Shopify Collection Page CRO: Cards, Filters and Quick Add
Written by Saif Ur Rehman, Lead CRO Auditor and founder of Cartvane. Updated .
Short answer: a collection page exists to get the shopper to the right product page. It fails when product cards hide what people decide on (image, title, price, variants), when filters are missing or awkward on a phone, when sort lacks the options people look for, or when quick add picks the wrong variant without saying so.
Most "collection page best practices" posts are design galleries. This guide pairs Baymard Institute's product list research with the Shopify-specific reasons filters quietly vanish: theme support, collections over 5,000 products, the 100-value cap, the price filter in other currencies, and tag filters in translated stores. It ends with a 10-minute test you can run on your own phone.
To see whether your collection step has an obvious leak before you start, run the free Cartvane scan. It walks your store at phone and desktop width, from homepage to collection to product to cart. No card needed.
The short version
- Every card shows image, title and price. Colour variations sit on one card with swatches, not as separate cards.
- Filters exist for what shoppers compare in your category, including anything already printed on the card.
- Shoppers can pick two values of one filter (blue or black) and see what they applied above the list.
- Sort includes price, rating, best selling and newest where your data supports them.
- Quick add on a product with sizes or colours asks which one before it adds anything.
- Back from a product page returns the shopper to the same spot in the list.
Industry research vs what Cartvane checks
Two kinds of information appear below. Keep them apart.
| Industry research | Cartvane product | |
|---|---|---|
| -- | -- | -- |
| Source | Baymard Institute public articles; Shopify Help Center and shopify.dev | An automated walk of your public pages |
| What it tells you | How shoppers use product lists and filters, and how Shopify filtering behaves | What one lab run found on your store |
| Type of data | Usability testing and benchmark scores across large ecommerce sites | Lab data: one browser, phone 390px and desktop 1366px |
| Scope | Ecommerce in general; Shopify setup docs | 91 checks; Collection group = 8 |
The research picture. Baymard's Product List UX 2025 (updated September 9, 2025) summarises more than 21,000 manually scored product list parameters across 170+ leading US and European ecommerce sites. It found that 58% of desktop sites and 78% of mobile sites perform "mediocre" or worse on product list UX. Those sites are large retailers rather than a Shopify sample, so treat the numbers as context about how hard product lists are to get right, and never as a forecast for your store.
What Cartvane checks. The homepage lists 8 collection checks in the 91. Three are named publicly: Product cards show price, image and title, Filters and sort exist and work on mobile, and Quick add handles sizes and colors correctly. Cartvane is independent of Baymard and Shopify.
Product cards: what each card must answer
A card has a few square centimetres on a phone to answer "is this worth a tap?". Start with the basics, then add what your category needs.
Image, title and price on every card. If a theme setting or an app hides price on some cards, or shows "From" prices without a range, shoppers have to open product pages just to compare. That is the first named Cartvane collection check.
Combine colour variations into one card. Baymard's 2025 product list article reports that 42% of benchmarked sites don't combine variations into one list item. In testing, separate cards per colour cluttered the list and made a specific colour harder to find. On Shopify, this is mostly a catalog decision: one product with a colour option shows as one card, while separate products per colour show as separate cards unless your theme or an app groups them.
More than one image. The same article says 80% of sites don't offer 3 or more thumbnails in product lists, and recommends at least 2 additional thumbnails, with apparel sometimes needing far more. On Shopify, put this in the test bucket. Extra card images add weight to a page that already loads many images. The speed side is covered in Shopify product page speed and mobile LCP.
Unit price when pack sizes vary. Baymard reports 67% of sites don't display a price per unit for products sold in varying volumes or quantities. Shopify has this built in: once you enter a unit price on a product or variant, it shows automatically on product, collection, cart and checkout pages of any Online Store 2.0 theme; vintage themes need code. Each product or variant can have one unit price, the unit type is the same across all markets, and some regions legally require it (Shopify Help: unit pricing). Preview the card after you add it, because Shopify notes some themes render units oddly.
Be honest about stock. Baymard's 2024 product finding update describes a participant who got excited about an item in the list, then found out on the product page that it was out of stock. Show sold-out on the card. If you use Search & Discovery, its settings can move empty filter values to the end of the list or hide them (Shopify Help: filters).
Filters shoppers expect
The five essentials. Baymard's 2025 product list article lists five filter types that should always be offered where they apply: price, user rating average, colour, size and brand. It reports 51% of sites don't provide all five. (Baymard's older standalone article on the same guideline carries a different percentage in its title; this guide uses the figure on the page linked here.)
