Quibi, Humane AI Pin, and Juicero collectively raised over $2 billion and still failed. What’s interesting to me is that all three had pretty impressive technology and funding, but were ultimately betting on assumptions about what people actually wanted. I feel like a few simple tests could’ve saved them a lot of money.
1. Quibi
- We believe: People will pay $5-8 per month to watch short, premium shows exclusively on their phones.
- To verify, we will: Release a few trailers on social media and link them to a fake subscription signup page.
- And measure: How many people actually attempt to subscribe.
- We are right if: At least 3% of 10,000 viewers attempt to sign up.
2. Humane AI Pin
- We believe: People would rather use a $699 ai wearable than pull out their phones for everyday tasks.
- To verify, we will: Give 20 people a basic wearable connected to a human operator for a week, simulating the AI experience without building the full product.
- And measure: How often they choose the wearable over their phones.
- We are right if: At least 60% of tasks happen through the wearable by day 7, and half the participants would preorder.
3. Juicero
- We believe: People need a $400+ machine to get significantly more value from our juice packs.
- To verify, we will: Give 50 people juice packs and have them try squeezing the juice by hand versus using the machine.
- And measure: Juice yield, taste, and whether people think the difference justifies the price.
- We are right if: The machine produces at least 30% more juice and users are willing to pay $400 for it.
My takeaways: The biggest lesson here is that building something impressive doesn’t really mean you’re solving a problem people care enough about. It’s a little crazy that Juicero’s biggest assumption was disproven by someone literally squeezing a bag.
As someone who wants to be building products, this is a good reminder to test the simplest version of an idea before getting too attached to the solution.
