Product data management: keeping product data complete, correct and consistent

What product data management means in retail and wholesale

In ecommerce and wholesale, product data management means making sure the data of every item is complete, correct and the same everywhere. That covers the product name, EAN and item number, prices, dimensions and weight, the description, attributes, category and images. This product data comes from several sources, such as suppliers, your POS or your ERP, and goes to several channels: your webshop, marketplaces, product feeds and business customers.

If you sell a hundred items in one webshop, you manage product data on the side. With thousands of items from several suppliers and three or four sales channels, managing product information becomes a job of its own. If an EAN is missing, the marketplace rejects your offer. If the weight is in grams instead of kilograms, your shipping costs are wrong. And if the webshop shows a different size than the marketplace, you get questions and returns.

The tool for this is usually called a PIM system. What a PIM is at its core is explained in what is a PIM system. This article is about the work itself: the building blocks of good product data management, the errors that occur most often and how to measure the quality of your product data.

PDM in manufacturing or product data in commerce?

Anyone searching for "PDM" or "pdm product data management" often ends up in manufacturing. There, PDM stands for managing engineering data: CAD models, drawings, bills of materials, versions and releases. A PDM system is then usually part of PLM (product lifecycle management) and is used by engineering and R&D. This page is about the other meaning: commercial product data that you sell with.

At manufacturers the two meet: the PLM supplies technical specifications that are later turned into sales-ready content in the PIM. How PLM, ERP, DAM and PIM work together is explained in PIM vs ERP. If you are looking for a CAD PDM for your engineering team, StoreLinkr is not the right tool. If you manage items for a webshop and marketplaces, read on.

The building blocks of good product data management

Good product data management is not a single program but a set of agreements. You need these eight building blocks in any case:

1. Sources: record where every product comes from: a supplier file, a wholesaler connection, the POS, the ERP or your own entry. Per source you then know what to expect and what is missing.
2. Data model: decide per category which attributes a product needs. A bike chain needs number of speeds, number of links and colour, a vase needs height, material and colour. One list of required fields for all categories only produces filler values.
3. Identifiers: every product has a unique SKU (your own item number) and, where available, an EAN. The EAN links you to marketplaces and suppliers, the SKU to your warehouse. Never use the same number for two products.
4. Ownership: per field there is one source that counts. Price and stock might come from the POS, texts and attributes from the PIM. That way systems do not overwrite each other.
5. Quality rules: agree when a product is finished, for example: valid EAN, brand, category, at least three images, weight in kilograms and a description of a few paragraphs.
6. Enrichment: add what the source does not supply: description, attributes, images, accessories and alternatives. Much of it can be fetched automatically from the EAN, see AI product enrichment.
7. Distribution: every channel has its own requirements for category, attributes and title. You manage the product once and send it to each channel in the right shape. More on that in PIM for ecommerce.
8. Maintenance: product data ages. Prices change, suppliers replace images, items are discontinued. Plan fixed time to check and clean up.

Common product data problems and how to fix them

Most product data errors do not come from carelessness but from several sources and channels meeting. These are the ones you see most often:

Example: two suppliers, one chain

A bike parts wholesaler gets the same chain from two suppliers. Supplier A calls it "CHAIN 11SP 116L" with a weight of 257, in grams. Supplier B calls it "Bike chain 11 speed" with a weight of 0.257, in kilograms. Without agreements there are now two products in the webshop, with different titles and weights a thousand times apart. With a match on the EAN, one unit for weight and a fixed source per field, one product remains, with the best title and a correct weight.

Product data quality: check and measure

Data quality you do not measure gets worse with every import. So track a few simple metrics you can compare every week:

• Completeness: how many products lack a brand, category, description, model, SKU, price, weight, dimensions or attributes?
• Images: how many products have no images or too few?
• Validity: how many products have an invalid EAN?
• Margin: how many products have a margin that is too low, for example after a supplier price increase?
• Availability: how many products have no stock but are still online?

Checklist for your weekly routine

1. Review new products from all sources and merge duplicates.
2. Bring down the number of products with a missing or invalid EAN.
3. Complete your best-selling categories first, then the rest.
4. Add images to products with none or only one.
5. Check margins after supplier price changes.
6. Take discontinued items without stock offline.
7. Write down the metrics and compare them with last week.

Using the StoreLinkr data quality dashboard

In StoreLinkr you do not have to count these numbers yourself. The data quality dashboard gives live feedback on your whole catalogue and flags products with no or few images, an invalid EAN, a low margin, no stock, or a missing brand, category, description, dimensions, attributes, model, prices, SKU or weight. Use it as a recurring check: fix one issue type at a time, for example all invalid EANs first and then the missing weights, and check again after every supplier import. That way you see straight away whether a new source improves your product data quality or makes it worse.

This routine is for day-to-day management. How to set up a PIM for the first time, with data model, import and connections, is covered in PIM implementation.

Product data management with StoreLinkr

StoreLinkr is a PIM with connections to webshops and marketplaces, built for webshops, wholesalers and retailers. This is how it covers the building blocks:

• Sources: connections with wholesalers, suppliers and POS systems, a one-time import of XLS, XLSX or CSV with column mapping and preview, and your own PIM products. An imported product becomes a PIM product in one action. See integrations.
• Ownership: product rules override imported fields per product, such as name, brand, model, descriptions, SKU, prices and category, so the next import keeps your corrections. Find and replace rules per import source clean up supplier texts.
• Data model: EAN plus extra barcodes, SKU, brand, model, prices, weight and dimensions, attributes, metafields (text, HTML, number, date, colour, URL, table, JSON and more), variants, product relations, your own category tree, images with order and alt text, and attachments.
• Enrichment: from the EAN via the product catalogue, Google Books for books, or your own AI (OpenAI, Claude or Gemini) with your own key.
• Quality: the data quality dashboard described above shows what is missing. Bulk actions and inline editing in the product overview help you fix it quickly.
• Distribution: you map categories to each channel's categories, including marketplace attributes. Per channel you choose its own product selection, price list and stock locations. Export goes to WooCommerce, Shopify, SitePack, CCV Shop, Lightspeed and HikaShop, to Bol, Kaufland and Mirakl marketplaces, and to feeds for Google Shopping, Meta, Microsoft Ads and TikTok.

What StoreLinkr does not do

StoreLinkr is not a CAD PDM and not a PLM. There is also no separate DAM, no CSV or Excel export of your PIM data, no translated product content per language, no different description per channel and no roles and permissions per user. If you need those, for example as a brand with a large range in many languages, an enterprise PIM is a better fit; see best PIM software. What StoreLinkr costs is on the pricing page.

Frequently asked questions