The rise of Google is a true Unicorn tale. The story of hardcore CS nerds rebuilding a product that was already ubiquitous with a crowded market and re-inventing it down to the algorithmic level, defying every norm on the way there. But most innovative products have some element of blinding obviousness about them. Uber? Maybe we could use the worldwide communications network that every has a constantly connected terminal to in their pockets at all times to dispatch cabs instead of, you know, telephones or just having cabs circulate in high-demand areas. Dropbox? Network file storage that leverages the speed of the modern internet and modern storage technologies to remove the hassle of dealing with networked files. Square? What if people could just run an app on their portable computing devices which are already vastly more powerful and constantly internet connected instead of using crusty custom-built PoS hardware and systems? Stripe? Imagine if banking embraced the technology of the late '90s. And so on.
90+% of the "Unicorns" are just the sensible application of technology to already well-known problems, and just doing it well and monetizing properly. That shouldn't be so exceptional, but somehow it is. I think that's more an indictment of the silicon valley way of doing things than anything else. And the thing is, the rate of return to VCs when a Unicorn hits is so big that it covers the losses from all the bozos.