Andreessen Horowitz’s Olivia Moore Sees Opportunity in Consumer AI’s Revenue Crisis
Andreessen Horowitz partner Olivia Moore says consumer AI needs ads, cheaper models and new categories as just 2.2% of U.S. households pay for AI today.
Summary
As of October 9, 2026, Andreessen Horowitz partner Olivia Moore’s Monday report on the top 100 consumer AI apps finds ChatGPT far ahead while Suno and ElevenLabs demonstrate staying power. Moore calls OpenAI’s renewed enterprise emphasis an expansion rather than a consumer retreat because consumer launches continue. Almost all AI revenue so far comes from subscriptions and token or API usage, concentrated among enterprises and prosumers, while only 2.2% of U.S. households pay for AI. She favors free, ad-supported access with optional subscriptions instead of relying on more households to pay directly.
AI services still cost more to operate than Facebook or Google Search, but ChatGPT’s Go plan costs $8 monthly, and Moore sees cheaper and open-source models serving consumer tasks that do not require frontier intelligence. Current revenue-driving uses, including coding and technical automation, still need advanced models. She considers most consumer AI prosumer software: Lovable, Replit and Fal support product building; Higgsfield and HeyGen handle product marketing; Manus, Fireflies AI and Granola manage work. Canva took six or seven years to add team or enterprise plans, but Gamma, ElevenLabs and Cursor became majority-enterprise businesses within 18 months. No top 100 entrant serves social, dating, marketplaces, retail, travel, finance or health, categories Moore wants populated within six months, showing true consumer AI remains early.
Positives
- ChatGPT remains the largest consumer AI app by a wide margin, confirming sustained mainstream demand.
- Suno and ElevenLabs show staying power despite ChatGPT’s commanding lead.
- ChatGPT Go’s $8 monthly price signals potential for lower-cost consumer access.
- Gamma, ElevenLabs and Cursor became majority-enterprise businesses within 18 months of starting as consumer products.
- Seven unrepresented categories, including social, retail, travel and health, offer substantial room for new consumer AI products.
- Open-source and cheaper models could reduce costs for consumer tasks that do not require frontier intelligence.
Risks & concerns
- Only 2.2% of U.S. households pay for AI, limiting direct consumer subscription revenue.
- AI services retain higher marginal costs than Facebook or Google Search.
- Almost all AI revenue remains concentrated in subscriptions and usage charges from enterprise and prosumer customers.
- Coding and technical automation still depend on expensive frontier intelligence because those users drive current revenue.
- No top 100 consumer AI app serves social, dating, marketplaces, retail, travel, finance or health.
- OpenAI’s increased enterprise emphasis reflects stronger monetization among business and prosumer users than ordinary consumers.