Case study · Artera · Health tech
Evolving a legacy nav on user evidence when marketing wanted sales language in the product.
The problem
Earlier promotions had moved Campaigns and Automations out of Settings onto the primary nav. Right direction, no room to scale, and Insights lived behind a separate login.
The product was built for one clinic at a time. Healthcare moved to centralized call centers, which meant staff were constantly flipping practices to see messages, and activity was easy to miss.
Branded terms, Conversation Builder and Conversation Manager, were too long for a nav, and icon-only meant we wouldn’t meet a future accessibility bar. We were in a standoff with PMM.
Role & scope
Drove decision rounds with Product Marketing, Sales and CX, and owned the downstream-impact framing for the ELT.
Moderated and unmoderated user testing (tree tests, preference tests, task-based usability, first-click) and collaborative workshops.
Information architecture and the primary nav system, scaffolding for future surfaces (including Enterprise Inbox), and low- and high-fidelity design.
Approach
Team-wide sessions proposing a scalable IA to win with Harmony.
Tree testing and preference testing for IA organization and brand treatment.
Two research cohorts to lock main nav and sub nav labels and the interaction pattern.
Release plan and downstream impacts; customer comms and engineering routes.
Released to all accounts; shared via webinar, email, and an in-product walkthrough.
Collected and responded to customer feedback; planned iterations with the roadmap.
Deep dive: design
Deep dive: research
Across tree testing and preference testing. Other data was mined from NPS, previous usability tests, and customer success.
As the name for branching message logic. Flow got 17%; Conversation, Workflow, and Sequence got 0%.
The scheduled-message task split almost evenly between Workflows and Triggers. Neither result was strong enough to decide, so we used product judgment and shipped Triggers.
44% dark grey, 41% dark teal. Only 15% chose white.
Outcomes & impact
Labels surfaced capabilities that icons had hidden. Campaigns is the proof: staff who never recognized the icon started asking what it was. That was the goal, not a support failure.
Text labels plus contrast that met accessibility guidelines in the primary nav. Not full compliance, but the first real step toward a future government contract.
Marketing owns buyer-facing language. Design owns in-product labels. That split was the outcome, so the next rename wouldn’t restart the conflict.
This phase was a stepping stone. The Inbox drawer was temporary. Enterprise Inbox would make Inbox a destination with its own nav, which is why the extra click to Mentions was acceptable to hold, not a reason to undo the system.
Feedback & next steps
Feedback
“I want to escalate the need to get the mentions added back to left hand column to make it easier to see when there is something being called. There was a change that moved it to an additional click.”Enterprise health system customer
Response
Already on the roadmap. Enterprise Inbox would make Inbox a destination, so Mentions sat with the other inbox views instead of behind a second click.
Feedback
Users expected the open main nav item to stay highlighted. Other users expected a second click on a given item to close the drawer. Valid feedback on interaction patterns.Shared via customer success team members
Response
Logged, not rolled back. Highlighting the open item and closing the drawer on a second click shipped as polish, after the IA was held.
How research changed what we built to give enterprise buyers proof they could act on.