
Summarize with AI
A large Shopify catalog is easy to upload and hard to browse. This guide covers how we plan collections, navigation and filters in 2026, what the Search & Discovery app can and can't do, how the standard product taxonomy and metafields feed filters, and how to keep collection pages fast, clear and indexable.
On most Shopify stores, shoppers see a collection page before they see a product. With a few hundred products that page can be simple. With thousands of products, several markets and variants in dozens of shades or sizes, it becomes the hardest page in the store to get right.
This guide is about that second case. It covers how to structure collections and categories for a large catalog, when to use navigation and when to use filters, what Shopify's Search & Discovery app can and can't do, how the standard product taxonomy and metafields feed your filters, what a good collection page looks like, and how to keep collection pages indexable without flooding Google with filtered URLs.
We checked every Shopify feature below in the Shopify Help Center and every store example live in November 2026. Some examples come from our own projects: Eveline Cosmetics, GANT, Monolith, The Odder Side, Home Carpet One, vetsak and OPTISOFA are wecanfly clients, and for each one we say what our part of the work was. Verve Coffee, Outdoor Voices and Art of Play are stores we only observe as shoppers.
Product categories on Shopify in 2026, in short
- Plan the structure before the theme. Decide which groups are menu items (collections) and which are product attributes (filters). Most catalog problems start when the two get mixed.
- Use automated collections for anything that changes. Conditions on category, tags, type, vendor, price, stock or metafields keep large collections up to date without manual work.
- Build filters on structured data. Shopify filters can use the product category, options, and product, variant and category metafields. Tags work, but they don't scale well.
- Know the limits. Up to 25 filters per store, up to 100 values shown per filter, and no filters on collections with more than 5,000 products.
- Treat the collection page as a landing page. Clear title, short intro, sensible default sort, honest badges and a filter panel that works with one thumb.
- Let Google index collections, not filter combinations. Shopify's default robots.txt and canonical tags already handle most of this. Don't undo them by accident.
What collections, categories, product types and tags are for
Shopify has four ways to group products, and they are easy to confuse:
- Collections are what shoppers browse: the pages under
/collections/that you link from menus. A product can be in many collections. - Product category is one standard category per product from Shopify's Standard Product Taxonomy, for example a specific type of sweater or face cream. It drives category metafields (attributes such as colour, fabric or fit), tax rates with Shopify Tax, and data for sales channels such as Google and Facebook.
- Product type is a free-text field unique to your store. Each product has one category and one product type, and Shopify treats them as separate fields.
- Tags are free-text labels. They are useful for internal rules and small campaigns, but a catalog filtered mostly by tags quickly collects typos and near-duplicates.
A practical rule we use in discovery workshops: if a shopper would say "show me all X", it's probably a collection. If they would say "only the ones that are Y", it's probably a filter. "Sweatshirts" is a collection. "Organic cotton" or "relaxed fit" is a filter on that collection.
Note that the Starter plan doesn't support collections at all, so everything here assumes the Basic plan or higher.
Category architecture for large catalogs
Keep the tree shallow and predictable
Shopify menus allow a top-level item with up to two levels of nested drop-down menus (three levels in total), and most themes support that depth. Three levels are enough for almost every catalog we have worked on: department, category, sub-category. If you need a fourth level, it is usually a filter in disguise.
Some patterns that work for large catalogs:
- One main path per product type, plus a few curated paths (new, bestsellers, sets, sale) that cut across it.
- Gender, room or use case at the top only when shoppers really start there. A fashion store usually starts with women/men; a furniture store often offers both product type and room.
- No dead ends. Every intermediate collection should show products, or at least clear links to its sub-collections and a "view all" link.
- Breadcrumbs that match the menu. On Shopify, a product page doesn't know which collection the shopper came from unless the theme passes it, so check that breadcrumbs make sense when someone lands on a product straight from Google.
Home Carpet One, a Chicago flooring and home decor retailer, shows why this matters. Its old site was overloaded with information and navigation was hard to use. In the Home Carpet One project we started with a discovery workshop, built a completely new sitemap, simplified the navigation to the most important links and a megamenu, and added numerous metafield-based filters. Not every product there can be bought online, so we also designed different types of product pages for products that need a quote or measurements.
Use automated collections for anything that changes
Shopify has manual and automated collections. Manual collections are right for small, curated sets such as a gift guide or a lookbook. For a large catalog, most collections should be automated:
- A collection can have up to 60 conditions, and you choose whether products must meet all of them or any of them.
- Conditions can use product title, variant title, type, vendor, category, tags, status, price, compare-at price, weight, inventory stock and metafields. Metafield conditions work with true/false, integer, decimal, rating and single line text metafields, and with metaobject references.
- Using the product category or a metafield as a condition is more robust than a tag. "Category is Sweatshirts and fabric is cotton" doesn't break when someone types "cottton" in a tag.
When we migrate a large store, we map the old category tree to automated collections before any design work starts. It's the moment to clean up categories that exist only because the old platform needed them.
Plan collections for several markets
If you sell in several countries from one store with Shopify Markets, the collection tree is usually shared, and only translations and availability differ. That's a big simplification compared with running separate stores. In the Eveline Cosmetics project, we first moved the stores from Magento to Shopify, and then consolidated 4 Shopify instances serving 5 stores (Poland, Germany, Ukraine, the Czech Republic and a pan-European store) into one Shopify Plus instance on Shopify Markets, in 4 months and with zero downtime. Some of Eveline's products have more than 100 variants, so the catalog structure had to be planned before the theme.
Separate stores per market still make sense when the assortment really differs. At The Odder Side, we migrated the fashion brand to Shopify Plus with separate instances for different markets and built a custom app that keeps product data in sync between them.
What goes in navigation and what goes in filters
A common question in our projects is whether better navigation or better filters sell more. The honest answer is that they do different jobs:
- Navigation answers "what kind of product do I want?" It should be short, stable and the same on every page.
- Filters answer "which of these fits me?" They should change with the collection, because a face cream and a sweater are chosen by different criteria.
The mistakes we see most often are both sides of the same problem: menus with 40 items that are really filter values ("blue sweaters", "sweaters under €50"), and collections so broad that shoppers need five filters to see anything relevant.
Live stores show the balance well:
- Eveline Cosmetics uses a three-level menu: product groups (face, body care, makeup and others), categories such as concealers or bronzers, and a few collections built around shopper needs (Calm & Soothe) and popular ingredients (Collagen, CICA). Inside the face collection, shoppers filter by skin type, active ingredients, effect, consistency, makeup coverage, SPF and age group. Note the overlap: an ingredient people search for by name deserves its own collection, while the full list of ingredients stays a filter.
- GANT in Poland splits the menu by gender and product type, with a few curated paths (the current season, bestsellers, linen, outlet). Inside a collection such as men's sweatshirts, filters cover category, colour, size, fit, fabric and availability.
- Verve Coffee, a US roaster, filters its All Coffee collection by roast (light, medium, dark), region and type (single origin, blend, decaf). These are exactly the decisions a coffee buyer makes, and none of them belong in the main menu.

