Catalogue · Filters and variants
A large product catalogue needs more than more categories
A large product catalogue in an online store: how to structure categories, filters, variants and imports so customers find products easily.
Alphabite · Ecommerce design and development team ·
Updated:
A large catalogue is organised in four layers — categories for the type of product, filters to narrow the choice, variants for the exact sellable version and collections for browsing by idea. It only works with consistent data: every variant has a permanent code, and every attribute has one meaning and the same units.
Which products should you check first?
The first problem in a large catalogue is often inconsistent data rather than missing filters. Choose a few dozen representative products: a best-seller, a product with many variants, an unavailable item and one that belongs to more than one collection. This is a practical test sample, not a catalogue-size requirement.
Every purchasable variant needs a permanent SKU. Decide whether “gold” means colour, material or finish. If those meanings are mixed in one field, a filter can mislead even when the store is technically working.
What is the difference between a category, a filter and a variant?
Customers should not have to decode your supplier spreadsheet to find a product. Organise navigation around the choice they are making. Prefer a shallow structure: instead of another level of categories, a filter is often clearer.
| Element | Job | Example |
|---|---|---|
| Category | Choose a product type | Necklaces |
| Filter | Narrow the choice | Material or adjustable length |
| Variant | Choose the exact sellable version | Length with its own stock |
| Collection | Browse by a shared idea | A gift edit across categories |
Jewellery catalogue example
Category
- Job
- Choose a product type
- Example
- Necklaces
Filter
- Job
- Narrow the choice
- Example
- Material or adjustable length
Variant
- Job
- Choose the exact sellable version
- Example
- Length with its own stock
Collection
- Job
- Browse by a shared idea
- Example
- A gift edit across categories
Which filters actually help?
A useful filter answers a question the customer asks while choosing: size, material, colour, price, availability. A filter with a single value, or with dozens of nearly identical values, only takes up space, especially on a phone.
Add a filter only when its data is filled in across the whole category. Otherwise products without a value disappear from the results, and the customer concludes you do not have what they want.
- Show only values that return results in the current category.
- Merge synonyms — “silver”, “silvery”, “Silver” — into one value.
- On a phone, show the selected filters and a clear-all button.
- Search should find products by name and by SKU.
Example: around 2,500 products at Lumies Jewelry
Lumies Jewelry is a jewellery and accessories store with around 2,500 products, organised into collections and categories for shopping on desktop and mobile. With a range like that, collection imagery starts the choice and accurate attributes help it continue to a specific piece.
At Amazon İş Market, customers find workwear and protective equipment by brand, category and specification and buy at retail or wholesale.
Who maintains the catalogue after launch?
A large catalogue changes every week: new products, seasonal collections, discontinued models. Without rules for who fills in attributes and in what format, data gradually drifts and filters lose their meaning.
Decide in advance what happens to the page of a product you no longer sell. If it will return, keep the page with clear availability information. If it is discontinued for good, point the customer to a similar product or the category instead of leaving an empty page.
How do you test product imports and updates?
Import the test sample twice. Change a price, remove an image and sell out a variant. Check that the second import updates the correct records instead of duplicating products or restoring old information. Failed rows should leave an understandable report.
For the public store, decide which product and category pages should be discoverable in search. Google documents ProductGroup for related variants; structured data must match what the visitor can actually see and choose. It does not fix missing product information.
Frequently asked questions
How should I organise categories in a large online store?
Organise categories by the type of product customers look for, not by your supplier spreadsheet. Keep the structure shallow and use filters for size, material or price instead of more category levels.
What is the difference between a variant and a separate product?
A variant is a sellable version of one product, such as a size or length, with its own SKU and stock. If versions differ in images, description and the way customers choose them, separate linked products are often clearer.
Which filters should an online store have?
The filters that answer real questions while choosing: size, material, colour, price, availability. Add a filter only if its data is filled in consistently across the whole category.
How do you import thousands of products without errors?
Start with a test sample of difficult products, import it twice and check that the second import updates the correct records. Every variant needs a permanent SKU, and failed rows need an understandable report.
Can a catalogue with thousands of products work well on a phone?
Yes, if categories are clear, filters show the selection and are easy to clear, and images and key information are visible without excessive scrolling. Lumies Jewelry, for example, organises around 2,500 products into collections and categories for desktop and mobile shopping.