CASE STUDY
Pinterest — A Scalable Navigation System
Hypothesis: Restructuring navigation around user intent would improve clarity and
increase engagement quality.
Desktop Web is Pinterest’s highest-intent surface. Desktop sessions are more deliberate and task-oriented—supporting behaviors like retrieving saved content and managing projects.
Pinterest Desktop Web (After): Side navigation with elevated Boards access and a simplified menu structure.
Improved Boards access, increasing revisit behavior strongly correlated with MAU/WAU retention
Reduced navigation-related support tickets (400+ at peak)
Launched fully supported RTL navigation (Arabic, Hebrew)
At the start of this work, navigation was fragmented across a top bar, profile menus, and a floating “Help” menu that obscured content. Key destinations like Boards, Settings, and Support were difficult to predict and often required multiple steps to reach—interrupting workflows and forcing users to navigate around the interface rather than through it.
Pinterest Desktop January 2025 (Before): Top navigation and legacy "?" menu in bottom right corner.
Users often arrive with a specific intent—retrieving saved content, managing ongoing projects, or resolving an issue—making navigation clarity especially important.
“When I’m on my desktop, I’m usually working on something specific. On my phone, I’m doing six things at once.”
— Desktop Pinner
This reinforced the need for a navigation system that prioritized continuity, predictability, and efficient access over novelty or discovery-driven patterns.
User need:
Get back to saved content quickly and reliably.
Boards were a critical destination for desktop users, but access was buried under Profile and inconsistent across navigation surfaces. Research and experimentation showed that users treated Boards as a primary retrieval tool rather than an expression of identity.
“I use my boards all the time. I just want a fast way to get back to them.”
— Desktop Pinner
Because revisiting saved content is strongly correlated with MAU/WAU retention, improving Boards access was treated as a retention problem—not just a usability improvement.
User need:
Resolve issues without interrupting their current task.
Users did not distinguish between “Settings” and “Support.” They approached both as part of a single problem-solving flow and were frustrated by navigation that forced them to guess where an action lived or obscured content mid-session.
“If something’s wrong, I don’t really care what it’s called — I just want to fix it.”
— Desktop Pinner
This insight directly informed the decision to consolidate Settings and Support into a single, predictable navigation surface and remove the floating menu that previously disrupted browsing.
User need:
Access identity-related information without clutter.
Profile was expected to represent identity and account state—not content retrieval or troubleshooting. Separating these concerns reduced cognitive load and clarified mental models across naviga
Across research, a consistent pattern emerged: desktop users valued clarity, continuity, and efficient retrieval over surface-level organization. These insights informed the information architecture and interaction decisions that followed.
With the problem framed around user intent, I conducted an audit of the existing Desktop Web navigation.
Architecture (Before): user flows included multiple, overlapping paths to the same destinations.
The diagram above maps how users navigated between Settings, Support, and Profile prior to the re-architecture. Related actions were spread across multiple entry points, requiring users to move between surfaces and menus to complete simple tasks.
The flow highlights how navigation had grown incrementally over time—introducing friction and requiring users to relearn where things lived.
Wireframe (Before): user flows included multiple, overlapping paths to the same destinations.
These wireframes show the existing Settings, Support, and Profile menus at the time of the audit. While each menu made sense in isolation, overlap across surfaces created inconsistent mental models.
This fragmentation surfaced in support feedback, where users often didn’t know where to go for basic account actions—such as logging out.
I reorganized navigation based on what users were trying to accomplish, rather than where features had historically lived.
This resulted in three primary intent groups:
Retrieve — getting back to saved content and ongoing projects
Resolve — addressing issues, preferences, and account needs
Identity — profile and account state
This reframing allowed each destination to live in a place that matched user expectation and reduced cognitive load across sessions.
Architecture (After): Grouping by intent
Boards were treated as a retrieval destination rather than a profile attribute.
By elevating Boards access within primary navigation, users could return to saved content more directly—supporting revisit behavior strongly correlated with MAU/WAU retention.
Boards (After): One-click to Your Boards.
Settings and Support were consolidated into a single problem-solving surface.
Research showed users did not distinguish between these concepts; they simply wanted to resolve an issue. Consolidation reduced confusion, simplified navigation paths, and allowed removal of the floating menu that previously obscured content.
Settings & Support (After): Settings and support actions are consolidated into a single, organized menu.
Profile was refocused on identity and account state.
Actions like logout were moved into Profile, reducing noise elsewhere in the navigation and reinforcing clearer mental models across surfaces.
Account & Profile (After): Key account actions, such as logout, are now located within Profile.
These changes simplified navigation structure, reduced fragmentation, and created clearer ownership across destinations—setting the stage for a more flexible and scalable navigation
Animation: Before and After: Navigation Re-architecture.
Designing navigation that mirrored and reflowed correctly for Arabic and Hebrew surfaced edge cases that strengthened the overall architecture.
RTL markets represent approximately 3% of Pinterest MAU, reinforcing the importance of predictable structure, flexible layout, and resilient interaction patterns across the system.
Before, opening Create brought up a panel on top of the home feed, covering the Pins people were browsing. The Side Drawer opens beside the feed instead: the grid reflows to make room, so Pinners can start a Pin, Board, or Collage while their content stays in view.
The Side Drawer launched in spring 2026.
Side Drawer (Before → After): Create opens beside the feed instead of covering it.
By prioritizing clarity, predictability, and system flexibility, the updated navigation supports both current needs and future capabilities across the platform.
What I’d carry forward:
Design navigation around user intent, not historical structure
Prioritize predictable structure over surface-level organization
Treat internationalization as a first-class system constraint
Build foundations that support future capability without re-architecture