Alliance for the Great Lakes
Redesigning Navigation for a Nonprofit
.png)
Reorganizing a nonprofit's website so donors, volunteers, and researchers can actually find what they need—increasing annual report findability by 35%.
Role
UX Researcher​
Team
4 UX Researchers​
Duration
10 weeks
Tools
Figma, Wireframing, UX Tweak, Tree Testing, Site Maps, Card Sorting
​
The Problem
The Alliance for the Great Lakes website had confusing navigation with duplicate content paths and unclear labels. Users struggled to find volunteer opportunities and donation information, with success rates as low as 33% for critical tasks like locating the annual impact report. The volunteer signup redirected to an entirely separate website with different navigation, breaking the user flow completely. This wasn't just frustrating for users. It directly prevented the organization from converting visitors into volunteers and donors, two things nonprofits really need.
Research Approach
We needed to understand not just where users got stuck, but why the current structure wasn't matching how they actually thought about the content.
Card Sorting​
2 Rounds
12 Participants Total ​
Revealed users couldn't tell the difference between 'Issues' content and 'News' content
Showed us which labels were actually intuitive vs which ones we just thought made sense
Tree Testing
2 Rounds
25 Participants Total ​
Only 33% could find the annual report in Round 1
Helped us figure out where users expected to find specific content
Heuristic Analysis
Evaluated 2 Core Tasks
Found duplicate paths, missing confirmation feedback, and confusing navigation for users
Key Findings
Three patterns kept showing up across all our testing:​
1
Users want action words, not category names.
When trying to volunteer, they looked for "Get Involved" not "Programs" or "Opportunities." The verb told them what they'd be able to do on that page.
2
Annual reports are publications, not org info.
We assumed the annual report belonged with other "About Us" content, but users went straight to News to find it. They treated it like any other recurring publication.
3
"Donate" signals individual giving.
Moving corporate partnerships under Donate actually made things worse. Success dropped from 89% to 37%. Users think of partnerships as collaboration, not donation. They expected to find it under Get Involved.
Iteration & Decision Making
Round 1 tree testing revealed an issue with the annual report placement. Only 33% of users could find it. Users were split between looking in News and looking in About, which told us the "About" label wasn't communicating what we thought it was.
​
For Round 2, we moved the annual report to Stay Informed (our renamed News section). Success rate jumped to 69%. Not perfect, but way better.
​
The corporate partnerships decision was harder. In Round 1, we had it under Support Us and users found it eventually (89% success), but 0% went there directly. Everyone started in Get Involved. So for Round 2, we moved it to Donate, thinking "this is still financial support." That was wrong.
​
Success dropped to 37%. Users overwhelmingly looked in Get Involved first (50% of first clicks). In hindsight, the data was telling us where it belonged all along. We just got attached to our own logic about what "donation" meant.
​
​
The Solution
The final sitemap consolidated scattered content and used action-oriented labels based on where users actually looked during testing.
.png)
Key Changes Implemented Include:
1
Consolidated volunteer opportunities under "Get Involved"
Everything action-oriented (cleanups, volunteering, youth programs, petitions) lives in one place. Beach cleanup task success improved from 89% to 94%.
2
Created action-oriented category labels
Changed "Issues" to "Explore Issues," "News" to "Stay Informed," "About" to "Learn About Us." Small change, but keeps category labels consistent.
3
Moved Annual Impact Report to "Stay Informed"
Treated it like the publication users expected it to be. Task success went from 33% to 69%.
4
Added persistent confirmation pages
Replaced disappearing pop-ups with actual confirmation pages users could screenshot and reference. Reduced uncertainty about whether donations and registrations actually went through.
Wireframes
Task 1: Register for a Beach Cleanup
Home Page & Drop Down
Tree testing showed 81% of users went directly to 'Get Involved' for volunteer tasks.The menu shows users five options for ways to get involved with the organization.
.png)
Cleanup Details & Sign Up
Once the user clicks on the “Join a clean up”, they will be directly navigated to a page that allows them to see the upcoming dates.
The original webpage had users navigate through different pages between clicking “Join a Cleanup” and the page to view upcoming cleanups.
Also, the navigation bar stays consistent and flows through each page, rather than disappearing.
.png)
Confirmation Page
Replaced brief pop-up with persistent confirmation page. Heuristic analysis found users missed temporary confirmation and weren't sure if registration completed.
.png)
Task 2: Make a Monthly Donation
Home Page & Donate Drop Down
Clear distinction between 'Donate Now' (direct to portal) and 'Other Ways to Give' (learn first). Reduces clicks for users who already know what they want.
.png)
Make a Monthly Donation
Portal branding matches main website design, reducing confusion when transitioning between platforms. ​
Donor Support contact visible during transaction to help prevent errors. ​
​
"Go to Checkout" button clearly labeled, presenting a clear next step for the user.
​​
.png)
Confirmation Page
Full confirmation page with transaction summary instead of disappearing pop-up. Addresses heuristic finding that users had no record or confidence their donation went through.
.png)
Reflection
Leading this project taught me that small label changes create massive differences in findability. The biggest surprise was corporate partnerships. Our logical placement under Donate completely backfired. Users treat partnerships as collaboration, not donation, and the data was showing us that from Round 1. We just got attached to our own reasoning.
​