If you weight companies by their market value, rather than a binary successful vs. unsuccessful[1] outcome, you'll come to a different conclusion. This is just a rationalization to avoid working hard.

[1] What's a "successful" company, by the standards of this and similar article? I don't want to inform my decisions with a study that considers Dropbox, on the one hand, and any sustainable small business, on the other, to be equally successful. There'll always be more of the latter.