Main Navigation Redesign

What's this all about?

  • Client: U.S. Federal Agency with 27.5k users
  • Problem: Poor navigation was costing time, and creating accessibility burdens
  • Outcome: Redesign launched with a 97% acceptance rate, meeting accessibility requirements, and future-proofed for upcoming additions
  • Note: Images have been anonymized due to Federal security requirements

“…honestly, my experience has been seamless. I forgot I was using it.”

–Federal User

“I do not want to go back to the old version!”

–Federal User
The Problem

Users struggled with a navigation system that proved to be more of a visual burden than a help.

We identified two key problems to solve.

  1. Historically, each time a new tool, page, or skill was needed, a tab had been added to the main navigation. As the navigation grew, it took up an increasingly large amount of screen real-estate. For those with the broadest access, it occupied half the screen.
  2. Tabs were not added in any particular order, ultimately slowing down users, and costing time and money.

This high-visibility project represented the largest single design change in 9 years, and required a careful, risk-managed roll-out to be successful.

The Solution

A navigation system that is easier, faster, and more accurate helps our users complete their work more quickly with less effort. This saves time and money for the organization as a whole.

The main navigation was rebuilt from scratch. I conducted user research, interview insights, and iterative testing, then partnered with development to ship safely in a Federal internal environment. Because of our risk-averse audience, a phased rollout was required, starting with one office and expanding to the national user base.

The new structure and labels reduced cognitive load on users and prioritized the destinations needed most often. Through each iteration, we measured success by a net reduction in misclicks and hesitation. During the 3-week roll-out, users opting out of the new system fell to 7% and kept trending downward as confidence increased.

The giant wall of tabs had to go. I created an overflow solution that felt familiar and easy to use for our officers.
General users had specific job duties and so fewer tools.
Supervisors had a higher level of access and many more tools and resources available to them.
The Research

We found that prior research on navigation didn't meet the needs of users across hundreds of offices with different functions. We needed to start from the beginning.

Variable Access

Users have widely variable navigation options, depending on their position and their job duties. We comprehensively audited all possibilities, in order to create our information architecture.

Non-Navigation Users

A significant number of users have limited task types and access. This means they do not see or use the many options from the main navigation. A simplified version was required.

Inconsistent Naming Conventions

With no clear conventions in place, navigation items had confusing and inconsistent names. We conducted a card-sort to guide decisions in defining menu items, future proofing for expansion.

Messy Middle

We delivered a clean design for users with the smallest, most common, and largest access levels. We used Matomo analytics to collect user screen sizes, ensuring the navigation would work for everyone.

Usability Sessions & Beta Test

We piloted first with a small group of users. I ran usability sessions, specifically looking for users' ability to reach key sections quickly and confidently.

For our sessions, realistic tasks were based on common activities. I observed time to destination, misclicks, and moments of uncertainty. Participants were asked if each label contained the expected topics and whether the information architecture matched their mental models.

Leadership is highly risk-averse, so the live beta test mitigated risk ahead of a wider release. This iterative approach led to strong adoption with few surprises.

The Challenges

Flexibility, low ego, and clear communication kept this project on track.

Tool Limitations

Due to government access and tool limitations, we created our card sort in Excel. This low-tech method successfully delivered the information we needed.

User Access

Our users' work days are heavily monitored, which adds constraints to user research. We relied heavily on supervisors, who had more flexibility in their day. We also took advantage of on-site clinic work to casually speak to general users.

Communication

We were part of a larger team spanning five organizations. The cross-team nature combined with challenging personalities required careful interactions and strong alignment within our team to avoid unnecessary delays in the project.

The Roll Out

Full roll out landed with a 7% opt-out rate that continued to drop over the first three weeks.

What Worked Well

  1. Design team communications — how we worked alongside the development team.
  2. Set up New Relic for data analysis of navigation items — commonly vs. infrequently used items.
  3. Great support from Federal partners to gather participants in multiple field offices.

What I'd Do Differently

  1. Pushing harder for earlier feedback from lower-level access users would have increased the speed of the simplified navigation build. The majority of users we spoke to were super users.
  2. Requesting a BA be assigned to the larger team, to clarify user flows or to ask content-specific questions.

Soft Launch

  • After a successful beta test, we enabled the main navigation for everyone as an opt-in, and promoted the opportunity in all communication channels.
  • We monitored for adoption and bugs, squashing them quickly to gain user confidence.
  • Analytics showed a steady increase in opt-ins, demonstrating that users preferred the new design enough to seek it out, and that the benefits were spreading via word-of-mouth in offices.