The received approach to a collection page is retrieval. The visitor knows roughly what they want, and the page's job is to get them to it in the fewest possible steps: four columns on desktop and two on mobile, filters open by default, sort by bestseller, quick add on the card, infinite scroll so nothing interrupts. Every one of these decisions is defensible and most of them are measurable. For a retailer with three thousand products and a customer who arrived by searching for a specific thing, the retrieval model is right, and the storefronts built on it convert well.
A premium brand is not in that position, and the collection page is where the difference costs the most. Its customer rarely arrives knowing which product they want; they arrive knowing which brand they trust, which is a different starting point with a different job attached. On a premium store the collection page is not an inventory to be searched. It is an edit: a sequence of products arranged by someone with a point of view, and the page either carries that point of view or it does not.
This is not an argument for hiding products or for ignoring what the numbers say. Every mechanism in the retrieval playbook has a place, and several belong on a luxury collection page unchanged. The argument is that the page's objective is different, and once the objective changes, the same components have to be arranged differently and judged against a different result. The collection page is where a good part of our merchandising work now sits, because it is the page most brands have never designed at all.
This article examines why the collection page is the most neglected page in premium eCommerce, how a retrieval page and an edited one are built from the same parts, why the first row carries the argument, how density and filters become brand decisions, which three theme defaults are worth reversing, and what a collection page is actually for once conversion rate is recognised as the wrong measure.
The Collection Page Is Where Most Premium Stores Hand Over to the Theme
Walk through the pages of a premium store in the order a brand designs them and the pattern is consistent. The homepage gets weeks of attention. The product page gets a template built with care, because it is where money changes hands. The campaign pages get shot, art directed and argued over. The collection page, which every one of those pages links to and which most visitors pass through more than once, gets whatever the theme ships with: a grid, a sort dropdown, a filter rail, and a heading that repeats the menu label.
The neglect is structural rather than careless. A collection page is generated rather than composed; it fills itself from the catalogue, so it never presents as empty and never demands a decision. A homepage with no content is obviously unfinished. A collection page with no design looks complete, because the products are there. That is the trap. The page looks done at exactly the point where the brand has contributed nothing to it, and the visitor experiences the theme designer's judgement rather than the brand's.
It also matters more than its neglect suggests. The product page is where a decision is confirmed, but the collection page is where the decision is formed: it determines which products a visitor sees at all, in what order, and alongside what. A weak product page loses a sale. A weak collection page loses the sales nobody knew were available, which is the larger number and the one that never appears in a report.
A Retrieval Page and an Edit Are Built from the Same Parts
It helps to be precise about the playbook before departing from it, because it is well constructed. In its mature form, seen at scaled contemporary retailers such as Aritzia, the collection page promotes adjacent subcategories through imagery at the top, surfaces the most-used filters without a click, keeps product cards above the fold, opens with a video card to start the eye moving, and interrupts the grid with editorial cards at a fixed scroll depth to hold attention. Every element is there for a reason, and the reasons are measured.
What makes a luxury collection page different from a standard product listing page?
A luxury collection page differs from a standard product listing page in its objective rather than its components. A standard listing page is optimised for retrieval: it assumes the visitor knows what they want and reduces the distance to it, through density, open filters, bestseller sorting and quick add. A luxury collection page is optimised for discovery: it assumes the visitor trusts the brand more than they know the range, and it uses sequence, pacing, imagery and restraint to show them something they did not arrive looking for. Both use grids, filters and product cards. They arrange them toward opposite outcomes.
The two pages are built from the same parts; what differs is what each is optimising for. Density, open filters and quick add all serve speed, and speed is the correct objective when the visitor's intent is fixed. A premium customer's intent is rarely fixed at the collection level. They have chosen the brand and are open on the product, which means the page's job is to shape a choice rather than to serve one. Every retrieval mechanism that shortens the path also shortens the exposure, and exposure is the thing a premium brand is trying to extend.
That reframing is the whole discipline, and it is why conversion treated as a brand quality discipline starts with the objective rather than the tactic. A team that tests quick add against no quick add without first asking what the page is for will find that quick add wins on the metric it was built to move, and will have optimised the wrong page for the wrong customer with complete statistical confidence.
The First Row of a Luxury Collection Page Is the Argument
On any collection page the first row is seen by nearly everyone and the fourth row by a minority. That is true whichever model the page follows; the difference is what each model does with the fact. Retrieval logic fills the first row with what already sells, on the reasoning that the highest-probability purchase deserves the highest-visibility slot. Edited logic fills the first row with what the brand wants the range to be known for, on the reasoning that the first row is the only part of the page that functions as a statement.
