GDSN / GS1: sync your product data with major retail chains

GDSN — Global Data Synchronisation Network. If you sell to major retailers (Carrefour, Leclerc, Auchan, Intermarché, etc.), you’ve probably come across requests for “GS1 synchronisation”. Here’s what that actually means, how the publication cycle works, and how your PIM interfaces with the GDSN network.

What is the GDSN?

The GDSN is a standardised global network for the exchange of product data between suppliers and retailers. It is based on standards defined by GS1, the international organisation that manages EAN/GTIN barcodes.

In practical terms, the GDSN enables a supplier to share their product data (specifications, packaging, images, logistics data) with their retailer customers in a standardised way — once, correctly, without exchanging Excel files via email.

The fundamental difference with a file drop on a server lies there: the GDSN does not merely carry data, it defines who publishes, who subscribes, what constitutes a unique record, and what reply the retailer must return. Major French retailers use data pools such as Agena3000 (Système U, Leclerc, Intermarché, Auchan, Carrefour) or Salsify to receive product data from their suppliers.

GDSN stakeholders

• GS1 — the organisation that defines the standards (GS1 France for France)
• Data pools — certified operators that store and synchronise data (Agena3000, Salsify, 1WorldSync, Syndigo…)
• Suppliers (Publishers) — publish their product data to a data pool
• Retailers (Subscribers) — subscribe to their suppliers’ data via the same data pool or via inter-data pool connections
• The global registry — the network’s central directory, which registers items and guarantees that a given record exists only once for a given supplier on a given market

The publication cycle: four messages

This is the part most presentations skip, and it is precisely what explains why a GDSN publication “doesn’t go through”. Four messages follow one another between the supplier, the data pools and the retailer.

• CIS — Catalogue Item Subscription: the retailer declares what it wants to receive. A subscription is defined by criteria (a supplier, a classification, a target market), not necessarily item by item.
• CIP — Catalogue Item Publication: the supplier declares to whom its items are made visible. A record published in a data pool is not automatically visible to all: publication is addressed.
• CIN — Catalogue Item Notification: the message that carries the data itself, in GS1 XML. That is the one your PIM produces.
• CIC — Catalogue Item Confirmation: the retailer’s reply. It is not binary — the item may be received, synchronised, put under review, or rejected with a reason. It is the most useful message in operations, and the most often ignored.

A supplier who does not read its CICs publishes blind: it believes it has listed five hundred items when some of them are under review at the retailer. Bringing those replies back into the product repository and treating them as anomalies is what separates a GDSN publication that holds from a one-off send.

The registration key: GTIN, GLN and target market

In the GDSN, a record is not identified by its GTIN alone. The triplet that makes it unique combines the item’s GTIN, the GLN of the information provider — the Global Location Number, the thirteen-digit identifier of the declaring entity — and the target market, expressed as a standardised country code.

The consequences are concrete. The same product sold in France and in Belgium corresponds to two distinct registrations, which may carry different statements and languages. A change of GLN — a merger, a legal reorganisation — is not fixed by an update: it creates new registrations. And two entities of the same group publishing the same GTIN on the same market conflict in the registry. Those three cases explain a great many unexplained blocks during a first listing.

Data exchanged via GDSN

The standard covers a very broad attribute set, organised around a product classification — GPC, Global Product Classification — which files each item in a hierarchy of segment, family, class and “brick”. The brick is not decorative: it determines which attributes are expected. A food brick calls for nutritional data and allergens; a power-tool brick, electrical characteristics. Picking the wrong brick means being asked for fields that make no sense for the product.

The attribute set mainly covers:

• Identification: GTIN (EAN), supplier GLN, brand, manufacturer reference
• Description: title, short description, long description
• Packaging: dimensions (L×W×H), weight, packaging type, order unit
• Packaging hierarchy: consumer unit → sales unit → pallet
• Images: product photos, technical drawings, packaging images
• Nutritional data (food)
• Regulatory data: customs codes, country of origin, allergens
• List prices and commercial terms (in some cases)

Those regulatory requirements largely overlap with the ones you already handle for REACH, RoHS, CE and WEEE compliance: it is the same data foundation, exposed in a different format.

Why it’s complex without PIM

The GDSN format is highly structured and demanding. Each attribute has a name, a type, a cardinality and validation rules defined by GS1. The main problems without PIM:

• Manual mapping — your internal product data does not have the same names as the GDSN fields. Manually mapping 200+ attributes for each retailer is time-consuming and prone to errors.
• Updates — when an attribute changes (weight, packaging, best-before date), you must republish to the data pool and notify all subscribed retailers.
• Multi-retailer — each retailer may have its own additional requirements beyond the GDSN standard (sector-specific attributes, additional images).

