When Uber started, we all thought it was a winner-takes-all market. The theory went something like this:
1. The more drivers are on a given platform, the more value users will derive from the platform in the form of cheaper rides, shorter waits, etc. This will result in more riders.
2. The more riders are on a platform, the more value drivers will derive in the form of higher utilization and more overall income.
3. Therefore, all drivers and riders will naturally gravitate to the winning platform. It's a winner takes all market. QED.
Honestly, at the time (and I'm old enough to remember that time), the theory made sense. But it was wrong.
The reason the theory was wrong is that there is actually zero cost for drivers in having multiple apps open at the same time. The switching cost is zero (actually, it's negative--we'll get to that in a second). And so point (3) doesn't happen.
The reason switching cost is negative is that it actually costs money to start driving (and to convince people to become drivers). People quit jobs, and get/fix cars, and invest in a bluetooth headset, whatever. So Uber was actually subsidizing every other platform's drivers.
So instead of an extremely defensible market you end up with a race to the bottom, with no barriers to competition.
While in retrospect the mistake and dynamics are obvious, at the time it seemed like a reasonable, risky, ambitious bet.
At some point these dynamics became obvious to Uber, which is when they started investing in self driving cars (never gonna work), uber eats, etc.