Resolution Layer Case Study · Read & Diagnose
A national foodservice-disposables manufacturer.
What a content-and-correctness read reveals — and where it leads.
This is a real EKOM catalog analysis with the client's identity removed — a national foodservice-disposables manufacturer (cutlery, napkins, tableware and bakery across house and licensed brands), read with no schema supplied. The headline isn't that fields are missing — most are filled. The point is that populated isn't the same as accurate: the catalog passes a completeness check while carrying high-consequence content defects — a certified-compostable line described as plastic, internal notes in buyer-facing copy, size data borrowed from the wrong product, and pack counts that contradict their own descriptions.
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 the source system. Even read conservatively, the live critical items alone warrant action.
What's inside
- The findings that carry real risk — the live sustainability-claim, buyer-content, and accuracy defects most likely to cost trust or a sale.
- Why these were invisible — how a field can be full and still wrong, and why that slips past every completeness check.
- How EKOM reads a catalog — profiling with no schema, validating fields against each other, and reading buyer-facing text for intent.
- What it means, and what's next — from diagnosis to an enriched, channel-ready catalog.
Standard validation checks whether a field is populated, not whether its value is right for that product. EKOM's resolution layer reads fields against each other and reads buyer-facing copy for intent — which is why these surface here rather than reaching a shopper.
The weight is in consequence, not count: a certified-compostable line is telling shoppers it's plastic. Many issues span more than one theme, so theme counts are approximate.
6
Certified items
mislabeled
9
Internal notes in
buyer copy
By finding area
| Area | ~Items | Who feels it |
|---|
| Certification vs. description conflict (compostable labeled plastic) | 6 | Shoppers + brand trust |
| Internal operations content in buyer-facing fields | 9 | Shoppers + brand |
| Misrouted sustainability claims & spelling errors | ~15 | Shoppers + search |
| Size data from the wrong product class | 5 | Shoppers + returns |
| Pack counts that contradict the item description | 3 | Ops + buyers |
| Dormant QA checkpoint & missing structured columns | catalog-wide | Ops + channels |
The findings that carry real risk.
The issues most likely to cost trust, a sale, or a return — live in the catalog now.
Live sustainability-claim conflict
The sustainability story is contradicting itself.
Six BPI-certified compostable cutlery items carry the word plastic in their buyer-facing descriptions — while the certification flag on every one reads BPI Certified = Y. A buyer filtering for compostable products gets zero results for items that literally qualify, and a buyer who lands on one reads a description that undercuts the certification. Two fields that must agree — the cert flag and the description — say opposite things.
Both fields are populated, so a completeness check passes. The conflict only surfaces when the two fields are read against each other.
Internal notes in buyer-facing fields
Content never meant for a shopper, sitting in shopper-facing copy. Nine "Extra Description" values contain internal operations content — a Pantone color-change schedule (PMS-209C will change to PMS-221U), repack instructions (repack from …), a retailer stock allocation, and internal reference codes. Each is a field a channel can publish straight to a product page — so the manufacturer's back-office notes become customer-facing text.
Wrong-product data & contradictions
Fields filled with the wrong content, not left blank. Five items carry dimensions from a different product class — a straw reading a napkin's measurements, a placemat reading a straw's length — so the size a buyer sees belongs to another product. And three items show pack counts that contradict their own descriptions (a description reading one quantity while the structured count says another, off by as much as 10×), throwing off pricing, shipping, and buyer expectations.
Why these were invisible.
Every one of these fields is full. The defect is in the content, not the coverage — which is exactly what a completeness check can't see.
The full-but-wrong effect
The catalog's own quality checkpoint was never run.
The catalog carries a built-in QA mechanism — a set of "Feature Confirmed" columns meant to record that each feature was verified. Across every row in the catalog they are 100% blank: the checkpoint exists but was never used. Meanwhile sustainability claims are routed to the wrong field — roughly 15 items carry a sustainability statement in the Occasion field while the Sustainability field sits empty — and spelling errors like Biodegradeable (10×) and Traditioinal (3×) quietly break search and filtering. None of these lower a completeness score; every field involved is populated.
A field that's full but wrong is worse than one that's empty — because nothing flags it, and it publishes with confidence.
This is the difference between checking a catalog and understanding one. A completeness check confirms a field is filled; it can't tell that the description contradicts the certification, that the dimensions belong to another product, or that a buyer is reading a back-office note. Reading fields against each other — and reading buyer copy for intent — is what makes both the defect and the fix legible.
How EKOM reads this catalog.
Why content problems surface here rather than passing a completeness check untouched.
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 recognizing the data as foodservice disposables.
No manual setup required.
→
2
Validate fields against each other
It compared the fields that must agree — certification flag vs. description, dimensions vs. product class, pack count vs. description math — so a full-but-contradictory field is caught by the field it disagrees with.
Correctness, not just presence.
→
3
Read buyer copy for intent
It read every buyer-facing value for whether it belongs to a buyer at all — catching internal ops notes, misrouted claims, and spelling errors that break search.
Does this text belong to a shopper?
Reading fields in the context of each other is how a "plastic" description gets caught by the compostable certification next to it, instead of publishing as-is. It's also what makes the next step precise: because EKOM knows which fields disagree and why, the correction targets the exact value that's wrong — not a blanket rewrite.
73%
Nearly three in four shoppers say they'd think less of a brand after finding incomplete or inaccurate product information — and for a manufacturer, the catalog is where that impression forms, at scale, across every distributor and marketplace that reads it. A certified-compostable line that reads "plastic," a back-office note on a product page, a spec that doesn't add up: each is a small crack in trust, published everywhere the catalog feeds.
Syndigo, State of Product Content, 2024
The consequential defects here aren't gaps — they're full fields carrying the wrong content. A certified-compostable line described as plastic, a straw wearing a napkin's dimensions, an internal repack note in buyer copy: each passes a completeness check and each misinforms a buyer. That's not a sign of a neglected catalog; it's the nature of one where every field got filled but no one could read every field against every other one.
None of these are coverage problems. A compostable fork labeled plastic never appears in a sustainability-filtered search no matter how complete its record. A pack count that contradicts its description breaks pricing and trust. A misrouted claim sits in a field no shopper filters on. Standard validation checks for presence; EKOM's resolution reads for correctness — validating each field against the others and against the buyer it's written for.
This pass read and diagnosed. The next step is where the catalog gets better, not just fuller.
From diagnosis to channel-ready.
The read in this report is the first phase. The same field-against-field understanding that surfaced these findings is what powers the work that follows — turning a diagnosed catalog into one that wins placement across every channel.
1 · Normalize what's wrong or misplaced
Align the compostable descriptions with their certification, move the internal ops notes out of buyer-facing fields, correct the wrong-product dimensions and the contradictory pack counts, route sustainability claims to the right field, and fix the spelling errors that break search.
2 · Enrich for discovery & compliance
Fill the missing structured columns channels depend on — unit of measure, case-pack quantity, GTIN/UPC — and stand the dormant "Feature Confirmed" checkpoint back up, so products pass validation and surface in the sustainability and pack-size facets buyers use.
3 · Stay channel-ready as the catalog grows
Ongoing catalog resolution so new SKUs and licensed-brand additions stay accurate, on-claim, and syndication-ready — the understanding compounds instead of eroding with every new item.
This is how EKOM moves a catalog from insight to impact —
and from a full catalog to a channel-winning one.
About this analysis
This read required no schema and no manual setup: the pipeline profiled the catalog on its own, validated fields against each other, and read buyer-facing copy for intent. The same read runs continuously as the catalog changes, so accuracy holds as new items and licensed brands are added.