CARPARTPICKER

Data policy

How we verify fitment

The promise of this site is that everything you see fits your car. That promise is only worth anything if we can say where each claim came from — so every fitment row in our database carries a source and a date, and nothing reaches this page until it has been confirmed twice.

01A fitment row is a claim, not a fact

We don’t assert that a part fits your car. We report that a named source claims it fits, and we show you which source and when it was published. The distinction matters, because the manufacturers themselves won’t assert certainty. Wilwood’s own installation sheet says it plainly:

“Due to OEM production differences and other variations from vehicle to vehicle, the fastener hardware and other components in this kit may not be suitable for a specific application or vehicle. It is the responsibility of the purchaser and installer of this kit to verify suitability / fitment of all components.”

If the company that machined the part hedges, a parts website inventing confidence on top of it is doing something worse than useless. So we carry their caveat forward instead of laundering it away.

02Two independent sources, or it doesn't render

A single catalog can be wrong, or stale, or silently miss a variant. So a claim needs two sources that were authored separately before we’ll show it.

verified

Two or more independent, current sources agree on coverage.

Shown

single_source

One authoritative source. Shown with its citation and the manufacturer's own caveat attached.

Shown, attributed

pending

Sourced but not yet confirmed. Sitting in the database awaiting a second source.

Never shown

rejected

Sources disagree. Kept as a tombstone so we don't re-derive the same bad claim.

Never shown

This is enforced in the query layer, not by policy or good intentions: pending and rejectedrows are structurally unable to reach a page. Showing you nothing is a worse experience than showing you a part — and a much better one than showing you the wrong part.

03What we cross-check on, and what we don't

Two catalogs printing the same industry part number is not two sources agreeing. Shape numbers are assigned upstream by FMSI, so every brand is reading the same reference — that’s one source wearing two hats. We treat matching part numbers as a checksum.

What we actually verify against is coverage: which years, trims, engines and qualifiers each manufacturer independently decided their part fits. Those lists are separately authored, they genuinely diverge, and their disagreements are the whole point.

04A real conflict this caught

The first part we ran through this process produced one. Hawk’s catalog lists HB532as the rear pad for the C6 Corvette Z06 and Grand Sport. An independent current source lists the same pad. On coverage they agree exactly — Z06 2006–2013, Grand Sport 2010–2013. Those rows are verified.

Then there’s the 2011 Z06 Carbon:

Source A — manufacturer catalog

2011 Z06, Ceramic Composite Brake → 19206975

An OE-only part. Explicitly not HB532.

Source B — current retail data

2011 Corvette, Z06 Carbon → HB532Z.570

Listed as fitting, rear, 0.570 thickness.

rejected

Two credible sources, directly contradicting each other, on a car whose rotors are carbon-ceramic. We show neither until a human resolves it.

This is not a pedantic edge case. The 2011 Z06 Carbon doesn’t have fancier rotors — it inherited the ZR1’s entire brake system: different calipers, different rotors, a different pad shape. HB532 is not the wrong compound for that car; it is the wrong shape, and it will not fit the caliper. Fitting a standard pad against a carbon-ceramic rotor also destroys a rotor worth more than most people’s cars.

The lesson that justifies this whole site:trim names hide different hardware. Same year, same model, same 7.0L V8, both badged “Z06” — different brakes. Year, make and model cannot see that. It takes trim and package qualifiers, and it takes more than one source willing to say so.

05Why enthusiast consensus doesn't count as a source

We read the forums. They will never promote a row to verified, for three reasons.

It isn’t independent.Someone running a given pad on their car almost certainly bought it by looking it up in the manufacturer’s catalog. “The catalog says it fits” plus “a person who read the catalog says it fits” is one source, twice.

It measures the wrong thing.Reviews describe noise, dust, and bite. They do not describe fitment — and people who buy the wrong part return it rather than reviewing it. Positive sentiment is filtered by survivorship to systematically exclude the exact failure we care about.

Silence isn’t agreement. If a catalog gets a year split wrong, nobody posts about the part they never bought.

So community reports are asymmetric here, and deliberately so: they can flag a row for review— “these don’t fit my car, the catalog is wrong” is real information — but they can never confirm one.

06Staleness, and what it actually means

A catalog printed in 2016 describing a 2006 Corvette isn’t stale. That car stopped changing in 2013; the catalog has ten years of hindsight and corrections baked into it. Fitment for a discontinued platform doesn’t rot.

What a source’s date really gives you is a coverage boundary— it can’t tell you about cars built after it was printed. So we record the publication date of every source and only make claims inside its boundary. We would rather tell you we don’t have data for your car than extrapolate a year range and guess. Extrapolating across a generation is precisely how the wrong pads end up on someone’s car.

If you manufacture or distribute parts

We’d rather list your parts from your data than infer fitment from a public catalog. We attribute claims to you rather than asserting them ourselves, we carry your suitability caveats through to the page, and we don’t publish applications you never authorised. If you’d like your ACES application data represented here accurately, we’d like to hear from you.

rohan.skaria@gmail.com

Start a build · Parts currently shown are demo seed data while verified fitment is loaded.