A filter for everything on the card. If the card shows an attribute, shoppers expect to filter by it. Baymard's Have Filters for All Displayed List Item Info (published September 17, 2019, so older) found 38% of benchmarked sites missing such filters, and describes test participants abandoning when a filter for a visible attribute didn't exist. On mobile it is worse: the filter panel is hidden, so finding out a filter doesn't exist can take many taps.
Two values of one filter. Baymard reports 14% of sites don't let users combine values of the same filter type. Shopify's default helps here: values from the same filter work as an or condition (red or green), and values from different filters work as an and condition (red and size 8) (Shopify Help: filters). If your theme or an app makes values single-select, fix it without waiting for a test.
Show what is applied. Baymard's Display "Applied Filters" in an Overview (updated May 13, 2026) says 28% of sites across its benchmarks show no overview at all. Without one, shoppers can't confirm what is applied, can't remove one filter quickly, and lose the context of the list. On mobile, Baymard found both a horizontally scrolling row and stacked rows of applied filters can work, provided shoppers can tell there are more. A button that only says "Filter (3)" performed poorly in testing, because shoppers still had to open the panel to see the three.
Be careful with horizontal filter bars. Baymard's 2025 article gives two reasons for caution: filter values are hidden until a shopper opens each type, and "All filters" buttons were overlooked. A horizontal bar can work with a small number of filter types; if you have many, test it against a sidebar or drawer.
For a first read on whether your filters and sort show up and work at phone width, scan your store free.
Why Shopify filters go missing
This is where Shopify stores differ from the retailers in Baymard's benchmark. A store owner creates filters in the Search & Discovery app and assumes shoppers see them. Often they don't. Every cause below comes from Shopify's own documentation, read on October 21, 2026.
| Symptom | Shopify cause | Where to check or fix |
|---|---|---|
| -- | -- | -- |
| No filters anywhere | Theme doesn't support storefront filtering. Filters you create aren't displayed. | Admin Content > Menus: a message appears in "Collection and search filters" if unsupported. Third-party themes may be unsupported without showing that message (Add storefront filtering) |
| Theme supports filters, still none on collections | Filtering switched off in the section | Shopify-developed themes: Product grid section of the collection template (and Search results section on search) |
| Filters on small collections, none on your biggest | Collection has more than 5,000 products | Split into smaller collections, e.g. Tops, Jeans, Boots instead of one Women collection (Shopify Help: filters; same limit on shopify.dev) |
| Some values never appear | A filter shows at most 100 values on the store | Group similar values (Onyx, Ebony, Midnight → Black) in Search & Discovery |
| Price filter gone for some shoppers | Price filter doesn't display for currencies other than the store's default currency | Known limitation. Make sure other filters still narrow the list for international shoppers |
| Tag filter missing in another language | Tag filter only shows in the default language; vendor values always use the default language | Use metafield filters for better translation support |
| Odd or duplicate option values | Option names differ across products ("Color" vs "Color:") | Make option names match exactly |
| New attribute not filtering yet | Search index not updated | Shopify suggests waiting 24 to 48 hours, or re-saving an affected product |
Two more limits apply. A store can have up to 25 filters in total, and filters on search results are based on up to the first 1,000 results. Applied filters are written into URL parameters (shopify.dev), which is useful for testing: a filtered URL you can paste into a message shows a developer exactly what you saw.
Sort: give shoppers the four they look for
Baymard's 2025 article names four sort types that test participants commonly sought: price, user rating, best selling and newest. Its benchmark shows 68% of desktop and 69% of mobile sites don't offer all four.
Two Shopify details matter here, both from the Help Center page on collection layout and sort order:
- New collections default to "Most relevant", based on sales performance. Other default options are best selling (all-time orders; newest to oldest if nothing has sold), title A–Z or Z–A, price high or low, newest or oldest, and manual. Check that the default suits each collection, especially seasonal ones.
- Price sort uses base prices. Market and B2B catalog prices are not used to decide the order. If those prices differ from base prices, a "price low to high" list can look out of order to international or wholesale shoppers.
The Help Center list of sort options doesn't include user rating. If rating sort matters in your category, check whether your theme or review app adds it, and confirm it works on a phone.
Quick add that doesn't guess
Quick add saves a trip to the product page when it works. On a product with sizes or colours, it has to ask which one. A button that adds the default variant without asking, or fails without a message, leaves a shopper with the wrong size in their cart or no idea whether anything happened. That is why one of Cartvane's named checks is Quick add handles sizes and colors correctly.
Research supports quick views for some catalogs only. In its 2024 product finding update, Baymard changed its guidance: quick views are now helpful for visually driven products such as apparel, but should still be avoided for spec-driven product types. A participant on American Eagle liked choosing a size without leaving the list.
