How to do e-commerce SEO: from category pages to reports

Making category pages your main landing pages, managing filter and pagination URLs, deciding what to do with products that leave stock, and watching the right reports.

· 12 min readSEO

Short answer

E-commerce SEO starts with category pages, because most demand lands there. Then filter and sort URLs are managed, pagination and variants get the right canonicals, and empty or out-of-stock pages get a decision. The reports to watch are Search Console’s Performance and Page indexing reports.

In one shop, the marketing lead opens the “shoes” category in the menu and tries the colour, size, brand and price filters. With every click the address bar gets a little longer. Nobody notices, but the number of pages offered to Google from the same catalogue has grown far beyond the number of products. This is where the real answer to how to do e-commerce SEO begins: first decide which pages answer a search, then decide which ones should never reach Google at all. This article takes those decisions in order and says only what Google Search Central’s documentation says.

You can find the product page itself in what a product page needs, and the general crawling checks in the technical SEO checklist. Here we only cover what is specific to a shop. To improve your shop’s own search box, see how to improve on-site search in an online shop.

1. Treat the category page as your main landing page

A buyer who types “running shoes” is not looking for one product, but for options. So most of the demand matches a product list rather than a single product, and the address of that list is the category page. A category page has to do three things:

  • Carry the searched phrase naturally in the title and the main heading (H1).
  • Describe the category in a short introduction, with the selection criteria and the subcategories.
  • List the products with real links.

Take the phrase for each category from the map in your keyword research: one phrase, one category. If two categories target the same search, they compete with each other; pick one.

Google works out much of your site’s structure from the links pages give each other. According to its documentation, you should link from menus to category pages, from category pages to subcategories, and from subcategories to all product pages. If a category page does not link directly to its products, Googlebot may not find some of them by crawling alone. An on-site search box does not fill the gap, because Googlebot generally does not type queries into a search box. For pages you cannot link to, a sitemap or a Merchant Center feed helps.

Three details make the difference:

  1. Use real links. Google’s documentation asks for <a href> rather than JavaScript when moving between pages.
  2. Let the link text say what it points to. The documentation suggests using, for example, the product’s name instead of generic phrases such as “click here”.
  3. Link more to the pages that matter. The documentation says Google infers a page’s relative importance from how many links point to it within the site. That is why it makes sense to link a best-selling product or category from the home page, blog posts or newsletters. This is what an internal link strategy looks like in a shop.

Let your guides be part of the same web: a “how to choose running shoes” article links to the relevant category and to the specific products it mentions with descriptive text. The category page links back to the guide where it helps.

3. Decide what to do with filter and parameter URLs

Filtering helps buyers, but Google’s documentation says its most common implementation, based on URL parameters, can produce an “infinite URL space”. It names two consequences: overcrawling (Google spends many requests on useless URLs) and slower discovery of new, useful URLs. Its advice is one of two paths:

  • If you do not need the filtered URLs to be indexed, prevent crawling of them.
  • If you do need them indexed, follow the best practices in the documentation.

The documented ways to prevent crawling are to disallow the filter parameters in robots.txt, allowing only the unfiltered listing page and the individual product pages, or to put the filter in a # fragment instead of a ? parameter. Google generally does not support what follows # for crawling and indexing. The documentation adds that rel="canonical" and rel="nofollow" can also signal a preference, but are generally less effective in the long term than those two methods.

If filtered URLs do need to be indexed, the documentation asks for the following: use the standard & as the parameter separator (commas, semicolons and brackets are hard for crawlers to recognise as separators). If filters sit in the path, it asks you to keep their logical order fixed and allow no duplicate filters. For a combination that returns no results, it asks for a 404 status code at that URL rather than a redirect to a generic error page. To see which address Google picks when one page opens at several URLs, read our article on canonical tags and duplicate content.

A filtered URL: should it be in the index? If not, block crawling; if so, build it properly; with no results, return 404 Should this filtered URL be indexed? No Yes No results Block crawling: a robots.txt rule or a # fragment Build the URL well: & separator, fixed order, its own canonical Return a 404, do not redirect to a generic error page URLs meant for the index are found through links and the sitemap A filtered URL: should it be in the index? If not, block crawling; if so, build it properly; with no results, return 404 Should this filtered URL be indexed? No Block crawling: robots.txt rule or a # fragment Yes Build the URL well: & separator, fixed order, its own canonical No results Return a 404, do not redirect to a generic error page URLs meant for the index are found through links and the sitemap
Figure: A filtered URL raises one question, whether it should be indexed; the answer decides if crawling is blocked or the URL is built properly, and a combination with no results returns a 404.

The table below sums up the types of URL a shop runs into and what the documentation says about each.

URL type Example What the documentation says to do
Filter /shoes?colour=green If it need not be indexed, block crawling with robots.txt; if it must be, use &, a fixed order and a 404 for empty combinations
Sort order /shoes?order=price The same list in another order should not be indexed: noindex or robots.txt
Pagination /shoes?page=2 Each page has its own URL and its own canonical; pages are linked to each other with <a href>
Variant /t-shirt?colour=green Each variant has its own URL; with an optional parameter, the canonical is the URL without it
Empty category /shoes/summer If empty, noindex; if the category is removed automatically, 404

4. Set up pagination properly

When a category has hundreds of products, you split the list into pages. Google names three patterns: pagination, a “load more” button and infinite scroll. The last two usually run on JavaScript, and Googlebot does not click buttons or generally trigger functions that need a user action. So all the products must also be reachable another way, through links or a sitemap.

