Planning Your Product Catalogue: Attributes, Variants and Data
In this article10 sections
Product catalog planning is the work of deciding, before the store is built, how every product will be described in data: which product types you have, which attributes each type needs, how variants such as size and colour are handled, which identifiers you store and where the data comes from. It sounds like a technical detail, but it shapes almost everything shoppers experience. Filters, search, product pages, comparison tables, marketplace feeds and ERP integrations all depend on it. A few days of planning at the start saves months of clean-up later.
Key takeaways
- Group products into types that share the same attributes, such as shoes, jackets or laptops.
- Choose attributes from the shopper's point of view, with fixed values and consistent units.
- Decide early how variants work: which options create separate stock items and which are just descriptions.
- Store standard identifiers such as GTINs, brands and manufacturer part numbers wherever they exist.
- Name one system and one person as the owner of product data, and write the rules down.
Why the catalogue structure matters
Imagine two stores selling the same jackets. In the first, each jacket has a long description and little else. In the second, every jacket has a type, a size range, a colour from a fixed list, a material, a waterproof rating and a fit. Shoppers in the second store can filter by size, colour and waterproofing, compare jackets side by side and find the right one through search. The first store relies on shoppers reading every description.
Good catalogue structure delivers:
- Useful filters on category pages, built from consistent attribute values.
- Better site search, because search engines can match attributes as well as text.
- Clear product pages with specification tables that are complete and consistent.
- Accurate feeds for Google Shopping and marketplaces, which require specific fields.
- Reliable integrations with stock, ERP and warehouse systems.
- Faster work when adding new products, because templates already exist.
Poor structure does the opposite. Every new feature, from filters to feeds, becomes a data clean-up project first. Our guide to category page UX shows how much filters depend on data quality.
Start with product types
A product type is a group of products that share the same set of attributes. Shoes need sizes, widths and materials. Laptops need screen size, processor, memory and storage. Sofas need dimensions, seat count and fabric.
- List everything you sell and group items that are described the same way.
- Keep types meaningful but not too many. If two groups need almost the same attributes, they can usually share a type.
- Separate types from categories. Categories are how shoppers browse; types are how data is structured. A "summer sale" category may contain many product types.
- Allow for growth. Think about products you plan to add in the next year or two.
In platforms such as Magento, product types are usually set up as attribute sets. Other platforms use templates, metafields or product types, but the principle is the same.
Choosing attributes
Attributes are the individual facts about a product: colour, size, material, voltage, capacity, compatibility. Choose them carefully.
Think like a shopper
Ask which facts customers use to choose and compare. Look at your site search terms, the questions customers ask your team, competitor stores and the attributes marketplaces require. Include those attributes; leave out internal details that shoppers do not care about, or keep them hidden from the storefront.
Use fixed values where possible
Free-text attributes lead to inconsistency: "navy", "Navy blue", "dark blue" and "NAVY" for the same colour. Where an attribute has a limited set of answers, use a fixed list. That keeps filters clean and makes reporting reliable.
Agree units and formats
Decide how measurements are stored and shown: centimetres or inches, kilograms or pounds, one decimal place or two. Store one standard unit and convert for display if you sell in markets with different conventions.
Decide which attributes do what
For each attribute, decide whether it should appear on the product page, be used as a filter, be searchable, appear in comparison tables or be sent to feeds. Not every attribute needs every role. Too many filters and searchable fields slow the store and confuse shoppers.
Variants: sizes, colours and options
Variants cause more catalogue problems than anything else. The key question is: which options create separate items with their own stock, price or image?
| Option | Usually a variant? | Why |
|---|---|---|
| Clothing size | Yes | Each size has its own stock |
| Colour | Yes | Different stock and images |
| Engraving text | No | A custom option, not a stock item |
| Gift wrapping | No | An add-on at checkout |
| Warranty length | Sometimes | May be a separate product or a paid option |
For each variant, plan its own identifier (SKU), stock, price if it differs, and images. Decide how many variants one product can have. Some platforms limit the number, which matters for products with many sizes and colours. If you sell bundles or kits, decide how they reduce the stock of their components; our guide to inventory management for online stores explains why this matters.
