Site Architecture for SEO: How to Structure a Website

If Google is slow to find your new pages, or your best pages sit five or six clicks away from the homepage, your site structure is probably the problem. Good site architecture SEO comes down to one idea: put your pages in clear groups, link them together sensibly, and keep every important page a few clicks from the homepage.

You don’t need a perfect diagram or a complex “silo” system. You need a structure that people understand at a glance and that Google can crawl through normal links. This guide shows you how to plan that for a new site, how to fix it on an existing one, and how to check the result.

What Site Architecture SEO Actually Means

Site architecture is how your pages are organised and connected. It covers three things that work together:

• Hierarchy: which pages are parents (categories, hubs) and which are children (articles, products, services).
• Internal links: the menus, breadcrumbs and in-text links that connect those pages.
• URLs: the folders and slugs that reflect the hierarchy in the address bar.

Google says it mainly finds pages by following links from pages it has already crawled. So the links are the part that does most of the work. Folders and diagrams only help if the links behind them are real.

A clear structure also helps Google understand how pages relate to each other. Google’s SEO Starter Guide says a logical organisation helps search engines and users see how a page fits into the rest of your site.

Flat vs Deep Site Structure

You will see two words used a lot: flat and deep. They describe how many clicks it takes to reach your pages from the homepage.

Flat Structure

In a flat structure, most pages are only one to three clicks from the homepage. The homepage links to main categories, the categories link to the pages inside them, and that’s it.

For example, a plumbing company might have: Homepage → Services → Boiler Repair. Three levels, and every service page is two clicks away.

Deep Structure

In a deep structure, pages sit many levels down: Homepage → Shop → Men → Clothing → Tops → T-shirts → Graphic T-shirts → Product. Each level adds a click. Pages at the bottom get fewer internal links and are found later, or not at all.

Which Is Better?

Flatter is usually better, but “as flat as possible” is not the goal. If you link every page from the homepage, the homepage becomes a wall of links and nothing looks more important than anything else.

My rule of thumb is simple: important pages should be reachable in about three clicks, and nothing you care about should be more than four or five clicks deep. The well-known “three-click rule” is a useful target, not a Google rule. A large shop with tens of thousands of products will naturally have a few more levels, and that’s fine if the categories are useful and well linked.

Click Depth: Why It Matters More Than Folders

Click depth is the number of clicks it takes to reach a page from the homepage using links. It is not the same as how many folders are in the URL.

A page at /blog/guides/technical/site-structure/ can be one click from the homepage if the homepage links to it. A page at /site-structure/ can be eight clicks deep if the only route to it is through old paginated archive pages.

Google has said for years that the number of folders in a URL is not what matters. What matters is how easily it can find the page through links. In my experience, click depth is one of the clearest signals of what a site treats as important. Pages linked from the homepage and main menu tend to get crawled more often. Pages buried deep tend to get crawled less and often perform worse.

The common causes of deep pages I see in audits:

• Blog posts that are only reachable through “Older posts” pagination.
• Products that only appear on page 12 of a category.
• Landing pages built for ads that were never linked from the site at all.
• Too many nested subcategories with only two or three items in each.

Hub and Category Pages

Hub pages (also called category pages, pillar pages or section pages) are the middle layer of your structure. They group related pages and link down to them. They are also often the pages that rank for broader searches.

A good hub page is a real page, not just a list of links. It should:

1. Explain the topic or category in a few useful sentences.
2. Link to every important child page with descriptive anchor text.
3. Highlight the most useful or popular children near the top.
4. Link sideways to closely related hubs where it helps the reader.

I always check that child pages link back up to their hub. That two-way link is what makes a group feel like a group, both to readers and to Google.

Topic Clusters and Silos

You’ll hear “topic clusters” and “silos” a lot. Both mean grouping related pages under a hub and linking them together. The idea is sound. Where I see it go wrong is strict siloing, where people refuse to link between sections even when it would help the reader. Google’s link guidance encourages linking related pages in context, so link across sections when it’s genuinely useful.

