Quibi’s riskiest assumption was that consumers would pay once their free trial ended, evidenced by the fact that Quibi drew 1M subscribers in its first week, but 90% of those left after their trials. Test card: we believe viewers who watch our shows free for 2 weeks will pay $4.99/ mo for a standalone app, even with Tiktok available for free. To verify, we’ll run a concierge pilot that gives 300 commuters 2 weeks of access to 3 inexpensive shorts, and measure the proportion who pay after it ends. We are right if at least 90 of 300 pay.
Humane’s riskiest assumption was that people would keep a screenless assistant after experiencing its limitations, evidenced by the fact that Humane sold 10% of their unit goal and they knew of serious problems before launch. Test card: we believe people will keep a $699 pin instead of reaching for their phones, despite its latency and errors. To verify, we’ll run a Wizard-of-Oz test where 30 users wear a clip-on microphone linked to an existing voice assistant for 2 weeks, with all the prototype’s issues, and measure how many keep the device when offered it at full price. We are right if at least 12 of 30 keep it.
Juicero’s riskiest assumption was that customers needed the press to extract juice from its packs, evidenced by the fact that Bloomberg reporters squeezed the packs by hand and got basically the same amount of juice. The finding reduced Juicero to a delivery business for packs. Test card: we believe our press extracts meaningfully more juice than human hands can. To verify, we’ll squeeze 15 packs by hand and run 15 through the press, and measure the juice yield. We are right if the press produces at least 30% more juice.
