Product Data for Bedding: Sizes, Fillings and Sets

Duvets, pillows and toppers live on measurements, fill weights and warmth classes — and they sell as sets. Here's how to model those attributes and the set logic in one system, and where the standards stop.

Jakob Feinböck, ProductbayJuly 4, 20267 min read
☝️Key takeaways
  • Bedding buys on three attributes: size, fill weight and warmth class — and each supplier labels them differently.
  • The retail hero is the set: a duvet plus matching pillow(s) sold as one bundle with its own EAN/GTIN.
  • No content standard fills warmth logic, fill data or descriptions — GDSN/ETIM only give you a category skeleton.
  • Productbay normalises measurements and fillings, models sets as linked attributes, and enriches the rest with AI.

A customer shopping for a duvet has already narrowed the decision to three numbers before they even read the product description: what size fits their bed, how warm they want it, and what it is filled with. Get any of those attributes wrong or missing, and the product simply doesn't get found — or gets returned. Bedding is one of those categories where the data is the product.

Product data for bedding is built on three structured attributes — size, fill weight and warmth class — plus set logic that binds a duvet to its matching pillows. This is a sub-category of the broader home textiles challenge, and it has a data profile all its own: heavily attribute-driven, variant-rich, and reliant on bundles that no supplier feed delivers cleanly.

Which attributes make bedding data hard to maintain?

Unlike a printed cushion cover, a duvet or pillow sells on measurable specs. The trouble is that every supplier records them differently:

  • Size: a duvet exists in 135x200, 155x220, 200x200, 200x220 and more; pillows in 40x80, 80x80, 40x40. Some feeds write "135 x 200 cm", some "135/200", some split into two columns.
  • Fill weight and warmth class: the same duvet ships as light, medium, warm or extra warm — driven by fill weight in grams. One supplier puts the warmth class in the title, another only in a PDF datasheet.
  • Filling material: down, feather, down-feather blend, microfibre, camel hair, virgin wool — a core filter that arrives as free text.
  • Care and cover: washable temperature, cover material (cotton, batiste, microfibre), certifications like Oeko-Tex land inconsistently or not at all.

Every one of these is a filter a customer uses. Normalising them into clean, comparable attributes across a dozen suppliers is exactly the manual work that doesn't scale — the same consolidate, normalize and enrich job every multi-supplier retailer faces.

How do you model a bedding set of duvet plus pillow?

The hero product in bedding is rarely a single item — it's the set: a duvet plus one or two matching pillows, sold as one bundle at a set price. That's where a flat product feed breaks down, because a set is not one row of data, it's several products stitched together:

  • The set needs its own EAN/GTIN, price and hero images, distinct from the components.
  • It has to reference its components so size, filling and stock stay in sync — a 155x220 duvet with an 80x80 pillow.
  • The same components often appear in several sets and as standalone articles, so the data can't just be duplicated.

Modelling this as linked attributes — rather than re-keying every spec into a flat bundle row — is what keeps a growing set catalog maintainable. Change the duvet's warmth class once, and every set that references it stays correct.

Which standards apply — and where do they stop?

Bedding retailers sometimes hope a data standard will fill the gap. Here is the honest picture of what the common ones do and don't deliver:

Data layerWhat the standard deliversWhere it stops
IdentifiersGTIN/EAN via GS1, GDSN for master data exchangeNo attributes, no content, no set logic
ClassificationETIM / eCl@ss give a category skeletonNot built for warmth class or fill weight
AttributesPartial, supplier-dependent in feedsSize, fill weight, warmth class rarely clean
Sales contentNot the job of any standardDescriptions, benefit copy, SEO text absent
Set / bundleNo standard models duvet-plus-pillow setsBuilt entirely by the retailer

In short: there is no content standard for bedding. The identifiers and classifications give you a scaffold, but the measurements, fill logic, set structure and the actual sales copy are all yours to build. That's the gap Productbay is designed to close.

How does Productbay help bedding retailers?

The job is the same three steps every multi-supplier retailer runs — and Productbay is built to run them across the specific attribute logic of bedding:

  • Consolidate: import every supplier source once — CSV, Excel, feed URL, FTP, API — and match by SKU or EAN/GTIN so existing products update and new ones are created.
  • Enrich: AI normalises sizes into a clean scheme, derives warmth class from fill weight, standardises filling materials, writes descriptions, assigns categories, translates via DeepL and can read specs out of PDF datasheets — always with a review queue before anything publishes.
  • Publish: two-way sync to Shopify and Shopware, ERP connections (Xentral, weclapp) and feed exports for Amazon, OTTO and Kaufland — each with per-channel transformations, sets included.

The set logic is where Productbay earns its keep in bedding: it models sets as linked attributes, so a duvet, its pillow and the bundle stay consistent without double maintenance. For the wider category context, see product data in home textiles. Productbay is built for specialist retailers running multi-supplier, multi-channel catalogs — from mid-sized shops to large chains.

Frequently Asked Questions

Let's look at your product data process

Sizes, fill weights, warmth classes and duvet-plus-pillow sets — bedding lives on structured attributes no standard fills for you. See how Productbay normalises fillings, models sets as linked attributes and enriches the rest in a 30-minute walkthrough.

Get started