Y CombinatorZynga Founder: Consumer Is Not Investible Right Now - Thats Why You Should Build It
At a glance
WHAT IT’S REALLY ABOUT
Mark Pincus on building consumer in the AI distribution winter
- Mark Pincus frames the internet as moving through three eras—web, social/mobile, and now AI/agents—where new “internet treasures” emerge once enabling infrastructure and cost curves mature.
- They argue investors are currently biased against consumer (often pushing founders toward enterprise) even when consumer metrics are strong, largely because consumer distribution playbooks are broken or unclear right now.
- Pincus introduces his “Proven, Better, New” framework: copy what’s already validated, improve with an indisputable benefit, and isolate the truly novel bet so it can be tested (and killed) quickly.
- The conversation emphasizes founder psychology and culture—killing ego to kill bad ideas, recognizing the rare “fish are running” moment, and practicing “Founder Mode” as presence, detail-orientation, and permission to change course.
- They predict a major consumer AI wave once inference becomes dramatically cheaper (a “business plan of free” for intelligence), unlocking always-on assistants and other agentic products that are currently too expensive for mass-market use.
IDEAS WORTH REMEMBERING
5 ideasConsumer may be unfundable, but the product opportunity can be greatest then.
Pincus argues that when investors sour on consumer, founders can still build the next “internet treasure” because new capabilities (AI/agents) unlock reinventions of seemingly generic services; the funding cycle doesn’t determine user desire.
Napster’s breakthrough was social awareness, not just file-sharing.
Seeing “millions of machines connected” and abundance of copies made the network feel like people, a visceral shift from connecting to corporate databases—an early clue for how new platforms create new user behaviors.
Trust is a first-order product requirement for social products.
Pincus attributes Tribe’s failure to getting trust wrong; Facebook’s early .edu container provided a credible, safer context that enabled identity and sharing—an enduring lesson for AI assistants handling personal data.
Use “Proven, Better, New” to focus innovation where it matters.
Start by legally copying proven mechanics to avoid wasting time, add “better” improvements that users universally agree on (e.g., cheaper/faster), and isolate the “new” hypothesis as the only real bet to test aggressively.
Assume the ‘new’ hook is probably wrong—and design to kill it fast.
Novel features may drive trial (“back of the box”) but often aren’t why users stay; founders should remain passionate about the vision but emotionally detached from any specific implementation so teams can pivot without demoralization.
WORDS WORTH SAVING
5 quotesEven though consumer is arguably not investable right now, the opportunity's never been greater to offer people a new, you know, internet treasure, reinvent some service that we thought was over or just generic, but it's enabled now because of AI and agents.
— Mark Pincus
So you remember that first moment that you loaded Napster and it said, "4.5 million machines are connected to you right now"?
— Mark Pincus
The thing I got wrong with Tribe was trust.
— Mark Pincus
When you have it, you know it. And when you don't, you don't know.
— Mark Pincus
Founder Mode is for every founder.
— Mark Pincus
High quality AI-generated summary created from speaker-labeled transcript.
