DU.edu: Information Architecture Case Study

Information is essential to any organization, and universities hold their sitemaps and page titles tightly. When we joined, DU.edu had not been updated in seven years, and its information architecture had not been revised in more than ten.

“The website is like a space, like a building. It represents the University to the outside world... our website needs to be the virtual space, the place, the information source, the image, the representation of everything wonderful that the University is, just like our physical campus does.”DU Steering Committee member

Our hypothesis was simple: by running a card sorting exercise, we believed the new information architecture would be truly informed by, and useful to, the people who use it.

Success meant a sitemap the university could lock in and use to restructure its content, and a navigation, for desktop and mobile, that was easy to find your way around. The new navigation structure became part of the final execution. It launched in 2018, and it is still in use at du.edu. Applications rose by 18%, and accessibility scores and SEO improved.

The work was the information architecture and the navigation that expressed it. Visual design, content and development ran alongside it with the rest of the team.

Working with university administrators, faculty representatives and the marketing department made this a tricky balancing act. The good news is that institutions of higher learning don’t shy away from the chance to learn, so we brought information design theory into the room to bring our clients up to speed and to explain the IA decisions.

• Steering committee and administratorsThe decision makers who owned what the site said about the university.
• Marketing and communicationsThe team that guarded the brand, the sitemap and the page titles.
• Marketing and admissions teamsAbout 18 people from marketing and 35 from admissions took part in the card sorting and shaped the groupings.
• Faculty representatives and studentsParticipants in the card sorting, and the people the site serves every day.
• The project teamOne project manager, one creative director, two senior designers, two developers, and me as the senior UX/IA designer.

Discovery and design ran 12 weeks across six sprints, as part of a redesign that launched in 2018. We moved from what DU had, to what peers did, to what users told us, to a structure.

1. S1
2. S2
3. S3
4. S4
5. S5
6. S6

• Challenge and competitive analysisReviewing what peer universities were doing, to bring DU.edu into the modern web.
• IA theoryTeaching information foraging and information scent, so decisions had a shared language.
• Analysis and card sortingAn Optimal Sort of the current IA with about 53 people from marketing and admissions, plus faculty and a small set of students.
• New IA and sitemapTurning the groupings into a sitemap, then locking it.
• Navigation UIDesigning the desktop and mobile navigation from the sitemap.

Competitive analysis

To bring DU.edu into the modern web, we first looked at what the university space was doing.

Information architecture theory

We introduced information design theory to stakeholders so we could justify IA decisions. The ideas behind it are information foraging and information scent.

Four information scent best practices guided the work:

• Be explicit. Labels should describe what users will find at their destination.
• Show where you are. Give feedback about the current location.
• Skip the jargon. Don’t use made-up words, institutional jargon or your own slogans as navigation items.
• Use plain language. It helps SEO, usability and accessibility.

Card sorting and analysis

Working with about 18 people from marketing and 35 from admissions, plus faculty representatives and a small set of students, we used Optimal Sort to card sort the current IA, then tested it and formulated a new path forward. We analyzed the results with a similarity matrix and dendrogram visualizations to show how the groups formed.

Sitemap and navigation strategy

An IA is not a sitemap, but it leads to one and, ultimately, a user interface. We grouped the card sorting results, tested them against hierarchy models, and drew a sitemap.

Final navigation UI

The sitemap became navigation for both desktop and mobile, designed as a mobile-first, responsive system.

The team recognized early that the organization could only move so fast on decisions. That led to very different outcomes than we first expected.

What I took from it

IA and navigation are never finished. They are improved over time. A website is a living entity, and our solution gave DU room to add new ideas while keeping a solid foundation for users.

Navigation is negotiated. It has to be flexible enough to serve both users’ needs and the institution’s needs.

Let data inform the IA, not dominate it. Card sorting gave us a starting point, and judgment shaped the rest.

Information architecture is not a luxury for big projects. DU discussed and weighed the navigation updates for longer than expected because changing it had such deep implications for the institution. Usability, user-centered language and clear logic were what we wanted to infuse into the solution, and working with large groups of stakeholders means being willing to push and pull ideas.

What I’d do differently. I would have tested the navigation update earlier, with working prototypes. I would also have let the university team co-create more, since they took longer to come around to owning the changes.