Identifiers, images and content
Identifiers
Give every sellable item a unique SKU that never changes. Store manufacturer identifiers too: the brand, the manufacturer part number and the GTIN barcode number where products have one. Shopping platforms use these to match your products, and Google lists them among the core fields in its product data specification.
Images
Agree image standards: size, background, angles, file naming and alt text. Link images to the right variant, so choosing a colour shows that colour straight away, on every device and in every feed.
Descriptions
Separate structured attributes from written descriptions. Attributes hold facts; descriptions explain benefits and use. Our guide to writing product descriptions shows how they work together.
Where product data lives and who owns it
As businesses grow, product data often comes from several places: the ERP for prices and stock, suppliers for specifications, the marketing team for descriptions and photographers for images. Without clear ownership, the same field gets edited in two places and the versions drift apart.
- Small catalogues: the e-commerce platform can be the main product database.
- Larger catalogues and several channels: a product information management (PIM) system can hold enriched product data and publish it to the store, marketplaces and print.
- Prices and stock usually come from the ERP or stock system, not the PIM.
Whatever the setup, write down which system owns each field and who is responsible for keeping it accurate. Include these decisions in your project brief; our ecommerce website requirements checklist has a section for catalogue and data.
Keeping the catalogue clean over time
Catalogues decay. New products arrive in a rush, suppliers send data in new formats and temporary fixes become permanent.
- Write a short data guide: product types, required attributes, allowed values, units and image rules.
- Use required fields so incomplete products cannot be published.
- Review new attribute values before they are added, so lists stay clean.
- Run regular quality reports: products missing images, attributes, identifiers or descriptions.
- Archive discontinued products properly, with redirects where pages had traffic.
Search and filters only stay useful while the data stays clean. Our guide to ecommerce site search shows how quickly poor data undermines search results.
Product catalog planning for several markets
If you sell in more than one country or language, plan for it now, even if you start with one market.
Translatable and fixed fields
Decide which fields change by language, such as names, descriptions and attribute labels, and which stay the same everywhere, such as SKUs, identifiers and dimensions. Attribute values from fixed lists, like colours, should be translated once in the list rather than typed again on every product.
Market-specific data
Some data differs by country: prices, tax classes, sizes, safety information, plug types or legal labelling. Plan where these live, so a product can carry different values for different storefronts without duplicating the whole product.
Ranges that differ
You may not sell every product everywhere. Plan how to switch products on or off by market, and how stock from different warehouses maps to each storefront. Our guide to selling online in Europe covers the wider requirements of multi-country selling.
How to run a catalogue planning workshop
A short workshop with the right people turns these decisions into a plan. Invite someone from buying or merchandising, marketing, customer service, operations and finance, plus your developer.
- Bring real examples: your best seller, your most complicated product and a product that often causes customer questions.
- Agree product types and list the attributes for each one, marking which are filters, which appear on the page and which feed other systems.
- Agree allowed values for fixed-list attributes, and the units for measurements.
- Decide variant rules for each type, including how images and stock attach to variants.
- Assign ownership for each group of fields and the system it lives in.
- Write it up as a short data guide that new staff and suppliers can follow.
The workshop usually surfaces disagreements, such as two teams using different names for the same thing. Resolving them now is far cheaper than discovering them after launch, when filters and feeds already depend on the data.
Frequently asked questions
When should catalogue planning happen?
Before the store build starts, ideally during discovery. Changing product structure after launch is possible but much harder, because data, templates, filters and integrations all depend on it.
Do I need a PIM system?
Not always. Small and mid-sized catalogues sold through one store can be managed in the platform. A PIM becomes valuable when you sell through many channels, in several languages or with large, frequently changing catalogues.
How many attributes should a product have?
As many as shoppers need to choose and compare, plus what feeds and integrations require. Avoid attributes that nobody uses, because each one adds maintenance work.
Can I fix a messy catalogue without rebuilding the store?
Usually, yes. Start with your best-selling product types: define attributes and values, clean the data in bulk and then switch on better filters. Work through the rest of the catalogue in stages.
Planning a new store or untangling an existing catalogue? Talk to our team about structuring your product data so filters, search, feeds and integrations work from day one.
Want an expert to look at your store?
Tell us what you are working on. A senior e-commerce consultant will reply within one working day with honest, practical advice. No obligation and no sales script.
- Reply within one working day
- Advice even if we are not the right fit
- Your details are never shared