Filters with Shopify Search & Discovery
On Shopify, storefront filters are set up in the free Search & Discovery app. Filters appear on collection pages and search results, and only if your theme supports them (or a custom storefront uses the filter API). Without theme support you can create filters, but shoppers won't see them.
What you can filter by
- Standard filters: availability, category, price, product type, tags and vendor.
- Product options: for example size or colour.
- Metafields: product, variant and category metafields. Supported types are single line text (also as a list), decimal, integer, true/false and metaobject reference (also as a list).
Limits worth knowing before you design
- Up to 25 filters per store.
- Collections with more than 5,000 products don't show filters at all. On a large catalog, "All products" or a huge "Women" collection may need to be split.
- Up to 100 filter values are shown per filter on the storefront.
- On search results, filters and counts use only the first 1,000 results, and filters are hidden when a search returns more than 100,000 products.
- Swatches and images in filters need a theme that supports the filter value API.
The 5,000-product limit is the one that most often changes the category architecture. We check product counts per collection early in a project, not after the theme is built.
Search settings in the same app
Search & Discovery also controls parts of on-site search. You can boost products for specific search terms (up to 10 terms per boost; boosts apply only to your online store search and to products available for sale), choose which result types search returns, and decide whether out-of-stock products are shown, hidden or placed last. Semantic understanding of queries is built into online store search and needs no setup.
At GANT, search and filtering were among the weak points of the brand's previous custom platform. In the migration to Shopify, we adapted GANT's existing global template and put our effort into custom listing components, dynamic size charts mapped to individual collections, and filtering and search tuned to get shoppers to the product faster. GANT's case study reports that product views more than doubled year over year (March–May 2026 against the same months on the old platform), together with 40% more revenue and 34% more purchases.
Product taxonomy, metafields and metaobjects as filter data
Good filters need good data. Here is how we structure it on Shopify in 2026.
Start with the standard product category
Assign every product a category from the standard taxonomy. You get category metafields (Shopify calls them product attributes in the taxonomy), such as colour, fabric or fit, with predefined values you can edit. According to the Help Center you can use category metafields as filters and in analytics reports, and colour entries linked to variant options can be displayed as swatches.
You can see this on live stores: GANT's fit and fabric filters and Outdoor Voices' colour filter use Shopify's standard attributes rather than custom tags. That keeps values consistent across thousands of products and is the same data Shopify sends to channels like Google.
Two limits to keep in mind: custom attributes aren't supported in the taxonomy, and unit-of-measurement attributes aren't available yet. Anything specific to your brand goes into your own metafields.
Use metafields for brand-specific attributes
Some attributes don't exist in any standard taxonomy: "resin transparency" for a table, "level of makeup coverage" for a foundation, "sleep function" for a sofa. Those are product or variant metafields, and they become filters in Search & Discovery.
- Eveline's face collection filters by skin type, active ingredients, effect and coverage, all stored as product metafields.
- vetsak, where we designed the UX/UI and developed the store for several regional markets, filters its sofas by colour, configuration and fabric.
- OPTISOFA, which we moved from WooCommerce to Shopify, filters its modular sofas by sleep function and bedding container.

