Skip to content
Faisal Mahmud TusherDigital marketing, SEO & analytics
with Faisal Tusher

Conversion Rate Optimisation › Funnel & Checkout

Category and Listing Pages

How category and listing pages help shoppers choose: product cards, filters and sorting, default order, load more or pages, empty results, phones, measurement.
Guideintermediate12 min readUpdated

What a listing page is for

A listing page sits between arrival and the product page. Shoppers land on it from search, from navigation and from campaigns, and they leave it either by choosing a product to look at or by giving up. The e-commerce conversion hub describes the conversion side of navigation and filters in a single test: whether a buyer can get from a category of hundreds of products to the few that suit them in a handful of taps on a phone. Everything on the page either shortens that path or lengthens it.

The same template serves several kinds of list: a category, a subcategory, a brand page, a sale, a search results page and a collection built for a campaign. Each is a list of products the shopper has to narrow, so the same principles apply, although the default order may differ. For a collection built for a campaign, the guide to merchandising the store for a campaign covers choosing its products, curating its order and taking it down when the campaign ends. A search results page, for instance, already knows what the shopper typed, which a category page has to infer.

The listing page also has a search engine audience, and the two jobs are related but separate. The category and product page guide covers what a category page needs to rank: a title that names the range, introductory content, crawlable links to products and subcategories, breadcrumbs and pagination handled as Google documents it. The faceted navigation guide decides which filter combinations deserve to be indexed. This guide takes those decisions as settled and asks what helps the shopper choose. Most of the time the answers agree; where they seem to conflict, both guides' requirements can usually be met at once.

Product cards that answer the first questions

A product card is a small advert for a product page. It has to give the shopper enough to decide whether the product is worth a click, and the attributes that matter are the ones that decide the purchase in that category: size and fit for clothing, capacity for storage, compatibility for spare parts, strength or grade for tools. The category SEO guide expects prices and availability in text on the grid, and as general practice a card carries:

  • An image that shows what distinguishes the product, taken the same way across the range so cards can be compared at a glance.
  • A name that says what the product is, not only a model code.
  • The price, including any saving shown honestly, and the price per unit where buyers compare that way.
  • The deciding attributes, one or two of them, in the units buyers use.
  • Availability, in words: in stock, low stock, back soon, out of stock.
  • A rating and review count where reviews exist, which the reviews and social proof guide places on category pages and search results as well as on product pages.

What a card should not carry is anything that competes with those: badges that every product has, promotional text that repeats the banner, or a colour swatch row so long it pushes the price out of view. When cards are too sparse, shoppers open product pages to find basic facts and return to the list, a pattern the product page optimisation guide recognises from the other side: the product page then has to answer questions the card could have settled.

Filters built around buying decisions

Filters do for browsers what search does for people who know what they want. The e-commerce conversion hub sets out what good filters do: they expose the attributes that decide the purchase, show counts, can be combined and removed, and do not hide items that are low in stock. Categories themselves should be shaped around how buyers think, by use, recipient, room or size, rather than how the warehouse is organised.

Filter propertyWhat it means for the shopperCommon failure
Built from buying attributesThe filters match the questions shoppers ask about the categoryFilters inherited from the product database, such as supplier or internal range code
Counts beside each valueThe shopper sees how many products remain before choosingValues that lead to zero results with no warning
CombinableTwo or more values narrow the range togetherChoosing one value clears the others
RemovableEach applied filter can be removed on its own, and all can be cleared at onceThe only way back is the browser's back button
Applied filters visibleThe shopper can see what is narrowing the listFilters applied out of sight, so the list looks short for no reason
Stock-awareLow-stock items stay visible; sold-out items can be hidden by choiceFilters that silently drop low-stock items

Which attributes become filters is a research question. The site search log shows the words shoppers use, and the site search analysis tutorial shows how to read it; searches for an attribute the filters lack are a direct request for a filter. Questions to customer service and returns reasons add the attributes buyers discover too late. The Faceted Navigation entry in the glossary explains the term, and the faceted navigation guide decides which filter pages search engines should see; a filter can be essential to shoppers and still be kept out of the index. Shoppers who know their situation but not the attribute names the filters use are better served by a product finder quiz, which asks about the situation and sets those attributes for them.