Internal Linking: The Part That Holds It Together

Your structure only exists in Google’s eyes through links. Google’s link best practices are short and worth following closely:

• Use real links. Google can reliably follow <a> elements with an href attribute. Links that only work through a JavaScript click event may not be found.
• Link every page you care about from at least one other page on your site.
• Write descriptive anchor text (the clickable words). “Boiler repair prices” tells Google and readers much more than “click here”.
• Link in context. A link inside a relevant paragraph helps people and gives Google more clues about the linked page.
• Don’t overdo it. Google says there is no ideal number of links per page, but an excessive number is likely too many.

The fastest internal linking win I know is this: when you publish a new page, go back to two or three older related pages and add a link to the new one. New pages are often orphans for weeks because nobody does this.

An orphan page is a page with no internal links pointing to it. Google may still find it through your sitemap, but it gets no signal from your own site that it matters. I treat orphan pages as a structure bug, not a content problem.

Navigation and Breadcrumbs

Main Navigation

Your main menu appears on every page, so its links carry a lot of weight. Put your main hubs there, not every single page. A menu with 150 links in mega-dropdowns is hard for people to use and spreads that weight thinly.

Check that your menu works without JavaScript tricks. If the menu items are built from buttons or <div> elements with click handlers instead of <a href> links, Google may not see them as links.

Also check your mobile menu. Google uses the mobile version of your site for indexing, so if your mobile menu hides links that the desktop menu shows, Google may never see them. I cover this in more detail in my guide to how mobile-first indexing works and how to check your site.

Breadcrumbs

Breadcrumbs are the small trail of links near the top of a page, such as Home › Services › Boiler Repair. They show where a page sits and give every page a link back up to its parents. On a deep site, they are one of the easiest ways to strengthen the structure.

You can add BreadcrumbList structured data (code that describes the trail to Google) as well. Keep your expectations realistic about how it looks in search results. In January 2025, Google announced it would stop showing breadcrumbs in the visible URL on mobile results and show only the domain there, while desktop results still show them. The on-page breadcrumb links remain useful either way, because they are real internal links.

URL Folders: Reflect the Structure, Don’t Drive It

Folders in your URLs should mirror your hierarchy where it makes sense, for example /services/boiler-repair/ or /shoes/running/. This helps people understand where they are. Google’s Starter Guide also notes that on sites with more than a few thousand URLs, grouping pages in directories can help Google learn how often each section changes.

A few practical rules I follow:

• Keep folders shallow. Two levels is usually enough. Long chains of folders don’t help.
• Don’t put a product in a folder for every category it belongs to. Give it one URL and link to it from each category.
• Avoid dates in blog URLs unless the date really matters to the content.
• Don’t change URLs just to make the folders look tidier. The risk is rarely worth it.

For slug and folder details, see my guide to SEO-friendly URL structure and slugs.

Faceted Navigation: The Biggest Structure Risk on Shops

Faceted navigation is the set of filters on category pages: size, colour, brand, price, rating and so on. It’s great for shoppers. But every filter combination can create a new URL, and the numbers multiply fast. Ten filters with a few options each can produce millions of URLs from one category.

Google’s documentation on faceted navigation warns that this can create a near-endless set of URLs. Googlebot can spend a lot of time on near-endless filter URLs before it realises they’re useless, which wastes your server resources and can slow the discovery of new pages.

Google’s recommendations depend on whether you need the filtered URLs in search.

If you don’t need filtered URLs indexed:

• Block the filter parameters in robots.txt, while keeping the main unfiltered category crawlable. In Google’s guidance, this works better than the other options.
• Or use URL fragments (the part after #) for filters, because Google generally ignores fragments.
• Canonical tags pointing to the unfiltered page, or nofollow on every filter link, can help but are less effective over time.

