Reliance Intelligence · 2026

~20% → ~28% D7 return · Experimentation & Retention

Problem Statement

• First-time AI users need clear, repeatable value before an assistant becomes part of their day.
• News and Self-care had to work across different economic and social contexts, not assume the usage patterns of existing power users.
• The team needed to learn which moments could build a daily habit during the internal beta.

Customer Persona(s)

• First-time AI users across Indian cities
• News users who want a concise, personalized daily briefing
• Self-care users who want a lightweight space for reflection and journaling

Impact

• Lifted the News D7 return rate from ~20% to ~28% by interlinking the daily, local and breaking news flows (A/B test on ~1K beta users)
• Users who personalized their breaking-news alerts returned at ~1.5× the D7 rate of those who didn't

Solution

• Owned both skills end to end: discovery, strategy, roadmap, design/UX, execution, GTM and iteration
• Ran field interviews with first-time AI users to understand how they perceive technology and where AI creates value
• Reframed the Self-care conversation model and defined the primary personas for both skills
• Built morning briefings for News and a daily diary for Self-care to create repeat usage
• A/B tested siloed flows (users find their own way) against conversational handoffs between flows, measured on D7 return rate

User Journey: Before -> After