For pagination the documentation’s requirements are short:

  1. Link the pages to each other in sequence with <a href>; you may also link each page back to the first.
  2. Give each page its own URL, such as ?page=2. Do not use # for page numbers, because Google ignores it.
  3. Do not make the first page the canonical of the whole series; each page carries its own canonical URL.
  4. The pages may share the same title and description; the documentation says pages in a paginated sequence do not need distinct titles.
  5. Google no longer uses rel="next" and rel="prev"; you do not need them, although other search engines may.

5. Variants and canonicals

When a product multiplies by colour and size, there are two routes. Google’s documentation recommends that each variant can be identified by its own URL: a path segment such as /t-shirt/green or a parameter such as /t-shirt?colour=green. If you use an optional query parameter, naming the URL without the parameter as the canonical helps Google understand the relationship. Write the same URL in internal links, the sitemap and the rel="canonical" tag, and let every indexable page carry a canonical pointing at itself. You can read what the tag is in the canonical tag entry.

Avoid temporary values in URLs as well: session IDs, tracking codes, values that depend on the user such as “nearby” or “last week”, and the current time. The documentation says these create short-lived or duplicate URLs.

6. Out-of-stock and discontinued products

Google has no single “do this” rule here; its documentation explains what status codes mean and what to do with empty categories. You decide according to the product’s situation:

  • Temporarily out of stock, coming back. Leaving the page in place, stating the stock status clearly and offering the buyer a way out such as “tell me when it is back” makes sense. A page that returns 200 and truly describes a product also keeps its message and its status code in agreement.
  • Permanently discontinued, with a replacement. Redirecting the buyer to the genuinely related replacement product or its category is the most useful route. Search Console Help says to use a 301 redirect when a page has moved; the redirecting URL is not indexed, and the target can be.
  • Permanently gone, no replacement. According to Search Console Help, a 404 is not necessarily a problem when a page has been removed without any replacement; such a page staying out of the index is expected.
  • A category whose products have run out. For a category with no items the documentation suggests a noindex tag; if your shop removes empty categories automatically, you can consider returning a 404.

If a page shows a “not found” message but returns 200, Search Console reports it as a soft 404 and recommends a 404 code for pages that are truly not found; the message and the status code must agree.

7. Structured data in one paragraph

Google recommends adding structured data because it helps Google understand the purpose of your pages. On a product page, Product markup is enough; the details are in what a product page needs. Do not mark up a category page as Product: according to Google’s product markup documentation, product rich results only support pages that focus on a single product (or variants of the same product).

8. Which reports to watch

Track a shop’s SEO with three reports, all inside Search Console and all free.

  1. Performance report. Switch to the “Pages” dimension, filter by a category folder in the URL and see which queries bring clicks and impressions. Clicks, impressions, click-through rate and average position are read here.
  2. Page indexing report. Its help page says not to expect every URL to be indexed; duplicate, filtered or redirecting URLs staying out is normal. Make sure the important category and product pages are indexed. The help page also notes that if you see many duplicate pages produced by filter or sort parameters, they probably should not be indexed.
  3. Sitemap. Include only the URLs you want indexed: categories, subcategories, products; leave out filtered, sorted and redirecting URLs.

Open them once a month in the same order and ask: did queries grow for category X, and is there a drop in a category whose products ran out?

Common mistakes

  • Leaving every filter combination open to indexing and never noticing.
  • Pointing the canonical of every paginated page at the first page.
  • Writing variant and product URLs differently in internal links, the sitemap and the canonical.
  • Redirecting a filter with no results to a generic “not found” page.
  • Redirecting every discontinued product to the home page.
  • Publishing a category page as just a product grid, with no heading and no introduction.

Common questions

Should I let filtered pages be indexed?

If you do not need them indexed, Google recommends preventing crawling. If you do, for example for a combination with its own search demand, build those URLs by the documentation’s rules. Those are the & separator, a fixed filter order, a 404 for empty combinations and consistent spelling across internal links and the sitemap.

Can I make the first page the canonical in pagination?

No. Google’s documentation says not to use the first page as the canonical and to give each page its own canonical URL.

Should I delete an out-of-stock product page?

There is no single answer. Keep the page of a product that is coming back, redirect a permanently discontinued product’s URL to its replacement with a 301, and return a 404 when there is no replacement. Google’s documentation gives no separate rule for this; the status code logic above applies.

One thing you can do today: open the Performance report in Search Console, look at the ten most-clicked URLs on the “Pages” tab, and check whether any of them is a filtered or parameterised URL. If one is, go back to step 3 of this article; if none is, you now have a ready list of category pages whose titles and introductions you can review. If you would like the category structure and the crawling of filtered pages handled together for your shop, our SEO service covers it; for the build side, see how to set up an online shop.

Sources

This article is general information, not legal, financial or technical advice. Platforms, their rules and the law change over time, so some details may have changed since 11 November 2026 or contain small errors. Before you make a decision for your business, check the current official sources or write to us.

Tell us about
your project.

Fill in the form or write to us directly, and we will agree on a day to talk.

Your first meeting is with Deniz Sezer, founder of Detmore.

First meeting free · Reply within one working day · Written proposal in 2 to 3 days · NDA ready

What do you need? pick as many as you like

Name, email and phone are required.

Contact

We reply within one working day.