❓ Problem

Car repair is one of the last consumer purchases where the price is set by a stranger, in person, after you have already handed over the keys. The check engine light comes on, you drive to whoever is nearest, and you accept the number. Only 17% of drivers believe they are always charged fairly, and 83% end up ringing someone they know to sanity-check the quote, which tells you the market has already decided the price is negotiable and has simply built a folk workaround instead of a mechanism.

The academic version of this is unusually clean. Researchers rang 2,778 shops for the identical repair and collected 4,603 quotes. Fewer than one in ten shops quoted the same price twice, and callers who signalled they expected a high price were quoted $35 more. That is not variance in cost structure. That is the shop reading the customer and pricing the read.

The dispersion has got more expensive to be on the wrong side of. Repair and maintenance costs are up 43.6% since January 2019, well ahead of general inflation, and the average car on US roads is now 12.6 years old, which means more unplanned, out-of-warranty work and less of it covered by anyone else. 82% of drivers say they delay repairs over cost. Some of that is genuine affordability. A lot of it is a refusal to walk into a negotiation you know you will lose.

The reason nobody has fixed it is not that the app is hard. It is that the supply side has never wanted to participate. Getting a quote requires a shop to stop working on a car, pick up a phone, and price a job it might not win. The average independent shop already misses between 25% and 40% of its inbound calls from customers who are actively trying to give it money. Asking that same shop to proactively bid on a stranger's brake job was never realistic.

✅ Solution

A reverse auction where the shops never have to log in.

• Post the job once, in a comparable spec. Make, model, reg, mileage, a photo of the warning light or the symptom described in plain language. The platform normalises it into a defined job so that every quote coming back is priced against the same thing rather than against the customer.
• A voice agent does the ringing, not the shop. This is the entire idea. The platform calls fifty local garages on the driver's behalf, describes the job, captures a price and an availability window, and returns bids inside ten minutes. The shop's participation cost falls to answering a phone call it was going to answer anyway. No signup, no dashboard, no onboarding, no marketplace liquidity problem on the supply side because the supply side does not know it is a marketplace.
• Two-stage bidding for anything undiagnosed. A check engine light is a question, not a job. Round one auctions the diagnostic at a fixed price. Round two auctions the actual repair off the returned fault code, which is now a genuinely standardised spec. Skipping this is how you end up with lowball quotes that triple once the car is on the lift.
• Payment through the platform, held until completion. Partly for the customer, mostly for you, because it is the only way a success fee survives contact with a shop that would rather take the booking directly.
• Cars first because the jobs are quotable over the phone. Then home repairs, where the tickets are bigger and the pricing is murkier, though the spec problem there is far worse and should not be treated as a free extension.

📊 Key Numbers

Market size

• The US automotive service market is roughly $211bn in 2026, but most of that is dealerships, tyres, collision and fleet. The independent mechanical repair slice that a quote-and-book marketplace can actually touch is closer to $80bn, spread across 253,000 repair shops.
• At a 6% success fee on booked work, that is a TAM of roughly $4.8bn. Treat it as a ceiling that assumes every repair in America goes to auction, which it will not.
• SAM: only unplanned, out-of-warranty, over-$300 jobs where the driver has no incumbent relationship. Routine servicing goes where it always went. Call it 35% of the spend, so $28bn of addressable GMV, or about $1.7bn in platform revenue.
• SOM: capturing 0.5% to 2% of that GMV within five years is $140m to $560m of bookings, or $8m to $34m in net revenue. The historical record says plan for the bottom of that band.