My intuition is that there are an infinite number of ways to fail, and the number of ways to succeed are much more scarce. So it’s good to study both, but studying success is often more informative. You definitely want to be intimately aware of the common failure modes.
I’d be very surprised if lots of failing companies are doing the _exact_ same as the most successful ones. Maybe cargo-culting all of the things, or actually doing _most_ of the same things but missing one necessary thing. There is an element of luck to it for sure, but the intangibles like business strategy, product market fit iterative discovery, hiring, managing, developing people, are all quite hard to pin down in pithy summaries. The best startups get these right either by meta-level luck of happening to have the right skill set for the problems they ended up facing (vs. rolling a natural 20 on a skill check with the same stats as other founders) or by actually working harder on acquiring the skills that they most need to develop. But doing these (hard, time-consuming) things well is uncommon.
> My intuition is that there are an infinite number of ways to fail, and the number of ways to succeed are much more scarce
This may well be true, but the factors for success are hyper-sensitive to the context of the time with a hefty dash of randomness thrown in. By the time we've observed the factors, the circumstances have changed so they're not entirely repeatable. Sticking to the formula may work, or may lead to wildly different results. But altering the formula is just as unpredictable. Context is critical.
I’d be very surprised if lots of failing companies are doing the _exact_ same as the most successful ones. Maybe cargo-culting all of the things, or actually doing _most_ of the same things but missing one necessary thing. There is an element of luck to it for sure, but the intangibles like business strategy, product market fit iterative discovery, hiring, managing, developing people, are all quite hard to pin down in pithy summaries. The best startups get these right either by meta-level luck of happening to have the right skill set for the problems they ended up facing (vs. rolling a natural 20 on a skill check with the same stats as other founders) or by actually working harder on acquiring the skills that they most need to develop. But doing these (hard, time-consuming) things well is uncommon.