Run this matrix on a phone before you trust your quick add:
| Product type | Tap quick add | Expected result |
|---|---|---|
| -- | -- | -- |
| Single variant, in stock | Once | Adds, cart count or drawer updates, clear confirmation |
| Multiple sizes or colours | Once | Picker opens; nothing added until a size or colour is chosen |
| One variant sold out | Choose that variant | Shown as unavailable before the tap; no silent failure |
| All variants sold out | Look at the card | Card says sold out; quick add hidden or disabled |
| After adding | Open the cart | Correct variant, quantity and price |
Cart drawer behaviour after the add belongs to the cart step, covered in the Shopify CRO audit checklist.
The 10-minute phone test
Baymard's mobile filtering examples (509 annotated screenshots from a 93-site benchmark) name the most common mobile problems: repeated trips back and forth between the product list and a separate filter screen, overlong lists of filter options, and filter screens that obscure or skew what is applied (Baymard mobile filtering examples). This test walks through each of them.
Open your largest collection on your own phone, on mobile data if you can.
- Find the filter button after scrolling. Scroll down two screens. Can you still reach Filter and Sort without scrolling back to the top?
- Apply two values of one filter (two sizes, or two colours). Do both stay selected? Does the count of products make sense?
- Close the filter screen. Can you see what is applied above the list without reopening it?
- Remove one filter from the list view. Does the other stay applied?
- Sort by price low to high. Does the order match the prices you see, in your main market's currency?
- Open a product, then tap Back. Are you returned to the same place in the list? Baymard's Return Users to the Same Place in the Product List (published November 17, 2020) found 13% of benchmarked sites send users back to the top, and observed this as a direct cause of abandonment. It is worse on mobile, where fewer items fit per screen.
- Quick add a product with sizes. Use the matrix above.
- Check tap targets on swatches, filter checkboxes and the "x" on applied filters. WCAG 2.2's minimum is 24 by 24 CSS pixels, or enough spacing around smaller targets (W3C: Target Size (Minimum)).
Write down every point where you hesitated. Those notes are your fix list.
Load more, pagination or infinite scroll? In its 2024 update, Baymard still prefers "load more", and now says pagination can work okay, with many details to get right (2024 product finding update). Whichever you use, step 6 matters more than the pattern. Baymard's return-to-place article notes that a "load more" list should rewrite the URL on each click (for example with history.pushState) so Back lands in the right place.
Fix now vs test
CXL sorts findings into five buckets: Test, Instrument, Hypothesize, Just Do It and Investigate (From Data to Test Hypotheses). For collection pages:
Just do it (no test needed): turn on filtering in the theme; split collections over 5,000 products; group filter values over the 100 cap; show price on every card; make quick add ask for size or colour; return shoppers to the same place after Back; show applied filters above the list.
Test: extra thumbnails on cards; two vs three cards per row on phones; horizontal filter bar vs drawer; which filter opens expanded by default.
Instrument: CXL's own example of an instrumentation gap is having no event tracking for "using filters on ecommerce category pages". If you can't see which filters people use, add events before you redesign the filter panel.
Investigate: a collection that gets traffic but few product page visits, when none of the causes above apply. Watch a few sessions or ask customers before you guess.
What Cartvane checks
Cartvane runs 91 checks on public pages at phone width (390px) and desktop width (1366px). The Collection group has 8 of them. The three named on the homepage are Product cards show price, image and title, Filters and sort exist and work on mobile, and Quick add handles sizes and colors correctly. The Speed group also includes Mobile LCP on home and collection.
A scan reports what one lab run could see. In the sample report, the scanned store's collection area scored 5.0 out of 5 with 6 collection checks observed and none open, so a healthy collection step is reported as healthy. The same report's speed finding on the cart drawer and filters says plainly "filter update not measured" when the scan couldn't time it. Use the scan to find leads, and confirm them with the phone test above.
Prices (re-checked live on pricing, October 23, 2026): free preview (health score, your 3 biggest leaks, full fix for one of them (a quick win); no card) · $19.99 full report, one time (the only plan; no subscription) · 14-day money back.
Scan your store free, then run the phone test on your biggest collection.
FAQ
Why aren't my Shopify filters showing?
Check in this order: does your theme support storefront filtering (Content > Menus shows a message if not, though some third-party themes don't show it); is filtering turned on in the Product grid section; does the collection have more than 5,000 products; and is the shopper using a currency other than your default (the price filter hides). After you add new attribute values, Shopify suggests allowing 24 to 48 hours for re-indexing.
How many filters should a collection page have?
