[dankohn1]

Uber is a classic two-sided network [0], like eBay or Craigslist. It is extremely difficult to displace a first-mover advantage. The hundreds of thousands of drivers are set up with Uber and will stick with it unless something new is meaningfully better (and Lyft is currently worse). Same with the users.

I think you can be legitimately critical of Uber's poor business ethics [1], but the sustainability of their current business model looks excellent, at least for each city in which they are established as the first mover.

[0] http://en.wikipedia.org/wiki/Two-sided_market [1] http://en.wikipedia.org/wiki/Uber_(company)#Sabotage_against...

[Retric]

The problem is you end up with network effects on the drivers side per city. So 2020's Ultra startup can just displace Uber in say NY and ignore every other city. After winning that they can fight for DC etc.

Also, drivers are a commodity market which is much easier to displace. Amazon marketplace vs eBay's auctions.

[baddox]

The product is a bit more commoditized than with eBay or Craigslist though, right? Above some reasonable level of service, a ride is a ride. Assuming that the friction for switching services for the two sides is relatively low, it seems like there's a clear path to a new competitor starting slow and building both sides of the market at the same rate.

But for eBay, the products people want vary vastly from each other. I want to buy something fairly niche, so to save time it helps if there's one massive site where all of the people on the other side of the market are all located.

[humanrebar]

There's some variety there. Pet rides. Rides for those with mobility problems. Options for extra trunk space. Options for large groups. Free WiFi. Drivers that speak specific languages. Pickup within five minutes or the ride is free.

And there is always differentiation based on customer service.

I guess I'm saying that there are plenty of niches to fill.

[prostoalex]

You just carve out a niche.

* Blacklane - cheaper than Uber Black, but the rides have to be pre-arranged (essentially a car service with UI)

* Wingz - cheaper than UberX, but only to/from airports

* Sidecar - cheaper than UberX, but with implied wait of up to an hour

* Leap buses - cheaper than UberX, but with set routes

If drivers are utilized at 100%, they will have very little reason to look around. But most of the services don't utilize (and hence don't pay) drivers for 100% of their time. Hence a myriad of phones with a bunch of identical apps on drivers' dashboards, and frantic search for 100% utilization (food delivery? shopping delivery? help with moving?) among the dispatcher services.