The Ultimate Guide to Site Structure (2021)
Site structure is about more than where your content will live, it’s about navigation, user experience and the communication of meaning and importance to the various algorithms and bots that crawl our sites – it’s the place where all ranking starts
There have been plenty of occasions when site structure has peaked in popularity within the SEO industry, but it has typically been one of the lesser concerns as agencies and consultants focused on more tangible parts of the SEO toolkit. In fact, the actual peak for search volume around related terms happened around 2004 and dropped as various CMS platforms began offering easy to build website options.
With the subject having become so niche, it would be understandable if some SEOs, and certainly most digital marketers more broadly, have missed the debates that have continued to rage behind the scenes on the best practice methodologies for organising information on the web.
In fact, site structure tends to be considered mostly or wholly within the domain of web development or user experience – but should we continue to overlook an aspect of all websites which plays such a large roll in how these sites are perceived by search engines? This guide will hopefully make a case that we should not, and explain why.
What is site structure?
Whether it’s referred to as site structure, information architecture or any other synonymous term, it refers to the manner in which information (content of any kind) is organised online. This includes not just page hierarchies, but internal linking, taxonomies, navigation, metadata and more. Essentially, site structure is the foundation upon which all of your digital marketing is built and upon which all your ranking efforts rely.
Why is site structure important?
Site structure is not only an important part of your customer journey, it also helps to determine how search engines crawl and index your content, how the link equity (authority, link juice or any other synonym) flows throughout your site. As one of the most fundamental parts of your site, it is also one of the most important, and errors here can undermine a lot of good work elsewhere.
You cannot accompany every consumer on their journey through your brand’s site, but you can make sure it is as easily navigated as possible and that it’s correctly signposted – for the user, and for the various bots that crawl your site and determine how you rank.
The anatomy of a website
If we were to think of a website as a body, the site structure would be the skeleton – it underpins every other part. If we take one of the simplest customer journeys – that of an average eCommerce user, we can see the following:
In the above behaviour flow, we can see that the customer arrives, identifies a product group, narrows it down, adds their product to the basket and then converts. However, while this is an ideal journey, a user (human or search engine bot) has to be able to make similar journeys from every possible entry point to a site.
Again, using the simplest possible example:
Here the homepage links to a small number of categories, which link to a few subcategories – but even in this simple diagram we can begin to see how a journey can be impacted by limited on-site navigation, or a confusing site structure. If a consumer enters the site above and wants to look at shirts following trousers, they would have to return to the ‘Men’ category page and start again; inconveniences such as this impact conversion to greater or lesser degrees.
Along with the underlying order of pages, internal linking, taxonomies (categories etc.) and meta-data all play a part in creating a structure that sees people enter at the right page or able to find it quickly and easily enough to complete their desired action.
It is possible to resolve internal navigation difficulties, as most sites do, with an ever-present, a navigation menu, making virtually all pages accessible from anywhere else on the site. This is where the category issues enter into things, however, especially for larger sites.
Clearly you can’t have every page available from anywhere, so determining the priorities of your menu will also take time. Does the list of menu items need to change depending on where a user is in the overall map of your site? Or is it possible to present them with category level pages? For the largest eCommerce sites, you’ll no doubt see multiple navigation options:
Amazon, here is providing several layers of navigation – the main navigation menu (top), the departmental navigation menu (left) and a secondary, related department menu (below the main nav).
In addition to being able to access the main site categories, you can also search by keyword, and navigate to subcategories and related categories. While most sites may not need to go to such lengths, knowing what could be looked for from where is important (we’ll take another look at this in the desire path section).
Taxonomy and how things are layered (more on this later, too) is where link equity (sometimes referred to as ‘link juice’) also play a part in planning a proper structure. While not as important these days, the depth of pages still communicates a level of importance comparative to the home page. While it’s difficult to guess exactly what elements go into how Google estimates the relative importance of pages, authority (made up of a variety of signals from external links to author expertise) has been pushed in a number of YMYL industries as a specific ranking factor and will likely impact sites outside of such industries too.
With this in mind, authority has a kind of trickledown effect (it’s more complicated in reality as it can work in the opposite direction, too, but it works for illustrative purposes) whereby authority is passed to pages lower in the sitemap. A simple way to think of this is as a wine glass/champagne pyramid:
Link equity, authority or your metaphor of choice, is represented by the wine – it goes in at the top and flows down through the layers to the bottom. Broken links at any level can place a board between layers, stopping the wine from reaching lower levels, while poor structure can leave glasses or even entire layers out of the pyramid altogether – this diminishes the page authority of these disconnected pages and can even prevent pages indexing entirely.
The below is a visualisation of a top-level site structure of and from Wikipedia.