Choose the type carefully: single line text for values like "cotton", integer or decimal for numbers such as seat height, true/false for yes/no features such as SPF.
Use metaobjects when a filter value needs more than a name
A metaobject reference lets one filter value carry an image, a colour, a description or a link. That's how you build colour swatches with exact hex values, material filters with a texture image, or a "collection" filter that links to a story page.
Monolith, a Swiss maker of solid-wood and epoxy-resin tables, is a good example of why this matters. We ran a discovery workshop, chose and customised the Reformation theme, and designed and built a catalog based on filters. The filters come with icons and tooltips that explain details such as resin transparency or the type of tabletop, so the shopper doesn't have to guess what a term means. In a catalog of unique products, the filter panel works almost like a product configurator.
Clean the data before you build the filters
Filters show every inconsistency in your data. Before launch we check:
- duplicate values that mean the same thing ("navy", "Navy", "dark blue");
- values that apply to too few products to be worth a filter;
- missing values on older products, which then disappear when a filter is used;
- filter names written for the warehouse instead of the shopper.
On migrations, this clean-up is part of moving the data, not an extra task. At GANT we structured and migrated the catalog with Matrixify, and the full catalog ended up cleaned and synchronised with the brand's ERP in near-real time.
What to put on a collection page for good UX
Title, intro and sub-collections
The top of a collection page should tell shoppers they are in the right place: a clear title, one or two sentences about the range (useful for SEO too), and links to sub-collections when the collection is broad. Keep long SEO text below the product grid, not above it.
A sensible default sort
Shopify lets you set a default sort order per collection: manual, best selling, alphabetical, price, date, or most relevant (the default for new collections, based on sales performance). Best selling uses the all-time number of orders, so a new collection with no sales shows the newest products first. Price sorting uses base prices, not Markets or B2B catalog prices, which is worth knowing if you sell in several currencies.
For shoppers, the sort menu should use plain words. GANT's Polish store offers popular, lowest price, highest price and new arrivals; Art of Play's playing cards collection offers best selling, price and date sorts next to filters by category and format.

Merchandising: manual order where it matters
Manual sorting lets merchandisers pin key products to the top. In the admin you can drag products or move a selection to the beginning, the end or a specific position; the list shows 60 products per page. In practice we keep manual order for the top rows of key collections and let an automatic sort handle the rest.
Product cards and badges
Product cards on a collection page should help compare, not repeat the product page:
- One image ratio across the grid, with a second image on hover only where it adds information.
- Titles that say what the product is. Outdoor Voices names products by product line and type (RecTrek Zip Pant, RecTrek Cargo Pant), so a grid of similar items is easy to skim. Put the colour in a swatch, not in a long title.
- Badges that mean something. Eveline's cards use badges such as Bestseller, New, New shade, Online exclusive and an extra discount label. Keep the set small and make sure each badge comes from data (a tag, metafield or date), not from manual edits.
- Variant information on the card, such as the number of shades or colour swatches, when shoppers choose by variant.
- Ratings where they exist, without stars on products that have no reviews yet.

Images that make products comparable
On visual categories, images do most of the work. vetsak's sofa collection uses wide lifestyle shots so shoppers can compare shapes and sizes at a glance. Whatever style you choose, keep it consistent across the grid.

