Hi Ibrahim! Thanks for breaking down the components of creating a strategy, as well as the importance of focusing on the "when". I'm curious to learn more about how to use hypothesis testing to determine the right sequence of the bets. For example, if a hypothesis is invalidated now, should it be revived down the road?
- I'd make sure the hypothesis is tight vs multi-dimensional
- I'd test multiple light bets and gather signals, then go deep
- if a hypothesis is invalidated, I'd be clear on why
- it could be the hypothesis OR the way it was tested OR the customers tested with OR timing / messaging / quality
- and yes, you can revise a hypothesis anytime for the following reasons
- you invalidated it too early / incorrectly
- the market is better receptive now
- you have a better bet to test it out
I tend not to think of it was valid / invalid hypotheses, but rather more in terms of "strategic relevance" / "testing ability" - focus your product delivery bandwidth on things you think are high-leverage AND where you have a real handle
Wow, that’s very insightful and actionable! I’ll think about mitigating multi-dimensional hypothesis testing through multiple light bets that one can have a good handle on. Thanks!
Hi Ibrahim! Thanks for breaking down the components of creating a strategy, as well as the importance of focusing on the "when". I'm curious to learn more about how to use hypothesis testing to determine the right sequence of the bets. For example, if a hypothesis is invalidated now, should it be revived down the road?
a few thoughts:
- I'd make sure the hypothesis is tight vs multi-dimensional
- I'd test multiple light bets and gather signals, then go deep
- if a hypothesis is invalidated, I'd be clear on why
- it could be the hypothesis OR the way it was tested OR the customers tested with OR timing / messaging / quality
- and yes, you can revise a hypothesis anytime for the following reasons
- you invalidated it too early / incorrectly
- the market is better receptive now
- you have a better bet to test it out
I tend not to think of it was valid / invalid hypotheses, but rather more in terms of "strategic relevance" / "testing ability" - focus your product delivery bandwidth on things you think are high-leverage AND where you have a real handle
Wow, that’s very insightful and actionable! I’ll think about mitigating multi-dimensional hypothesis testing through multiple light bets that one can have a good handle on. Thanks!