Launching a new product or feature in an existing company is almost the same as launching a new startup.
Product-Market Fit for new things is hard. Don’t make it harder.
When launching anything new that is big, observe new customer cohorts more vs. your existing customers.
There are diminishing returns when we think about how much data we can gather without building something to test it.
Don’t deploy big bets without properly thinking about tracking and acceptance of failure culturally and as part of the company’s strategy.
New bets should be limited to as little changes for the customer as possible to isolate improvements in data
Predictability for big bets is inherently limited but it does not absolve you from making a proper business case.
A low success rate doesn’t mean zero success rate. Successes can still pay for other failures, which needs to factor into ROI calculations.
Reward learning, not just success. A culture that accepts failure supports innovation and helps teams maintain motivation in the face of failed bets.
Designating time for big bets explicitly, and being upfront about the potential risks and rewards, creates a more open culture and supports innovation.