If some filtered pages deserve to rank (for example “women’s waterproof hiking boots”):

• Use the standard & to separate parameters.
• Keep filters in a consistent order so the same combination always makes the same URL.
• Return a 404 status for filter combinations with no results, instead of redirecting them or showing an empty page with a 200 status.

My approach is to pick a small number of filter pages that match real searches, make them proper landing pages with clean links, and keep the rest out of Google’s crawl. If you’re unsure what a 404 or 200 status tells Google, my guide to HTTP status codes and SEO explains each one.

How to Plan the Structure for a New Site

Planning before you build is far easier than fixing later. This is the process I use:

1. List the topics, services or products you need to cover. Use keyword research and customer questions, not just what you feel like writing about.
2. Group them into a handful of main sections. Most small and medium sites need four to eight. Each becomes a hub.
3. Sketch the tree. Homepage → hubs → child pages. If any branch goes deeper than three or four levels, merge some categories.
4. Decide the URL pattern. Usually one folder per hub, such as /services/ or /guides/, then a short slug.
5. Plan the links. Main menu to hubs, hubs to children, children back to hubs, breadcrumbs on every page below the homepage, and in-text links between related pages.
6. Leave room to grow. Choose section names broad enough that new pages fit without a restructure next year.

Restructuring an Existing Site (With 301 Redirects)

Restructuring a live site carries real risk. Before you move anything, ask whether you can fix the problem with links alone. Often you can. Adding hub pages, breadcrumbs and better in-text links reduces click depth without changing a single URL.

If you do need to change URLs, here’s what I’d do:

1. Crawl the current site and export every URL, plus organic landing page data from Search Console and analytics.
2. Map every old URL to its closest new URL. Send each one to the most relevant new page, never in bulk to the homepage.
3. Set up permanent 301 redirects, one hop from old to new, with no chains.
4. Update internal links to point straight at the new URLs, not through redirects.
5. Update your XML sitemap with only the new URLs and submit it in Search Console.
6. Keep the redirects for at least a year. Google recommends keeping them as long as possible, because other sites and bookmarks still use the old URLs.

Expect some ranking movement for a few weeks while Google recrawls. I’d avoid restructuring right before your busiest season, and I’d never combine a restructure with a redesign, a CMS change and a domain move all at once. If something drops, you won’t know which change caused it.

How to Check Your Site Architecture

Run a Crawl

A desktop crawler (Screaming Frog, Sitebulb and similar tools) follows your links the way a search engine does. When I audit a site’s structure, I look at these reports first:

• Crawl depth: how many pages sit at each click level. If a large share is at depth five or more, you have a structure problem.
• Inlinks per page: which pages get the most internal links. Your key pages should be near the top.
• Orphan pages: connect the crawler to your sitemap and Search Console, then look for URLs that exist but get no internal links.
• Redirects and broken links: internal links that hit 301s or 404s.
• Parameter URLs: a sudden jump in URLs with ? usually points to faceted navigation leaking.

Use Search Console

Google Search Console shows how Google sees your structure:

• Links report: the internal links section lists your most linked pages. If your key hubs aren’t near the top, your linking doesn’t match your priorities.
• Page indexing report: a long list of “Discovered – currently not indexed” or “Crawled – currently not indexed” can point to weak internal linking or too many low-value URLs.
• Crawl stats: under Settings, this shows how many requests Googlebot makes. Lots of requests to parameter URLs suggest a faceted navigation problem.
• URL Inspection: check a deep page to see whether Google knows about it and which page referred it.

I re-run a crawl after any big change, and at least once a quarter on sites that publish often. Structures drift over time as pages get added without links.

Final Thoughts

Site architecture SEO is not about clever diagrams. It’s about clear groups, real links and short paths to the pages that matter. Plan your hubs, link them properly, keep filters under control, and check the result with a crawl and Search Console.

If you’re starting from scratch, sketch your hubs before you write a single URL. If your site already exists, start by crawling it and fixing click depth with links before you think about moving anything. When you’re ready to settle the URL side, my guide to writing clean, readable URLs is a good next step.