What does a missed call cost an auto repair shop?

One missed call costs you one repair order, and often the customer behind it. The driver whose car will not start does not wait for a callback. They ring you, hear voicemail, and ring the next shop in the results. Marchex, which listens to service department calls for national operators, found up to 21% of inbound calls missed or failed, and the best performing locations still missed nearly 10%. Those are chains with people whose only job is the phone. An independent with one advisor at the counter and the phone on the wall is not beating that. The rest of this page puts a number on your own shop and compares five ways to answer when every hand is busy.

How many calls does an independent auto repair shop miss during business hours?

Somewhere between one in ten and one in five, and nobody has published a clean number for independents. The only source that shows its method is Marchex, a call analytics company that records calls for large multi-location service operators. Its 2026 analysis counts unanswered, dropped and mishandled calls together, finds rates over 20% across locations, and says a healthy rate is 10% or below.

The vendor blogs that rank for this question quote 23% or 38% for independents and credit the Automotive Service Association. None link to a report, and I could not find one. Treat those figures as marketing until someone shows the data.

The useful number is your own. Count missed calls in your carrier app for two weeks. That count feeds the arithmetic below.

Why does the shop phone go unanswered?

Because the person who should answer it is doing the work that pays. In a one to three bay shop the phone is picked up by whoever is closest, and whoever is closest is under a car, at the counter with a customer, or writing up an estimate. The morning drop-off rush and the late afternoon pick-up rush are when the phone rings most and when the counter is busiest. Add lunch. Add every minute after you lock the door.

The after-hours share is the one owners underestimate. A car that dies on the way home produces a call at 6pm. A check engine light noticed on Saturday produces a call on Saturday. Those calls go to voicemail by design, and the driver books wherever a person or an agent picked up. The same curve that decides who wins an online lead decides who wins a ringing phone, only faster, because the next number is already on the caller’s screen.

Does a driver with a broken down car leave a voicemail or call the next shop?

They call the next shop. No study I could find measures this for auto repair, so what follows is reasoning, not data. The figure repeated across the web, that 80% of callers who reach voicemail hang up, has no traceable source, so it is not on this page.

What you can rely on is the shape of the decision. A driver with a working car and a service due in two weeks might leave a message. A driver whose car is on the shoulder, or who needs it for work tomorrow, is looking at a list of shops with a call button next to each one. Voicemail is the signal to press the next button.

What is one missed call worth at your average repair order?

Missed calls per week, times the share that were real jobs, times your average repair order, times 50 weeks. That is the whole formula, and every number in it should be yours.

Say you count 12 missed calls a week. Not all of them are jobs. Some are suppliers, wrong numbers, or the same customer twice. If a third were people who wanted work done, that is four jobs a week. The Paar Melis 2025 benchmark, built from the financials of hundreds of US auto repair shops, puts the average repair order at $702. At that figure four jobs a week is about $140,000 a year of work you never quoted. Even if only one of the four would have booked with you, that is $35,000 against a phone answering option that costs a few thousand a year at most. Put your own average in if you know it. The arithmetic is the same.

Two honest caveats. Some of those callers would have called back. And a missed call is not a lost customer when the caller already knows you and trusts you. The formula is for new callers, which is why the rush-hour and after-hours calls matter most. They are the people you have never met, and they are the ones the homepage numbers are about.

How do you catch the call when nobody is free?

Five options, and they are not equal. The table shows what each does with a call that comes in at 4:45pm while you are under a car. Prices were checked in September 2026.

Option	What happens to the call	Captures the car and the problem	Puts you on the line	Covers after hours	Cost
Voicemail	Caller hears a greeting and decides whether to wait for you	Only if they leave a message	No	Same voicemail	Free
Missed-call text back	Caller gets an automatic text within a minute asking what they need	Only if they reply	No	Yes, by text	Depends on your phone or CRM plan
Live answering service	A person at a call centre answers in your shop’s name and takes a message	Yes, as a message	Rarely, most plans take messages	Yes	$250 a month for 50 minutes on Ruby’s entry plan
AI voice agent	An agent answers, asks the questions you gave it, and transfers, takes a message, or books if you run a calendar	Yes, as structured fields	Yes, a warm transfer in the hours you set	Yes	Cents per minute in usage (figures below), plus the build
Front desk hire	A person answers during their shift	Yes	Yes, when they are at the desk	No, unless you pay for it	Their wage, plus the hours the phone rings while they are at lunch

