A product variant is one specific, sellable version of a product. A T-shirt sold in three sizes and four colors is one product with twelve variants: "M, black" is one of them. The product carries what is shared, the name, the description, the main photos. Each variant carries what differs: its SKU, its price if it differs, its weight, its stock and, when it looks different, its own photo.

It concerns anyone selling a physical product that comes in more than one form: clothing, jewelry, prints in several sizes, candles in several scents, digital products in several editions. Getting variants right is the difference between a product page where a customer picks "M, black" in two taps and a store with twelve near-identical products nobody can find.

What is a product variant?

The structure has three levels.

  • The product is the thing you sell: "Organic cotton T-shirt".
  • The options are the dimensions along which it varies: size and color. Each option has values: S, M, L for size; black, white, navy, sand for color.
  • The variants are the combinations that actually exist: every pair of one size and one color you sell. Three sizes times four colors gives twelve variants, unless some combinations do not exist, in which case you simply do not create them.

Each variant is what the customer buys and what you count. It has its own SKU, its own stock number, and it appears as one line on an order. Price, weight and image can be shared with the product or set per variant. A larger size may cost more; a heavier material may ship differently.

A variant is not an accessory or a bundle. "T-shirt plus tote bag" is a different product or a bundle, not a variant. It is not a personalization either; "engrave a name" is an option that does not create separate stock. And it is not a synonym for SKU: the SKU is the code that identifies the variant, the variant is the thing itself.

Why it matters

Two people care about variants: the customer and you.

For the customer, variants turn a decision into a choice. Instead of twelve product tiles labeled "T-shirt black M", "T-shirt black L" and so on, they see one product, pick a size, pick a color, and see the price and availability for that exact combination. A well-built variant selector also tells them what is out of stock before they add it to the cart, which prevents the most common cause of abandoned checkouts on clothing stores.

For you, variants are the unit of stock. Take a small brand that receives 120 T-shirts: 10 units of each of the 12 variants. If stock is tracked at the product level, the store shows "120 in stock" long after "M, black", the most popular one, is gone. Customers order it, you cannot ship it, and you refund 15 orders a month at $28 each while the store keeps saying it is available. Tracked per variant, "M, black" shows sold out at the eleventh order, and the other eleven variants keep selling. The same brand also learns which sizes and colors to reorder: 10 of "M, black" sold in four days, 10 of "S, sand" in four months.

How it works

Setting up variants is mostly a matter of deciding options and values before creating anything.

  • List the options that change stock. Size, color, material, scent, edition. If a choice does not change what you ship or what you count, it is not a variant option.
  • List the values for each option. Keep them short and in the order customers expect: XS to XL, not alphabetical.
  • Generate the combinations. Most tools create every combination automatically. Delete the ones you do not sell.
  • Give each variant a SKU. A simple pattern works: TSH-BLK-M for T-shirt, black, medium. The SKU is what your orders, stock and reports use.
  • Set stock per variant. Enter what you actually have for each combination. This is the number that decides "add to cart" versus "sold out".
  • Set price and weight where they differ. A $2 surcharge on XL, a heavier weight for the ceramic version. Leave the rest inherited from the product.
  • Add photos for variants that look different. Colors need their own image; sizes usually do not.
  • Let every channel read the same variants. The link-in-bio page and the store should show identical options and stock, which happens on its own when both use one shared catalog.

Benchmarks and examples

Numbers that help decide how far to go:

  • Options per product: 1 or 2 is ideal, 3 is the practical maximum. Beyond that, the selector becomes a puzzle.
  • Values per option: up to 6 or 7 for colors, up to 8 for sizes. Past that, split the product into two.
  • Variants per product: 12 to 24 is common for apparel; 50 and above is a warning sign that a product should be split.
  • Stock per variant at launch: for a first run, 5 to 15 units of each variant, with more of the middle sizes. Reorder from real sales.
  • Sold-out rate by variant: a brand where one variant sells out in days while another lasts months has learned more from variants than from any survey.

Examples. A print seller offers one image in three sizes: A4 at $25, A3 at $40, A2 at $65. One product, one option, three variants with three prices and three stock counts. A candle maker has four scents in two jar sizes: eight variants, with the large jar priced higher and weighing more for shipping. A creator sells a preset pack as a digital product in two editions, "Lightroom" and "Mobile", which are two variants with different files delivered after payment and no stock at all. A ceramics studio makes each piece by hand in small batches; they use one product per shape and a "glaze" option, with stock of 1 to 3 per glaze, updated after each kiln.

Common mistakes

  • One product per variant. Twelve products called "Tee black M", "Tee black L" and so on. The catalog is unreadable, the store is unbrowsable, and stock is right only by accident.
  • Variants for choices that do not change stock. "Gift wrap: yes/no" doubles every variant for nothing. Use an option that does not create variants, or a separate add-on product.
  • Inconsistent option order and naming. "Colour" on one product, "Color" on another; sizes listed S, L, M. Customers slow down, and filters break.
  • Tracking stock at the product level. The store says "in stock" while the popular size is gone. This is the most expensive mistake on the list.
  • No photo for color variants. A customer choosing "sand" with a black T-shirt still shown hesitates, and hesitation is a lost sale.

Best practices

  • Name options the way customers say them. "Size", "Color", "Scent". Values like "Medium" or "M", not "Size 2".
  • Keep the same option order across the store. Size first, then color, on every product. Customers learn it once.
  • Show availability in the selector. Grey out sold-out combinations before the cart, not at checkout.
  • Use a SKU pattern and stick to it. Product code, then option codes, in the same order everywhere. It makes counting and reordering fast.
  • Photograph what differs. Every color, every material. Skip separate photos for sizes unless proportions change.
  • Split products that get too big. A product with 60 variants is two or three products. "Classic Tee" and "Classic Tee Kids" convert better than one tee with 8 sizes and 8 colors.
  • Read sales by variant monthly. Reorder what sells out, drop what does not move, and adjust the quantities of the next batch. See inventory.

In Roctify

Products in Roctify support options and variants, and each variant carries its own SKU, price and stock. Stock is tracked per variant and deducted when an order is paid, on whichever channel took it, so the link-in-bio page and the online store always show the same availability for "M, black". Both read the same shared catalog, so there is nothing to sync by hand.

Physical variants use your shipping and tax settings like any other product. Each order line shows the exact variant and its SKU, which is what your picking and your reorders rely on, and Pro exports let you work with that data in a spreadsheet.

FAQ

What is the difference between a variant and a SKU?

The variant is the version of the product, "M, black". The SKU is the code you give it, "TSH-BLK-M", so that orders, stock counts and reports can refer to it without ambiguity. Every variant should have exactly one SKU. A product with no options has one variant, and therefore one SKU.

How many variants is too many?

When the selector stops being helpful. Two options with up to six or seven values each is comfortable. Three options is the practical limit. If a product needs 50 or more variants, split it: by audience, by material or by range. Smaller products with fewer variants are easier to browse, photograph and count.

Should each variant have its own price?

Only when the cost genuinely differs. Larger sizes, heavier materials or premium finishes justify a different price and customers accept it. Pricing every variant differently for no reason confuses the page. Set the price on the product and override it on the few variants that need it.