Customer Service and AI Receptionist Glossary: Plain-English Definitions
The terms you’ll meet when choosing how to answer your business phone, from IVR and warm transfer to concurrency and per-minute billing, each explained in a few plain sentences.
What's Inside
What does this customer service and AI receptionist glossary cover?
It defines 54 terms you’ll meet when deciding how your business phone gets answered: who answers, how calls are routed and handed off, how an AI receptionist works, the software it connects to, and how it’s priced and measured. Each definition is two or three plain sentences, with a link to the page that covers the topic in depth.
Vendors don’t always use these words the same way. Where a term means different things to different people, such as virtual receptionist or conversational IVR, the definition says so. Jump to any term:
These definitions are written by Morgan Systems in plain language. The glossary compares no products and carries no affiliate links. Product names appear only as examples, and prices are kept in one place: our AI voice agent pricing comparison.
This glossary is part of a three-page series on phone customer service basics:
What are the different ways a business phone gets answered?
A call can be answered by a person at the business, a person at another company, a recording, a phone menu, or an AI. These are the names for each.
AI receptionist
Software that answers a business’s phone calls in a natural voice. It answers common questions, books appointments, takes messages, and passes calls to a person when it should. Some are sold on their own; others come built into field service or CRM software.
AI voice agent
The broader name for AI that holds a spoken conversation, on the phone or on a website. An AI receptionist is an AI voice agent set up to answer a business’s calls; the same technology can also run support lines or make outbound calls.
Answering service
A company whose staff answer your calls in your business’s name, usually to take messages and sometimes to book appointments. You forward your calls to them for some or all hours.
Virtual receptionist
A term used in two ways: sometimes a real person answering calls remotely for several businesses, sometimes an automated menu or AI. Check which one a provider means before comparing it with anything else.
Receptionist
A person at the business who answers calls and greets visitors. A receptionist can do things no software can, but covers only the hours they work and one call at a time.
Auto attendant
The simplest phone menu: a recorded greeting that sends callers to an extension or phone when they press a key. It is usually a feature of a business phone system rather than a separate product.
IVR (interactive voice response)
An automated phone system that lets callers choose options by pressing keys or saying short phrases, and that can look up simple information such as a balance. “Press 1 for billing” is IVR.
Conversational IVR
A phone menu that understands short spoken answers instead of key presses but still follows set branches. Some vendors also use the name for AI voice agents, which hold an open conversation.
Voicemail
A recorded message left when nobody answers. It is cheap and always on, but the caller has to wait for a callback, and some hang up instead of leaving a message.
Call center and contact center
A team of agents answering calls in volume, with software that queues callers and measures each call. A contact center handles email, chat and texts as well as calls.
How do calls get routed and handed to a person?
Most of the words in this group describe where a call goes next. They matter most when you set up the rules for when an AI hands a call to your team.
Call forwarding
Sending calls from your business number to another number, such as an answering service or an AI receptionist. You can usually forward all calls, only when the line is busy, after a set number of rings, or after hours, depending on your phone carrier.
Call overflow
Calls that arrive when everyone is already on the phone or busy. Overflow coverage sends them to a backup, such as an answering service or AI receptionist, instead of voicemail.
After-hours answering
Coverage for calls outside your business hours. It is a common first job for an AI receptionist, because those calls otherwise go to voicemail.
Call routing
Deciding where each call goes: to a person, a department, a menu or an AI. Routing can depend on the key a caller presses, what they say, the time of day, or the number they called from.
Automatic call distribution (ACD)
Software that queues incoming calls and hands each one to the next free agent. It is what made large call centers possible.
Warm transfer
Passing a call to a colleague with an introduction: who is calling and why. The caller doesn’t have to repeat themselves, which is the main thing people dislike about being transferred.
Cold transfer
Also called a blind transfer: the call goes straight to the new person with no introduction. It is faster, but the caller may have to explain everything again.
Escalation
Handing a call or message up to a person when the AI or first person can’t resolve it. Good setups escalate on clear rules, such as an emergency, an upset caller, or anyone who asks for a person.
Callback
Returning a call later instead of answering it live. Some systems offer callers a callback instead of waiting on hold.
Missed call
A call nobody answered. For a small business, a missed call from a new customer is often a lost job, because many callers try the next business instead.
Concurrency
How many calls a system can handle at the same time. A person handles one; an AI receptionist can take several at once, though some plans cap the number.
What are the parts of an AI receptionist?
An AI receptionist listens, decides what to say, and speaks, while following the information and rules you give it. These terms come up when you set one up or compare products.
Speech recognition
Turning the caller’s words into text, also called speech-to-text. Accents, background noise and poor phone lines are where it is most likely to slip.
