OkCupid meets Grouper. Friends, Networking, or Dates.

People tell the app who they would like to meet in a standard survey and link their social media. Additionally, OkCupid style Quizzes and tests can be created and shared by anyone. The AI reads all content and sets up small group dinners or activities to introduce like-minded people. The AI books the restaurant or other venue, does the scheduling, gives everyone a personalized memo of who else is attending and any common connections to you, appoints the most gregarious attendee to be the “host” who moderates (VIP points), and suggests the first three conversation topics. The AI follows up and collects info on who each participant enjoyed meeting.

• Downside: OkCupid got beat by tinder. Mass market humans are far more motivated by hot pictures and lazily swipe, even if it’s bad for them.
  • Counterpoint: Could be valuable to a more niche but more hardcore/lucrative audience likely more serious about a meaningful relationship - dating or business. That was OkCupid. You knew they were serious. But they stopped innovating when the founders sold.
• I think this is the best bet for a $20m-$40m acquisition in 12-24 months in a business that doesn’t require more than $2m funding. It’s easy to do on a small scale. Focus on just NYC, LA, and SF. If it has any traction in those cities, Match.com would buy it early just to say they are keeping up with the latest and greatest AI cool sauce.
• Maybe it could be bigger company if the AI is actually really good - to the point where I’d actually want to use it each week just for my professional career because it is worth its weight in gold in making valuable connections. If it’s a novelty (people use a few times than churn) then I sell quickly to Match. If I get monthly re-engagement, grow.
• Grouper Social Club was the most fun and natural way to do app dates in the 2011-14 era. Maybe AI can scale it economically with less Ops burden, while also making superior matches so the novelty doesn’t fade as fast like it did with Grouper. I did Grouper 4 times before it seemed like the novelty faded. The matches were not great and it was hard to organize to schedule with 2 other friends.
• My previous experience in social apps was that the people who are most driven by the marketing message ‘We’ll help you find friends!’ are the people you LEAST want to be friends with. This has to be branded to appeal to worthy people pursing something that is very valuable to them.
  • Find soulmate
  • Debate about a particular topic (a salon)
  • Pay it forward (everyone is a mentor and mentee in some way)
• Other lessons from Grouper: small revenue stream from restaurants and bars who want the business. Incredibly good gamification and rewards - status is key for a social club. Only valuable people get invited to the really cool shit.
• Other business model of the matches are really good: bounties. Contracts for real relationships or marriage. Commissions on business deals. You bribe the AI and the AI will prioritize you. Feel free to bribe the AI with $100k. (Feels less sleazy than hiring a matchmaker)
• To test it out, it would only take 2-3 months to code a very good MVP (marketing and branding heavy) and get a few hundred people to do this in SF. Worst case, I will meet some cool people.
  • Use Justin to do some world-class branding. Branding is key. Going for Epic.
  • Use Dinki & Gadi to be the host and do the “AI scheduling”.
  • Use Omar to code a survey and a quiz system to collect data for matchmaker
  • Bribe Makena, Jiwon, Cindy, and a few other friends to create quizzes
  • Spend a week to code a basic AI matchmaker myself cause it’ll be fun - makes matches and produces the personalized memo’s of who accepts the dinner invites
  • To launch, post on Bookface. Go for a Raya-like exclusive word of mouth invite system. Get the cool kids to sign up first. The people I would like to meet.