One reason some startups exist: how do you allocate engineers to problems where incumbents either can’t achieve organizational conviction to allocate engineers or couldn’t attract sufficiently productive engineers to matter.
To which a partial answer is you silo them in a separate company with a culture not inimical to engineering, you compensate them with equity, and you eat a lot of free cultural lunches. (e.g. Revealed preferences mean some people value dress codes at literally billions.)
Repeatable way to find startup opportunities: look for large organizations with important problems that really need to hire 10 people who know what a for loop is and implement their recommendations but have a deeply felt reason why they can’t do that.
Five years from now they’ve got three options: rethink how they make decisions, buy the startup and almost inevitably kill it for the same reason they couldn’t do it internally, or compete with the startup. (Not a given they’d lose the competition but risk-adjusted ROI is great.)
