Facebook's Home feed is one of the most used interfaces in the world—billions of people start their day scrolling through it. But by 2022, younger users were reporting that it felt "out of touch" with how they lived their lives. The feed showed content randomly, without considering what users needed in different moments.
A commuter might want practical updates during their morning commute, connection with friends after work, or inspiration during creative blocks. But the feed treated all moments the same—missing opportunities to serve users when they needed it most.
This case study documents my work as the sole UX Research Intern on Facebook's Top of Home team, where I led generative research to understand how the feed could evolve for the future. With limited traditional research methods available for exploring "what if" scenarios, I developed two novel research approaches that would reveal how users wanted to engage with Facebook across different moments and emotional states.
Select design outcomes informed by the generative research findings.
The Problem
"The feed felt generic—it showed the same content at 8am as it did at 8pm, ignoring that users have different needs throughout their day."— Research Finding
Facebook needed to evolve to stay relevant with younger users. By 2022, Facebook faced a critical challenge: its core demographic was aging, while younger users (Gen Z and millennials) were spending more time on TikTok, Instagram, and other platforms.
The feed lacked temporal and emotional context. Traditional feeds show content based on relevance and recency, but they don't consider when content is most useful or how users feel when they open the app. A weather update might be valuable first thing in the morning, but not at 9pm. An inspirational quote might resonate when someone's feeling unmotivated, but feel trivial when they're celebrating.
The research challenge: Product teams wanted to understand future opportunities, but traditional user research excels at testing existing concepts, not generating new ones. How do you research something that doesn't exist yet? How do you understand what users might want before they can articulate it?
My Approach
Novel Method Development
Created "Time Loop" for understanding temporal needs and "Feeling Cues" for emotional contexts—approaches tailored for exploring future scenarios that users couldn't yet articulate.
In-Depth User Research
Conducted 15+ in-depth interviews with diverse users (students, working professionals, parents) to uncover latent needs and desires through experiential simulation.
Experience Prototyping
Created 21 experience prototypes in Figma based on research insights, translating abstract needs into concrete design concepts that product teams could evaluate.
Synthesis & Presentation
Synthesized findings across methods to identify patterns and themes, then presented insights to product teams to influence roadmap and strategic direction.
Key Outcomes
Research insights directly informed product development and created reusable methodological frameworks.
Why Traditional Methods Fell Short
Traditional user research excels at answering questions like "Do users understand this interface?" or "Would they click this button?" But when exploring future opportunities for a product as mature and universal as Facebook Home, these methods hit limitations.
The challenge: Users are anchored to their current experience. If you ask them "What do you want from Facebook Home?", they'll describe incremental improvements to what exists. But Facebook needed to understand entirely new paradigms—content organized by emotional state, feeds that adapt to time of day, features that anticipate needs rather than react to them.
This required methods that could simulate future experiences, not just ask about current ones. I needed to develop research approaches specifically designed for exploring future experiences through experiential simulation rather than direct questioning.
The Two Novel Research Methods
To uncover forward-looking user needs related to context (time) and emotion, I developed and applied the following two methods:
Method 1: Time Loop
Understanding Temporal Needs
Users don't need the same content at 8am as they do at 8pm. Rather than asking "What do you want from Facebook?", I used visual cues to simulate different times of day: morning alarm, commute, midday break, evening wind-down.
For each scenario, participants imagined opening Facebook and described what would be most helpful. This revealed patterns: users wanted practical, action-oriented content in the morning and emotional, connection-focused content in the evening.
Method 2: Feeling Cues
Connecting Content to Emotion
Facebook serves emotional needs beyond information or entertainment. Users come to Facebook when feeling lonely, seeking inspiration, wanting to relax, or needing motivation.
I presented participants with visual cues representing emotional states (connected, informed, inspired, nostalgic, motivated). For each feeling, we explored what content could evoke or support that emotion—revealing emotional content patterns that linked specific content types to specific outcomes.
Why These Methods Worked
Visual Simulation
Instead of asking users to describe future experiences they'd never had, these methods simulated scenarios they could evaluate and react to—producing far richer insights.
Context-Rich Scenarios
By grounding prompts in specific contexts (time of day, emotional state), participants could imagine concrete, contextualized desires rather than abstract preferences.
Pattern Identification
While individual responses varied, consistent patterns emerged across participants—enabling identification of universal needs that could inform product design.
Research Impact
The Time Loop and Feeling Cues methods generated rich insights that directly informed product development. While specific findings remain confidential due to NDA, the research contributed to features that reached over 1 billion users.
Features informed by this research: The Sports scorecard and Spotify connect features—currently deployed across Meta products—emerged directly from patterns identified through these methods. Temporal patterns revealed users wanted sports updates during specific times of day, while emotional context research identified music as content supporting multiple emotional states (relaxation, motivation, celebration).
Beyond shipped features: The 21 prototypes created from this research informed product direction beyond what ultimately launched, shaping Facebook's understanding of contextual and emotional engagement for future development.
Methodological Contributions
- Time Loop method: A reusable framework for understanding temporal patterns in user needs that other researchers could apply to future-focused projects
- Feeling Cues method: A systematic approach to mapping emotional content associations that could guide emotionally-aware interface design
- Visual cue approach: Demonstrated how visual simulation could overcome limitations of direct questioning in generative research
- Hybrid methodology: Showed how combining temporal and emotional lenses could provide comprehensive understanding of user needs
Key Learnings
For future-focused research: Generative research requires patience and systematic methodology. Users can't articulate needs for experiences they've never had—research methods must simulate future scenarios rather than asking about abstract possibilities.
Multiple perspectives matter: Combining temporal (Time Loop) and emotional (Feeling Cues) lenses revealed different aspects of user needs. Neither method alone would have provided complete understanding—together they created a comprehensive picture of how users wanted to engage with Facebook across contexts.
Conclusion
While specific project outcomes remain confidential, the development and application of the "Time Loop" and "Feeling Cues" methods provided valuable frameworks for exploring future-facing user needs in a nuanced way.
These methods allowed us to move beyond current usage patterns and delve into how Facebook Home could proactively integrate into users' lives based on temporal context and emotional state, generating rich qualitative data even when exploring abstract future possibilities.
They represent adaptable techniques for generative research in product development—approaches that other researchers and teams could apply when traditional methods fall short of revealing what users might want from experiences they've never had.
Skills & Methods Demonstrated
Generative Research, Method Development, Semi-Structured Interviews, Insight Synthesis
Conceptual Design, Prototyping, Figma, UX Writing, Co-Design Workshops
Novel Method Creation, Research Strategy, Future-Focused Research, Pattern Recognition
Stakeholder Presentation, Product Strategy, Cross-Functional Collaboration