We’ve all been on websites that make it difficult to find what you’re looking for, and it’s usually a frustrating experience. On the other hand, a well-structured website can be a joy to use. That’s why Information Architecture (IA) – organising information in a way that simplifies how people navigate and use it – is so important online. Good IA helps users navigate quickly, defines the structure of a website, and has significant overarching impacts on its UX.
This blog is the first in a three part series that reveals the process we followed to redesign a massive information architecture. We’ll also include some tips to help you overcome your own IA challenges. This first blog will cover the IA design process at a high level, and the second two posts will delve deeper into card sorting and tree testing.
IA redesign: starting the journey
We recently reviewed the information architecture of the website owned by one of our clients, Barnardo’s. As the Barnardo’s site had grown, its IA had become fragmented, making the site difficult to navigate. With over 185 unique labels to name and organise, improving IA was a top priority if we were to ensure a good UX for this large and complex website.
So, how did we do it? While the standard UX principles of user research, design, and usability testing always apply, there are special tools to use and considerations to make for IA.
Collate the existing content
The first step is to review the pages of your website to understand their context and message. Create a label for each page and write it on a card. The end result is a stack of cards that represent the current content of the website. These cards can then be used for a card sorting exercise.
Do a card sort to organise your content
In a card sort, participants organise topics written on cards into groups. This helps to define categories, group related content, and verify content labels. The exercise removes the content from the context of the existing IA, providing a blank canvas.
You can then either sort the cards with people that represent real users, or do the card sort yourself. Use your card sort output to propose a new IA. This is the starting point for your redesign, not the final solution. If you have two valid propositions, you can A/B test these in the next step: tree testing.
Conduct tree tests to verify your assumptions
Tree testing is a form of usability testing that evaluates how easy it is to find content in the information architecture (i.e. navigation). During tree tests, participants undertake a set of tasks that ask the user where they would look to find content.
The output of the tree test reveals how quickly users find content, where they look, and where they get lost. It’s the perfect way to test the proposed IA you developed with your card sort.
You can measure the success of your tree test with metrics that indicate how easy or difficult a task was to complete (most tree testing software provide such metrics). Also, review user paths to see how users navigated the IA. This will help to validate the IA design or guide improvements.
Like all usability testing, tree tests should be conducted with participants that represent the user groups of your site.
When you know what worked and what didn’t work for your proposed IA, you can improve it. If you tested multiple versions of a tree, you’ll be able to see which performed the best. If you noticed participants consistently looked in the same ‘wrong’ place for certain content, you could move your label there to make it the ‘right’ place.
With these steps, you will be able redesign your IA so that it works better for your users. But it doesn’t end there. When your IA is up and running on your website, you should continuously monitor how well it is performing by checking your visitor site analytics.
Do you need help creating or redesigning your information architecture? Contact Matthew, our Technical Director, today on +44 207 125 0160 or drop him a line on [email protected] for a free consultation.