Case study · Service texting · Sept 28 to Oct 22, 2026
50 service appointments in four weeks, from customers who had stopped coming in.
Winner Chevrolet in Wilmington, DE handed us a list of lapsed service customers. We checked every vehicle's history, texted each owner personally, answered every reply by hand, and booked the visits into the store's own scheduler.
Winner Chevrolet, Wilmington, DEThe problem
Every store has thousands of customers who quietly stopped coming in.
They bought the car from you. They came in for the first oil change, maybe the second. Then they drifted to the quick lube down the street, or nowhere at all. Nobody noticed, because nobody is looking at the customers who are not on the schedule.
The usual answer is a mass blast or an automated reminder. Those get ignored, and the ones that don't get ignored get reported as spam. Once carriers flag a store's number, every message it sends afterward is at risk, including the ones that matter.
What brings people back is a message that is accurate about their specific vehicle, sent by a person who will answer when they reply. That is slower to do. It is also what works.
Illustration. Each dot is a service customer. The dim ones are the people who have not been back in a year or more. The store still has their number. Nobody has texted it.
What we did
Five steps, in this order.
- 01 Build the list
- 02 Carfax every VIN
- 03 Clean and segment
- 04 Send paced, personally
- 05 Answer and book by hand
-
Build the list from the store's own service history
We started with the dealer's service export and pulled customers whose last visit was months past due. Nothing purchased, no third-party lists.
About 1,650 customers made the final list -
Run every VIN through Carfax before anyone is texted
If a vehicle had been serviced anywhere recently, had been sold, or had a new owner, the name came off the list. The store never texted people who didn't need the visit.
On the 13 to 24 month segment, this removed about 4 of every 5 namesRemoved: 0 of 10Texted: 0Reasons: serviced elsewhere · sold · new owner -
Clean and segment
Remaining customers were grouped by how far past due they were, whether the vehicle had an open safety recall, and whether they had prepaid maintenance to use. Each group got a message that fit its situation.
-
Send personally, in small paced batches
Messages went out in small batches every 8 to 12 minutes from a number dedicated to the store. Before every batch we checked the opt-out and spam rates, with an automatic stop if either rose.
Opt-out rate, runningwithin limitSpam reports, runningwithin limit -
Answer every reply by hand and book the visit
A real person answered each reply, usually within minutes, and booked the appointment directly in the store's scheduler. Anything that needed a manager, like an unhappy customer or a safety concern, was escalated to a manager.
293 replies, zero botsThreads are recreated from the campaign with details changed. No real names, numbers, or dates.
Service teamtexting on behalf of the dealer
Results
The numbers, and how each one was measured.
From roughly 1,650 customers texted to 50 booked visits. The bars are to scale. Every figure below says where it came from.
- Customers textedfinal list~1,650
- Deliveredconfirmed status1,38391%
- Repliedat least once29321% of delivered
- Bookedin the scheduler501 per 28 delivered
Who booked
- Mostly customers 4 to 12 months past due
- Customers 20 or more months lapsed
- Vehicles with open safety recalls
- Prepaid maintenance visits the customer had not used
| Metric | Result | How it was measured |
|---|---|---|
| Customers texted | ~1,650 | Unique customers on the final send list after Carfax verification and cleaning. |
| Delivery rate | 91%1,383 of 1,517 | Messages with a confirmed delivery status from the carrier. Messages without a confirmed status are not counted either way. |
| Reply rate | 21%293 people | Delivered customers who sent at least one reply. Each one was answered by a person. |
| Opt-out rate | 7% | Customers who opted out. Each was removed immediately and not contacted again. |
| Appointments booked | 50about 1 per 28 delivered | Visits booked into the store's scheduler during the campaign window, Sept 28 to Oct 22, 2026. |
| Est. service revenue | $8,800 to $22,400estimate, not actuals | 50 visits multiplied by the store's own repair-order history. See the calculation below. |
Revenue estimate
50 visits × the store's own RO figuresMedian and average come from 918 Chevrolet repair orders in the store's own history. The median gives the low end ($8,800), the average gives the high end (about $22,400). Neither is an actual: it is what 50 typical visits have been worth at this store.
Projection, not a promise
What the same rates would mean at your store.
Drag the slider to the number of lapsed service customers you think you have. The math below applies this campaign's delivery, reply, and booking rates, and this store's RO figures. Your numbers will differ.
This is arithmetic, not a forecast. It assumes your customers reply and book at the same rates as Winner Chevrolet over four weeks, and that your repair orders look like theirs. The honest version is: we run it, and you get your own numbers.
What the replies revealed
It is more than appointments.
When a person answers the text, customers tell you things a reminder would never hear.
Extra work on the same ticket
Customers who booked a routine visit mentioned problems they wanted looked at while they were in:
- An oil leak
- A broken horn
- A tire pressure sensor warning
A second vehicle
One customer asked to bring a second vehicle on the same visit. That is two repair orders from one text, and it only happened because someone was there to say yes.
Win-backs you would never hear about
Some replies came from customers who were unhappy. One owner said their ABS and electrical issues weren't fixed last time. Replies like that were escalated to a manager as win-back opportunities instead of being lost.
Why it is safe
Your customer list, your phone number, and your reputation come first.
The fastest way to lose a store's trust is to get its number flagged or its customers annoyed. Every part of the process is built to prevent that.
Opt-outs are handled before anything else
A customer who opts out is removed on the spot and never texted again. This is checked before any reply is written.
Sending is paced, with an automatic stop
Small batches every 8 to 12 minutes. Opt-out and spam rates are checked before each batch, and sending stops on its own if they rise.
No bots
Every reply is read and answered by a real person. Nothing goes to a customer that a person did not write.
The store's data stays private
The service export is used for this store's campaign and nothing else. It is not shared with other dealers, resold, or merged into another list.
No customer information on this page
No names, phone numbers, VINs, or real message contents appear here. The threads and panels above are recreated on purpose.
One dedicated number per store
Your campaign runs on its own number. Another dealer's sending never touches your reputation with the carriers.
What it takes to start
Four things from you. Launch in days.
No software to install, nothing for your advisors to learn. The appointments land in the scheduler they already use.
A service export
Customers and repair orders from your DMS. We use it to find who is overdue and what their vehicle has had done.
Scheduler access
A login or booking process for your service scheduler, so appointments land where your advisors already work.
A texting number
A number dedicated to the store. We handle the setup and the pacing.
A manager for escalations
One person we can hand unhappy customers, safety concerns, and win-back opportunities to when they come up.
Next step
See what this would look like at your store.
A 20-minute demo. Bring a rough idea of how many service customers you have, and we will walk through how a first campaign is built, what we check before anyone is texted, and what we report back.