Three things I believed early that did not hold. All three are common beliefs, which is partly why I held them.
Getting the product right is necessary and it is not the constraint.
Distribution is. A good product that nobody can find loses to an adequate one already inside the buyer's workflow, and that is not a market failure — it is what switching costs mean.
It does, at a rate far below what the work deserves and far below what a competitor with a mediocre product and a real distribution effort achieves.
This is uncomfortable for anyone who prefers building to selling, which is most people who build things.
The best product does not win. The best product that people can find, evaluate, and switch to without pain wins.
The belief that documenting how things are done is bureaucracy that slows a small team down.
What actually happens is that undocumented processes become dependent on the people who hold them, and the cost arrives all at once, later, when one of those people leaves.
Building distribution alongside the product rather than after it. Writing things down earlier than feels necessary. And treating "this obviously good thing will be recognised" as a hypothesis rather than an assumption.
Knowing when a thing is not working versus not working yet. See deciding when to stop. I do not have a good answer and I am suspicious of people who claim one.
Because the retrospective version of any company's history is tidier than the actual one, and the tidy version is not useful to anyone trying to do it.
Read next
Deciding when to stop →