Eduardo Martinez
Bite-sized case studies designed as quick snapshots to start bigger conversations.
I'm a Staff Product Designer with 16+ years of experience. I've worked at the intersection of scale and genuine complexity: launching first-of-their-kind products and creating design patterns that became industry standards.
I’m happiest when I’m building something that makes life a little easier for the person on the other side of the screen. The best outcomes I've seen come from respecting intent, not chasing clicks. When you meet people where they are, the metrics follow naturally.
Glassdoor
Practice Fusion
Earlier Roles
Originate
Meta
Facebook App
Risk Foundations
Past Work
These are quick, digestible overviews of three key projects that showcase different aspects of my design approach, from tactical navigation improvements to strategic IA transformation to product redesign. Each "bite" highlights the problem, solution, and results in just a few slides.
I'm happy to dive deeper into any of these projects or walk through the complete case studies during our conversation.
Meta - Facebook App
Facebook App's 3 billion users were getting lost and frustrated with navigation, leading to force-quit behaviors. I developed a holistic "Enter, Immerse, Escape" framework that restored user confidence while generating $1.9M in additional daily revenue and earning direct C-suite recognition.
My Role
Product Design Lead
Collaborators
PM Lead, Eng Lead, User Research, Data Science
Timeline
H2 2022 - H1 2023
The Problem
Facebook's navigation had become a maze of competing elements that left users struggling to find their way after clicking notifications. These minor pain points compounded across billions of daily interactions affecting satisfaction and engagement and manifesting in behaviors like force-quitting.
I developed an "Enter, Immerse, Escape" framework treating navigation as a complete user journey, supported by 20+ prototypes and research to understand how small frictions multiplied across billions of interactions.
Land users in the right place with clear context from the first moment.
Remove visual noise so users can focus on content without distraction.
Ensure users can always find their way back, no matter how deep they explore.
Key Solutions
I created a holistic system that guided users through their complete journey, from landing in the right place, to focusing on content without distractions, to confidently finding their way back home.
Built personalized splash screens, contextual deep linking, and intelligent state restoration so users always knew where they were and how they got there.
Created dynamic scrollaway navigation and floating tab bars that disappeared when users wanted to focus, putting content first instead of competing interface elements.
Designed contextual navigation with thumb-friendly back/home buttons and omnidirectional gestures so users could confidently explore knowing they could always return to familiar ground.
Results
We made a strategic bet that respecting user intent outperforms metric manipulation tactics and it paid off. We demonstrated that user-centered design can drive business wins while building user trust. The work earned leadership recognition and received visibility across Meta's product families, inspiring teams across Instagram, WhatsApp and Facebook Lite to adopt key patterns and insights.
The most immediate impact came directly from users.
Across user research, feedback was clear: the experience felt cleaner, more modern, and easier to use. Users called it "less confusing" and "more intuitive," praising how "everything just makes more sense now." The design restored confidence for both profile management and content exploration.
$1.9M
Additional daily revenue
+4.8%
Content views per visit
+1.5%
YA Sessions
+5%
iOS revenue
"We have a good design where we decided to get out of the way when people are trying to view content, and guess what, people like it... all of the engagement metrics are up and to the right."
— Head of Facebook
Meta - Facebook App
Enter/Immerse/Escape work revealed we could only push navigation clarity so far with an IA that presented main use cases in a decentralized way that reflects our org chart more so than how people think about them. Leveraging past learnings and research insights, I developed the Core Use Cases concept: an IA framework that intentionally redesigns the navigation around use case priorities, further decluttering the navigation experience and focusing attention on what people find important.
My Role
Product Design Lead
Collaborators
PM Lead, Eng Lead, User Research
Timeline
H2 2023 - H1 2024
The Problem & Opportunity
Over the years, Facebook grew from a vibrant town square to a dense metropolis with confusing roadways and attention-grabbing billboards. People see twice the number of buttons than other social apps, with no coherent hierarchy making everything seem equally valuable, creating barriers between people and social discovery.
Research revealed three core navigation problems limiting user engagement.
Less than 40% of Video Watch Time comes from the tab. Dynamic tab bar breaks muscle memory, leading to app abandonment.
