He's building an AI media empire
TL;DR: Ricardo Vice Santos, early Spotify employee with deep personalization experience, built DreamStories - AI-powered personalized children's books - reaching ~$3M/year revenue with 70,000+ characters created on platform. The insight came from reading to his son and noticing lack of representation in children's books. Uses paid acquisition as 'level playing field' across Facebook, Google, TikTok with focus on unit economics and retention rather than viral growth. Studied DTC brands (Dollar Shave Club, Kinko) to understand gifting dynamics - most books purchased as gifts (not by end consumer), requiring different marketing approach (sell to gift-giver's aspirations, not child's preferences). Vision extends beyond books to AI-customized media (movies, TV) where each person gets bespoke content, but starting with books because that's what AI can do well today and what people will buy now. Background from Spotify informed understanding of long-tail content and personalization at scale. Refuses to focus on 'get rich quick' revenue story, emphasizes building toward larger vision of personalized media future.
Key Insights
- Paid acquisition (Facebook, Google, TikTok) is level playing field anyone can access - difficulty is making unit economics work in auction model
- For gifted products, market to gift-giver's aspirations not end user's preferences - different messaging than direct-to-consumer
- Start with narrow AI use case that works today (personalized books) while building toward bigger vision (all media customized)
- Study successful DTC brands in adjacent categories to understand retention and customer behavior patterns
- Former Spotify experience with personalization and long-tail content directly applicable to AI consumer products
Actionable Takeaways
- For AI products: Start with use case that's both technically feasible today AND people will pay for now, even if vision is bigger
- For gift products: Research shows ~80% of razors bought by women for men - market to purchaser's aspirations, not recipient's
- For paid ads: Focus on retention and lifetime value to make auction economics work - everyone can buy ads, not everyone profits
- For consumer AI: Study how people consume existing media (books, music, video) to find personalization opportunities
- For validation: Use personal pain point (lack of representation in kids books) to identify market need