Reading between the lines of a used car listing.

Every used car ad is a document with four layers. Most buyers skim one. The other three are where the price is hiding.

The listing genome

A used car listing encodes information in four places: the title, the structured fields, the free-text description, and the photos. Each one is a different channel, and each one is gameable in a different way. The reason buyers get burned isn't that the information isn't there — it's that they only read one channel out of four.

Structured fields are the comfortable part. Mileage, year, fuel type, gearbox, power. They render in a clean grid on every platform, they're easy to filter on, and they feel objective. Most people short-circuit there and skip straight to "fair price?" That's the mistake. Structured fields are what the seller chose to declare. The other three channels are where you find what they chose not to.

Title-line tells

The title is the only sentence the seller writes that they know will be read. It's also the part they spend the most time second-guessing. Read it twice.

The trim mismatch

"BMW 320d Sport" on a car that's clearly an SE in the photos. "Audi A4 S-Line" on a body that's missing the badge. A title that names a trim level richer than the car actually wears either means the seller doesn't know what they own (bad sign, probably not the original owner who would) or they're deliberately fishing for search traffic. Either way, you trust the rest of the ad less.

The missing trim

More telling, in some ways, is when no trim is named at all. "VW Golf 1.6 TDI 2014, 145,000 km." On a car platform where trim level changes resale value meaningfully, omission usually means base trim. Not a deal-breaker; just useful priors before you've read another word.

The vocabulary of suspicion

A short field guide. "Rare" almost never is. "Must see" is a substitute for facts. "No time wasters" means the seller has had a bad week of viewings — ask yourself why. "Quick sale" means a deadline you don't know about. "Mechanic special" or "needs TLC" is honest, at least, but translates to: budget for a repair you can't yet scope. None of these phrases are red flags on their own. They're priors. They shift the burden of proof.

The structured fields nobody questions

Here the trick is to read for what is missing more carefully than what is present. A listing that says "1st owner" is making a specific claim. A listing that says nothing about ownership history is making a different one — by omission.

• "Service book" vs "full service history". A service book is a physical artifact; it may or may not contain stamps. Full service history is a claim about completeness. If only the former is mentioned, assume gaps.
• Inspection valid until [date]. If the date is close to expiry, the seller is passing the next inspection cost — and any work needed to pass it — to you. If the date is suspiciously fresh, ask which station performed it.
• Owner count. "2nd owner" on a six-year-old car is normal. "4th owner" on the same car is a sign it's been through hands quickly, which usually means each owner found something they didn't want to fix.
• Accident-free. Often a self-declaration with no document behind it. Treat as sentiment, not data, until proven by a history report.

Across a hundred listings for the same model, you start to see what the median ad declares. The ones that deviate downward — fewer fields, vaguer language, missing dates — are not always worse cars. But they're worse listings, and you should price the friction in.

Decoding the description

The free-text description is where a seller writes in their own voice, and it's where the most information per character lives. It also rewards close reading more than any other part of the ad.

Euphemisms

Used car descriptions have their own dialect. "Driven by a non-smoker" is a useful detail and also a tell — the seller is reaching for things to say. "Minor scratch" is the seller's adjective, not yours; assume it's a panel. "Needs love" is a warning. "Garage-kept" is reassuring on paint and interior; it says nothing about the mechanical condition. "Reluctant sale" is filler. "Engine runs strong" usually means the engine starts.

Omissions

On a 150,000 km timing-belt engine, the description should mention the timing belt. If it doesn't, you ask. On a car older than the warranty, the description should mention recent service. If service history stops at year four of a ten-year-old car, that's not "no information" — that's information. For diesels with DPFs, EGR valves, or AdBlue, silence on any of these components after high mileage is conspicuous.

Over-justification

A short reason for sale is healthy. "Selling because moving abroad," "company car arriving," "downsizing" — these are fine. A paragraph of life events is not. When a seller volunteers an emotional arc for why they're parting with the car, they're managing your perception of a sale they expect you to hesitate on. The longer the explanation, the more carefully you read the rest.

