Google search was the first one to find a business model for pure search (adsense i.e. advertising based on your search terms).
That allowed them to build the best pure search experience they could and was so profitable that they could spend more money building it than anyone else. That's a virtuous cycle.
Their competition at the time was moving away from pure search (because it was unprofitable for them) into yahoo-like portals, because that's where the money was (also ads but non-contextual display ads).
Which is why that's not a good analogy for uber.
In U.S. there's really Uber and Lyft. Every time Uber is discussed, "no lock-in" comes up but why no one even tries to compete with Uber and Lyft in US?
It does seem like people who actually have the money to potentially mount an attack on Uber or Lyft came to the conclusion that it would be a foolish waste of money.
Uber and Lyft have a clear business model: they are better and cheaper than taxis and have global presence (while taxi companies are historically local and don't have the capital to go global).
Customers in this market do care about price and better service, so it's rather obvious that Uber and Lyft will win over taxis (arguably they already won given that they're 100x bigger than any taxi company).
Both Uber and Lyft are still in the phase of aggressive expansion and growth, they still enter new markets, subsidize the service to increase usage density (which makes the user experience better), they experiment with things like Uber Pool, being alternative to busses etc.
None of the above is cheap so it's understandable that they loose money today.
But at some point the growth will stop, their costs will go down and one of them (or even both) will be very profitable companies.
The only thread for Uber and Lyft are self-driving cars (i.e. Google, Tesla, GM) because they'll drive costs down by 5x. Whoever wins that race wins the taxi business, which is why Uber spent $680 million on Otto and GM $1billion on Cruise and Google was apparently paying $120 million to a single individual in charge of self-driving cars.