Sorting and the default order

Most shoppers never change the sort order, so the default order is a merchandising decision with large consequences. It should follow a rule the business can state, such as relevance, best sellers in the category, newest first for a range bought for novelty, or a curated order a merchandiser maintains. A default with no rule drifts towards whatever the platform does when nobody decides.

Whatever the rule, products that cannot be bought should not occupy the first positions. The e-commerce conversion hub names results that put out-of-stock items first among the mundane failures of on-site search, and the same failure on a category page wastes the most valuable positions on the page. As general practice, products that are out of stock stay in the list, clearly labelled, but move to the end of the default order.

Evidence for the order comes from the list itself. The merchandising analysis guide reads selections by position: a list with steep decay by position shows that shoppers rarely look past the first row, which makes the sort order the merchandising decision rather than the list's contents. The sort options offered to shoppers should match how they compare in that category, such as price, rating or newest, and should be labelled in plain words.

Load more, infinite scroll or numbered pages

A long range has to be split, and the shopper's experience is the deciding factor here. Each option has a character, described in the table as general practice.

OptionSuitsShopper's experienceWatch for
Numbered pagesRanges shoppers compare carefully and return toA known position, easy to come back toSmall page sizes that force many page loads
Load more buttonMost rangesControl over when more products appear, with the footer still reachableThe button lost below a long grid on phones
Infinite scrollVisual ranges browsed for inspirationNo interruption while browsingA footer that cannot be reached; a lost position after visiting a product

Two experience points apply whatever the choice. A shopper who opens a product and returns should land where they were, not at the top of the list. And the shopper should be able to tell roughly how large the range is, from a count of results, so they know whether to keep scrolling or narrow it with a filter.

The search side has a documented answer that leaves the shopper's choice open. Google's guidance on pagination and incremental page loading, as the category SEO guide states it, is that each page of a list has its own address, pages link to each other with crawlable links, later pages are not canonicalised to the first, and a load more button or infinite scroll also updates the address and exposes the paginated addresses so every product can be reached. The Pagination glossary entry summarises the same points. What must not happen is for products to be reachable only through script.

Empty results and unavailable products

Every listing page eventually shows nothing: a combination of filters that matches no product, a category whose stock has run out, a search for something the store does not carry. A blank grid ends the journey, so as general practice an empty result offers a way forward: say which filters caused it, offer to remove the most restrictive one, show the nearest matches or the most popular products in the parent category, and keep the search box and the navigation in view.

Unavailable products need the same care. A product that is out of stock for now stays visible with its status in words and, where the business offers one, a way to be told when it returns; one that is gone for good gives way to alternatives. The status codes and indexing side of empty categories and sold-out products belongs to the soft not-found errors guide, which recommends filling, merging or removing empty listings; the conversion side is that the shopper who reaches one is offered something to do next.

Listing pages on phones

On a phone the listing page is narrow, filters compete with products for the screen, and most shopping happens with a thumb. The speed and mobile guide sets the layout rules: actions within reach of the thumb, tap targets large and separated, nothing that depends on hover, and a layout that does not shift as images load. For a listing page that translates into a few practical choices.

  • A filter button that is always reachable, usually fixed near the top or bottom of the screen, opening a panel that shows counts and applies without losing the shopper's position.
  • Applied filters shown above the grid as removable labels, so the shopper can see and undo them.
  • Cards that keep the deciding facts at phone width, rather than dropping the attribute line to save space.
  • Reserved space for images, so the grid does not jump while the shopper is trying to tap a product.

Phones also expose weight. A grid of large images loaded all at once is slow on a mobile connection, and the speed guide covers page weight and third-party tags; on listing pages the usual culprits are oversized images and scripts that load before the products do.

Measuring a listing page

The GA4 e-commerce developer guide documents two events for lists. view_item_list records that a list of products was shown, with the list's identifier and name and each item's position, and select_item records that a shopper chose a product from it. The e-commerce tracking tutorial sets them up on category pages, search results and carousels, and recommends that lists rendering more items as the shopper scrolls send another view_item_list for each newly shown batch, named the way a buyer would recognise the list.

From those events the listing page has its own small funnel: list views, selections from the list, products from the list added to the basket, and purchases credited to the list. The merchandising guide's measures are the ones to read: selection rate, basket rate from the list, purchases from the list, and selections by position. The Add-to-Cart Rate for sessions that included the category, read beside them, connects the list to the rest of the journey. The funnel analysis tutorial builds the wider funnel from the same events, and segmenting by device usually shows whether phones are the problem. The funnel and checkout section sets out the biggest-leak rule for deciding whether the listing step is the place to work at all.

