When variants aren't enough: linked products and swatches
Sometimes a colour should be a variant. Sometimes it should be a separate product that behaves like one. Choosing wrong is expensive.
A brand sells a jacket in five colours. Should that be one product with five variants, or five products linked together?
The instinct is variants. It is often wrong, and the cost of finding out late is high.
When separate products win
Different content per colour. If each colour has its own photography set, description, or fabric detail, variants force you to compromise. Shopify gives one description per product.
SEO. Five products means five URLs, five titles, five sets of ranking signals. If people search "black leather jacket", a dedicated page can rank for it in a way a variant selection cannot.
Different lifecycles. One colour discontinues, another launches. As variants that is edits to a live product with real risk of disturbing the others. As products it is independent.
Variant limits. Colour × size × material multiplies fast and there are caps. Split on the axis that behaves most independently.
The problem this creates
Five separate products means five separate pages with no relationship. The shopper on the black jacket has no idea the brown one exists. You have solved a content problem and created a discovery problem.
That is what linked products fix: separate products that present as one family, with a swatch row that navigates between them.
Swatches are not just squares
A swatch row seems trivial until you handle real catalogs:
Colour is not always a hex value. Patterns, prints and textures need an image swatch. A single beige square does not represent a floral print.
Unavailable options still need to appear. Hiding a sold-out colour makes the range look smaller than it is. Showing it struck through communicates "this exists, just not now" — which is information the shopper wants.
Order matters. Alphabetical is almost never right. Merchants want their hero colour first.
Labels do work. "Silk", "Matte Black", "Limited Edition" carry meaning a colour chip cannot.
How to decide
Ask: does each option need its own content or its own URL?
- No → variants. Simpler, and the native option is genuinely better when it fits.
- Yes → separate products, linked, with swatch navigation.
The trap is picking variants because it is easier to set up, then discovering six months in that you need per-colour descriptions and SEO. Restructuring a live catalog with existing URLs, reviews and ad spend pointed at it is far more painful than choosing correctly at the start.
Decide on the catalog you expect in two years, not the one you are launching with.