Internal linking in ecommerce: structure, links and mistakes to avoid

By Julián, founder of Rankedora · 2 November 2026 · 9 min read

Google discovers a store’s pages by following links, and uses the text of those links to understand what the destination page is about. In an ecommerce store with hundreds or thousands of URLs, internal linking decides which pages are crawled easily, which stay hidden and which receive more relevance within the site. It is one of the levers with the most return and the least cost, because it depends only on you.

What a link has to meet to count

Before talking about structure, a technical requirement: Google only reliably follows links that are an <a> element with an href attribute. A JavaScript button that navigates on click, or an element without href, may work for the user and not exist for the search engine.

Also check the link text: it should describe the destination naturally. If the reader understands what they will find before clicking, the search engine does too.

The base structure: home page, categories, subcategories and product pages

Most stores work well with a simple hierarchy in which each level links to the next and each page links upwards. As a rule of thumb, which is not a Google rule, the pages you want to sell should be a few clicks from the home page. The more buried a product page is, the later it is discovered and the less it is reinforced.

What to link from each type of page

The blog as a source of links to what you sell

In many stores the blog receives traffic and the categories don’t, and there are almost no links between the two. It is the most common opportunity we see when reviewing a site. Some guidelines:

• Every guide should link to at least one relevant category or subcategory, with descriptive text and within the body, not only in a closing box.
• Every important category should link back to a guide that helps people choose, for example “how to choose a size” or “differences between fabrics”.
• If the article targets an informational search, don’t force the commercial link into every paragraph. One or two well placed ones perform better.
• Review old guides: they often link to sold-out products or to categories that no longer exist.

Pagination and filters

Long categories are split into pages. Google explains that each page must be reachable through crawlable links, with its own URL, and that you shouldn’t point the canonical of every page to the first one, because content that only appears on the following pages would stop being discovered. Filters are another chapter: they generate almost endless combinations, and not all deserve to be crawled or indexed. We have explained how to decide in filters and facets in ecommerce.

Orphan pages and buried pages

An orphan page exists but no other page on the site links to it. They are usually old product pages, categories from past campaigns or pages created outside the menu. To find them:

• Crawl the site with a tool such as Screaming Frog, Sitebulb or similar and export the URLs it reaches by following links.
• Compare them with the sitemap and with the URLs that appear in Search Console or in your analytics.
• Those that are in the sitemap or receive visits but don’t show up in the crawl are candidates for orphans.
• Decide for each one: link to it from a relevant category or guide, merge it with another, redirect it or remove it.

Common mistakes

• Huge drop-down menus that link to everything and prioritise nothing.
• Loading products or pagination only with JavaScript and no crawlable links.
• Linking to URLs that redirect or return an error instead of the final URL.
• Repeated, generic link text (“see more”, “here”).
• Filling the footer with dozens of keyword links.
• Random related products that have no relation to the product page.
• Using nofollow on important internal links for no reason.

What internal linking contributes to AI answers

There is no official metric saying how much internal linking weighs in ChatGPT’s or other assistants’ recommendations, so it is best not to promise specific effects. What is reasonable is that a website where every concept has a clear page, with consistent links and consistent names, is easier to crawl, read and summarise. That order also helps guides and categories reinforce each other, as we explain in SEO vs GEO for ecommerce.

How to audit it in an afternoon

1. Crawl the site and export click depth and number of incoming links per URL.
2. Sort by business importance: categories and product pages with the most margin first.
3. Detect the important ones with few incoming links or high depth.
4. List the orphans by comparing with the sitemap and with Search Console.
5. Review the link text of your main categories.
6. Choose ten new high-value links (guide to category, category to guide, product page to product page) and publish them this week.
7. Crawl again after two weeks and check that there are no broken links or redirects.

Quick checklist

• Links as <a href> and not just JavaScript buttons.
• Clear hierarchy: home page, categories, subcategories, product pages.
• Visible breadcrumbs with BreadcrumbList.
• Every guide links to a category and every important category links to a guide.
• Pagination with crawlable links and its own canonical on each page.
• No important orphan pages.
• Descriptive and varied link text.

Frequently asked questions

Sources consulted

• Google Search Central: crawlable links
• Google Search Central: pagination and incremental loading in ecommerce
• Google Search Central: breadcrumb structured data