AI in Sales
The $1,340 Nobody Called Back About: A Deferred-Work Audit for Repair Shops
Your close rate measures what customers approved that day. It says nothing about the declined line items sitting in your inspection records. Here is a five-step audit to find out how much deferred work your shop is carrying and what it is worth.
by Jerrod Anthraper
A Camry comes in Tuesday for a check-engine light. Your tech finds the misfire, and the inspection also flags a leaking water pump and rear brakes at 3mm. You text the customer a $1,340 estimate at 10:40 a.m. She approves the misfire repair, declines the rest, and says "let me think about the other stuff." She picks up at 4. Everyone is polite. The car leaves.
Nobody calls her again. Six weeks later the water pump lets go in a grocery store parking lot and a tow truck takes it to whoever was open.
That is not a diagnostic failure. That is a follow-up failure, and most shops cannot see it because it never shows up as a lost job.
The leak: your close rate is hiding your callback rate
Ask a shop owner their close rate and they will quote you the number on the estimate they presented — approved dollars over presented dollars, that day, that visit. It is a real number and it is usually decent.
What it does not measure is everything that got deferred. "Let me think about it," "I'll do it next paycheck," "let me check with my husband," "call me when you get the part." Those are not losses on the board. They sit in a gray zone where nobody owns them, and in most shops nobody ever touches them again.
The number that should scare you is not your close rate. It is how long it takes anyone at your shop to respond to a customer who is actively deciding. Across industries, average response time to an inbound lead runs about 42 hours — and a deferred estimate gets treated with even less urgency than a new lead, because it already feels half-handled.
Why it happens in a shop specifically
The service advisor is the bottleneck, and the bottleneck is structural, not personal.
Your advisor is running the phone, the counter, parts calls, warranty approvals, and three customers who all want a status update on cars that are still on the lift. Every one of those is a person standing in front of them right now. A deferred estimate from Tuesday is a person who is not standing there. It loses every time, all day, forever.
The second reason is that there is no artifact. An approved job becomes a repair order — it lives somewhere, it has a status, someone chases it. A declined line item becomes a note in the inspection that nobody queries. There is no list called "people deciding right now," so there is no list to work. The work does not get skipped because it is unimportant. It gets skipped because it is invisible.
A callback audit you can run this week
This is a measurement exercise, not a sales push. The goal is to find out how much deferred work your shop is actually sitting on before you decide whether it is worth fixing.
1. Pull 30 days of declined line items
Export every inspection or estimate from the last 30 days and isolate the lines that were presented and not approved. Most modern shop management and DVI systems will do this in a few clicks; if yours will not, pull 40 repair orders by hand — a sample is enough to see the shape.
Total the dollars. Do not adjust it, do not talk yourself out of it. That raw number is your deferred pipeline. In most shops it is somewhere between one and three weeks of revenue, sitting in a database.
2. Count how many of those customers were contacted again
For each declined line, find any record of a follow-up after the car left — a call log, a text, an email. Not a marketing blast. A specific contact about that specific deferred job.
The percentage is usually in the single digits. Write it on the whiteboard. This one number will tell your team more than any sales training you could book.
Be strict about what counts. A generic "time for your oil change" reminder is not a follow-up on a $600 water pump. If the contact does not name the specific deferred job, it does not go in the column.
3. Sort the list by consequence, not by dollar value
Do not chase the biggest ticket. Chase the job most likely to become someone else's emergency.
Rank every deferred line into three buckets: safety and imminent failure (brakes at 3mm, tie rod play, coolant leak), schedulable maintenance (fluids, filters, belts approaching interval), and cosmetic or optional. The first bucket is where your callbacks belong, because those customers are going to spend that money in the next 90 days regardless. The only question is whether they spend it with you or with whoever is closest when it fails.
4. Time-stamp your first response on new inquiries too
While you are in the data, check the other end of the funnel. Pull your missed calls, web form submissions, and text messages for the same 30 days, and log how long until someone responded.
Do this honestly, including nights and weekends. If a form came in at 6:10 p.m. Friday and someone replied Monday at 9, that is 63 hours, not "Monday morning."
5. Pick one bucket and work it manually for two weeks
Take the safety-and-imminent-failure list. Have your advisor make five calls a day — not a pitch, just: "You had rear brakes at 3mm when you were in on the 12th. Wanted to check where you landed on that, and I can hold a slot Thursday."
Track how many book. Two weeks of this gives you a real conversion rate on deferred work, which is the only way to know what your pipeline is genuinely worth. It also proves the constraint: after ten days your advisor will tell you they cannot keep it up, and they will be right.
What the first-responder data says
Research on inbound sales consistently finds that 78% of sales go to whoever responds first. That is the whole argument for working the deferred list.
A customer with 3mm brakes is not deciding whether to fix them. She is deciding when, and who. If she gets a text from you on Thursday, you are the first responder to a decision she has already half-made. If she does not, the first responder is whoever is open the day the pedal goes soft — and that shop gets a job you already diagnosed, already quoted, and already paid to acquire.
Where the system takes over
The audit above will tell you the size of the leak. It will not close it, because the constraint is the same one that created it: your advisor has a counter to run.
That is the part our AI sales agent, James, handles. He answers inbound in under 60 seconds around the clock, works the deferred list on a schedule instead of when someone remembers, and books the return visit straight onto your calendar — then asks for the review after the work is done. Your advisor keeps the counter. The follow-up stops depending on a quiet afternoon that never comes.
Start with step one. Export 30 days of declined line items and total the dollars. If that number is small, ignore all of this. If it is not, you now know exactly what the callback problem is costing you.