As we can see, even this top-level diagram is much more complicated than our original map – though your actual sitemap will, in reality, look more like the below (which we’ll come to a little later):
Desire paths and internal linking
When you (or your web developer) initially built your site, there would have been a fairly strict hierarchy in place – through your main nav menu, for example, and through category and other pages that have been stocked with well researched copy, are keyword focused and which pointed to all appropriate pages. However, over time – no matter how careful we are – this structure will begin to fray at the edges and, if left unchecked, the whole thing can unravel altogether.
For this reason, it’s vital that part of your organic search (SEO) strategy is a periodic assessment of your internal linking. This is not to say all pages need to lead to all other pages – just that as a site expands (and if you’re successful, they almost all do), it is important to nurture the consumer journey and allow for authority to flow through the site.
Where to start
If you still have your initial plan, then this can be an ideal place to start. If, however, like 90% of website owners, you never had one, or it’s long lost, then you would be best to start with an overview (like the one above). For this you can generally skip the nav menu (which should, hopefully, have been updated as you go) and leave yourself with the marginally less daunting task of on-page links.
By starting here, you can follow the initial part of most journeys – from the homepage to various product or service pages, and auditing these links will allow you to spot potential gaps between where the journey takes you and where you would like potential consumers to go. While doing this, you can keep track of links you’re missing, potential new pages etc. which you can record fairly simply (as below from a previous review of the Click site).
Once you have these links in place, you can move on and do the same from sub-pages – in Click’s case, the next step is to approach the service pages, resource pages and main blog page separately in a similar manner.
User behaviour and customer journeys – next steps
While you may have your own ideas about how the consumer journey should progress, they may have other ideas – and if they’re not catered to, it can lead to a hefty drop off rate and number of single page sessions (while these are not too bad for regularly updated blogs, if it doesn’t correlate with a healthy returning visitor rate, you can safely assume you’re not doing all you can).
This is where desire paths come in. Desire paths (represented by the image below) are the well-trodden short-cuts between more organised pathways and, as the saying goes, represent the difference between user experience (UX) and design.
Unfortunately, for a website, there is no way for the consumer to form these paths themselves and so you’ll need to infer what users want from the data available – you can find the best data on this in Google Analytics in the ‘Behaviour>Site Search>Search Term’ menu options and in the ‘behaviour flow’ diagrams under ‘Behaviour>Behaviour Flow’.
With these behaviour flow diagrams, you can see where consumers are entering your site, and where they drop out, giving you the opportunity to take a closer look at the links on these pages and the possibility of improving user access to other parts of the site.
The search terms and pages will offer some insight as to where consumers are when they feel the need to search your site (provided you have on-site search) – what they’re searching for when they do.
Using this information, it is possible to get at least an initial impression of the direction your consumers want to travel. You can also add search term as a secondary dimension to the search page for ease of visibility of exactly what was searched where – if you don’t have what they were looking for, add it to your content plan and, if you do, think about making it available without requiring the search.
Going with the flow
The above will help you with the first efforts to cater to consumer intent (don’t forget that consistent testing is the foundation of any digital marketing strategy), but it does overlook the need to structure your pages to encourage the flow of authority.
The easiest way to do this is to begin with your site diagrams and then move on to a content report from one of the many search marketing tools that offer link reports (Search Console, Ahrefs, SearchMetrics, Semrush etc.). These tools will often give you a report which details the most linked to pages on your site and, from here, you can plot out the pages that can be linked to from those pages to best serve both the user journey and the flow of link equity.
By arranging your content report descending by links (or, in Ahrefs, by URL rating), you’ll be able to see at a glance where your linking exercises need to begin. While it would be preferable to go through your content in its entirety, it is less time intensive to select your top performing 20 pages for your first run through.
This is not an exercise in simply linking to wherever you need a boost, however – internal linking should be as organic as your standard link building exercises, focusing on relevance. You can safely think about trying to have a link to one up the chain and one down (content permitting) so your sub-services could link to your, main service page, related resources or similar (think of the Amazon example earlier).
How to structure a website
Your website structure impacts a number of things – the path that crawlers will take, that users will take and also impacts the perceived meaning and relevance of sections of your site to various topics and searches. The previous sections have dealt with why you should do think of these aspects (authority, relevance, user and crawler journeys) and offered a few strategies to start thinking about. However, if you’re about to begin a new business website, or are looking to design a new site for an existing brand, there is the question of how we can approach site structure in the best way to serve all of the main concerns.
Types of website structure
For most sites, the optimal structure will be a hierarchy, so we’ll deal with that in more detail, but as a brief overview, the following are typical site structures:
- Database – the database structure sees the website operate as any other database, in that it allows for the easy storage and searching of information at scale. These sites aren’t really intended to be navigated as such, but users will search for and be presented with specific information. This kind of structure can work for real-estate sites, for example, where users will expect to search and refine searches to reach their goal.
