PRODUCT DATA AS AN ONGOING SERVICE
Complete.
And kept complete.
Skuply fills the missing attributes of webshops from 5,000 products up, writes the copy that follows from them, and keeps it filled with every new SKU. Connected through the API, fixed amount per month.
No login needed for the measurement. We read what sits publicly on your product pages.
MATERIAL
full-grain leather · 0.94
EN STANDARD
EN ISO 20345 S3 · 0.97
SOLE
rubber, slip resistant · 0.88
GTIN
no match · stays empty
34% FILLED
93% FILLED
ONE PRODUCT, WORKED OUT
A work boot as it came in, and as it goes out.
On the left is what the supplier sent along: 34 percent of the fields filled. On the right is the same boot after one round: 93 percent. Every field that gets added carries a source and a number that says how direct that source is.
MATERIAL · 0.94
Full-grain leather. Taken literally from the manufacturer datasheet, the strongest source there is.
EN STANDARD · 0.97
EN ISO 20345 S3, normalised against the EN table. That table says one thing only, so the number is high.
SOLE · 0.88
Rubber, slip resistant. Derived from your own title and the standard. A derivation scores lower than a literal source.
GTIN · no match
Not in the supplier feed, not matched in a manufacturer database. Then the field stays empty.
Below 0.85 nothing goes live; it goes to the review list with the source attached. A GTIN is never generated, only matched.
THE SCORE
One number, and four places where it makes a difference straight away.
CHANNEL
WHAT IT GIVES YOU
Your own search
Anyone filtering on leather now actually sees the product.
Google Shopping
Attributes are matching input. Missing required fields get suppressed.
Marketplaces
Amazon and eBay rank partly on data completeness.
AI assistants
An assistant reasons about properties. Without attributes there is nothing to derive, so the product is never mentioned. These days that is called GEO and it is still open ground.
To be straight about it: organic search traffic is a portfolio bet, not a promise per product. Of 10,000 enriched products, 10,000 do not climb.
So at the start we record a holdout: 2,000 products stay untouched, so the difference can be measured.
COMPLETENESS SCORE
EXAMPLE91%
was 34%
AFTER THE FIRST PASS
PER ATTRIBUTE
material
96%
size
99%
colour
94%
en standard
88%
gtin
61%
12,400 SKUS · 15 ATTRIBUTES · MEASURED 20 AUG 2026
THE PROBLEM, IN NUMBERS
An empty attribute is not cosmetic. The product does not exist in the place where the buyer is looking.
38%
of the attributes are empty in a catalogue built up over years by different people
5
spellings of the same size, inside the same category: 43, 43 EU, size 43, EU43, 43.0
150,000
manual entries to fill 10,000 products with 15 fields once
0
chance that a product with an empty field appears in a filtered search
GLOVE, 3 OF 15 FIELDS
WORK TROUSERS, 4 OF 15 FIELDS
WORK JACKET, 2 OF 15 FIELDS
This is what a product with empty fields looks like to a filter, a marketplace or an assistant. Recognisable enough to exist, too empty to be chosen.
WHAT WE FILL
Three zones, and one line that is fixed.
Attributes
Derived from datasheets, titles and photos, with source and confidence attached.
material, colour, size, dimensions, weight
the Google product category as a taxonomy ID
variant normalisation into one term
alt text from image recognition
Copy
Derived from the attributes. If the attribute is not there, it is not in the copy.
short and long descriptions from the fields
USP lines and long-tail question answering
Schema.org markup per attribute
Never invented
Below the threshold nothing goes live. A GTIN is never generated, only matched.
GTIN and EAN are only matched
no match means: the field stays empty
price, stock and warranty stay untouched
The copy is a consequence of the data, not the other way round. So a description never states a property that cannot be checked in a field.
PRICE
One package, no variants. The meter is the number of SKUs.
One-off first pass: € 0.10 per SKU. All amounts are exclusive of VAT. At 12,400 products that is € 1,240 once and € 499 per month.
THE METER
The tier counts the number of SKUs in your sitemap, not the number of variants. Grow past a boundary and the rate follows on the first of the next month.
You do not pick a tier. You connect your shop, we count the SKUs in your sitemap, and the rate follows from that. Want to see what there is to gain first? Measure without connecting.
START BY MEASURING
Put your address in. Within a minute you see how many fields are empty, per category.
No name, no email, no login. We read your product pages the way any visitor can, and send the outcome as a link you can share.