◄ BACK TO QUESTS
QUEST 05◆ SIDE QUEST+1,500 XP
Prompt Mine —
Prompt Mine —
PERIODMarch 2024 – November 2024
ROLELead Product Designer
PLATFORMiOS, Web App.
TEAMCore Team
Lead Product DesignerUX ResearchRetentionGuerrilla Testing

Prompt Mine is an AI prompt marketplace — a sister product to Memo'd — built on the latest AI market discoveries. With only a year in production and 3 months in market, it was already showing promise. But users were dropping off on first visit before discovering any value. I led the research to understand exactly why — and designed targeted interventions to change it.

[ 01 ]

Dropping off before finding value

Users were entering the app and dropping off without ever coming back. Those who stayed avoided the feed entirely and retreated to their own profiles — never exploring what made Prompt Mine interesting. Fast-paced agile sprints and a tight budget made carving out time for proper research feel nearly impossible.

SCREENSHOT 1
[ 02 ]

Guerrilla impression testing

Designed an impression testing protocol built for maximum insight with minimal setup. Prepared lo-fi paper prototypes with real content, crafted targeted first-impression test questions, and ran guerrilla usability testing with 6 participants from the target audience. Synthesized findings into clear problem themes around discoverability and value clarity.

SCREENSHOT 2
DETAIL 2
[ 03 ]

+3.12% retention

3.12% increase in post-sign-up retention following the redesign launch. Identified that users needed descriptive UI, social proof, and creator personality signals to feel confident exploring the feed. Secured cross-functional buy-in to reimagine social interactions — initiating a product shift toward deeper user engagement and community-driven discovery.

SCREENSHOT 3
✦ VISUAL RESULTS5 IMAGES
Visual result 11 / 5
✦ LEARNINGS & FUTURE VISION
→ WHAT I LEARNED

Understanding how users experience a product for the first time has an outsized impact on everything that follows. Simple, strategic UI tweaks can build confidence for both sides — users and business — without requiring a full rebuild. Agile does not mean fast: moving too quickly without empathy creates problems that take far longer to untangle.

→ WHAT I'D DO NEXT

Future directions I'd pursue: a personalized prompt recommendation engine surfacing prompts based on the user's role and adoption history, and a collaborative layer, shared prompt libraries for teams building AI workflows together. The future of prompt management is social, not solitary.