Bumble BFF
Moving online conversations to real-world friendships
- Role
- Product Designer
- Timeline
- 5 weeks
- Skills
- UX Design, UI Design, Design systems, Cross-platform Design, Vibe Coding
- Team
- Individual Project
Project Overview
Problem
Bumble has over 50 million active users, but only a small fraction are paying (*Data provided from the course).
Goal
Increase conversion among non-paying users — 90% of whom are under 35.
Research
Original Hypothesis
The original brief suggested users would convert to paid subscribers if they were notified of matches within 0.5km via “timely notifications”.
The data challenged the original hypothesis
The initial concept centered on proximity notifications, based on the assumption that users would be more likely to connect when they discovered a match nearby. However, analysis of Bumble’s engagement data revealed a different story: proximity alerts were not driving meaningful engagement.
Rather than optimizing the notification experience, I stepped back to understand the underlying problem. If users weren’t acting on nearby matches, what was actually preventing them from meeting in person?
Reframing the opportunity
To answer this question, I conducted guerrilla interviews with app users and analyzed competing products in the space. Through these conversations, a consistent pattern emerged: users weren’t struggling to find matches—they were struggling to move conversations into real-world interactions.
Three themes repeatedly surfaced:
Too much chatting, not enough meeting
Users described endless conversations that rarely translated into plans. While matching was easy, taking the next step felt difficult and often stalled.
Planning creates friction
Coordinating schedules, deciding where to meet, and determining who should initiate plans introduced enough effort that many promising connections never progressed.
Most apps optimize for matching, not meeting
Competitive analysis revealed that most platforms focused heavily on discovery and messaging, while offering little support for helping users connect in person.
A new design direction
These findings led me to redefine the opportunity. Instead of helping users discover nearby matches, I shifted the focus toward helping users transition from digital conversations to real-world interactions.
This insight became the foundation for exploring structured, in-person meetups as a way to reduce planning friction and create more opportunities for meaningful connections.
Opportunity Statement
“How might we enable Bumble BFF users to connect with others in real life?”
User Scenarios
- 01The user finds & attends a group event on Bumble BFF and later makes a new friend connection after the event (happy path).
- 02The user creates an event to host (happy path).
- 03The user finds & attends a group event on Bumble BFF and later blocks or reports another user that attended the event (unhappy path).
Final Design
Retrospect
Feedback
“Great work on this. You built a really well thought out UX flow, and your reflection (especially your discussion of vibe coding) shows strong self-awareness and learning.”
— Professor Neumann
Lessons Learned
- AI-assisted (“vibe coding”) tools significantly accelerated the design process by quickly generating screens, flows, and functional prototypes.
- Rapid prototyping allowed me to explore and iterate on ideas much faster than traditional workflows.
- Speed sometimes came at the cost of intentionality, as I found myself reacting to AI-generated outputs instead of driving design decisions.
- The experience reinforced the importance of maintaining a clear design vision when collaborating with AI.
- I learned that AI is a powerful accelerator, but designers must remain critical, intentional, and actively guide the final outcome.
With more time I'd like to
- Explore workflows that better balance AI-generated outputs with intentional design decision-making.
- Experiment with establishing stronger design principles, user flows, and visual direction before introducing AI.
- Use AI to accelerate execution rather than drive key design decisions.
- Refine AI-generated designs to better reflect my own design judgment and craft.
- Investigate how AI-assisted prototyping impacts creativity, exploration, and overall design quality over time.