QuestionMeasureWhat a weak reading suggests
Do shoppers find the list at all?List viewsNavigation or search does not lead there
Do the cards earn a click?Selections over list viewsCards lack the deciding facts, or the order is wrong
Does the click match what they wanted?Basket adds from the listCards promise something the product page does not deliver
How far down do shoppers look?Selections by positionThe default order decides what sells
Do filters help?Filter use, and selections after filteringFilters built on the wrong attributes

Compare periods carefully. A listing page's results move with the season, the range and the campaigns pointing at it, so before-and-after readings are directional, and a controlled test is the way to prove a change where traffic allows. The conversion rate calculator turns the counts into comparable rates (CVR) for each period or arm.

A stationery shop rebuilds its pen category (illustrative)

Illustrative example — not a real result

Inkthimble Stationery rebuilds its pen category

Inkthimble Stationery, an invented online stationery retailer, lists 260 pens in one category. The cards show a photo, a name and a price; nib size and ink type, the two questions customers ask most, appear only on product pages. The filters are brand, price band and colour. The default order is "newest", and because restocks count as new, 11 of the first 24 positions hold sold-out pens. Over four weeks the list records 31,000 list views, a selection rate of 16%, and 1,180 products added to the basket from the list. The site search log shows 640 searches for "fine nib", "fountain" and "gel" in the same period. The team adds nib size and ink type to every card, rebuilds the filters around nib size, ink type, refillable or not and price band, each with counts, keeps low-stock pens visible and moves sold-out pens to the end of the default order. It also adds a "back soon" label. In the four weeks after the change, list views are similar at 30,400, the selection rate is 21%, and basket adds from the list reach 1,520. Selections by position flatten slightly, with more shoppers reaching the third row. The team treats the comparison as directional because a back-to-school campaign started in the second period, and plans a controlled test of the card layout on its notebook category.

Common mistakes

  • Cards without the deciding facts. Shoppers open product pages to find basic attributes and give up after a few.
  • Filters from the product database. Build them from the questions buyers ask, not the fields the system has.
  • Values that lead nowhere. Show counts so shoppers do not choose a filter that returns nothing.
  • Sold-out products in the first rows. Keep them visible and labelled, at the end of the default order.
  • A default order nobody chose. Write down the rule and check it against selections by position.
  • Losing the shopper's place. Returning from a product page should land where the shopper left the list.
  • A blank page for an empty result. Offer the nearest matches and the filters to remove.
  • Hidden filters on phones. Keep the filter button within reach and the applied filters visible.
  • Products reachable only through script. Pair any scrolling or load more pattern with real paginated addresses.
  • Measuring clicks without baskets. A list that earns clicks but not basket adds is attracting curiosity rather than intent.

Frequently asked questions

Is infinite scroll better than numbered pages for a category page?

Neither is better in general. Infinite scroll suits browsing ranges where shoppers look at many similar items, numbered pages suit shoppers who compare and come back to a position, and a load more button sits between them. Whatever the shopper sees, Google's pagination guidance, as the category SEO guide states it, expects each page of the list to have its own address reachable through crawlable links, so the choice for shoppers does not have to compromise how the products are found.

Should out-of-stock products appear on category pages?

Usually yes, but not at the top. Removing them hides products that will come back and confuses shoppers who saw them before, while leaving them in the first rows fills the most valuable positions with things nobody can buy. As general practice, keep them visible and clearly labelled, move them to the end of the default order, and give shoppers a way to hide them with a filter. Products that are gone for good follow the availability rules in the category SEO guide.

How many filters should a category page have?

As many as there are attributes that genuinely decide the purchase in that category, and no more. The test is whether shoppers use the filter to narrow the range towards what they want; a filter almost nobody uses, or one that returns the same products whichever value is chosen, adds clutter. The site search log and questions to customer service show which attributes shoppers ask about, and filter usage in analytics shows which ones earn their place.

Terms used on this page

Tools to try

Where this page is used

Sources

  1. Measure ecommerce (GA4 developer guide) — Google for Developers
  2. Pagination and incremental page loading — Google Search Central

Related