It’s easy to mistake excitement about a product for evidence that people will use it. Looking at Quibi, the Humane AI Pin, and Juicero, I’d want each team to test one assumption before committing to a big launch. These are the tests and thresholds I would propose.
Quibi
Quibi’s biggest gamble was that people would pay for short videos when they already had so many free options.
We believe… viewers will pay $5 a month for premium short episodes on their phones.
To verify, we will… give 50 target viewers a free week of a small, licensed collection on a basic mobile website, then ask them to pay.
And measure… how many buy the next month.
We are right if… at least 15 pay.
The cheapest test was selling a small collection of shows before building an entire streaming service.
Humane AI Pin
I’d question whether using the pin would actually feel easier than pulling out a phone.
We believe… people will choose a wearable voice assistant for everyday tasks.
To verify, we will… lend 20 people rough prototypes using the actual AI backend for a week, leaving their phones available.
And measure… the share of eligible tasks they voluntarily finish on the wearable without switching.
We are right if… that share exceeds 60% over the final three days.
A basic prototype could expose frustrating delays before investing in custom hardware.
Juicero
The biggest question was whether the machine added enough value to justify its price.
We believe… juice buyers will pay $700 for automated pressing.
To verify, we will… let 20 buyers compare hand-squeezing with a basic press, disclose packet costs, and offer refundable $50 reservations.
And measure… deposits placed.
We are right if… at least five reserve.
The cheapest test: squeeze a packet by hand, then see who still wants the machine.
