I would like to see this from another angle though - not why they grow so big - but what it takes to kill such a company (i.e. cause it to fail). One of the bi

I would like to see this from another angle though - not why they grow so big - but what it takes to kill such a company (i.e. cause it to fail). One of the big advantages the giants have is that while no doubt they hire very capable engineers, the number of engineers who need to leave to make a dent on their performance is a really large number, and gives them a really long notice to correct things. In other words, at any given point of time they have many more 'dispensable' employees. A second benefit of being a giant is that you can forego revenue streams which would be the size of an entire business at a competitor to gain some advantages in the market - for e.g. Google made Android free. I don't think they could have pulled off such a move when they were younger and poorer. This diversity of income streams gives them plenty of time to leave behind lagging products/services and concentrate on the cash cows - that is, they can apply the Pareto principle at a much much larger scale. Contrast that with the potential disruptors who simultaneously compete with the large companies for employees, could sometimes get killed if just a few key hires leave, rarely have diversity of huge income streams to allow them to change course, and then have to bear the burden of regulation without the cushion of the cash pile. It lends a more stable environment for the large companies to move in the direction where the market expects them to go without having the tremors which the small companies face when they try to do similar things. During those tremors - engineers at the smaller companies - don't usually like to stick around and see what happens. This has been compounded of late because of the huge salary difference between working at a giant versus working at a smaller company.