The middle three are the real choices. Text back is cheap, GoHighLevel and most business phone systems have it as a setting, and it catches the patient caller. The answering service catches everyone but hands you a stack of messages to return, so the work has moved, not gone. The agent is the only option that asks the questions you would ask, tells the caller when to bring the car in, and puts you on the line when it matters. Its running cost is small: Vapi charges a $0.05 per minute platform fee plus speech and model costs at provider rates, and a Twilio number is $1.15 a month plus $0.0085 per inbound minute. The real cost is the build, which is what I do and what the next two sections are about.

Can an AI answer the shop phone and book the car in?

It can answer, ask the right questions and connect you. Whether it can book depends on whether your shop runs on a slot calendar, and many one-owner shops do not.

Every vendor page for this search promises to book straight into your shop management system. That is true when you run Tekmetric, Shop-Ware or Mitchell with bays scheduled by the hour, and only when the vendor has a two-way integration with your specific system. Ask for that in writing before you sign. It is not true, and not useful, when the shop takes walk-ins during opening hours and the owner decides the order of work after seeing the car. Forcing a calendar onto that shop creates bookings nobody keeps.

For that shop the agent’s job is simpler and worth more. Capture who is calling, which car, and what is wrong in the caller’s own words. Say when to come. Put the owner on the line if they are free and the call warrants it. Take a message with a callback time if they are not. That is what I am building right now, and the decisions are below.

What I see when I build this for a car repair workshop

The agent I am building now answers for a one-owner car repair workshop in Poland. Open Monday to Friday, eight to four, walk-ins welcome, the owner’s mobile is the shop phone. It runs on Vapi with n8n behind it. It has taken test calls and the warm transfer to the owner’s mobile works. It has not been live long enough to give you call counts, so what I can show you is the decisions, which are the part most shops get wrong.

It asks for the caller’s name first and uses it for the rest of the call. Then it asks which car, make, model and roughly the year, and what is happening with it, and it writes down the answer in the caller’s words. “Grinding noise when I brake” stays “grinding noise when I brake”. It never becomes “probably the pads”.

It never diagnoses, never quotes a price, and never confirms a diagnosis from another workshop. The line it uses, translated: “I cannot judge that over the phone and I do not want to guess. The mechanic needs to see the car, and then we will tell you straight away what needs doing and what it will cost.” Every “what is it, how much, how long” gets that answer and an invitation to come in. The owner wanted it that way because a price given on the phone by anyone, person or agent, becomes a promise the mechanic has to argue with later.

There is no booking tool. The shop takes walk-ins, so the agent says when the workshop is open and asks roughly which day the caller plans to come. If someone wants a fixed hour, that is the owner’s call, so the agent transfers during opening hours or takes a message with the preferred time outside them.

The transfer is a warm one, and it only happens Monday to Friday between eight and four. The caller holds, the agent dials the owner’s mobile, briefs the owner in one breath (who, which car, what is wrong, what they want), and the owner accepts or declines. If nobody picks up, the agent is still on the line, says so, and takes a message with a callback time.

Two smaller rules do a lot of work. If the caller describes something dangerous, brakes failing, smoke, the engine cutting out on the road, the agent tells them not to drive the car and escalates. If the caller asks for something the shop does not do, tyres, motorcycles, trucks over 3.5 tonnes, registration inspections, it says so in one sentence and offers help with anything else. It does not recommend another business, because the owner does not want to vouch for one.

The last decision is about the owner’s inbox. After each call an n8n workflow reads the structured summary and emails the owner only when there is something to do: a transfer that was missed, a callback that is owed, a caller who said they are coming in. The email carries the caller’s name and number, the car, the problem in their words and the best time to call back, with an urgent flag on top when it applies. Calls about opening hours or the address send nothing. An owner who gets an email for every call stops reading them by Thursday.

Where to start

Count your missed calls for two weeks and multiply by your own average repair order. If the result is smaller than a text-back plan, turn on text back and stop reading. If it is bigger than a year of an answering service, and for most shops it will be, decide what you want the phone to do at 4:45pm while you are under a car: take a message, or ask the questions you would ask and put you on the line. That decision, more than the vendor, is what you are buying.