) confirming that the Telephone Consumer Protection Act's restrictions on "artificial or prerecorded voice" calls cover AI-generated voices, which means such calls generally need the called party's prior express consent unless there's an emergency purpose or an exemption applies. Under the FCC's rules, a telemarketing call that uses an artificial voice requires prior express written consent when it goes to a cell phone or a residential line. Answering inbound calls is a different situation, but don't let an AI start dialing your old customer list without legal advice.
How to set one up
- Decide which calls it takes. After hours only, overflow when nobody answers in a set number of rings, or all calls first with a transfer to staff. Overflow and after-hours are the easiest places to start.
- Write your intake script. The questions your best office person asks, in order. Keep it short.
- List services, area and exclusions. What you do, the towns and ZIP codes you serve, and what you don't do, so it can say no politely.
- Define emergencies and escalation. Which situations ring the on-call phone, which get booked first thing, and who gets alerted.
- Connect the calendar and CRM. Without this, it can only take messages. With it, it can book real slots and create lead records.
- Set the greeting and disclosures. Company name, that it's an automated assistant, and a recording notice where needed.
- Test it yourself. Call as a confused customer, an angry one, someone outside your area, someone with an emergency. Fix what breaks.
- Review and tune. Read summaries weekly at first. Adjust the script where it misunderstands callers or asks the wrong things.
What does an AI receptionist cost?
Pricing models vary widely, so compare the structure, not just the headline number. Some providers charge a flat monthly fee with a cap on minutes or calls. Others charge per minute or per call. Some field-service platforms sell it as an add-on; for example, Jobber lists an AI Receptionist add-on at $29 a month, as listed on its pricing page in October 2026, on top of a Jobber plan.
The costs to ask about:
- Monthly platform fee, and what usage it includes.
- Per-minute or per-call charges beyond the included amount.
- Setup and script-writing, if someone does it for you.
- Phone number or porting costs, and texting registration for confirmation messages.
- Calendar and CRM integration, and whether it works with the tools you already use.
- Contract length and what happens to your call records if you leave.
Weigh those against the calls you currently miss. Our missed-call calculator shows the arithmetic with your own numbers, and the guide on what missed calls cost walks through it. If you want a receptionist built into your phone system and CRM, that's what our AI reception service does. If you'd rather start smaller, a missed-call text back covers some of the same gap.
Frequently asked questions
Will customers know they're talking to AI?
Some will guess. We recommend having it say it's your company's automated assistant. What a caller needs from it is a quick, accurate answer and a booked time.
Can an AI receptionist transfer a call to me?
Yes, if your phone system supports it. You set the rules, such as transferring emergencies or any caller who asks for a person.
Does an AI receptionist replace an answering service?
It can cover the same hours. A human service handles unusual and emotional calls better; an AI handles volume, speed and booking into your systems. Some businesses use both.
Can it give prices over the phone?
You decide. Fixed prices for simple services can work. Anything that needs a site visit should be booked, not guessed.
Sources
- Thumbtack Help: Refund policy (on responding quickly)
- 18 U.S.C. § 2511, Cornell Legal Information Institute
- California Business and Professions Code § 17941, California Legislative Information
- FCC Declaratory Ruling FCC 24-17 on AI-generated voices, February 2024
- 47 CFR § 64.1200, Cornell Legal Information Institute
- Jobber Receptionist feature page, read October 2026
- Jobber pricing page, read October 2026
General information for service-business owners, not legal, tax or financial advice. Rules vary by state; check with a qualified professional for your situation.