Sources: the call volume is AAA's own, from its 2025 release on staying proactive about repair and maintenance, which reports the 27 million emergency roadside calls it took across the United States in 2024 and the share of them that were a tow or a battery. The cost figure is the US Bureau of Labor Statistics, Consumer Price Index table 2, motor vehicle maintenance and repair, for the 12 months to July 2026. The 3 figures that used to sit here are gone rather than dressed up: the 1 in 4 unanswered calls figure was credited to 2 industry bodies that publish no such number and its only trail was a marketing blog that is now dead, and the $428 average repair order is published by nobody a reader can reach without paying $2,950 for a trade factbook. Read your own answered call rate off your own phone system before you act on anybody's average, including ours.
Why this hits Detroit, MI shops in particular
The local shape of the problem here is knowledgeable owners describing symptoms precisely, on a fleet worked hard by salt and winter. That matters because the shops losing this fight are almost never losing on the work. They are losing in the gap between a customer needing help and somebody answering.
The pattern repeats in 3 ways, and all 3 are worse for an independent shop:
- A stranded driver calls everybody until somebody picks up. Somebody whose car would not start this morning is not comparing labour rates. They are calling down a list, and the list stops at the first shop that answers and can say what a diagnostic costs and when it could be looked at. Most of them never leave a voicemail.
- Your service advisor is also your estimator, your parts caller and your cashier. The phone rings hardest during morning drop-off and evening pick-up, which are the same 2 windows the counter is physically busiest. The calls that ring out are the ones nobody had a hand free for.
- The biggest leak is work the customer already saw and declined. It is already diagnosed, already priced and already photographed, and it sits in the shop management system as a list nobody owns. Closing it does not need a better recommendation, it needs a follow-up that does not skip a day when the counter gets busy.
The honest arithmetic: an answering service covering overflow and after hours runs roughly $140 to $500 a month. At an average repair order in the low hundreds, it pays for itself on 1 or 2 vehicles a month that would otherwise have gone down the road. That is why the shops that have done it rarely go back.
What this looks like on a real Detroit, MI call
Before
A commuter's car will not start at 6 in the morning. They ring you at 7:20 while the shop is dark, get a recorded greeting, and do not leave a message. By 7:40 they have reached a shop 2 miles away that answers overflow calls around the clock, and that vehicle, its 90,000 mile service and the family's second car went with it.
After
The same call reaches an AI front desk agent. It takes the year, make, model and the symptom in the caller's own words, says what your diagnostic costs and that you can look at it today, offers the 2 drop-off windows you actually have, and texts the advisor a written summary before the doors open. The car is on the lift by 10.
The right build for each part of the shop
Declined work follow-up
- Every declined line item is worked as a list rather than remembered by whoever wrote the estimate
- The follow-up carries the photo or video from the inspection, the price and the reason it was recommended
- A customer who declined in March is asked again in June, before the noise sends them somewhere else
Phone and web intake
- Overflow and after-hours calls are answered rather than left to a recorded greeting
- Year, make, model and symptom are captured in the customer's own words and handed to the advisor in writing
- Anything that sounds like a safety concern reaches a person the same hour
Service reminders and reactivation
- Reminders fire on that vehicle's own mileage and date rather than as a monthly blast to the whole list
- A lapsed customer gets 1 specific invitation naming their own vehicle and what is due
- State inspection and registration dates get their own cadence instead of waiting for a breakdown
Fleet and commercial accounts
- A fleet enquiry reaches an owner the same day rather than joining the retail queue
- Scheduled maintenance across a whole fleet is planned rather than reacted to
- The follow-up sequence runs for weeks, because that is how long these accounts take to move
The tools doing the work
| What it does | Tools | Monthly cost | Setup |
|---|---|---|---|
| AI phone and web intake, answering overflow and after hours and escalating | Smith.ai, Goodcall, Numa, Retell for a custom build | $140 to $500 | Low |
| Shop management with digital inspections and declined work tracking | Tekmetric, Shop-Ware, AutoLeap, Mitchell 1 | $200 to $700 | Low |
| Two-way texting, estimate approvals and photo or video approval links | Numa, Podium, Weave, or the texting already inside your shop management system | $150 to $500 | Low |
| Service reminders on the vehicle's own mileage and date | Kukui, Steer, Mudlick, or the reminder module inside Tekmetric or Shop-Ware | $200 to $600 | Low |
| Review generation after the vehicle is picked up | Birdeye, Podium, NiceJob | $75 to $300 | Low |
| Custom front desk and declined work agent across voice, text and your shop management system | Built by OpsJuice on Retell, n8n and Tekmetric or Shop-Ware | Project based | Managed |
The first 30 days, in order
- Day 1. Put an answering service on your main line for overflow and after hours, with a written rule naming what goes straight to a person. Nothing custom, nothing integrated. It recovers the breakdowns you are losing this week.
- Day 7. Turn on two-way texting for estimate approvals and send the inspection photos with the estimate. Approvals get faster, arguments at pick-up get rarer, and this usually needs no new software at all.
- Day 21. Work the declined work list every single week, and set service reminders to fire on each vehicle's own mileage and date. Declined work is the highest return automation in this business because the selling was already done.
Only after those 3 are running does a custom build make sense, and it makes sense for a specific reason rather than as an upgrade: several locations with different labour rates, a fleet book that needs its own scheduling rules, or an intake flow your shop management system cannot represent without somebody retyping it into a second screen.