In practice the first six to eight products are the argument and everything below is supporting evidence. A well-edited first row establishes the price architecture of the collection, the material or silhouette that defines the season, the hero piece a visitor should leave remembering, and something unexpected enough to prove the range has depth. It reads as a curated shelf rather than a leaderboard, and it is done by hand, by a merchandiser or a founder, because no automated sort can know what the brand wants said.
One idea from the retrieval playbook survives this intact: a moving image in the first position does start the eye. A short clip in the top-left card, fabric moving, a garment being worn, works as an entry point for an edited grid as well as for a dense one. The difference is what it leads into. On a retrieval page it leads into the bestsellers. On an edited page it leads into the argument.
"The first row of a collection page is the brand's argument. Everything below it is evidence."
Density Is a Brand Decision, Not a Benchmark
Grid density is the decision most often made by default and most often defended with a number. Four columns on desktop and two on mobile is the received answer, and for a broad catalogue it is probably right: more products per screen means more products seen per scroll, and when the products are interchangeable, seen is what matters. The question is whether a premium brand's products are interchangeable, and whether being seen at thumbnail size counts as being seen at all.
How many products per row should a luxury collection page show?
A luxury collection page should show as many products per row as the imagery can carry without losing detail, which in practice means two or three on desktop and one or two on mobile, rather than the four and two that most themes default to. The reason is that a premium product's value lives in detail: construction, texture, drape, finish. At four columns the image is a thumbnail and the detail is gone, so the grid communicates quantity rather than quality. Fewer, larger images make each product an argument rather than an option, and the slower scroll is the point rather than a cost.
Density also governs rhythm, which is the quality most collection pages lack entirely. A grid of identical cards at identical intervals is a spreadsheet with pictures. An edited grid varies: a full-width image breaking the sequence, a pair of products presented together because they are worn together, a landscape shot among portraits. This is the discipline of pacing a storefront the way an editor paces a magazine, applied to the one page where it is most often absent.
The current instinct is to give the visitor a density toggle and let them choose, which is a reasonable answer for a retailer and the wrong one for a brand. Offering the control concedes that the layout is arbitrary. A brand with a point of view about its own products should also have one about how they sit on a page, and should hold it.
Filters Should Speak the Brand's Taxonomy, Not the Theme's
Filters are where a collection page quietly reveals whether the brand or the platform is in charge. Most theme filter rails expose whatever product attributes exist in the catalogue: size, colour, price, product type. They are accurate, comprehensive and identical to every other store on the platform, and they describe the range in the language of a database rather than the language of the brand.
Should a luxury brand show filters on its collection page?
A luxury brand should show filters on its collection page when they express the brand's own logic and hide them when they merely expose the catalogue's attributes. Size and colour are useful and should stay. Beyond those, the filters worth showing are the ones a knowledgeable customer would use to think about the range: fabric, occasion, construction, collection, origin. Filters that reproduce the platform's default taxonomy, and especially a prominent sort-by-bestseller, push a visitor toward the average choice and undo the editing above them. Fewer filters, named in the brand's own vocabulary, is the discipline.
The most damaging filter is the one that looks most helpful: sort by popularity. Surfaced by default it converts an edit back into a leaderboard with one click, and because it is the sort most visitors reach for, the merchandising effort above it is undone for the majority of sessions. The brand has designed a sequence and then offered a button that deletes it. Promoting the most popular subcategories at the top of the page is the same move in a different costume, and applied across a category it is the mechanism that makes every store converge.
None of this argues against navigation within a collection. A visitor should be able to move toward what interests them. It argues that the routes offered should be the brand's routes, named in the brand's words, and that the default view a visitor lands on should be the edited one, with sorting a deliberate second step rather than an open invitation.
Three Defaults to Reverse: Quick Add, Infinite Scroll and Sold Out
Three theme defaults deserve reversing on almost every premium store, and each is defended by a metric that is real and beside the point. All three shorten the path from grid to cart, which is exactly the path an edited collection page is trying to make worth walking.
Quick add lets a visitor put a product in the cart from the grid without opening the product page. On a retailer it lifts cart rate, because the product page adds nothing the visitor needs. On a premium store the product page is where the material, the construction, the fit notes and the price justification live; skipping it produces a cart full of products the visitor has not been sold. The result is a higher add-to-cart rate and a lower purchase rate, with the gap appearing as abandonment that gets blamed on checkout.