50% of friend requests go unseen in 24 hours, and 28% of pending requests have only 1 impression monthly.
Creation receives inconsistent treatment across the app. People don't notice existing creation affordances, as evidenced by small font size increases lifting sharing.
We learned where people go, what they value, and where they get stuck. High-value spaces like Video are underserved while other spaces have too direct access. We can now be more opinionated about navigation structure and build a smart, predictable city where people move efficiently to value.
Framework & Solution
This concept prioritizes core use cases through a Tab Bar serving as a central hub. The IA is designed with simplicity and intuition in mind and is structured around fundamental uses to help users form a clear and lasting understanding. This universal categorization also allows for addition of new services without disrupting the user's learned mental model, replacing the existing dynamic Tab Bar that constantly shifted and confused users.
By clearly organizing the app around core use cases we can reduce complexity and streamline navigation to deliver a user experience where content and services are intuitively organized.
Content Consumption - Feed, Video, Stories
Connections - Friends, Dating, Following
Creation - Unified creation hub
Communication - Notifications, Messages, Community Chats
Commerce - Marketplace, Shop
2 Utility Entry Points (Top Nav)
Search & Discover - Universal search and feature discovery
My Stuff - Profile, Saved, Settings via side panel
Results & Strategic Impact
The project successfully secured crucial leadership buy-in and additional resources to continue this extensive strategic work. Implementation proceeds with careful sequencing as we view this IA change as an evolution rather than revolution. Early implemented pieces have already demonstrated impactful wins through more intentional navigation design.
The Core Use Cases concept convinced executive leadership to invest in long-term navigation transformation, with strategies now embedded in Facebook's roadmap for the foreseeable future. This represents a fundamental shift from tactical navigation fixes to systematic, user-centered IA evolution.
$1.06B
Revenue Impact
+0.76%
Billing Revenue
Early IA implementations delivered measurable impact through through the following changes: streamlining tab bar behavior for high-value surfaces like Feed, testing side panel concepts to reduce More Menu friction and optimizing entry point hierarchy for Video and Friending. This shows how focusing on navigation people problems (content discovery barriers, clutter, and antiquated structures) drives user and business results. The work spans near-term IA optimizations to longer-term transformations while the Core Use Cases vision shapes Facebook's navigation roadmap.
Meta - Risk Foundations
The team was running global regulatory work across spreadsheets and ad hoc chats. No single source of truth, no visibility into what was coming until work landed on your desk with a short deadline. There was no AI in the process at all. I built the first AI-augmented version from scratch, solving problems with no established playbook: how to surface what the model generated, when to keep humans in the loop, when to hand the work to the agent entirely.
My Role
Sr Staff Product Designer
Collaborators
PM, Front End Eng, Back End, QA
Timeline
2025-2026
The Problem & Opportunity
Risk Foundations is responsible for how Meta identifies, interprets and responds to global privacy regulations, keeping Meta compliant and operational worldwide. Meta conducts tens of thousands of risk and compliance reviews each year. The org was stood up overnight to handle it, and the processes never scaled. Work fragmented across ad hoc chats and competing spreadsheets. Downstream teams had no visibility into incoming work until it arrived on their desk with a regulatory deadline already close.
Research revealed three systemic problems making the process unsustainable.
Experts spent hours gathering information and filling out standardized forms just to get a review started, before any actual compliance thinking could happen. At the scale Meta operates, that wasn't a workflow problem. It was a structural one.
Work lived in ad hoc chats and competing spreadsheets. Downstream teams had no notice until work landed with a short deadline.
Internal tooling existed but adoption was near zero. Everyone kept their own copy of everything.
AI was built for exactly this kind of problem: document processing, interpretation, triage, pattern matching, summarization. But redesigning a regulated compliance process through an AI lens meant answering questions the field hadn't encountered before. How to surface what the model generated, when to keep humans accountable, and how to build toward a world where manual processes are the fallback, not the default.
The Solution
I designed an AI-first compliance experience from scratch. AI handles triage, interpretation and the first draft of every decision, running continuously in the background. The core design question was when to keep users in the conversation and when to step out to a structured artifact. The answer became a framework: the interface signals decision weight. Chat for routine work. A structured page for anything consequential. Issues get caught while decisions are still being shaped, not after they ship.