Pagination or "load more"
For large collections we prefer numbered pages or a "load more" button over endless scrolling. Shoppers can come back to the same place with the back button, the footer stays reachable, and every page has its own URL. Whatever you choose, test that the back button returns the shopper to the same position in the list.
Mobile is where most collection pages are used
Many shoppers browse collections on a phone, so check in your analytics how much of your collection traffic is mobile and design the page to work with one thumb:
- A sticky "Filter and sort" button that opens a full-screen panel, with the number of results on the apply button.
- Selected filters shown as removable chips above the grid, with a "clear all" link.
- Two products per row on most categories, one on categories where the image matters most.
- Light pages. Collection pages load many images, so lazy-load everything below the first rows and keep app scripts to a minimum.
Speed is part of category UX. On Eveline's five stores, the new architecture reached a CLS of 0 and an LCP under 3 seconds on mobile, which matters on a collection page full of product images.

SEO for collection pages
Collection pages often rank better than product pages for category searches such as "linen shirts" or "modular sofa". A few rules keep them in good shape.
Let collections be the indexed pages
Shopify's default robots.txt blocks filtered collection pages, because they duplicate the collection itself. On the stores we checked in November 2026, the default rules disallow collection URLs with sort parameters (sort_by), with more than one filter, and with tag combinations (+ in the URL). Shopify also adds canonical tags automatically. On the stores we checked, a filtered or sorted collection pointed to the plain collection URL, and a product opened from a collection (/collections/…/products/…) pointed to the main product URL. Check this on your own theme, especially if it is heavily customised.
That means:
- Don't create SEO landing pages out of filters. If "linen shirts" deserves to rank, make it a real collection with its own title, intro and handle.
- Edit robots.txt only if you really have to. You can customise it with a
robots.txt.liquidtemplate, but Shopify calls it an unsupported customisation and warns that incorrect use can result in the loss of all traffic. An old template can also keep outdated rules. - Avoid near-duplicate collections (for example "sale" and "outlet" with the same products). Choose one and redirect the other.
Titles, descriptions and handles
Every collection has its own SEO title and meta description. Shopify's guidance is a title of up to 70 characters (60 to avoid truncation) and a meta description of about 160 characters. Keep handles short and descriptive, and add a redirect whenever you change one.
Collections in several languages and markets
With Shopify Markets, each language version of a collection is its own page with its own URL (Shopify adds hreflang tags automatically), and titles and descriptions are translated with the free Translate & Adapt app or another translation app. Translate collection names and filter values together, otherwise shoppers get a translated menu with filters in the source language.
More on technical SEO for Shopify stores is in our Shopify SEO guide and on our Shopify SEO optimisation service page.
Working in the theme editor
Most collection page changes don't need a developer. In the theme editor you can create a separate collection template, assign it to selected collections, change the number of products per row and add sections such as banners or sub-collection links. Filters and sort options appear automatically once the theme supports them and Search & Discovery is set up.
Custom work starts where the theme stops: custom listing components (as at GANT), filters with icons and tooltips (as at Monolith), or badges and swatches driven by metafields and metaobjects. If you edit theme code, work on a duplicate theme and test it before publishing. See also our no-code and low-code Shopify services and Shopify theme development.
What we learned from large-catalog projects
- Data decides more than design. Filters, collections, swatches and badges are only as good as the product data behind them. We plan categories, metafields and metaobjects before wireframes.
- Check the 5,000-product limit early. It changes the category tree more often than any design decision.
- Fast filters and search show up in behaviour. At GANT, product views more than doubled year over year after the migration, which the case study links to the new listing components, size charts and search.
- Filters can explain, not just narrow. Monolith's icons and tooltips turned a filter panel into a guide through unique products.
- One catalog for many markets is easier to run. After Eveline's consolidation, one team manages prices, content and promotions across markets from one admin.
- Navigation is often the first thing to simplify. At Home Carpet One, a new sitemap, a megamenu with only the key links and metafield-based filters came before any visual changes.
For more examples by industry, see our guides to UX for home and decor stores, building a beauty store on Shopify, clothing stores with inspiring UX and the UI/UX features a Shopify product page needs. To measure how collection changes affect sales, see how to use Shopify analytics and reports.

Planning categories for a large catalog?
wecanfly is a Shopify Premier Partner. For brands with large catalogs we run discovery workshops, plan category architecture and product data, design and develop Shopify and Shopify Plus stores, and migrate stores from other platforms. If your collections or filters need a second look, start with a Shopify UX and usability audit or contact us to talk about your store.
