A buyer emails to ask how deep the shelf unit is. Your sales rep says 32 centimetres, because that is what the line sheet says. The product page on your webshop says 30. The spec the factory sent says 31.5, and the retailer measured one in their stockroom at 32 and a bit.

Nobody is lying. The number was correct at some point in each of those places, and then the product changed slightly, or someone rounded, or a file was copied before the correction went in. There is no version anyone can point to and say: that one is right.

That is a product data problem, and almost every wholesale brand has one. It stays invisible while you sell forty products to twenty stockists, and it starts costing real money somewhere around the point you stop being able to hold the range in your head.

This article covers what wholesale product data actually consists of, the five ways it reliably breaks, and what a structure looks like that holds past a few hundred SKUs.

What counts as product data

More than most people assume. It helps to see it as five layers, because they come from different places and fail differently.

Most brands manage the first three reasonably and treat the last two as paperwork to find when somebody asks. That is the part that is changing, and we will come back to it.

Five ways product data goes wrong

1. One SKU doing the work of four

A style exists in four colourways and carries a single code. It feels tidier. It is the most expensive shortcut in wholesale.

Everything downstream inherits the ambiguity: the pick is a guess, stock figures are meaningless per colour, you cannot tell which variant actually sells, and a wrong-item claim becomes somebody's word against somebody else's. One SKU per sellable variant is the foundation everything else sits on, and retrofitting it is painful enough that most brands do it exactly once, under duress.

2. The master file that is nobody's job

There is a spreadsheet. It was accurate in March. Since then three people have added products to it, one has filtered and sorted it without undoing the sort, and a copy went to the warehouse that has since diverged.

The question that exposes this is simple: if a dimension is wrong, where do you fix it so that it is fixed everywhere? If the honest answer is a list of places, you do not have a master file. You have several files of equal authority and no way to choose between them.

3. Product data living inside documents

For a lot of brands the real product database is the line sheet. It holds the images, the prices, the minimums and the descriptions, and each season it gets rebuilt from the last one.

The trouble is that a document is a snapshot. Change a price and you have changed one file, not the price. Everything that was ever sent out stays wrong, and a PDF cannot be corrected after it lands in an inbox. Data belongs in records that documents are generated from, not the other way around.

4. Every channel keeping its own copy

Your webshop has product content. Your B2B portal has product content. The marketplace listing has its own. A big stockist asked for a spreadsheet in their format and someone built it by hand.

Four copies means four maintenance jobs, and in practice three of them stop being done. The buyer who finds a discontinued colour on your site and orders it is not making a mistake; they are reading something you published. That is also why adding channels feels harder than it should — covered in more detail in our guide to multi-channel B2B sales.

5. Supplier data that was never collected

Composition, origin, carton dimensions, HS codes. Somebody knew at the time. It was in an email, or a PDF spec, or a conversation with a factory contact who has since moved on.

This one is different from the others because you cannot fix it from your own desk. It has to be collected from suppliers, which takes production cycles, which is why it is worth starting before anyone demands it.

The structure that holds

Four rules. They are not complicated, and the difficulty is entirely in the discipline rather than the concept.

One record per sellable variant. If a retailer can order it on its own, it has its own record and its own identifier. Not per style. Per thing that ships.

One place is authoritative. Not the most complete place, not the newest file — a nominated system that everything else copies from. Everywhere else is a view of it, including your line sheets.

Each field has an owner. Someone owns composition. Someone owns carton weights. Someone owns copy and images. Unowned fields are the ones that are blank in nine hundred rows.

Collect at the point of ordering. Ask suppliers for composition, origin and carton data when you place the purchase order, not in a data project next spring. It arrives reliably when it is a condition of the order and almost never when it is a favour.

How much structure do you actually need

Not every brand needs dedicated product information management software, and plenty have bought one and used it as an expensive spreadsheet.

Under about a hundred variants with one sales channel, a well-disciplined single sheet is genuinely fine, as long as it really is single. The signals that you have outgrown it are specific: you sell through more than two channels, your range turns over seasonally, you have variants in multiple dimensions such as size and colour together, or somebody has to be asked before a product fact can be confirmed.

The more common path is not buying a separate PIM but making your ordering platform the authoritative home, since that is where the data has to end up anyway to be sold from. If you are mapping this out across your systems, our guide to a B2B tech stack covers how the pieces fit, and API integration covers keeping them in step.

What it costs when it is wrong

Worth being concrete, because product data work is easy to postpone indefinitely.

Wrong dimensions and weights produce shipping surprises and credit notes. Ambiguous SKUs produce wrong-item claims, which are credit notes with a side order of goodwill damage. Stale descriptions produce buyers ordering things that no longer exist. Missing compliance data produces a scramble when a retailer's buying team sends a supplier questionnaire with a deadline on it.

And the quiet one: time. Every season spent rebuilding the same catalog from the same spreadsheet is a week that nobody counts as a cost, because it has always been like that.

Where this is heading

The direction of travel in Europe is toward product information being structured, granular and provable rather than descriptive. Packaging rules already ask brands to know what their materials are. Retailers increasingly ask for composition and origin as a condition of listing. The proposed Digital Product Passport for textiles would attach a data record to the individual product.

None of that requires action this quarter. All of it rewards the same groundwork: a record per variant, fields that exist, and supplier data collected as a matter of routine rather than retrieved under pressure.

How Orderchamp Cloud helps

Orderchamp Cloud gives your product data one authoritative home and then uses it everywhere, which is the part that matters. A record per variant, with images, descriptions, barcodes, stock, minimums and customer-specific prices attached to it.

Your B2B Portal presents that catalog to each retailer at their own prices, the Sales App works from the same records in the field, and line sheets become something you generate rather than maintain. Orders then reference the same product records, which is what keeps wholesale order management accurate without re-entry. Update a product once and every channel reflects it. For the wider picture, see our guide to a B2B e-commerce solution.

Want one version of your product data everywhere? Explore Orderchamp Cloud or book a demo.

Frequently asked questions

What is product data management in wholesale?

Do I need PIM software?

When have I outgrown a spreadsheet?

What’s the first thing to fix?

Who should own product data?

How do I get composition and origin data from suppliers?

What does bad product data actually cost?