Product data enrichment: from EAN to complete product

What is product data enrichment?

Product data enrichment, or product enrichment for short, means completing the product information a channel needs to show your product well: a clear product name, a long and a short description, attributes, images and a category. The goal is simple. A product should be complete on every channel where you sell, whether that is your own webshop or a marketplace.

Why supplier and POS data is often incomplete

Most product data arrives from a supplier, a wholesaler or a point of sale system. Those sources are built for purchasing and stock, not for selling. That is why you often see:

• Short or abbreviated names, such as "PANNIER BLK 20L" instead of a readable name for a bike bag.
• No description, or one line that is the same for a hundred products.
• No images, or just a small thumbnail.
• No category that fits your own structure.

Filling those gaps by hand easily takes minutes per product, and with thousands of items that is not realistic. So enrichment starts from something almost every product does have: the EAN. Read more about that number in what an EAN code is, and about where enrichment fits in your product data at manage product data.

How product enrichment works from the EAN

In StoreLinkr every enrichment starts from the product's EAN. A product without an EAN cannot be enriched, and the EAN itself is never changed. StoreLinkr looks in this order:

1. The product catalogue. A large catalogue of product data by EAN: name, description, images and category.
2. Google Books for books. If the code starts with 978 or 979, it is an ISBN and the data comes from Google Books.
3. Your own import data. What your supplier or POS already delivers counts as a source.
4. Your own AI. With your own key for Gemini, OpenAI or Claude, the AI rewrites the texts in your tone. How that works is explained in AI product descriptions.

Then you get a preview with the old and the new value per field. You accept field by field. Text the AI wrote is marked as AI generated, so you can still refuse it.

Per field: source and control

The texts follow one fixed rule. If there is a source text, the AI rewrites it faithfully: every fact stays, no new claims are added. If there is no description anywhere, the AI only writes one when you switch that on, and then only from what is known, such as title, brand, attributes and category. What else a good text needs is covered in how to write product descriptions and SEO product descriptions.

Imported products and your own PIM products

Not every product comes in the same way, and that matters for what happens after enrichment.

Products from a supplier, wholesaler or POS

An imported product is updated with every new import. Without a safeguard, the next import would replace your new text with the supplier's bare text again. That is why StoreLinkr stores accepted texts as product rules. The next import then does not overwrite them. Attributes, weight and dimensions stay as the supplier sends them: the supplier knows best how heavy a coffee grinder is or how big a box is.

Products you create yourself in the PIM

A product you create yourself in the PIM has no supplier that overwrites it. What you accept is simply stored on the product. The same rule applies here: no EAN, no enrichment, so give your own products an EAN if they have one. What a PIM does exactly is explained in what a PIM system is.

Automatic and bulk enrichment

New products from an import

You choose from which import sources new products are enriched automatically. When such an import creates a new product, it is enriched once. Only empty fields are filled: by default the name, the descriptions, the category and the images. Whatever is already there stays.

Enrich first, then to the shop

Do not want a bare product to go live for a while? Then switch on "enrich first, then to the shop". A new product then waits for its enrichment before it goes to your sales channels, such as WooCommerce, Shopify, SitePack or a marketplace like Kaufland. That wait is at most 15 minutes after the import.

Existing products in bulk

For your existing range you select products in the product overview and enrich them in one go. Start with a small group, for example twenty children's books or one brand of coffee grinders, and check what happens before you do the rest.

The enrichment log and reverting

Every enrichment ends up on the product enrichment page, whether you ran it manually, in bulk, through a file import or automatically. Per enrichment you see:

• the value before and after, per field;
• which sources were used, with the source URLs;
• whether the AI wrote the text.

You can filter on texts written by AI. Useful to read those first. If an enrichment is wrong, one click reverts it.

Checklist: how to start with product enrichment

1. Put an EAN on every product, and on every variant. No EAN, no enrichment.
2. Connect your AI with your own key. Choose Gemini, OpenAI or Claude and the model. A daily token limit keeps costs in check, and you pay the AI provider directly.
3. Set your voice: tone, audience, keywords, language and, if you like, own instructions such as "Always mention the material and the capacity in litres".
4. Pick your import sources for automatic enrichment and decide whether new products are enriched before they go to the shop.
5. Check the first results in the log, filter on AI text and revert whatever is off.

Everything about the feature itself is on the AI product enrichment page. A more complete product helps your channels, but search engines and AI assistants decide for themselves what they show.

Frequently asked questions