Resolution Layer Case Study  ·  Read & Diagnose

A performance-automotive parts distributor.

A supplier-by-supplier read of the parts catalog — and where it leads.

Scope
Eight supplier lines  ·  turbo, diesel, performance
Live critical findings
26  ·  wrong-part, compliance, fitment
Total issues surfaced
81 distinct
Method
8 per-supplier passes  ·  split auto-selected
Vertical
Automotive Parts — Turbo, Diesel, Performance
This pass
Read & diagnose — step one of the arc
Prepared by
EKOM
Type
Case study — client anonymized

What this is.

This is a real EKOM catalog analysis, with the client's identity removed. The client is a performance-automotive parts distributor — turbo, diesel, and performance parts sourced from multiple supplier feeds. EKOM read the catalog the way a distributor catalog is actually built: supplier by supplier. Every product shares the same fields, but completeness and quality vary by the supplier whose feed each product came from. Reading each brand's feed on its own surfaced 81 distinct data-quality issues — and, more importantly, the systemic ones: whole-feed gaps a blended scan would scatter into noise.

These are machine-surfaced findings meant as a triage signal, not a verified defect list — a small share may be intentional or may resolve against source data. Even read conservatively, the live critical items — wrong-part risk, compliance gaps, fitment blind spots — warrant action.

What's inside

Standard catalog validation checks whether a field is populated, not whether its value is correct. A turbo can pass every completeness check while its description names the wrong engine, its barcode doesn't match its part number, or its Prop 65 status reads "Unknown." EKOM's resolution layer reads structure and meaning together, supplier by supplier — which is why these surface here rather than reaching a buyer or a marketplace feed.

At a glance.

The weight is in severity: 26 issues are critical and live now. Many issues span more than one theme, so theme counts are approximate.

26
Critical
& live
32
High
severity
81
Distinct
issues
8
Supplier
lines

By supplier line

SupplierFindingsRepresentative issue
HKS14Fitment IDs blank on every item; Prop 65 "Unknown" across the line.
BD Diesel13New active items missing stock status, fitment, description, and Prop 65.
BorgWarner11Part numbers in the barcode field; marketing copy blank across the line.
Forced Performance10Barcode ≠ part number (transposition); Prop 65 "Unknown" on every item.
ACL10Bearings whose descriptions name the wrong engine; missing barcodes.
Garrett827 assembly kits with malformed IDs and no born-on date.
Wehrli8Two parts with descriptions exactly swapped; a Steel part called "Aluminum."
Industrial Injection7Fitment IDs missing on 107 of 110 items.

By theme

Theme~CountWho feels it
Miscategorization (parts filed under the wrong type)~15Buyers
Compliance gaps (Prop 65 "Unknown", missing CARB EO)~14Compliance + channels
Malformed / mismatched identifiers (barcode ≠ part #)~12Ops + channels
Fitment blind spots (missing vehicle fitment IDs)~12Buyers
Wrong, swapped, or wrong-engine descriptions~10Buyers
Missing barcodes / EDI syndication blockers (ACES/PIES)~10Retail channels
Missing or placeholder content & images~8Buyers

The findings that carry real risk.

The issues most likely to cost a return, a rejected feed, or compliance exposure — live in the catalog now.

Wrong-part risk
Descriptions that will send a buyer the wrong part.
Two Wehrli Turbo Pedestal items have their descriptions exactly swapped — each names the engine family belonging to the other. Several ACL bearings carry long copy naming the wrong engine. A Wehrli LML Billet Steel flange opens by calling itself Aluminum. In a fitment-driven category, a wrong-engine description isn't a typo — it's a return, and a dented reputation with the installer who ordered it.
None of these fail a completeness check — every field is filled. They only surface when the description is read against the part's own attributes.

Compliance exposure, by the whole feed

Prop 65 reads Unknown on every item for Forced Performance, HKS, Industrial Injection, and BD Diesel's new items — non-compliant for California sale and grounds for rejection by major marketplaces. CARB-approved items across HKS and Garrett are missing the required CARB EO number, and some catback items are flagged both CARB-approved and requiring acknowledgement — a contradiction on a regulated part.

Invisible to fitment search

Vehicle fitment IDs are blank for all HKS items and 107 of 110 Industrial Injection items — so they never appear when a buyer searches by year, make, and model, which is how auto parts are actually shopped. The part is in the catalog, in stock, correctly priced — and unfindable by the person who needs it.

Identifiers that break receiving and EDI

A Forced Performance turbo's barcode doesn't match its part number (a two-digit transposition, repeated on a related drop-ship item); several brands store the MFR part number in the barcode field instead of a valid UPC; 27 Garrett kits carry malformed "G"-suffix IDs; and Autocare Brand IDs are blank across Forced Performance — blocking ACES/PIES EDI syndication.

Why these were invisible.

In a distributor catalog, the most serious problems are systemic to a single supplier's feed — and a blended scan hides exactly that.