Text-to-speech
Turning the AI’s reply into a spoken voice. The quality of the voice is often the first thing callers notice.
Large language model (LLM)
The AI model that works out what the caller means and what to say back. Some AI voice agent products let you choose which model to use, and that choice can affect both quality and cost.
Speech-to-speech model
A newer approach where one model listens and speaks directly, instead of converting speech to text and back. It cuts the delay before the AI answers and handles interruptions more naturally.
Latency
The pause between the caller finishing and the AI replying. Long pauses make an AI sound robotic, so lower latency feels more like talking to a person.
Interruption handling
Whether the AI stops and listens when the caller talks over it, sometimes called barge-in. Without it, callers end up waiting for the AI to finish a sentence they already know.
Knowledge base
The information you give the AI about your business: services, hours, service area, policies and common questions. An AI receptionist answers only as well as its knowledge base.
Prompt
The written instructions that tell an AI how to behave: its role, tone, what to ask, and what never to do. Build-your-own platforms give you the full prompt; built-in receptionists usually turn it into settings.
Guardrails
Rules that limit what an AI can say or do, such as never quoting a price or always transferring a gas leak. They keep the AI inside what you have approved.
Hallucination
When an AI states something confidently that isn’t true. A good knowledge base, an instruction to say “I don’t know,” and testing before going live all reduce it.
Intent
What the caller wants, such as booking, rescheduling, a price question or an emergency. The AI works out the intent and then follows the matching rule.
Transcript and call summary
A written record of the call and a short summary of it. Reading transcripts for the first few weeks is the quickest way to find and fix wrong answers.
What software does an AI receptionist connect to?
An AI receptionist is most useful when it can see your calendar and write to your records. These are the systems it connects to.
Online booking
Letting customers book appointments themselves within rules you set, such as services, hours and travel time. Many AI receptionists book through the same rules.
See: How Jobber online booking works with the AI Receptionist
Calendar integration
A connection that lets an AI see when you’re free and add appointments. Without one, most AI receptionists can only take a message.
CRM (customer relationship management)
Software that keeps your customer list, their history and your follow-ups. An AI receptionist connected to a CRM can recognize returning callers and log every call.
Field service management software
Software for trades and home services that runs scheduling, quotes, jobs, invoices and payments. Jobber and Housecall Pro are examples, and both now include an AI receptionist.
See: Jobber AI Receptionist review · Housecall Pro CSR AI review
Help desk
Software where support requests from phone, email and chat become tickets that a team works through. Zendesk, Freshdesk and Zoho Desk are examples.
Ticket
A record of one support request, from the first contact to the fix. An AI voice agent on a support line can create the ticket so nothing is lost.
Integration
A connection that lets two pieces of software share information, such as an AI receptionist adding a booking to your calendar. Built-in receptionists need fewer integrations because they live inside your software.
Webhook
A message one system sends to another the moment something happens, such as a call ending. It is a common way to pass call details to software that has no direct integration.
What do the pricing terms and call measures mean?
AI receptionists are priced in a few different ways, and support teams measure calls in a few standard ways. These terms help you compare like with like.
Per-minute billing
Paying for each minute the AI spends on calls. Quiet months cost less and busy months cost more.
Flat-rate plan
A fixed monthly price that includes a set number of calls or minutes. Usage over the limit is charged as overage.
Overage
The charge for usage beyond what a plan includes. It is the line to check before choosing a plan for a business with busy seasons.
Add-on
A feature sold on top of software you already pay for. AI receptionists built into field service and CRM software are often priced this way.
Setup fee
A one-time charge to get started, sometimes for the service provider to build and test the agent for you.
Phone number and carrier fees
Charges for the phone number and the calls themselves, sometimes billed separately from the AI. Some products include them; others pass them through.
Answer rate
The share of calls answered by a person or an AI rather than missed or sent to voicemail.
Abandonment rate
The share of callers who hang up before anyone answers, often while on hold or in a phone menu.
Average handle time (AHT)
The average length of a call, including any hold time and the work done after it. Support teams use it to estimate how much staff time calls take.
First contact resolution
The share of problems solved on the first call or message, with no follow-up needed.
Deflection
The share of contacts resolved by self-service or AI without reaching a person. A higher number only helps if those customers really got what they needed.
Containment rate
The share of calls an AI handles from start to finish without transferring. Watch it alongside complaints: high containment is a problem if callers who needed a person couldn’t reach one.
Cost per call
What it costs to handle one call, from staff time or AI charges. It is the simplest way to compare a person, an answering service and an AI on the same calls.
See: Is AI worth it for customer support? · Receptionist cost calculator
The bottom line: most of these terms describe one question: who answers when your team can’t, and what happens to the call next. If you are weighing your options, our guide to the best AI voice agents for small business puts them side by side by business situation.