WebGoodPeople

← All guides

Auto parts search worksheet · 8 min

Auto Parts Search Pilot: Fitment and SKU Test Checklist

A parts search result can look plausible and still be wrong. Use this pilot to check identifier matching, replacement relationships and vehicle fitment against your own catalog evidence before approving a search implementation.

8 min read · Updated: 9 November 2026

Prepare a trustworthy sample

Ask a catalog specialist to select 20–30 real parts from your store: common sellers, similar codes, superseded numbers, items with fitment qualifiers and products with incomplete data. Keep the source part ID, brand, code, category, current stock and the evidence for each vehicle application. Use a staging catalog if the production feed changes while you test. Record its export date and the search index date.

Separate an exact identifier match, a documented interchange and a verified vehicle application. They answer different questions. A part can be discoverable by a related number without being confirmed to fit the buyer's vehicle.

Data boundary: a VIN decoder identifies vehicle attributes; it does not, by itself, prove that a particular SKU fits. The US NHTSA vPIC API documents vehicle specifications. Your application still needs a licensed or otherwise verified part-to-vehicle relationship.

Run the parts query matrix

Fill the worksheet with your own part numbers and vehicle examples. For each row, save the result IDs and evidence shown to a shopper, not only a pass/fail flag. Run the same set on desktop and mobile where those interfaces differ.

CaseExample query typeWhat to inspect
Exact codeFull OEM or supplier numberRequested item first; code and brand visible
Code variantSame code with spaces or separators changedSame item, with normalization recorded
Near collisionA code differing by one characterNever silently substitute the other part
InterchangeKnown replacement numberRelationship and its source are visible
Vehicle + partYear, make, model and part typeOnly verified applications presented as compatible
QualifierA vehicle with an engine or trim exceptionQualifier retained; uncertain fitment marked clearly
Stock changeKnown item after an availability updateShown stock matches the agreed source and delay
No fitmentA vehicle with no verified matching partNo invented compatible result

Also test a deleted SKU, an out-of-stock part, a brand spelling variant and a generic category query. For a code with punctuation, compare exact input, spaces removed and separators changed. Preserve the original query in the test record so a normalization bug is reproducible.

Verify fitment separately

For every vehicle result, inspect the underlying application record: year, make, model, engine, position and qualifiers as applicable to that product. Ask the catalog owner to resolve ambiguous and negative cases. Do not convert “possible match” into a confirmed-fit badge to make the pilot look better.

In North American aftermarket catalogs, ACES and PIES provide standardized exchange formats for fitment and product data. The Auto Care Association explicitly says that ACES and its vehicle database do not create your product's fitment data: you must research, create or obtain it. Check the source, coverage, license and update cycle before proposing a year/make/model or VIN-led interface.

Set acceptance rules

Agree on rules before viewing the results. The numbers below are suggested pilot gates, not measured performance or a universal SLA. Adjust the speed and coverage thresholds to your catalog and device mix, but keep the safety rules explicit.

  • All selected exact-code queries return the requested item first.
  • No near-collision or missing-fitment case is labelled as confirmed fit.
  • Every interchange shown has a traceable source and relationship type.
  • Stock and price reflect the agreed source within the signed-off update window.
  • Measure response time on the same device and network; record the median and slowest case.
  • Catalog specialists sign off on the false positives, not just the search team.

Keep two failure lists: search behavior (normalization, ranking, filters or stale index) and catalog data (missing interchange, fitment or qualifier). This distinction determines who can fix the issue and what a vendor can responsibly promise.

Download the worksheet

Download the auto parts search pilot CSV. It contains 16 test cases and columns for expected IDs, fitment evidence, observed results, index age, owner and decision. Replace each placeholder query with a real catalog example. Review the sheet with the person responsible for applications data.

Turn findings into a build scope

Send the completed worksheet, a small catalog sample and your platform details to the team implementing search. Mark which requirements are simple identifier matching, which require catalog cleanup, and which need a licensed fitment feed or custom interface. Our auto parts search page explains how we evaluate those requirements during a catalog pilot. For a cross-industry baseline, use the general site search pilot guide.