The validation rules, and why they reject

GS1 publishes a set of validation rules that data pools apply before accepting a record. They fall into three families, and it is useful to know which one is blocking you.

• Format rules — a country code that does not exist, a malformed date, a unit of measure outside the permitted list. These are the easiest to fix, and the ones a well-typed product repository eliminates at source.
• Dependency rules — an attribute becomes mandatory as soon as another is filled in, or as soon as the item belongs to a given GPC brick. These are the ones that surprise: the record was valid yesterday, it no longer is because the classification changed.
• Consistency rules — gross weight must exceed net weight, a case’s dimensions must be compatible with the number of units it holds, the levels of the hierarchy must add up. These are the most expensive to fix after the fact, because they expose wrong logistics data upstream, in the ERP.

The rules in force, their numbering and their evolutions are published by GS1 on gs1.org; data pools publish a reading of them adapted to their market.

How Pixee PIM interfaces with GDSN

Pixee PIM offers two GDSN integration modes depending on your situation:

Data pool export mode

If you are a supplier and need to publish to a data pool (Agena3000, Salsify, 1WorldSync):

1. You maintain your product data in the PIM — a single source repository
2. Pixee PIM generates GS1 XML files (CIN format — Catalogue Item Notification) compliant with the GDSN standard
3. The file is uploaded to your data pool via API or FTP (depending on the operator)
4. Subscribed retailers automatically receive updates

Import mode from data pool

If you are a retailer and receive GDSN data from your suppliers:

1. Pixee PIM connects to your data pool as a subscriber
2. New supplier publications are automatically imported into the PIM
3. GDSN data is mapped to your internal PIM data model
4. You enrich and complete the data before publishing it to your sales channels

GDSN and barcodes: understanding GTINs

The GTIN (Global Trade Item Number) is the central product identifier in the GDSN. A single product may have several GTINs depending on its packaging level:

• GTIN-13 (EAN-13): consumer unit
• GTIN-14: sales unit or pallet (with a different prefix)
• GTIN-8 (EAN-8): small products with limited space for the barcode

The hierarchy is not a display convenience: each level carries its own dimensions, its own weight and its own number of contained units, and those are the values the retailer’s warehouses work from. A case whose gross weight is declared at the net weight value passes format validation and skews the palletisation calculation. Pixee PIM manages that hierarchy and synchronises it to the data pool. Checking the codes themselves (check digit, duplicates, prefixes) is handled by the EAN Manager module.

What to prepare before the first publication

A first listing rarely fails on the technology: it fails because data nobody kept up to date suddenly becomes binding. Four points to settle before connecting anything:

1. The GLN of the publishing entity — it is requested from GS1 and must match the legal entity the retailer knows.
2. The GPC classification of each item — it drives the required attributes. Doing it once at product-family level costs less than correcting it item by item after rejection.
3. Measured logistics data — dimensions and weights per packaging level, measured rather than estimated. This is the area that produces the most consistency rejections.
4. Target markets — the list of countries where the item is sold, with the corresponding regulatory statements and languages.

Who needs to implement GDSN?

Major retailers are increasingly requiring their suppliers to adopt GDSN. If you sell or wish to sell to Carrefour, Leclerc, Intermarché or Système U, you will likely receive a request for GDSN listing as part of your supplier onboarding process.

For SME suppliers who do not yet have a PIM, this is often the trigger that justifies the implementation of a structured product repository — rather than managing the GDSN via Excel files. The same logic applies upstream, to the FTP and EDI feeds from your own suppliers.

Frequently asked questions

Do you need GDSN if you only sell to one retailer?

Not necessarily. A single retailer may accept a file drop or its own supplier portal, and a data pool subscription for a single recipient is rarely worth it. The trade-off tips at the second or third retailer: what costs then is the number of proprietary formats to maintain, not the data pool.

What happens when a value changes after publication?

You republish the record, and the network propagates the update to every subscriber. Two precautions: a change affecting the identification of the item calls for a new GTIN, not an update; and some retailers impose a contractual notice period before any change to logistics data. Republishing does not exempt you from warning them.

Does GDSN replace EDI?

No, the two coexist and do not handle the same thing. GDSN carries the reference product data — what the item is, how it is packaged, what it contains. EDI carries transactions: orders, despatch advices, invoices. A supplier listed in grocery retail generally uses both, with different teams and different timelines.

How long does a first listing take?

The lead time depends far less on the technology than on the state of your logistics data: connecting to the data pool and generating the messages is quick, collecting reliable dimensions and weights per packaging level is the real critical path. Start with a narrow batch of items until you get clean confirmations, then widen.