SEO Architecture: How to Structure a Site for Google and AI

SEO architecture is the way your site's pages are organized, linked and addressed so that search engines and AI crawlers can find every important page, understand how pages relate and decide which ones matter most. Good architecture means a shallow, logical hierarchy, plain HTML links between related pages, clean and consistent URLs, and no endless filter or duplicate URLs soaking up crawl activity.

This guide is for e-commerce and Shopify teams planning a new store, a redesign or a cleanup. You'll get what Google actually says about structure, why AI crawlers raise the stakes, a model of how click depth works, a page-type blueprint for online stores, a 7-step planning process and a symptom, cause and fix table you can use as an audit checklist.

What is SEO architecture?

Site architecture covers three connected layers:

• Hierarchy: how pages are grouped, for example home, then categories, then subcategories, then products, plus guides and support content.
• Internal links: the navigation, breadcrumbs and in-content links that connect those pages and tell crawlers which ones you consider important.
• URLs: the addresses for each page, which should be readable, consistent and free of duplicates.

Google's SEO Starter Guide puts the goal simply: a logical organization helps search engines and users understand how your pages relate to the rest of your site. It also notes that grouping similar pages in directories matters most for sites with more than a few thousand URLs, because Google can learn how often each directory changes and crawl them at different frequencies. Most e-commerce stores with large catalogs fall into that group.

Why architecture matters for Google and AI crawlers

Crawlers find pages through links

Google says it uses links both as a relevance signal and to find new pages to crawl, and that every page you want indexed should get at least one link from another page on your site (see Google's link best practices). It can generally only follow links that are <a> elements with an href attribute. A product reachable only through a search box or a script-driven menu may never be discovered.

AI crawlers are less forgiving of messy sites

AI search tools send their own crawlers, and the early data shows they stumble on broken structure far more than Googlebot does. In one month of traffic on its network, Vercel found that about 35% of ChatGPT and Claude crawler fetches hit 404 pages, compared with about 8% for Googlebot. ChatGPT's crawler also followed redirects on 14.36% of fetches, versus 1.49% for Googlebot. Vercel's advice was to keep redirects correct, keep sitemaps up to date and use consistent URL patterns.

The same study found that none of the major AI crawlers it measured render JavaScript, including OpenAI's crawlers, ClaudeBot and PerplexityBot. If your navigation, product grid or internal links only appear after JavaScript runs, those crawlers may see a page with no paths to the rest of your site. Server-rendered HTML links are the safe default for both Google and AI tools.

Flat vs deep: how many clicks is too many?

Click depth is the number of clicks it takes to reach a page from the homepage. Pages close to the homepage tend to be discovered sooner and receive more internal link value. Google does not publish a maximum depth, but a common rule of thumb is to keep important pages within about three clicks of the homepage.

A shallow structure is easier to reach than most people expect, because each level multiplies. If every page links to 20 unique pages one level down, here's how many pages you can reach:

That's why a well-built category tree can put thousands of products within three clicks without a cluttered menu. The opposite problem, a "flat" site where everything links to everything, removes the hierarchy that tells crawlers which pages matter and which ones belong together.

The ideal e-commerce site structure

For most online stores, a clean architecture uses five page types, each with a clear job. Here is how they map to Shopify's URL patterns:

Guides matter more than they used to. AI tools answer detailed buyer questions, and Google says AI Overviews and AI Mode may run several related searches, which it calls query fan-out, to build one answer. A guide that links into the right collections gives both shoppers and crawlers a path from the question to the product. Our guide to topical authority covers how to plan those clusters.

URL structure best practices

Google's URL structure guidelines (updated October 2026) come down to a few rules:

• Use readable words, not long ID strings, in the language your audience uses.
• Separate words with hyphens, not underscores.
• Pick one case. Google treats /Shoes and /shoes as different URLs.
• Trim parameters that don't change the content, such as tracking, session or sort parameters.
• Don't use fragments (the part after #) to change page content, since Google generally doesn't support them.
• Use root-relative links rather than parent-relative ones, which can create endless bogus URLs.

On Shopify, the /products/, /collections/, /pages/ and /blogs/ prefixes are fixed on standard stores. You control the handle after the prefix, so make each one short, descriptive and stable. If you change a handle, keep the redirect option on so the old URL forwards to the new one. Our redirect chains guide shows how to keep those redirects clean over time.

