Site architecture SEO is the practice of arranging and linking your pages so people and Google can reach every important page in a few clicks, and can tell which pages matter most. In practice that means a shallow hierarchy, a hub page for each main topic or service, plain HTML links between levels, and no page left without a link pointing at it.
That’s the short answer. Architecture is much easier to see than to read about, so this guide is built around diagrams, one per site type, and every section points back to a label in them. All of them are illustrative examples I drew for this post, not client sites.
One boundary first. If you run an online store, product and category trees have their own rules, and I cover those in my ecommerce site structure guide. This post is for every other kind of site.
What Does Google Actually Say About Site Architecture?
Less than people think, and that’s useful to know. Google’s SEO starter guide says search engines “will likely understand your pages as they are right now, regardless of how your site is organized.” Architecture is not a magic ranking lever.
What Google does say is mostly about links. Its link best practices page says Google can reliably crawl a link only when it’s an <a> element with an href attribute. The same page adds: “Every page you care about should have a link from at least one other page on your site.”
Google’s ecommerce documentation states the depth part plainly. Google “can use information such as 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.” That sentence was written for stores, but nothing about it is store-specific.
The starter guide also notes that folders start to matter once you have “more than a few thousand URLs,” because Google can learn how often each directory changes. So my working summary is this. Links decide discovery and importance. Folders help big sites. Everything else is for your visitors.
The Shape Every Good Architecture Shares
Here’s the skeleton I sketch before touching any site. Labels in brackets are the ones I refer to through the rest of this guide.
[H] Home /
├── [HUB] Main topic or service A /topic-a/
│ ├── [S] Supporting page A1 /topic-a/subtopic-1/
│ ├── [S] Supporting page A2 /topic-a/subtopic-2/
│ └── [S] Supporting page A3 /topic-a/subtopic-3/
├── [HUB] Main topic or service B /topic-b/
│ ├── [S] Supporting page B1
│ └── [S] Supporting page B2 ··· contextual link ···> [S] A2
├── [U] About, Contact, Pricing (utility pages)
└── [X] Orphan page: no links in (the problem we hunt for)
Read it top to bottom. Home links to every HUB, each HUB links to its own S pages, and every S page links back up. The dotted line from B2 to A2 is a contextual link, written inside body copy because the topics genuinely overlap.
The X node is the one that hurts. It exists, maybe it’s even in the sitemap, but no page links to it. By Google’s own rule above, that page has no internal path at all.
How Deep Is Too Deep? Click Depth vs URL Depth
People mix these up constantly. Compare the two paths below to the same page.
| Path | URL depth (folders) | Click depth (links from home) | Verdict |
|---|---|---|---|
| Home > Services > Roof Repair | 2 (/services/roof-repair/) | 2 | Healthy |
| Home > Blog > Page 9 of archive > Post | 1 (/roof-leak-signs/) | 4+ | Too deep, despite the flat URL |
| Home > Roof Repair (linked from main menu) | 2 | 1 | Ideal for a money page |
The second row is the trap. A flat URL looks shallow, but if the only route to that post runs through page 9 of a blog archive, Google has to follow many links to reach it. Google’s quote above is about links followed, not slashes in the address.
My working rule is three clicks or fewer for anything I want ranking, and one click for the pages that make money. That’s my habit from audits, not a Google number. Google doesn’t publish one. URL folders are a separate decision, and my URL structure SEO guide covers when they help.
Site Architecture Diagrams by Site Type
The skeleton stays the same everywhere. What changes is which pages play HUB, which play S, and how big each layer grows.
Service Business
[H] Home
├── [HUB] Services /services/
│ ├── [S] Drain Cleaning /services/drain-cleaning/
│ ├── [S] Water Heater Repair /services/water-heater-repair/
│ └── [S] Emergency Plumbing /services/emergency-plumbing/
├── [HUB] Guides /guides/
│ └── [S] How to Fix a Slow Drain ──> links up to Drain Cleaning
└── [U] About · Reviews · Contact
Small service sites usually have the opposite problem from big ones. They put every service on one long page, so Google sees one URL trying to rank for ten different jobs. In my experience, giving each service its own S page is the biggest single architecture win for a local business.
The guide in the diagram does one job. It links up to the matching service, so informational traffic flows toward the page that books work.
Blog or Publisher
[H] Home
├── [HUB] Category: Email Marketing /email-marketing/
│ ├── [S] Post: Subject Line Tips
│ ├── [S] Post: Email Deliverability
│ └── [S] Post: Welcome Sequences
├── [HUB] Category: SEO /seo/
└── Archive pages /page/2/ ... /page/40/ (depth leak)
Blogs leak depth through chronology. Every new post pushes older ones further down the archive, so a great post from two years ago can drift dozens of clicks away from the homepage.
The fix is treating category pages as real HUBs, with an intro and hand-picked links to the best posts, instead of an endless date-sorted feed. Whether to organise those clusters as strict silos or linked clusters is a bigger debate, and I give my verdict in content silo vs topic cluster.
SaaS
[H] Home
├── [HUB] Product /product/ ──> [S] Feature pages
├── [HUB] Solutions /solutions/ ──> [S] Use case and industry pages
├── [U] Pricing /pricing/
├── [HUB] Blog /blog/ ──> [S] Posts linking to features
└── [U] Docs /docs/ (often a separate subdomain)
SaaS sites tend to grow as islands. Marketing owns the product pages, content owns the blog, and support owns the docs, often on a separate subdomain and a separate CMS. Nobody links across.
Honestly, I’d start by connecting the islands. Feature pages should link to the docs that explain them, and blog posts should link to the feature they describe. My SaaS landing page SEO guide covers what those feature and solution pages need once they’re linked.
