Definition
Faceted navigation is the set of filters and sort controls a store puts on its category pages so shoppers can narrow a range by size, colour, price, brand, rating or availability. Each selection usually changes the URL, and because filters combine freely, a category with a handful of facets can generate more addressable URLs than the store has products. Handling it well is one of the defining technical problems in ecommerce search.
How one category page multiplies
- Base category
- /collections/running-shoes, a page worth having in search results
- One filter applied
- /collections/running-shoes?colour=navy
- Two filters applied
- /collections/running-shoes?colour=navy&size=9
- Sort order added
- The same set of products in a different sequence, at another distinct URL
- Facets available
- Six filters with several values each, all combinable and order independent
- URLs a crawler can reach
- Tens of thousands of near-identical pages, most of them listing a handful of products or none at all
- Pages worth indexing
- The category itself, plus a short list of combinations people actually search for, such as brand within category
An illustrative structure. The arithmetic is the whole problem: nothing is broken on any individual URL, and the volume of them is what consumes crawling and dilutes the pages that matter.
Why it matters
Faceted navigation matters because it decides how much of a store's crawl and index budget is spent on pages that will never earn a visit. Left open, filter combinations flood the index with thin variations of the same list, and the versions Google keeps are often not the ones the store wants ranking. Handled deliberately, the same system becomes an asset: a small number of filter combinations map to real demand, such as a brand inside a category or a size range shoppers search by name, and those deserve proper pages with their own titles and copy. The rest should be usable by shoppers and invisible to crawlers.
Where it goes wrong
- Relying on noindex to control crawling, when a crawler has to fetch the page to read the tag, so the crawl cost is paid before the instruction is seen
- Treating the canonical link as a directive: it consolidates signals when Google agrees with it, and it does not stop a URL being requested in the first place
- Blocking parameter patterns in robots.txt without checking which filtered pages already rank, which can remove pages that were earning traffic
- Building facets on path segments rather than query parameters, which makes it far harder to describe the unwanted patterns to a crawler in one rule
- Letting filters generate links to combinations that return nothing, so crawlers spend requests on empty result pages that no shopper would ever open
Questions about faceted navigation
How should a store stop crawlers reaching every filter combination?
Decide first which combinations deserve to exist in search, then keep the rest out of the crawl path. Standard query parameters in a consistent format make the unwanted patterns describable in robots.txt, and filter controls that do not produce crawlable links reduce the problem at source. The goal is that a crawler sees the category and the handful of valuable combinations, and never discovers the long tail.
Which filtered pages are worth indexing?
Combinations with genuine search demand and enough products to be a satisfying result. Brand within a category, and attributes people search by name such as a shoe width or a specific finish, are the usual candidates. Give those a real page with its own title, heading and short introduction rather than leaving them as a filtered view that happens to be reachable.
Does faceted navigation hurt rankings on its own?
The filters themselves do not. The damage comes from what they generate: many near-duplicate URLs competing for the same terms, crawl requests spent on pages of no value, and internal link equity spread across combinations instead of concentrating on the category. Stores with small catalogues often see no effect at all, while large catalogues feel it quickly.