What the WheelsAI condition score actually measures
A condition score isn't a mystical AI rating — it's a breakdown of what the photos show, what the MOT history implies, and what the dealer has declared. Here's what each number means.
Published
Three components, plain-English weighting
Every WheelsAI condition score is the weighted average of three independent checks. Each one is scored out of 100 and contributes equally to the headline number. When one component is missing data — usually because a dealer hasn't uploaded photos or the MOT history is too young to be useful — the score reflects that with a lower confidence rating.
Visual condition (photos)
AI vision on every uploaded photo: dents, paint defects, panel gaps, wheel kerbing, interior wear and cabin cleanliness. A car with nine photos and no visible defects rates high; one with three distant photos rates low regardless of the actual car.
MOT health
Drawn from DVSA records: frequency of recent advisories, weight of dangerous-category items, and mileage consistency across tests. A car with a clean recent MOT outranks one with a fresh advisory on brakes, even if the second one looks cleaner in photos.
Service compliance
If the dealer declares a service record — main dealer, independent, or self — and the declared service intervals align with age and mileage, this component scores high. If the record is 'service history unknown', this component drops the overall score and surfaces as an orange bar on the listing.
What the score is not
It isn't a guarantee. It doesn't diagnose engine condition, gearbox health, or anything underneath the car that a photo can't see. Treat it as a triage tool — a way to shortlist which cars are worth a viewing — not a substitute for an inspection.
The takeaway
If a listing has a condition score above 80 with high confidence, it's worth booking. Below 60, or with low confidence, ask the dealer specific questions about the weak component before you drive out.
Frequently asked questions
Frequently asked questions
Why is the confidence low on a car with great photos?
Usually because the MOT history is short (car under three years old) or the service record is undeclared. The score is still correct; it just has fewer data points behind it.
Can dealers game the score?
Only by uploading better photos of a genuinely better car. The AI runs on the actual images; there's no way to hide wear by re-framing a shot.