The stakes: one in three visitors couldn’t finish what they came for
A prospective patient could arrive with a simple question about treatment time, insurance, or finding a provider and leave without an answer. Usability testing showed that only 67% of visitors could complete these core tasks. Years of organic growth had turned the navigation into a record of the company's structure instead of the patient's journey.
I was asked to restructure the full experience without sacrificing its search equity, while keeping the model flexible enough for more than 90 markets with different regulatory requirements.
"I clicked on 'For Doctors' by accident and got completely lost. I'm just trying to find out if my insurance covers this."
Patient participant
A site that outgrew its own map
The site had become an ecosystem for patients, parents researching for teens, and dental providers. Each audience brought different questions and every market brought different rules. The old structure made those paths cross too early, burying critical content four or five clicks deep and sending visitors into sections meant for someone else.
The Scale of the Problem
Where Users Got Lost
Session analysis showed exactly where the structure broke down: the main menu blurred audiences, the provider finder lost people before results, and treatment answers sat too deep in the tree.
The discovery: patients sort by goal, not by business unit
I separated the company's content model from the visitor's mental model. A 300-page audit exposed duplication and drift. Card sorting with patients, parents, and providers revealed how people grouped questions. Tree testing then showed whether the new language held up under real tasks.
Content Audit
Inventoried all pages across markets, identifying redundancies, gaps, and content that had drifted from user needs.
Card Sorting
Open and closed card sorts with patients, parents, and providers to understand mental models across user types.
Tree Testing
Validated proposed taxonomy with task-based tree testing before committing to development resources.
Key Insights
The evidence converged on one idea: visitors organized the experience by goal, not by department. They wanted to understand treatment, judge fit, find a provider, or get help. The navigation had to begin there.
Navigation Confusion
"I clicked on 'For Doctors' by accident and got completely lost. I'm just trying to find out if my insurance covers this.", Patient participant
Content Depth
"I gave up after clicking through five pages trying to find treatment time information. Why is this so hard?", Parent participant
Mobile Experience
"The menu takes up my whole screen and I still can't find what I need. I'll just call my dentist instead.", Mobile user participant
The turn: rebuilding the map around patient tasks
The turning point was rejecting another layer of labels on the old tree. We rebuilt the information architecture from the ground up, using the card-sort language to create a task-first taxonomy that could scale across audiences and markets.
Three Design Principles
The solution was grounded in three core principles that directly addressed the research findings:
Task-First Navigation
Reorganize around user tasks ("Find a doctor", "Learn about treatment") rather than business units. Clear audience separation with persistent wayfinding so users always know where they are.
Progressive Disclosure
Surface key information upfront with clear paths to detail. Reduce maximum depth from 5 clicks to 3 for critical content while maintaining comprehensive information for those who want it.
Mobile-First Redesign
Design for the 60% first. Simplified navigation patterns, thumb-friendly tap targets, and streamlined provider finder optimized for on-the-go research.
Implementation Process
Content Audit
Inventory and gap analysis
Card Sorting
User mental model mapping
Tree Testing
Validate new taxonomy
Global Rollout
Phased deployment
The proof: task success from 67% to 94%
The redesigned site was rolled out over six months with continuous measurement against baseline metrics. Results exceeded targets across all key performance indicators.
Rollout Strategy
Measured Impact
2x Site Traffic Growth
Better UX improved engagement metrics that search engines reward. SEO and UX aligned to drive organic growth without paid acquisition.
What it taught me
-
IA is the Foundation
No amount of visual polish can fix broken information architecture. Getting the structure right enabled every other improvement.
-
SEO and UX Can Align
The fear was that restructuring would hurt search rankings. In reality, better UX improved engagement metrics that search engines reward.
-
Global Requires Local
A structure that works in the US may fail in markets with different healthcare regulations. We built flexibility into the IA from the start.