Infinite scroll removes the natural end of a page, and with it the moment where a visitor decides to look at something properly rather than look at more. For a retailer, more is the goal. For an edited collection the end matters: it is where the sequence concludes and the visitor's attention returns to what they saw. Pagination, or a load-more control, restores that pause. It costs a click and buys a decision.
Sold-out products are, by default, hidden, pushed to the bottom, or greyed out as failures. On a premium store a sold-out piece is evidence: of demand, of a range that moves, of a season that existed. It belongs on the page, marked honestly, with a route to be notified or to see what replaced it. Removing it removes the archive one product at a time, and the archive is what a returning customer uses to judge whether the brand has been consistent or merely fashionable.
We have built collection pages on these principles for Rat & Boa, as part of a mobile-first rebuild of the whole storefront, and for Ramy Brook, within an ongoing merchandising and UX retainer during which the brand's sales grew 200 per cent in twelve months and new customers by 45 per cent year on year. Neither figure belongs to the collection page alone; a retainer touches everything, and honest attribution says so. What the collection page contributes on both stores is the first decision a visitor makes about the range, and on both it was designed rather than defaulted.
"A collection page is measured by which products get seen, not by how quickly a visitor leaves it."
What a Collection Page Is Actually For
Everything above depends on a single change of measure, and without it none of the rest survives contact with a reporting dashboard. Judged by its own conversion rate, an edited collection page will usually look worse than a retrieval page, because it deliberately slows the visitor down and sends them to product pages rather than the cart. Judged by what it is for, it looks very different.
How should a premium brand measure collection page performance?
A premium brand should measure collection page performance by reach and depth rather than by conversion rate. The measures that describe the page's actual job are the share of the range receiving product page views, the number of product pages visited per session that began on the collection, the scroll depth at which visitors leave, and the full-price sell-through of products positioned deliberately rather than by bestseller sort. Conversion rate on the collection page itself measures how quickly visitors left it, which for an edited page is close to the opposite of the intent. The page succeeds when more of the range gets seen by the right customer.
The clearest recent example is Mina Baie, a premium bag brand for new parents, where a new collection page went live in March 2026. In the five months that followed, sessions landing on collection pages rose 83 per cent on the same months a year earlier, and orders from those sessions rose 90 per cent. Conversion rates normally dilute as traffic scales, since each additional visitor arrives colder. Here both add-to-cart and conversion held, and the additional volume converted at a slightly better rate than the old page managed with a third less traffic. The page did not create the demand. Its job was to carry it without losing quality, and it did.
This is the same logic as optimising for customer quality over conversion rate, applied to one page. A collection page that converts efficiently by sending everyone to the same six products has served the average visitor. A collection page that spreads attention across the range has served the customer the brand exists for, and the difference shows in what they buy, what they pay for it, and whether they come back.
Merchandising the Grid by Hand
How should a luxury brand design its collection page?
A luxury brand should design its collection page as an edit rather than a catalogue. That means sequencing the first row by hand so it makes an argument about the range, setting grid density by what the imagery can carry rather than by a benchmark, naming filters in the brand's own vocabulary and keeping the default view edited, reversing quick add, infinite scroll and hidden sold-out products, and measuring the page by how much of the range gets seen rather than by its own conversion rate. Each decision is small. Together they are the difference between a page the brand designed and a page the theme generated.
1. Sequence the first row by hand. The first six to eight products are the only part of the page that functions as a statement. Choose them for what they say about the range, not for what sold last month, and revisit them with every drop.
2. Set density by the imagery. Two or three columns on desktop, one or two on mobile, at whatever size lets the detail read. Vary the rhythm with full-width breaks and paired products, and hold the layout rather than offering a toggle.
3. Name filters in the brand's vocabulary. Keep size and colour, add the routes a knowledgeable customer would actually take, and make sorting a deliberate second step rather than a default that deletes the edit.
4. Reverse the three defaults. Route every add through the product page, replace infinite scroll with a pause, and keep sold-out products visible as evidence rather than hiding them as failures.
5. Measure reach, not exit. Share of range viewed, product pages per session, scroll depth and full-price sell-through describe what the page is for. Its own conversion rate describes how fast people left, which is the wrong question.
At DBCO the collection page is where a good part of our merchandising work now happens, because it is the page most likely to have been left to the theme and the one with the most to give back. Brand-led Shopify Plus design and build for premium fashion, beauty and lifestyle brands that would rather their range was seen than sorted. If you are rebuilding yours, we would welcome a conversation.