- Hierarchical – this organises information taxonomically and by importance, starting with a single page and branching below that to categories and subcategories. This is the most common and most useful (generally) structure, so we’ll deal with this in more detail separately.
- Matrices – an interlinked, database-like approach, there is no sequential navigation to matrix site structures, instead information is linked in a variety of ways, allowing for users to choose their own adventure (Wikipedia is the most often referenced example).
- Sequential – this is the ‘wizard’ approach, useful for exams, onboarding and similar sites which demand a step-by-step process from start to finish.
Hierarchical website structures
This is the kind of structure we’ve referred to throughout this article – and is the kind of structure most people will be familiar with. It is arranged with fewer pages at the top and more as you descend through the hierarchy, beginning with a top-level or ‘home’ page. This kind of structure falls into three further categories:
Flat hierarchies limit the depth of the information in favour of a broad approach. The aim of this kind of hierarchy is to maintain an (often artificial) click-through depth for relevant information. They tend to have expansive top-level navigation which can make it easier to navigate from a first page to a destination page, but can lead to frustration if long lists of similar options lead users to the wrong destination.
In addition, Googler John Mueller had this to say about flat structures:
[We] don’t really know how these URLs are related to each other and it makes it really hard for us to be able to understand how relevant is this piece of content in the context of your website.
The most common type of hierarchy online, the deep hierarchy presents users with a journey – refining and narrowing the focus of information as a user proceeds through a website. While this tends to be the most common kind of structure, it’s important to remember why the ‘4-clicks-deep’ rule became a benchmark for SEO.
John Mueller had this to say about number of clicks to reach a page:
What does matter for us a little bit is how easy it is to actually find the content. So especially if your homepage is generally the strongest page on your website, and from the homepage it takes multiple clicks to actually get to one of these stores, then that makes it a lot harder for us to understand that these stores are actually pretty important.
On the other hand, if it’s one click from the home page to one of these stores then that tells us that these stores are probably pretty relevant, and that probably we should be giving them a little bit of weight in the search results as well.
So it’s more a matter of how many links you have to click through to actually get to that content rather than what the URL structure itself looks like.
While this seems to contradict the advice we quoted regarding the flat structure, this is actually just a little flesh on the bones. It does not imply that deep hierarchy is bad, just that it should be easy to navigate between levels of the hierarchy. Again, consider the multiple navigation options available in the Amazon example.
Often overlooked, a silo hierarchy is useful for brands that offer more than one type of service which have their own ontological framework. Historically, brands have used subdomains to make these kinds of differentiation – see insurance companies, for example, which have done this to separate sections of their offering (car.insurance-brand.com vs. life.insurance-brand.com for example).
However, while subdomains can dilute the authority of a site and impact the ability of pages on those subdomains to rank, a siloed structure maintains the site’s integrity and overall authority while allowing for clear delineation between subjects, topics, services etc.
While you want to allow navigation between siloes, the aim here is to keep them distinct in meaning.
Best site structure for SEO
There is no definitive ‘best’ site structure for SEO – like so many things in the field, ‘it depends’. The most common is definitely the hierarchy, but one of the best ranked sites on the web is Wikipedia – a matrix. The important thing, when choosing the structure for your site is to remember that it needs to provide the best flow of information for meaning, user journeys and crawler navigation.
For most sites, this will be one of the hierarchies, which allows you to guide the user journey from entrance page through to desired goal completion, to direct and communicate importance to search engines and to group or cluster information into topics and themes. While it is eventually the quality of content and its earned authority through links that will determine where your pages rank, the best site structure for your needs will simply improve your site’s ability to communicate the right signals in the best way.
Site structure visualisation
There are a number of ways to map an existing site, but (as far as I’m aware at the time of writing) all of them will require a paid tool or another (though there’s a 30 day trial of PowerMapper available which claims to generate maps for up to 10,000 pages), but the one most SEO professionals will be familiar with is the Screaming Frog forced crawl visualisation:
While they aren’t as tidy as we might like (when dealing with thousands of individual nodes, how could it be?), even a top-level overview like the above can offer real insight into how your site structure is performing and with the ability to zoom in to each node, you can see whether they’re performing as you’d hope.
You’re looking to craft a structure that, in these visualisations will see multiple subordinate nodes radiating from the homepage, each with their own branching set of related nodes. Each of these, lovely clean (largely imaginary, things will get messy) groupings represents levels in your hierarchy – so make sure they match your plan and that nothing is going wrong with how the structure is being interpreted.
You can also use the crawl tree graph if your site is small enough to give you a more standard hierarchy view (but this gets messy quickly).
Site structure checklist