Home Pattern: AI-curated triage feed that pre-fills context and surfaces relevant requirements upfront, so experts arrive at a review already oriented, not starting from zero. Persistent chat for initiating new work.
Object Widget: Embeds structured artifacts directly in the conversation thread. Sized by decision cost and information cost, from a quick confirmation to a full review.
Canvas Mode: Full-screen workspace triggered when task complexity exceeds what inline widgets can handle. Left panel for reasoning, right panel for structured workspace and version history.
Results
The work landed. The direction it pointed to is now Meta's public commitment. The agentic risk review experience is in pilot, with the agent handling the process end to end and humans brought in only for clarification or final attestation. Expert attention now goes where it belongs: the novel cases, the high-stakes calls, the things that actually require human judgment. Technology handles the scale. People drive its direction. Less a process to be managed, more a system that runs itself.
The patterns built for this work didn't stay inside the product. I ran a sprint, pulled the patterns together and built a canonical reference site with machine-readable files teams could plug directly into their prototyping workflows. Adoption was organic. The patterns solved a problem teams were already experiencing.
~50%
of teams contributing back
All
teams aligned on shared patterns
The shared source of truth is still growing. Teams download the pattern files, apply them directly in their prototyping setup and propose new ones back into the system. The contribution model means no single point of ownership. The system improves itself as more teams build with it.
Glassdoor
Glassdoor users loved our reviews but ignored our job search entirely. As sole designer on the jobs team, I led a complete redesign that consolidated all decision-making information into a seamless dual-pane experience, resulting in a 63% increase in job applications and inspiring industry-wide adoption of similar patterns.
My Role
Senior Product Designer
Collaborators
PM, Front End Eng, Back End, QA
Timeline
2015-2016
The Problem & Insights
Users came to Glassdoor for company insights but avoided our job search page. Organizational neglect left a fragmented and overwhelming job search experience that saw minimal usage despite engaged users elsewhere on the platform.
As sole designer without UX researchers on the team, I personally organized and conducted comprehensive research: design audits, user interviews, usability testing and user flow analysis to understand why our users weren't using our job search.
Users found the experience overwhelming and confusing due to scattered information and poor visual hierarchy. The tabular layout was difficult to scan and process, while users consistently wanted more details about roles but had to navigate through 4+ separate pages just to gather basic decision-making information.
The fragmented multi-page flow created massive friction. Users would abandon their search rather than hunt across multiple pages for the information needed to evaluate opportunities.
Strategy & Solution
I unified the fragmented 4+ page flow into a single search experience where all decision-making information became accessible. A rapid MVP delivered 20% improvement, validating unlocking the path to the full dual-pane redesign.
Compact List View: Mobile-inspired layout displaying more results per page with dramatically improved visual hierarchy.
Integrated Information Panel: Comprehensive right pane with full job descriptions, company information, and reviews; eliminating the 4+ page navigation flow.
Seamless Application Flow: Users could research, evaluate and apply to jobs without leaving search results, removing all friction from job-seeking.
Results
Through comprehensive design research and iterative testing, I transformed a broken and neglected experience into an industry-leading solution. The 63% increase in job applications proved that understanding user pain points and addressing core friction can unlock massive engagement gains, while the dual-pane pattern's widespread adoption across Indeed, LinkedIn, and Monster validated the strength of the user-centered approach.
Users embraced the intuitive dual-pane design, appreciating immediate access to comprehensive job information without navigation friction. The streamlined experience allowed users to finally see the value in Glassdoor's job search capabilities, directly driving our significant engagement gains.
+20%
Job clicks - MVP
+63%
Job clicks - Dual Pane
The redesign not only generated impressive metric gains but also sparked industry-wide emulation across job search platforms. Indeed, LinkedIn, and Monster all implemented similar dual-pane approaches within years of Glassdoor's launch, validating the strength and influence of the solution in establishing new standards for job search experiences.
What you’ve seen here is just a glimpse; bite sized snapshots of a few projects I’ve had the privilege to lead. Each case study runs deeper: from user research insights and messy problem spaces to the strategic decisions and hard won outcomes behind every screen.
If anything sparked your interest, I’d love to walk you through the full story.