Ecommerce Site Architecture
Last updated
Architecture decisions made at the platform level affect every page on a store. Category depth determines how many internal links a product page receives and how much link equity reaches it. URL structure determines whether pages consolidate authority or fragment it. Internal linking patterns determine how crawlers move through the catalogue and how users navigate from discovery to purchase.
Getting these decisions right at the start is considerably easier than restructuring an established store, where every URL change requires a 301 redirect and carries migration risk.
What is the optimal category depth for ecommerce?
Standard practice limits category depth to three or four click levels from homepage to any product page: Homepage > Main Category > Subcategory > Product. Beyond four levels, pages at the bottom of the hierarchy receive progressively fewer internal links and less link equity. The mechanism is Google’s own: it says it can use “the number of links it needs to follow to reach a page and the number of links to a page to infer the relative importance of a page over the rest of your site”.1 The three-or-four figure is practitioner convention rather than a Google threshold, but what it is trying to protect is documented.
The practical goal is to keep structure as flat as meaningful organisation allows. A store with 50,000 products in 200 categories cannot be entirely flat (some hierarchy is necessary), but adding an extra layer of sub-subcategories to better organise a small product range usually costs more in crawl efficiency than it gains in organisational clarity.
When assessing depth, count the number of clicks from the homepage to the deepest product page, not the number of URL segments. A URL like /footwear/womens/trainers/running/nike-air-max/ is five segments deep but may only be four clicks from the homepage if there is a direct link path. URL structure and navigable depth are distinct.
How should ecommerce URLs be structured?
Clean, hierarchical URLs that reflect the category structure are worth having, but be clear about what they do. Google asks for descriptive, readable URLs mainly for people: it recommends organising content “so that URLs are constructed logically and in a manner that is most intelligible to humans”, and using readable words rather than long ID numbers.2 What they are not is how Google works out your hierarchy. Google states that it “generally doesn’t look at the structure of URLs to work out the structure of a site” and analyses the links between pages instead.1 Words in a URL remain a low-weight relevance signal, covered on URL structure; the structure that carries this page’s argument is the link structure, not the path.
The recommended pattern: /category/subcategory/product-name/
Shopify’s default URL structure uses /collections/ for categories and /products/ for product pages: domain.com/collections/womens-trainers/products/product-name. This is a functional structure. The /collections/ prefix does not harm search performance. Moving an established Shopify store away from that structure is not an option: Shopify reserves /products and /collections as fixed paths and does not allow redirects from them.3 The default structure is fine, and the question does not arise.
What to avoid:
- Session IDs and user-specific parameters in URLs (
?sessionid=abc123). Google lists these among the URL patterns that cause inefficient crawling and says to “avoid the use of session IDs in URLs and consider using cookies instead”.2 - Repetitive keyword stacking (
/shoes/trainers/womens-trainers/womens-running-trainers/) - Sort and filter parameters that create near-duplicate URLs. Google names shopping sorting parameters in the same list,2 and they are covered separately in faceted navigation SEO
How do breadcrumbs distribute link equity?
Breadcrumb navigation (Home > Trainers > Women’s Trainers > Nike Air Max) creates internal links from every product page to each category above it. This bidirectional linking matters for link equity: the product page receives a link from the category via the product listing; the product page links back up to the category via its breadcrumb.
This bidirectional structure means category pages accumulate link equity from all the product pages beneath them, in addition to any external links they earn. Internal links redistribute the equity a site already has rather than adding to it: external backlinks are what increase the total available. Product pages receive link equity from the categories above them, even in the absence of external links to individual product URLs.
Breadcrumbs should use BreadcrumbList structured data in JSON-LD. Google uses the markup to categorise the page’s position in the site hierarchy and can show that trail in the result, though it lists the breadcrumb rich result as a desktop feature.4 The internal linking benefit described above applies regardless of how the result displays, and is the more reliable reason to implement it.
How does internal linking work across large catalogues?
On a large store, internal linking strategy is a question of priority. Every page cannot link to every other. The structure should direct the most link equity to the highest-value pages, and Google describes the same rule from its side: “as a general rule, the more links a page has to it within a site, the higher the relative importance of the page to other pages on your site”.1
Homepage to top-level category pages. The homepage is the most linked-to page on most ecommerce sites. It should link directly to primary categories. Linking from the homepage to individual product pages is appropriate only for flagship products, which is Google’s own example: for a best seller, “consider linking to it from the home page or in other content”.1
Category pages to subcategories and featured products. Category pages should link to child subcategories and to bestselling or featured products. These links push link equity from the relatively well-linked category level into the product level. Google recommends the full chain, “from menus to category pages, from category pages to sub-category pages, and finally from sub-category pages to all product pages”, and warns that where category pages do not link to every product, “Googlebot might not find all of your products by crawling alone”.1 Where linking to everything is impractical, it names an XML sitemap or a Merchant Center feed as the fallback.1
Product pages to related products and parent category. Related products widgets are link equity redistribution, not only a UX feature. Products that link to related and complementary items create a web of connections that improves crawlability across the catalogue.
Editorial content to category pages. Blog posts, buying guides, and editorial content on the site should link to category pages. Categories are durable: individual products go out of stock and are removed; category URLs persist. External links earned by editorial content that point to a category page continue to accrue value even after the specific products in that category change.
Why are category pages the correct link building target?
Earning external links to product pages is structurally difficult. Products get discontinued, renamed, and removed. An external link pointing to a specific product URL is often wasted within a couple of years when the product leaves the catalogue.
Category pages are durable destinations. “Women’s running shoes” is a stable URL that can rank for commercial queries over years without change. Digital PR campaigns, outreach, and editorial mentions should be targeted at category page URLs as the landing destination, not individual product SKUs.
For the same reason, when evaluating which pages deserve SEO investment (content improvement, link building, structured data completeness), category pages typically offer a better return than product pages for most queries.
Does architecture change for AI shopping?
The conclusion above holds for links and for commercial rankings, and needs scoping for AI surfaces. Category pages remain the right target for link building, because they are the durable URLs that accumulate equity and rank for head commercial queries. AI shopping surfaces retrieve at a different granularity: the unit is the product listing, assembled from feed and page data, rather than the category page that leads to it.
Both are true at once, and they do not compete for the same work. Architecture serves the first; catalogue and listing completeness serves the second, and is covered in AI shopping agents and ecommerce visibility. The failure mode is assuming a strong category structure carries a store into AI recommendations on its own.
