When a shop hands you two or three references, remember what you're holding: names the shop chose to give you. That doesn't make them worthless — it makes them a starting point that needs work. The work is to ask those references different questions than you asked the shop, for a different purpose, and then to read the pattern in how they answer rather than the words themselves.
Asking a reference the same questions you asked the shop just confirms the shop's story. The point is to surface what a shop-supplied reference won't volunteer: the surprise that went badly, the thing the shop didn't do well, the part the customer would contract differently if they did it again.
The eight questions to ask each reference
- What kind of project did they do for you? You're verifying the scope matches the portfolio the shop claimed.
- How did it actually go — on time, on budget, on scope? The most honest answer is usually “mostly, with one issue.”
- What was the worst surprise during the project, and how did the shop handle it? Every project has one. The question is whether the shop owned it or hid it.
- Would you use this shop again? Listen for hesitation. A “yes, but…” is a no.
- Did the shop do everything they said they'd do? Compare it to the shop's own answers to your forty-seven questions.
- What did the shop NOT do well? Every honest reference has something. “Nothing” means they're being polite — push gently.
- Has the owner or operator changed since your project? Rotating ownership is a fraud-prone pattern; a reference for “the old owner” doesn't validate the new one.
- Would you have contracted with them any differently? This surfaces the lesson the reference already paid for, so you don't have to.
Three follow-ups are worth the small awkwardness: ask to see receipts or invoices, which check actual costs against the shop's claims; ask whether they knew the owner as a friend or family before the job, because fraud-prone shops lean on network references; and ask whether you can see the finished car, because the proof of quality is the vehicle, not the testimonial.
Read the pattern, not the praise
The answers matter less individually than as a shape. A few patterns to weigh: when every reference raves with no negatives, you're probably looking at a cherry-picked list — push for a real customer the shop didn't hand-select. When references are hard to schedule, evasive, or suddenly short on time, treat them as possibly fake or coached and ask for more. When two or more references mention the same issue — communication was slow, say — that's a real pattern, and it belongs in your score.
“All rave is a warning. “Yes, but…” is a no. A reference you can't pin down is one you can't trust.”
It's tempting to think a shop with great reviews and a long history has earned a pass on this. It hasn't. Every documented deposit-loss case I've read started with a shop that had good reviews and years on the sign. Longevity isn't honesty, and a five-star page isn't a reference you controlled. Even the trade press is careful here — Hagerty's own guide is titled “How to (try to) protect against a car restoration rip-off.” Their word is try to. The reference check is one of the few moves where you, not the shop, decide what gets asked. Use it that way.
Sources & notes
- The eight reference questions, the three follow-ups, and the pattern-reading guidance are from Chapter 5 of the Shop Vetting System Playbook (“The Reference-Check Protocol”), including “don't ask the same questions you asked the shop” and “‘yes, but…’ is a no.”
- Hagerty article title “How to (try to) protect against a car restoration rip-off” — Hagerty.
Dorian Quispe
Restoring a '67 in Los Angeles, and writing down what it actually costs. Author of the Shop Vetting System.