Optimal Workshop
Figma
The YSPH’s site featured an overcrowded mega menu panel that required excessive clicking among accordions — frustrating students, faculty and staff.
How might we create a primary nav that is less overwhelming for users?
A more traditional experience with multiple mega menu panels/categories — and without accordions — that reduces the overall cognitive burden on users while also better orienting new and returning users.
How did we get there? Read on.
Key stakeholders were asked directly in discovery interviews, “Why do you have so many links in your menu?” The answer was shocking. Their shared understanding was that every page of a site must be part of “the” menu — not just “a” menu. That nuanced difference had monumental implications for the site.
Instead of relying on section nav or in-page nav elements throughout the site to drive users deeper into content, they flooded a singular menu with every link from five levels of content site wide. And in order to manage that overwhelming volume, an excessive number of accordions were added.
To reassure stakeholders that essential content would still be accessible to users, we shared a variety of section nav options. The new shared understanding opened the door to significantly reducing the number of links. However, there wasn’t clear alignment on narrowing down the 455 links, what content should remain at the highest levels, and how it should all be organized.
As a way to move forward, I recommended we get input from users via a card sorting research activity.
Understand the mental models behind how students, faculty and staff naturally group YSPH information and label categories — so secondary and tertiary content could be removed from the site’s primary nav
Format: Open card sort – online
Participants:
- 47 faculty (10), staff (15) and current students (22)
- 74 prospective students of all degree levels and programs
Timing: 1/9-1/25
Platform: Optimal Workshop
- Which cards are grouped together by participants (this similarity matrix)
- How many categories they create and what they label them (Participants created anywhere from 2-12)
- Which category users sort the cards into
- Finding: 28 of 49 cards consistently put in similarly named categories
- Implications:
- Very clear picture of the mental model and expectations of the audiences
- Very solid foundation for the new YSPH information architecture and nav
- Finding: 4 cards had more variety in placement
- Implication: Experience provides a clear path forward for placement and labeling
- Finding: 8 cards had much greater variety in placement among a larger set of categories
- Implications:
- The topics don’t fit neatly in one category, so users put them in many different categories based on their own mental model
- Placement to be heavily influenced by the overall set of final categories, anticipated content, and strategic priority
- Finding: 9 cards were placed in categories that will likely not exist on the eventual site
- Implications:
- Final placement heavily influenced by experience which shows that users will be better served by including those topics in other categories
The card sort research resulted in a clarified architecture for the site and the nav structure, so we presented a recommended mega menu — with an 80% reduction in links from the original. Stakeholders were extremely concerned that users wouldn’t be able to find or get to key content at levels 3, 4, and 5 of the site.
Just sharing key UX concepts like “red routes” and “information scent” and restating the addition of section nav wasn’t resulting in the positive agreement necessary.
I made the case to do tree testing with internal and external audiences to validate whether or not the new architecture and labels would allow users to efficiently find and get to key content at deeper levels. We came to an agreement on the eight highest priority user tasks for testing.
- Validate the findability of key information within the new content categories and hierarchy
- Confirm that the navigation is efficient and aligns with user expectations
Format: Online
Participants: 70 anonymous responses from faculty, staff, and current students
Timing: 2/1-2/9
Platform: Treejack from Optimal Workshop
- Success rate: % of participants who found the right category or categories to complete a task
- Directness: % of participants who went straight to the correct category without backtracking or trying different categories
- First click: % of participants whose first click headed towards the right category or categories first. This is important because it is a key predictor of task success
1. Find the overall cost of attendance for a Master of Science in Public Health (MS) student for the upcoming year.
2. Find the last day of classes for the current semester.
3. Verify what is required for a student to graduate from the two-year Master of Public Health (MPH) program.
4. You’ll be attending a public health conference in a couple weeks. Make sure your talking points and key facts about YSPH are up to date.
5. Find coverage of Dean Ranney’s, MD, testimony before the U.S. Senate Judiciary Committee in November on gun violence.
6. Find a relevant internship opportunity for an MPH student that came to your office hours.
7. Someone you know asked for an introduction to someone in Health Policy & Management. Find the contact info for a specific colleague so you can make that introduction.
8. Find out where graduates of YSPH end up working after leaving the School.
The tree test provided clear and compelling evidence to the YSPH faculty and staff that key content would still be findable and navigable in the completely transformed primary navigation. The alignment and confidence from this set of activities also contributed to a positive client/agency partnership on other site improvements as well. And overall, the primary nav will benefit from an 80% reduction in links and reduce the cognitive load currently being put on users.
With the complete new enterprise site scheduled to launch in Fall 2025, we anticipate conducting more research in the future to validate that the site will be better serving internal and external users. We also anticipate comparisons between past and future analytics to identify any areas of the site that may need immediate refinements.
