logoPandorLabs
High street & fast fashion

COS Product Data

Prices, variants, availability, and images from cos.com — normalised, validated, and delivered into your stack. No collector to build, and none to maintain.

source specification

Source
cos.com
Segment
High street & fast fashion
Origin
United Kingdom — launched 2007, H&M Group
Categories
Womenswear, Menswear, Shoes, Accessories
Formats
JSON · CSV · Parquet
Refresh
Continuous to daily

about the source

What COS sells, and how its catalogue behaves

COS — Collection of Style — was H&M Group's first move upmarket, and it defined the architectural minimalism that a decade of contemporary labels went on to copy. The assortment is deliberately small, heavily repeated season to season, and priced at a premium that the group protects carefully. Very little of it is ever deeply discounted.

why teams track it

Why COS data is worth having

COS is the reference point for minimalist contemporary pricing, and because its core styles persist across seasons its catalogue makes an unusually clean longitudinal series. Brands positioning in the same space use it to check whether their own price ladder reads as premium or as high street.

Visit cos.com

collection method

How we collect COS

COS is collected through COS's internal search API plus the Next.js data endpoints behind its product pages, so the structured payload is read directly. Every source gets a dedicated collector rather than a generic crawler, because the field detail that makes this data useful only survives if the extraction is built for the site it runs against.

Structured at the source

Records are read from cos.com's own structured responses wherever they exist, rather than reconstructed from page markup. That keeps the feed stable across visual redesigns.

Validated every run

Each field is checked against expected types and historical ranges. A collector producing anomalies is quarantined and repaired upstream instead of emitting bad prices into your pipeline.

Normalised to one schema

Every brand in the catalogue lands on the same schema, so adding a source is a configuration change on your side rather than another integration to write.

15min
Fastest refresh
99.9%
Uptime SLA
3
Delivery formats
49+
Fashion sources

what you get back

Fields in the COS feed

A normalised core that is identical across every source, plus the attributes that are specific to this one.

Standard across every source

  • Product name, brand, and source URL
  • Current price, original price, and currency
  • Category and subcategory as the source classifies them
  • Colour, size, and variant availability
  • Product images, deduplicated across variants
  • Description, composition, and care text
  • Collection timestamp on every record

Specific to COS

  • Stable product identifiers across seasonal carry-over styles
  • Composition and care detail from the structured page data
  • Colourway variants with per-colour image sets
  • Category taxonomy as COS classifies it

SOC 2 Type II

Audited controls across security, availability, and confidentiality. Report available under NDA.

GDPR & CCPA

Public catalogue data only. No personal data collected, and a DPA is available on request.

99.9% Uptime SLA

Contractual availability with monitored collectors and a public status page.

Data residency

Choose EU or US processing and storage regions to match your obligations.

questions

Frequently Asked Questions

Everything you need to know before you send us your first request.

They answer different questions. H&M tells you where the volume market sits; COS tells you what the same group charges when it removes price from the pitch. Customers benchmarking premium basics almost always want COS specifically, because H&M pricing is not a useful comparison at that tier.
Yes, and it is one of the more useful properties of this source. COS repeats core styles year after year under stable identifiers, so a collection run on a regular cadence produces a genuine multi-year price and availability series rather than a set of disconnected seasonal snapshots.
Into the systems you already run, rather than through an API you have to integrate against. Normalised records land as JSON, CSV, or Parquet in S3, GCS, or Azure Blob, or straight into Snowflake or BigQuery, on whatever cadence you set. Webhooks can push price and availability changes as they are detected. A solutions engineer fixes the schema, cadence, and destination with you during onboarding, so the first delivery already matches your pipeline.
Maintaining the collector is our job, not yours. Every field is validated against expected types and historical ranges on each run, and a collector that starts producing anomalies is quarantined rather than allowed to emit bad records. We repair it upstream, and the schema we deliver to you does not move — which is the entire reason to buy this rather than run a scraper in-house.
We collect only publicly visible catalogue data — the prices, descriptions, images, and availability any shopper sees without logging in. We do not bypass authentication, we do not collect personal data, and we honour rate limits so the source is never disrupted. Our infrastructure is SOC 2 Type II certified, our processing is GDPR and CCPA compliant, and we sign DPAs as part of procurement.
Pricing is scoped per engagement, driven by catalogue size, refresh cadence, the number of locales you need, and the delivery destinations involved. Historical backfill is quoted separately. Every engagement starts with a free sample pulled from the live cos.com catalogue, so you can check the data against your own benchmarks before committing to anything.

Still have questions?

Talk to an engineer

Ready to Get Started?

Talk to us about your sources and volume. We'll return a sample dataset from your target sites before you commit to anything.

SOC 2 Type II
GDPR & CCPA compliant
99.9% uptime SLA

© 2026 PandorLabs, Inc. All rights reserved.