Developing an Information Architecture

Try Loop11 free. Sign up now to get started. Get Started

Course Chapters

Based on the findings from the prototype user testing, we felt justified to take another look at some areas of content classification and labeling within the prototype design. Although task success was quite high in the previous round of user testing, the process of discovery could have been more efficient for tasks such finding sales items.

We chose to use Optimal Workshop’s tool, OptimalSort, for setting up the exercise and analysing the results.

Purpose of card sorting

Card sorting is a well-established research technique for discovering how people intuitively categorize information. You can use card sorting results to group and label your website information in a way that makes the most sense to your audience.

Card sorting involves creating a set of cards that each represent a concept or item, and then ask participants to group the cards in a way that makes sense to them.

Create cards and categories

To create our card sort, we first started by collecting all major content items which exist in the current EverWear prototype design.

Next, we transposed individual concepts onto cards, ensuring we applied common and easily understood language. Through this process we also tried to make sure the concepts we created were mutually exclusive, so that there was little, or no perceived overlap in content.

We chose to undertake an open-sort activity where participants sorted the cards into logical groups, and then labeled these groups themselves.

26 participants were sourced from an online panel to undertake the card sorting exercise.

Analysing your card sorts

The results of the sort produced some interesting key insights which including the following:

  1. The majority grouped all specials and sales discounts together
  2. Several participants thought there was room for a blog, or a “News and events”, category, for competitions and news related items
  3. There were two common approaches to grouping clothing:
    A. by category (such as hoodies, hats)
    B. gender, which most participants thought made the most sense

The obvious categories revealed through the card sort were: Mens / Womens / Sale and About us. To address issues found in the prototype testing, we felt the revised architecture should group sale items together, regardless of what the “special” was. This should result in less confusion and a more direct navigation.

Sustainability and “Relax & Win” were also two areas which participants found difficult to categorise. In light of these results, the team debated whether it would make sense for Everwear to have a blog which could provide an active vehicle to engage with the customers and positively impact the business.

Through this process, two potential new IA structures were created, containing variances in the form of labeling and grouping. The team then set about to understand which of the two was more intuitive to potential users.

Version A: About us, Men, Women, Sale, Sustainability, Relax & Win

Version B: About us, Men, Women, Sale, Everwear blog

Tree Testing

To evaluate which version of the IA was most effective, the team carried out a Tree Testing exercise using Optimal Workshop’s tool, TreeJack.

Tree testing removes the UI from the equation and lets you see how easily people can find information on your website, and exactly where they are getting lost. It can also answer questions like:

  • Do my labels make sense?
  • Is my content grouped logically?
  • And, if people are having difficulties navigating, what’s stopping them?

Version A – Overall performance – Success 65% – Directness 76%
Version B – Overall performance – Success 67% – Directness 86%

The second IA performed better in testing, however the difference was marginal.

Some key findings of the Tree Testing included the following:

  • Evidence that it would be beneficial to include a “sales” subcategory within the “men” and “women” categories of the site in addition to the key “sales” category. Tree testing revealed that a number of participants drilled down to “womens” and “mens” categories in search of sale items.
  • A number of participants expected to find sustainability content within the “About Us” category, whilst others expected to find it within specific clothing pages. As such, additional information on sustainability practices could be added to both “About Us” and individual product pages to reinforce the origins of the fabrics and the brand’s philosophy.

Finalising the Information Architecture

Armed with learnings from the card sort and tree testing, the team developed a finished IA containing the following main categories:

About us, Men, Women, Sale, Everwear blog

We’ll then take these categories and apply the new IA to our design to be used in the next chapter, live site user testing.

Want more inspiration?
Join the Fab-UX 5 newsletter!

Five links to amazing UX articles,sent to you once a week.

No SPAM, just pure UX gold!

No Thanks