usdigitalmarketingindex.com SEO
When should ecommerce product variants have separately indexable URLs?
Updated November 15, 2026
When a variant is, for a shopper, a different thing to search for and to buy: it is looked for by its own name, it differs in more than appearance, and its page can carry content the other variants’ pages do not. In every other case the variants should each have a URL, and those URLs should point to one canonical product page. That second arrangement is Google’s stated recommendation. Its ecommerce guidance says to “make sure that each variant can be identified by a separate URL”, and, for products with unique URLs per variant, to include the canonical product URL on all variant pages in a canonical link tag (Google’s guidance on URL structure for ecommerce sites). Having a URL and being indexed separately are different decisions. The first is the default. The second needs a reason.
Meaningful independent demand
The first test is whether people look for the variant itself.
| Evidence | Where it comes from | Points toward a separate indexable page when |
|---|---|---|
| Search queries naming the variant | Search Console query data for the product page | Queries include the variant's distinguishing term and are specific to it |
| On-site search terms | The store's own search logs | Shoppers type the variant's name |
| Sales distribution | Order data | The variant sells as though it were its own product |
| How the manufacturer lists it | The manufacturer's catalog | It has its own model name or number |
| Customer questions | Support records | Questions concern the variant specifically |
Use the store’s own data. A variant that nobody asks for by name gains nothing from an indexable page of its own.
Variant-specific content and availability
The second test is whether the variant’s page would differ in substance.
| Difference between variants | Substantive? |
|---|---|
| Color only; same description, same specifications, same price | No |
| Size only, in clothing | No |
| Storage capacity, with different prices | Partly; specification and price differ, description does not |
| Different function: a corded and a cordless version | Yes |
| Different compatibility: a part made for different vehicle models | Yes |
| Different material with different care, performance, or use | Yes |
| Different regulatory status or age restriction | Yes |
A page that would repeat its sibling word for word, apart from one attribute, is a near-duplicate. Making many of those indexable multiplies URLs without adding anything a searcher could choose between.
Availability matters as well. A variant that is in stock for a week each year is a weak candidate for a permanent indexed page of its own.
Canonical and navigation choices
Google’s guidance offers two URL structures for variants: a path segment, such as /t-shirt/green, or a query parameter, such as /t-shirt?color=green. For the parameter form it adds: “use the URL with the query parameter omitted as the canonical URL”.
The two arrangements, set side by side:
| Element | Consolidated (default) | Separately indexable (exception) |
|---|---|---|
| Variant URL | Exists; selects the variant | Exists; a full page for the variant |
| Canonical tag on the variant URL | Points to the main product URL | Points to itself |
| Sitemap | Main product URL only | Variant URL included |
| Internal links | To the main product URL | To the variant URL where the variant is the subject |
| Title and description | The product's | The variant's own |
| Structured data | A ProductGroup with variants | A Product for the variant, related to its group with isVariantOf or inProductGroupWithID |
| Content | Shared | Distinct sections for what differs |
Whichever is chosen, the signals must agree. The guidance asks sites to “Use the same URL in internal links, sitemap files, and” canonical tags. A variant URL that carries a self-referencing canonical while the sitemap lists only the parent, and the structured data nests it under the parent’s URL, tells Google three different things.
Navigation follows the same decision. Consolidated variants are reached through the selector on the product page. Separately indexed variants also appear as their own tiles in category listings where a shopper would expect to find them.
Examples of color versus functionality
A product variant ownership matrix for a hypothetical catalog:
| Product | Variant attribute | Independent demand | Substantive difference | Decision | Canonical |
|---|---|---|---|---|---|
| Cotton T-shirt | Color (8) | No | No | Consolidate | Main product URL |
| Cotton T-shirt | Size (6) | No | No | Consolidate | Main product URL |
| Laptop model | Storage: 256 GB, 512 GB, 1 TB | Some | Specification and price | Consolidate; show each configuration clearly | Main product URL |
| Drill | Corded and cordless | Yes | Function, power source, use | Separate products | Each its own |
| Brake pad | Vehicle fitment (40) | Yes, by vehicle | Compatibility | Separate pages per fitment group, if each has its own fitment data | Each its own |
| Paint | Color (200) | For a few named colors | Appearance only | Consolidate; consider separate pages only for colors searched by name, with real content | Main, with named exceptions |
| Phone | Color | No | No | Consolidate | Main product URL |
| Phone | Model tier: standard and pro | Yes | Specification, camera, size | Separate products | Each its own |
Conditional decision rules:
- Every variant gets a URL that selects it.
- If the variant differs only in appearance or size, consolidate.
- If the variant differs in function, compatibility, or specification, and shoppers search for it by name, treat it as a product with its own indexable page.
- If demand exists and the page would have nothing distinct to say, consolidate and make the variant easy to find on the main page.
- If the count of variants is large, the default is consolidation. Exceptions are named one by one, not generated by rule.
- Whatever the decision, the canonical tag, sitemap, internal links, and structured data all express it.
Checks after implementation:
| Check | Pass condition |
|---|---|
| Canonical on each variant URL | Matches the decision for that variant |
| Sitemap | Contains only the URLs meant to be indexed |
| Structured data | Group and variant URLs consistent with canonicals |
| URL Inspection on a consolidated variant | Google's selected canonical is the main product URL |
| URL Inspection on a separate variant | Google's selected canonical is the variant URL |
| Index coverage over time | No growth in near-duplicate variant URLs |
A canonical tag is a signal Google weighs with others, so the inspection step is the confirmation. If Google selects a different canonical from the one declared, the signals disagree somewhere.
These criteria, independent demand and substantive difference, are this guide’s reading of when to depart from the default. Google’s documentation states the default and the mechanisms. It does not prescribe which variants a store should index.
The reverse reorganization, several separate SKU records becoming one product with variants, raises its own questions about URLs and redirects, covered in the guide to managing SEO when several individual SKUs become one consolidated product.
The rule: a URL for every variant, one canonical product page by default, and a separate indexable page only for a variant that shoppers seek by name and that has something of its own to say.
Add your comment to this answer