There is no right number. Offer the ones shoppers in your category compare on, starting with price, rating, colour, size and brand where they apply, plus a filter for every attribute printed on your cards. Shopify allows up to 25 filters per store and shows only the filters that apply to the products in each collection.
Should I use infinite scroll or pagination?
Baymard's 2024 research still prefers "load more", and says pagination can work if done carefully. Whatever you choose, make sure Back from a product page returns shoppers to the same spot in the list.
Does quick add help conversion?
It depends on your products. Baymard now recommends quick views for visually driven products like apparel and still advises against them for spec-driven products. A quick add that adds the wrong variant is worse than none. Test it with your own traffic.
Why does the price filter disappear for some customers?
Shopify's price filter doesn't display when customers shop in any currency other than your store's default currency. This is a documented limitation, so make sure your other filters still let international shoppers narrow the list.
About the author
Saif Ur Rehman is the founder and Lead CRO Auditor of Cartvane, based in Lahore, Pakistan, with 5+ years focused on Shopify conversion rate optimization and 110+ Shopify CRO projects. On a collection page he starts with the phone test: open the biggest collection, apply two filters, tap into a product and come back. More on the About page.
Related: Shopify CRO audit checklist (91 checks) · Shopify product page CRO · Shopify product page speed and mobile LCP · Sample report · Pricing · Free scan
Sources
Baymard Institute
- Product List UX 2025: 8 Common Pitfalls & Best Practices (58% desktop / 78% mobile mediocre or worse; 21,000+ parameters, 170+ sites; 42% variations; 80% thumbnails; 67% unit price; horizontal toolbars; 14% multi-value; 51% five filters; 20% applied overview; 68% desktop / 69% mobile four sorts; updated Sep 9, 2025): https://baymard.com/research-articles/current-state-product-list-and-filtering
- Filter List Design: Have Filters for All Displayed List Item Info (38%; published Sep 17, 2019): https://baymard.com/research-articles/have-filters-for-list-item-info
- Display "Applied Filters" in an Overview (28%; mobile scroll vs stacked; avoid count-only; updated May 13, 2026): https://baymard.com/research-articles/how-to-design-applied-filters
- Return Users to the Same Place in the Product List (13%; 87%; load more and history.pushState; published Nov 17, 2020): https://baymard.com/research-articles/return-same-place
- 2024 E-Commerce Product Finding update (quick views for visually driven vs spec-driven; load more preferred, pagination okay; out-of-stock participant; published Sep 17, 2024): https://baymard.com/research-articles/product-finding-2024-launch
- 509 Mobile "Filtering Options" examples (93-site benchmark; common mobile filtering issues): https://baymard.com/mcommerce-usability/benchmark/mobile-page-types/filtering-options
Shopify
- Adding filters with Shopify Search & Discovery (theme support; 5,000 products; 100 values; price filter default currency; tag/vendor language; 25 filters; or/and logic; empty values; 1,000 search results; indexing 24–48 h; option name mismatch): https://help.shopify.com/en/manual/online-store/search-and-discovery/filters
- Add storefront filtering (Content > Menus check; third-party themes may not show the message; Product grid / Search results section setting): https://help.shopify.com/en/manual/online-store/themes/customizing-themes/common-customizations/storefront-filters
- shopify.dev, Support storefront filtering (applied filters as URL parameters; filters created in admin first; 5,000-product limit): https://shopify.dev/docs/storefronts/themes/navigation-search/filtering/storefront-filtering/support-storefront-filtering
- Editing collection details and layout (default "Most relevant"; sort options; price sort uses base prices, not Market/B2B prices): https://help.shopify.com/en/manual/products/collections/collection-layout
- Displaying unit prices for products (auto on collection pages in Online Store 2.0 themes; one unit price per product/variant; same unit across markets; legal requirement in some regions; preview rendering): https://help.shopify.com/en/manual/products/details/product-pricing/unit-pricing
Standards and method
- W3C, Understanding SC 2.5.8 Target Size (Minimum) (24 by 24 CSS pixels or spacing): https://www.w3.org/WAI/WCAG22/Understanding/target-size-minimum.html
- CXL, From Data to Test Hypotheses (five buckets; instrumentation example on filters on category pages): https://cxl.com/conversion-optimization/from-data-to-test-hypotheses/
Cartvane
- Homepage (91 checks; Collection 8 with three named checks; Speed check "Mobile LCP on home and collection"): https://www.cartvane.com
- Sample report (collection area 5.0/5, 6 observed, 0 open; SPD-07 "filter update not measured"): https://www.cartvane.com/sample
- Pricing: https://www.cartvane.com/pricing
Get your free Shopify CRO audit
Paste your store URL and see your biggest conversion leaks in minutes. 91 checks, no card required.