Positive signals

Not everything is a red flag. Real green flags exist, and they're specific. Named-shop service stamps ("serviced at Müller & Söhne, Stuttgart, since new"). Recent invoices itemized by part ("new front discs and pads, March 2026 — 480 EUR, invoice available"). Replacement parts named by OEM or quality brand rather than "new." Mention of consumables — wipers, brake fluid, coolant — implies a seller who understands what maintenance is. These are signals you weight upward.

Photo forensics

Photos are the part of the listing the seller controls most precisely. Which means the absences are deliberate. A fifteen-photo ad of a ten-year-old car that contains zero photos of the rocker panels, wheel arches, under-bumper, or engine bay is telling you exactly where it doesn't want you to look.

• Angles deliberately avoided. Sill rust lives at the rocker. Bumper damage lives at the corners. Underbody damage lives, predictably, underneath. If a private seller can shoot fifteen glamour angles but not one of the engine bay, that's a choice.
• Tire age and wear. Sidewall DOT codes give you the manufacture date. New tires on a car otherwise advertised as "barely driven" deserve a question. Uneven wear across an axle suggests alignment or suspension trouble.
• Panel gap asymmetry. Compare the gap between the bonnet and each fender. Between each door and the body. Asymmetry between the left and right side, especially at the front, is a fingerprint of past frontal damage.
• Suspicious lighting. Heavily filtered photos at golden hour can hide colour mismatches between repainted panels. A flash photo or a photo in flat overcast light is more honest than a glossy sunset shot.
• Background details. Carpet tiles, numbered parking bays, fluorescent lighting, rows of identical cars behind — these are dealer lots, even when the listing claims "private seller." A driveway, a garage, a residential street with another car's mirror in frame is more plausibly private. Neither is automatically better, but the dealer/private distinction changes your legal protections in most markets, so it matters that the listing is honest about it.

The free-text gold problem

Here's the practical issue. You can read one listing carefully. You can read ten. By listing fifty, you've forgotten which Audi mentioned a new clutch at 90,000 km, which one said "clutch recently replaced" without a date, and which one stayed silent. You start filtering on price and mileage again because the structured fields are the only thing your memory can hold.

That's the channel where pattern-matching across listings creates real leverage. Tracking, across a model and engine combination, what percentage of listings at this mileage mention a timing belt change, a DSG service, an injector replacement — that's a statistical view of what's normal, and any given ad's silence becomes meaningful only against that baseline.

What an AI like Torque actually does with a listing

Torque reads the whole listing — title, structured fields, description, photos — and extracts a structured picture: what was explicitly claimed, what was conspicuously missing, which photos didn't appear, and how the result compares to the median ad for the same model, year and engine. The output isn't a verdict. It's the briefing you'd write yourself if you had time to read two hundred ads for the same car. It's also evaluated against your profile — what you actually care about — so that "high mileage" means something different for a long-commute buyer than for a weekend-only one.

A worked example

Consider this listing, which is fictional but assembled from patterns you'll find on any classifieds site.

The twelve-item mental checklist

Before you message a seller, extract these from every used car listing:

1. What does the title overclaim or omit?
2. Is the trim level explicitly named?
3. Service book or full service history — which word did they use?
4. When does the next inspection fall, and who passed the last one?
5. How many owners, and over how many years?
6. Are accident and ownership claims self-declared or document-backed?
7. Does the description mention the components known to fail on this engine at this mileage?
8. Are repairs named by part and date, or only by adjective?
9. Is the reason for sale short and ordinary, or over-justified?
10. Which angles are missing from the photo set?
11. Do the photos place the car at a private location or a dealer lot, and does the listing agree?
12. Against the median ad for this model, year and engine — is this listing richer in detail, or thinner?

Read any listing through those twelve, and you've already done more than the buyer the seller is hoping for. The point of decoding a car listing isn't to catch the seller out. It's to know which questions are worth your Saturday — and which ones aren't.