URLs are one of the primary ways of user interaction with a website. In many cases, influencing Usability and how the user perceives the website’s structure. A URL Structure for SEO can, and should, also meet Usability principles.
When the subject is URL structure, SEOs and Webmasters tend to look at it, mostly, from an technical and isolated standpoint. I lost count of how many times professionals chase flat architecture myths. Or the “What’s the best URL structure for SEO” myth. We rely only on superficial technical advice, and forget most of the usable aspects of designing for users AND machines.
When presented with a question regarding URL structure, SEOs tend to leave on the table most considerations regarding usability and user interface. After all, we still chase around beliefs that sites should have a flat architecture. Or that URL directories hinder the ability for a page (and ultimately a website) to rank well in search results. Let me break the news for you, and tell you that’s completely false!
Like a good information architecture, a well designed URL structure is the foundation of any website. It will pave the way for a good website performance, but most importantly it should also factor in and contribute towards usability and Accessibility aspects. From a reputation standpoint, URL changes are a website’s worse enemy. Basic PageRank math says that with every hop from one document to another, a decay should be factored in, otherwise one could just increase PageRank of a document to infinity by linking to itself in a loop; the same applies to URL changes, since technically you’re hoping from one document to another. Therefore one of the most desired qualities of a well designed site architecture is perenniality.
That’s also one of the reasons why you should always fix internal redirections, and make sure your internal linking always points to canonical URLs.
Long lasting URL architecture
When designing a new URL architecture, the main question we should ask ourselves is: How long will my site architecture last without needing to be changed?
When we think about URL structure only from a technical standpoint, we often pass on the opportunity of getting ahead of the herd, and planning a well thought and long-lasting foundation. First and foremost, let’s not mix URL structure with website and breadcrumb navigation. These 3 things can be completely aligned, but can also be crafted differently from each other to better serve user intent and interaction with a website. Let’s go with a couple of simple and practical examples.
URL architecture for a small online store
Let’s imagine I’m running a very specific sub vertical online store. I might have a restricted stock and focus on a specific item like Jedi Lightsabers. In that context I would probably have product categories like:
Then my product URLs could inherit the category structure:
Or even have a more flat and non-hierarchical structure without much compromise to usability principles:
Since I’m running a niche site, and niche sites tend to be just about a very specific and sub vertical market it might be OK to lean towards a flat structure, mostly because topicality wouldn’t really be lost.
URL architecture for a big online store
I’m running a very large site… No wait, I’m running an Imperial Intergalactic Megapod. I’m no longer a niche website; I sell everything, from Jedi Robes and Lightsabers, passing the Droids, X-Wings and Tie Fighters section, to Fuel Cells. In that context I would have product categories like:
http://www.imperialmegapod.int/robes/sithlord-cloaks http://www.imperialmegapod.int/lightsabers/crossguard-blade http://www.imperialmegapod.int/spaceships/x-wing http://www.imperialmegapod.int/droids/mark-I http://www.imperialmegapod.int/helmets/stormtrooper
Since the variety of products I’m selling are of much wider in diversity, a flat URL structure would probably compromise the ability for my users to navigate the website without losing track of where they are, and most importantly the perception of content organization and distribution across the website. To better illustrate this, let’s look at how a flat URL structure would look in this case:
http://www.imperialmegapod.int/jedi-hoodie-cloak-cotton-brown-beige-w0773 http://www.imperialmegapod.int/x-wing-t-70-white-s0880 http://www.imperialmegapod.int/r2d2-multi-function-utility-interface-arm-titanium-r2910 http://www.imperialmegapod.int/tiex1-fuel-cell-ics-fuel-to-laser-c8893 http://www.imperialmegapod.int/first-order-stormtrooper-helmet-white-0033
URLs are not tied to their hierarchical directory structure, there is no topic really defined. This is something that, from a usability standpoint matters, even more if URLs are less descriptive and lean towards a more simple nomenclature.
Now imagine you receive a list of those URLs in a message, email from a friend or any other non-visually crafted way. Some might not even have a hierarchical organization for subcategories, and all URLs will just stem from the root domain like a spaghetti incident.
Wouldn’t it be helpful for you to understand the website if those products were organized and structured inside their respective categories and subcategories?
http://www.imperialmegapod.int/robes/sithlord-cloaks/jedi-hoodie-cloak-cotton-brown-beige-w0773 http://www.imperialmegapod.int/spaceships/x-wing/x-wing-t-70-white-s0880 http://www.imperialmegapod.int/droids/r2-type/r2d2-multi-function-utility-interface-arm-titanium-r2910 http://www.imperialmegapod.int/spaceships/tie-fighter/tiex1-fuel-cell-ics-fuel-to-laser-c8893 http://www.imperialmegapod.int/helmets/stormtrooper/first-order-stormtrooper-helmet-white-0033
The person receiving these URLs might not want what you’re sending them, but if you have an hierarchical product URL, they will be lead to perceive that this website might have other similar products inside each category, or sub-category. As Jakob Nielsen points out in the linked article above, a usable site requires:
- a domain name that is easy to remember and easy to spell;
- short URLs
My take here: This shouldn’t be taken as blanket statement. You should always prefer shorter URLs, as they are less prone to errors, but it doesn’t mean you should avoid having a structured URL path.
- easy-to-type URLs;
My take here: I will type more easily the path to category or sub-category, and then navigate to the product, than a path of very long URL in a flat structured website.
- URLs that visualize the site structure;
- URLs that are “hackable” to allow users to move to higher levels of the information architecture by hacking off the end of the URL;
My take here: you can only do this with well structured URL paths.
- persistent URLs that don’t change
My take here: perenniality.
Sometimes going after the easy recipes like “Google gives more weight to URL paths closer to the root” and not thinking further, might not be the best approach. Google oversees their interests and often gives broad advice as an attempt to fit most of the possible different scenarios. As Webmasters and SEOs, we should always try to bring in usability principles that will contribute to a longer perenniality of the website’s architecture. Moreover it becomes easier to plug in more categories and expand without losing topicality.
Best practices for choosing a great URL structure
Consider that going with too much towards extremes might no be a good idea. Having extremely long URLs is as bad as having a flat architecture sitewide.
When Google says stuff like “Google gives more weight to URL paths closer to the root”; they will usually assume you link 2nd level deep URLs from the root and 3rd level deep URLs from 2nd level deep URLs, and so on. This usually is not interpreted by the masses as it should, and often people tend to take this kind of advice in a very literal way.
Search engine users more often than not look at a URL when they’re assessing the credibility of a website. If URLs look messy and disorganized, your chances to attract a click from users decreases considerably. So, whatever is the configuration you chose, make sure your URL structure meets most usability standards and isn’t just machine friendly.
Don’t stick with the first draft… Most of the time, I find out it’s great to sleep on it a week or two before deciding on a URL architecture. Usually my process on designing URL structure goes over several steps and attempts. I usually try to draft 2 or 3 choices, and in each of them I emphasize a determined behavior where I factor in user intent and usability principles.
Next time you have to think about a new website, try to involve people from outside pure technical teams. It’s probably a good idea to let someone with experience in usability and information architecture to chime-in. In the end, always try to go with a balanced approach.