Group options as variants when customers see them as the same product in different forms, and split them into separate products when somebody would search for one specifically.

The decision made once and lived with

A blue jumper in four sizes: one product with a size option, or four products.

The answer seems obvious until you try it with harder cases, and the structure chosen at setup determines how the shop behaves for years.

It affects whether people find things in search, whether stock figures are meaningful, how many pages exist, and how often somebody orders the wrong thing.

It is also awkward to change later, because product addresses, stock records, and order history all depend on it.

The rule that resolves most cases

Ask whether a customer would search for the option specifically.

Nobody searches for a large blue jumper as a distinct thing. They search for the jumper and then choose a size, so size is a variant.

Somebody does search for a specific colour when the colour is why they are buying, and somebody searches for a particular capacity when that is the deciding attribute rather than a preference.

Where the option is the reason for the purchase, it deserves its own page. Where it is a detail chosen after deciding, it is a variant.

Why the distinction matters practically

The first and last pull in opposite directions, which is the whole tension. Separate products are more findable and considerably more work.

Too many options is a real cost

A product with five option types, each with several choices, produces a page where a customer has to make five decisions before adding anything to a basket.

Each decision is a place to hesitate, and hesitation at the point of purchase is expensive.

It is worse when the options are unfamiliar, because the customer now needs to learn something in order to buy, and many will decide to think about it and not return.

Where five options genuinely exist, set sensible defaults so the product can be bought without changing anything, and let people who care adjust.

The option nobody picks

Every established shop has them, and they are invisible until somebody looks.

A size that has sold twice in three years. A colour ordered once. A configuration added because a supplier offered it.

Each carries ongoing cost: stock to hold or a supplier order to place, a line in every stock check, an image to maintain, and one more decision imposed on every customer who reaches that page.

Look at sales by variant, not by product. The distribution is usually far more concentrated than anybody expects, and removing the tail simplifies the page for everybody while losing almost nothing.

Where an unpopular variant matters to a few good customers, keeping it as a made-to-order option rather than a stocked one preserves the sale without the cost.

A worked example

A workwear supplier listed each garment as a single product with three option types: size, colour, and whether it was branded.

Sales were flat and customers were phoning to ask whether particular colours existed.

Two problems emerged. Only the main product page could appear in search, so somebody looking for a specific colour found nothing, and the page showed the default colour's photograph regardless of which option was selected.

They split colour into separate products, each with its own page, name, photograph, and address, while keeping size as a variant within each.

Branding became a checkbox with a clear price addition rather than a third option type.

Search traffic to product pages rose substantially, because there were now pages matching what people actually typed, and the phone calls stopped.

Show the option's effect immediately

A common fault that costs orders and is easy to fix.

When somebody selects a variant, the page should update: the photograph, the price, the stock status, and the delivery estimate.

A page that keeps showing the default image after a colour has been chosen makes the customer unsure what they are ordering, and uncertainty at that moment produces either a phone call or an abandonment.

Stock status per variant matters as much. Telling somebody an item is in stock and then discovering at checkout that their size is not is worse than saying so on the page.

Name variants the way customers do

Internal codes leak into option lists constantly.

A dropdown offering choices like 4X-STD-BL asks the customer to decode a scheme designed for a warehouse.

Use the words a customer would use, and where a technical code genuinely matters, show both: the plain description first and the code alongside it.

This also applies to sizes that are not self-explanatory. Where your medium is not what people expect, a measurement beside the label prevents returns, which are the most expensive possible consequence of an unclear option.

The counter-case

Splitting everything into separate products is not the safe default, and doing it wholesale causes different problems.

Four sizes as four products means four pages competing for the same search, four stock records, four sets of photographs, and a customer landing on the wrong size with no obvious way to switch.

It also multiplies maintenance: a price change or a description correction now has to be made four times, and eventually they diverge.

The judgement is per attribute rather than per product. It is entirely normal for colour to be a separate product and size to be a variant within it, which is what the example above arrived at.

Getting it right

  1. For each option, ask whether somebody searches for it specifically.
  2. Split those, and keep the rest as variants.
  3. Set a default so the product can be bought unchanged.
  4. Update image, price and stock when a variant is chosen.
  5. Rename options in customer words, with measurements.
  6. Review sales by variant and remove the tail.

Step six is the one nobody runs, and it usually simplifies several pages at once.

Writing the description itself is covered in writing a product description somebody reads.

Frequently asked questions

Should sizes be variants or separate products?

Variants, in almost all cases. Nobody searches for a large blue jumper as a distinct thing; they find the jumper and then choose a size.

When should an option be its own product?

When somebody would search for it specifically, or when it is the reason for the purchase rather than a detail chosen afterwards. Colour often qualifies; size rarely does.

What is the cost of too many options?

Each option is a decision before the basket, and each is a place to hesitate. Set sensible defaults so the product can be bought without changing anything.

How do I find options nobody picks?

Look at sales by variant rather than by product. The distribution is usually far more concentrated than expected, and the tail can often become made-to-order instead of stocked.

What should happen when a variant is selected?

The photograph, price, stock status, and delivery estimate should all update. A page still showing the default image leaves the customer unsure what they are ordering.

Should I split everything into separate products?

No. That multiplies pages competing for the same search, stock records, and maintenance, and a price change then has to be made several times. Judge per attribute.

West Coast Media Solutions Inc. provides web design, web development, hosting, digital marketing, and business consulting to organisations across Canada, drawing on more than twenty-five years in the field.

Options set up years ago?

Check sales by variant, not by product. The tail is usually longer and thinner than anybody expects.

Start a Conversation