How to plan your site architecture, step by step

1. Export what you have. Crawl the site and pull every indexable URL, plus Search Console pages with impressions. This is your baseline and your redirect source list.
2. Research how shoppers search and browse. Group keywords by intent: broad category, specific subcategory, product and research question. Our guide to keyword mapping walks through this.
3. Draw the hierarchy. Map categories and subcategories so each one targets a distinct search, then decide which products belong where. Avoid two collections chasing the same query.
4. Assign one URL per intent. Write the handle for every collection, product and guide, following Google's URL rules. Note any URL that changes.
5. Plan the links. Decide what the main menu, footer, breadcrumbs, collection pages and guides will link to. Every important page needs at least one contextual link.
6. Control filters and duplicates. Decide which filtered views deserve their own indexable pages and block or canonicalize the rest.
7. Launch with redirects and verify. Redirect every changed URL to its closest match, resubmit your sitemap and crawl again to check depth, orphans and broken links.

Google's own starter guide cautions against reorganizing a working site for its own sake, since search engines likely already understand your pages as they are. Restructure when the current setup blocks growth, such as buried products, duplicate collections or a new product line with no home.

Internal linking that supports the structure

Hierarchy sets the skeleton. Internal links carry the weight. Four link types do most of the work on an online store:

• Main navigation: links to your top collections. Keep it focused on the categories that matter commercially, and make sure it renders as HTML links.
• Breadcrumbs: show each page's place in the hierarchy (Home, Women, Running shoes) and link back up. Add breadcrumb structured data so Google can show the path in results.
• Contextual links: links inside collection descriptions, product copy and guides. Use descriptive anchor text that fits the target page, not "click here."
• Related links: related products, "shop the collection" blocks and links from guides to products. These connect siblings and pass shoppers along the buying path.

Mega menus deserve a second look. A menu that links to every subcollection on every page can make the site feel flat to crawlers and shoppers alike, because the same few hundred links appear everywhere and the real priorities disappear. Keep the menu to the collections that earn revenue, and let collection pages and breadcrumbs handle the deeper levels. On collection pages, make sure pagination uses real links to each page, so products on page three or four are still discoverable without infinite scroll or a "load more" button that only works with JavaScript.

Your XML sitemap supports this structure but doesn't replace it. A sitemap tells crawlers which URLs exist; internal links tell them how those URLs relate and which ones matter. Shopify generates a sitemap automatically, so the main job is making sure the pages it lists are live, canonical and linked from somewhere on the site.

Pages with no internal links at all, called orphan pages, are a common problem after product launches and migrations. Our internal link checker shows the links on any page, and our guide to orphan pages covers how to find and fix them across a site.

Faceted navigation, filters and duplicate URLs

Filters for size, color, price and brand are great for shoppers and risky for crawlers. Google's faceted navigation guidance warns that filter URLs can create a near-endless number of pages, which wastes crawl resources and slows discovery of new pages. Its options, in rough order of strength:

• If you don't need filtered pages in search: block the filter parameters in robots.txt, or use URL fragments for filters. Canonical tags and nofollow links also help, but Google says they are generally less effective over time.
• If you do want some filtered pages indexed: use a consistent filter order, prevent duplicate filters and return a 404 for empty or nonsensical combinations.

The practical middle ground for most stores: turn the few filter combinations with real search demand, such as "black running shoes," into proper subcollections with their own copy, and keep the rest out of the index. Our faceted navigation guide and crawl budget optimization guide go deeper.

Common architecture problems and how to fix them

A technical SEO audit checklist will catch most of these in one pass. Our sitemap checker is a quick way to confirm your sitemap lists live, canonical URLs.

Where to start with your store's structure

Begin with a crawl and three questions. Can every important product be reached within about three clicks through plain HTML links? Does each collection target a distinct search, with no duplicates competing? Are filter and tag URLs kept out of the index unless they have real demand? Fix those, keep redirects tidy whenever URLs change and add guides that link into your collections.

If you're planning a redesign or migration and want the structure right from day one, our technical SEO team can map it with you, or you can get in touch for a plan.