Color names, size runs, and the Shopify filters that hold both

Why does one color arrive under a dozen names?

Because four different people named it. Your buyer wrote Navy, one supplier wrote Midnight, another wrote Indigo, and a third sent a spreadsheet saying Dark Blue. Every one of those becomes its own row in the filter, and the rail dutifully offers all four.

A shopper who wants a navy dress ticks Navy, sees a quarter of what you stock, and decides you do not have much in navy. Nothing in that sentence is a bug. The filter did exactly what it was told.

Shopify solves this. You can group values, and Shopify's own documentation uses the example a dress shop would pick: Onyx, Ebony and Midnight, collapsed into one entry named Black. Do that once per family and the panel starts telling the truth about your range.

Two things it will not do. The Category filter cannot be grouped at all. And grouping is manual, so a new supplier with a new word for red is a new value sitting outside the family until somebody notices.

How many colors should a shopper actually see?

Fewer than you have. A single filter shows a shopper at most 100 values. The admin shows you up to 1,000, so a color list can look complete on your screen and quietly stop short on the storefront.

The ceilings on grouping are generous: 200 values inside one group, 1,000 groups across the store. You will run out of judgement before you run out of groups. Ten or twelve color families is a rail a person can read. Forty entries is a second catalog.

Why does picking a size still leave a thousand dresses?

Because the run is lopsided. Put the number of styles beside each size and the shape appears: say a middle size carries a thousand styles and the last size in the run carries three. One filter row is a department. The row beneath it is a shelf.

That shape has two consequences. A shopper in a middle size needs more filters after the size, not fewer, because size alone still leaves them a thousand dresses. A shopper at the end of the run needs the opposite: all three on one screen, and some sign of which are still in stock before clicking.

Sizing itself Shopify handles well. A filter named Size gets a sort order built for clothing, so extra small comes before small instead of alphabetically after it, and you do not drag anything into place. Sold-out sizes are a different problem, with their own page on the size filter and stock: that fix is about availability, not about the run.

Where do occasion and silhouette live?

Nowhere, until you put them somewhere. Nobody shops a formal store by fabric composition. They shop it by the event they are dressing for, the shape they want, the neckline, the length, and sometimes the designer. None of those is a product option like size or color, so none of them exists as a filter until you build it from a field you fill in yourself or from tags.

This is where the 25 goes, standard and custom counted together. Price, availability, size, color and designer is five. Occasion, silhouette, neckline, length, sleeve, train, embellishment and fabric is eight more, and you are at thirteen before a single other category of product has asked for anything.

So choose. The facets worth the budget are the ones shoppers reach for first, and the order they sit in is a merchandising decision, not a technical one. A rail of six facets people use beats a rail of twenty they scroll past.

What happens when the collection passes five thousand?

The filters go. Not slower, not fewer: a collection holding more than 5,000 products displays none at all, and nothing in the admin warns you first. Shopify's own advice is to split the collection into smaller ones.

For most catalogs that advice is sound. For a dress shop it is expensive, because the big collection is usually the one that ranks and the one every campaign points at. Splitting it into four means four thinner pages, four sets of links to rebuild, and a shopper who has to guess which one holds the dress.

You have three honest options. Split, and accept the cost. Point shoppers at pre-filtered smaller collections instead, which is the per-collection approach taken to its limit. Or move the filtering somewhere that has no cliff.

What changes if the filters do not come from Shopify?

Three things, in order of how much they matter to a dress catalog. The 25 goes, so occasion and silhouette stop competing with size and color for a slot. The 5,000-product cliff goes, so your largest collection keeps its filters and keeps its ranking. And counts stay live as a shopper narrows, so a combination with nothing behind it reads as a zero before anybody clicks into an empty page.

Sledge filters read sizes, colors, tags and the details you fill in yourself as equal sources, and they group values the way Search and Discovery does. They draw the grid as well as the rail, which is how they keep working on a collection Shopify would leave bare. Personalization is included from the entry plan, so the order of the rail can follow what a shopper has already looked at.

None of that removes the naming work. Somebody still has to decide that Onyx and Ebony are both black. The tooling decides how many of those decisions you are allowed to make.