Ecommerce filter SEO: which pages should Google index?

Decide which product filters deserve search landing pages, which should stay browsing tools, and how to check crawling, noindex and canonical rules.

8 October 2026

Areza Digital

A shopper chooses a product type, a colour and a price range. Your shop gives them a useful shortlist. Should that combination also become a page in Google?

Sometimes. A useful starting rule for ecommerce filter SEO is to give a search landing page a clear job: help someone choose products for a recognisable need. A temporary combination of browsing preferences does not automatically need that role.

Start with one category. Write down which pages you would confidently send to a new customer and which controls simply help that customer narrow the list after arriving.

Choose the customer question before the URL

For each proposed landing page, ask what the customer is trying to buy. “A backpack that fits my laptop” suggests a different selection task from “show the same backpacks in ascending price order”.

Review customer enquiries, on-site searches and the search queries already bringing people to your shop. Treat a proposed category as a hypothesis until those sources support it. A keyword tool’s suggestion alone does not tell you whether your catalogue can answer the question well.

Before promoting a filter to a landing page, check that:

• The product selection matches a need you can describe plainly.
• Product attributes are reliable enough to support that selection.
• The page helps someone compare the available options.
• Someone owns the page when the range changes.

If “fits a laptop” depends on measurements you have not collected, fix the product information first. Generating more category text will not fix the selection.

An illustrative example: a backpack shop

Imagine a small shop selling backpacks. Its filters include intended use, colour and capacity, plus a price-sort control. This is an illustrative example, not an Areza client case or a claim about search demand.

The team reviews three possible entry points:

Swipe or scroll to compare.

For the first candidate, the team would write a short explanation of fit and show the relevant dimensions alongside products. It would also link to that selection where shoppers naturally look for it.

The colour and sort controls would remain available. The team would decide their search treatment separately, rather than removing useful shopping controls to simplify SEO.

Give each technical setting one job

Faceted navigation means combining filters to narrow a list. Google’s guidance warns about excessive crawling of filter URLs and recommends preventing crawling where those URLs are not needed in search. Source: Google on faceted navigation.

That is a design decision, not a ready-made cleanup plan for URLs already in search. Separate these questions in the implementation brief:

Should this page appear in search? A noindex rule can exclude a page, but Google must be able to crawl it and read the rule. Blocking it in robots.txt at the same time prevents that check; a blocked URL can still appear in search. Source: Google on noindex.

Is this a duplicate of another page? A canonical link expresses the preferred URL for duplicate or very similar content. It is a signal, not a guaranteed selection. Avoid declaring a useful, distinct product category to be a duplicate merely because both pages contain products. Source: Google on canonical URLs.

Does the crawler need to visit these combinations? Review crawl prevention separately from removing indexed pages. Google’s faceted-navigation guidance describes canonical links as less effective for controlling crawling than blocking unwanted URL patterns.

For an existing shop, have the implementation team record the current indexing state and intended outcome before changing these rules. A blanket rule for every URL containing a question mark is too broad a brief.

Keep the rest of the catalogue reachable

Pagination has a different job from colour or sorting. Google recommends separate URLs for pages in a sequence, crawlable links between them and each page’s own canonical URL. Its crawler does not click a “load more” button to reveal products. Source: Google on ecommerce pagination.

In the backpack example, test a product that appears only after the first batch. Reach it through the catalogue without relying on a filter combination you intend to block. Repeat on a phone, including clearing filters and returning from a product page.

Review five real URLs before rolling out a rule

Ask the team to show the response and rendered page for one example of each:

1. Chosen landing page: the heading, products, internal links and search settings support its intended role.
2. Ordinary filter combination: the agreed crawling or indexing treatment is present and the shopper’s controls still work.
3. Sorted duplicate: the preferred version is consistent with the duplication decision.
4. Later catalogue page: its products remain discoverable through the intended navigation.
5. Impossible combination: the page gives a useful way back instead of pretending to be a valid category. For crawlable faceted URLs, Google recommends an HTTP 404 for empty or nonsensical combinations. Source: Google’s filter URL guidance.

Record the expected outcome next to the observed one. After launch, review whether chosen landing pages appear for relevant queries and whether unwanted patterns still dominate crawling. Do not assume every small catalogue has a crawl budget problem, or treat fewer indexed URLs as a success on its own.

Areza works on websites and ecommerce platforms and the catalogue and shopping journeys behind online stores. A useful starting brief is one category, a few actual filter URLs and the customer question each proposed landing page should answer. Bring us those examples to define a manageable first change.