Multi-Location Business
[H] Home
├── [HUB] Locations /locations/
│ ├── [S] Austin /locations/austin/
│ ├── [S] Dallas /locations/dallas/
│ └── [S] Houston /locations/houston/
├── [HUB] Services /services/
│ └── [S] each service ──> links to the locations that offer it
└── [U] Store locator (JavaScript map + plain HTML list of links)
The note on the store locator matters. Many locator widgets load location pages only through a JavaScript map, which gives Google no <a href> links to follow. I always ask for a plain list of location links under the map.
Location pages also need real local content, not a city name swapped into a template. My local SEO for multiple locations guide covers that side.
Navigation Links vs Contextual Links: Which Does What?
Both are internal links, but they do different jobs in the diagrams.
| Link type | Where it lives | Job in the architecture | Diagram example |
|---|---|---|---|
| Main navigation | Header menu, every page | Defines the HUB layer, keeps money pages one click deep | Home to HUB |
| Breadcrumbs | Top of the page | Links each S page up to its HUB | S A1 to HUB A |
| Footer | Bottom of every page | Utility pages, legal, contact | U nodes |
| Contextual | Inside body copy | Connects related topics across HUBs | B2 to A2 |
Navigation tells Google what your site is about at the top level. Contextual links tell it which specific pages relate, and the anchor text explains how. I keep menus short, around the HUBs only, and let contextual links do the fine-grained work. The wording of those anchors is a separate skill, and I break it down in anchor text SEO.
Why Do Hub Pages Matter So Much?
Look back at the skeleton. Each HUB is the one page that links to every S page in its topic and receives a link back from each. That makes it the natural target for your broadest keyword and the page with the most internal links in its group.
Most sites I audit are missing this layer. They have a homepage and 200 posts, with nothing in between except a tag cloud. Building five or six real hub pages, each with a short intro and curated links, usually fixes depth and orphan problems in one move.
How Do You Find Orphan Pages?
An orphan is the X node: a live page with no internal links pointing to it. Sitemaps can hide them, because Google may still find the URL there, but without links it has no path and no internal signal of importance.
The method is a comparison. Crawl the site by following links, then list every URL you know exists from the XML sitemap, your CMS and the Search Console page indexing report. Any URL in the second list but not the first is an orphan. My XML sitemap best practices guide explains what belongs in that file.
Then decide each orphan’s fate. Link it from its HUB if it’s useful, merge it if it duplicates something, or remove it.
How to Run a Site Architecture SEO Audit
This is the routine I run on a new site, and it works with free tools.
- Crawl the site. Screaming Frog’s SEO Spider is free for up to 500 URLs and reports crawl depth for every page it finds. Sort by depth and look at anything at 4 or deeper.
- Open the Search Console Links report. The internal links table shows which pages your site links to most. Your HUBs and money pages should sit near the top. Google’s help page notes the tables cap at 1,000 rows and don’t show whether a link is nofollow.
- List orphans with the comparison method above.
- Inspect a menu page with URL Inspection. Check the rendered HTML for real
<a href>links in the navigation. - Redraw the diagram. Sketch the site as it really is, then as it should be, and list the links that close the gap.
If the crawl shows thousands of near-identical parameter URLs, that’s a crawling problem rather than an architecture one, and my crawl budget guide is the next stop.
The Detail Most People Get Wrong: JavaScript Links
Every line in my diagrams assumes a real link. Here’s what that means in code, using the examples from Google’s link documentation.
<!-- Google can follow these -->
<a href="/services/drain-cleaning/">Drain cleaning</a>
<a href="https://example.com/locations/austin/">Austin</a>
<!-- Google can't reliably follow these -->
<span href="/services/drain-cleaning/">Drain cleaning</span>
<a onclick="goto('/locations/austin/')">Austin</a>
The page looks identical to a visitor in both cases. That’s why this one slips through. A mega menu, a tab widget or a “load more” button built on click events can quietly cut whole branches out of the tree. I check this first on any site built with a JavaScript framework.
What the Diagrams Add Up To
Every diagram in this guide shares the same skeleton: a homepage, a handful of hubs, supporting pages linked up to them, and contextual links across. The click depth table showed that links, not folders, decide depth. The site type diagrams showed where each kind of site leaks, from blog archives to SaaS islands to JavaScript locators.
If you want a second pair of eyes on your own tree, our technical SEO service starts with exactly this crawl and redraw. You can also request a free audit. Running a store with filters? Read my companion guide to faceted navigation SEO next.
Frequently Asked Questions
What Is a Good Site Architecture for SEO?
A shallow hierarchy where every important page is reachable in about three clicks through plain HTML links. Each main topic or service gets a hub page, supporting pages link up to their hub, related pages link to each other in body copy, and no page is left without an internal link.
Does Site Architecture Affect Rankings?
Indirectly. Google’s starter guide says it will likely understand your pages regardless of how the site is organized. But Google also says it uses the number of links needed to reach a page, and the number of links to it, to judge relative importance. Architecture shapes both.
How Many Clicks From the Homepage Should a Page Be?
Google publishes no limit. My rule is three clicks or fewer for any page you want ranking, and one click for your main money pages. Click depth counts links followed, so a flat URL can still sit very deep.
Do I Need Folders in My URLs for Good Site Architecture?
No. Google’s starter guide suggests folders mainly help sites with more than a few thousand URLs, by letting Google learn how often each section changes. On smaller sites, links do the real work, and folders are a choice for readability.
Last updated: October 2026 by Mizanur Rahman