The per-supplier effect
The worst issues aren't scattered — they're whole-feed, one supplier at a time.
Because every product shares the same fields, a single blended pass averages the catalog toward "mostly fine" and buries the pattern. Reading each supplier's feed on its own exposes it: every HKS item missing fitment; an entire brand's Prop 65 statuses "Unknown"; 27 Garrett kits with the same malformed ID and missing born-on date; a whole ACL line of bearings without barcodes. These aren't 81 random errors — they're a handful of systemic feed gaps, each repeating across a supplier's products.
That's the difference between "clean up some records" and "fix how this feed comes in." The systemic ones are the cheapest to fix and the most expensive to leave.

It's also what makes the fix precise. Once you can see that a defect is a whole-feed pattern rather than a scattering of one-offs, the correction is a rule applied to that supplier's intake — not a manual pass over hundreds of records. Reading supplier by supplier is what makes both the pattern and its remedy legible.

How EKOM reads this catalog.

Why supplier-specific problems surface here rather than averaging away in a whole-catalog scan.

1
Profile the catalog
EKOM profiled the catalog with no schema supplied — inferring the vertical and every field's role, fill rate, and quality signals, and detecting that quality tracks with the supplier each product came from.
No manual setup required.
2
Partition by supplier
The catalog's own structure pointed to the split: partition by supplier brand and read each feed in full, so whole-feed gaps surface instead of averaging out.
The catalog chose its own split.
3
De-duplicate & group
Findings from every supplier are unioned onto the single catalog, duplicates removed, then grouped by theme and by who feels it.
Signal, not noise.

A single blended scan of a distributor catalog reports "mostly complete" and misses that one supplier's entire feed has no fitment. Reading each feed in full is how a whole-brand Prop 65 gap, or 27 kits sharing a malformed ID, gets caught instead of buried — and it's what makes the next step precise: EKOM knows exactly which fields each supplier's feed is missing, so the enrichment targets the right gaps.

What this means.

more than 25%
More than one in four organizations facing data-quality challenges report losing over $5 million annually as a result — and for an automotive-parts catalog, that figure doesn't capture the channel cost: every fitment search, marketplace feed, and EDI validation gate relies on clean structured data. Catalog defects don't announce themselves; they surface as wrong-part returns, products missing from year/make/model search, and feeds rejected before they reach the shelf.
Forrester Data Culture and Literacy Survey, 2023 (RES181258)
The consequential defects here concentrate by supplier feed — which is both the risk and the opportunity. A whole-feed fitment gap or an entire brand's "Unknown" Prop 65 status is a systemic hole; but because it's systemic, it's fixable as a rule at intake rather than a manual crawl through records. That's leverage a blended, record-by-record view never exposes.
None of these are content problems. An HKS part with no fitment IDs never appears in a year/make/model search no matter how good its copy. A "Unknown" Prop 65 status is a compliance and marketplace-rejection risk no description can fix. A barcode that doesn't match its part number breaks receiving and EDI. Standard validation checks for presence; EKOM's resolution reads for correctness — reading each supplier's feed in full.
This pass read and diagnosed. The next step is where the catalog gets better, not just cleaner.

From diagnosis to channel-ready.

The read in this report is the first phase. The same supplier-by-supplier understanding that surfaced these findings is what powers the work that follows — turning a diagnosed catalog into one that sells across fitment search, marketplaces, and EDI without friction.

1  ·  Complete fitment & compliance
Fill vehicle fitment (year/make/model) where feeds are blank, resolve "Unknown" Prop 65 and missing CARB EO numbers, and reconcile the wrong-engine and swapped descriptions — so parts are findable, compliant, and correct at the point of sale.
2  ·  Standardize for the channel
Normalize identifiers (barcodes matched to part numbers, valid UPCs, clean IDs) and complete ACES/PIES brand and category data — so every supplier feed passes marketplace and EDI validation the first time.
3  ·  Hold the line at intake
Because the defects are systemic to each supplier's feed, ongoing intelligence can catch them as products arrive — so new brands and new SKUs come in clean and channel-ready instead of accumulating the same gaps.
This is how EKOM moves a catalog from insight to impact —
and from a diagnosed catalog to a channel-winning one.
EKOM
The Resolution Layer
ekom.ai
About this analysis

This is EKOM's second pass on the same automotive-parts catalog first analyzed earlier. The input did not change — the method did. The first pass read the catalog against the target data architecture it was meant to feed — checking each record against itself and against the fitment/PIES schema, which the analysis needed as context. This pass required no schema and no manual setup: the pipeline profiled the catalog on its own, recognized that quality tracks with each product's supplier feed, chose to partition by supplier, and read each feed in full context — surfacing 81 distinct, severity-ranked defects across eight supplier lines. Same catalog, read without a schema and organized the way it's actually built, with nothing configured by hand.

Case Study  —  Client anonymized  ·  The Resolution Layer
EKOM