5 ms·

Sometimes it's different, sometimes it's not. Tech allowed a company like Amazon to scale. Tech allows a company like Uber to allocate rides vastly more efficie

by rconti 8y ago

Sometimes it's different, sometimes it's not. Tech allowed a company like Amazon to scale. Tech allows a company like Uber to allocate rides vastly more efficiently, tech allows a company like AirBnB to enable property owners to extract more value from their property. But tech doesn't allow Tesla to sell 10x more cars than Toyota (for example).

AlexandrB 8y ago
> Tech allows a company like Uber to allocate rides vastly more efficiently It's more accurate to say that venture capital allows Uber to subsidize rides more aggressively. I see no evidence that Uber is substantially more efficient (when you include the driver's costs) than traditional taxi companies. Edit: Uber has also managed to skirt a bunch of labor laws by claiming their employees are contractors. I guess that's a kind of "efficiency".
- rconti 8y ago
  I think it's unquestionable that tech like Uber's can/does allocate cars and rides more efficiently. The taxi model relies on drivers circling looking for a fare and phone dispatch with, at best, a dispatcher seeing GPS cab location on a screen and trying to eyeball-match it to the caller's location. (or, at worst, caller-reported location, and radioing a bunch of cabs asking them "hey, where you at?") At its most efficient, an Uber model can automatically and predictively allocate rides based on realtime rider and driver current and predicted locations with realtime traffic/travel data. There's simply no comparison. The extent to which rides are cheaper due to skirting of regulations, extracting money from VCs, pushing costs onto drivers, is effectively a separate issue. That is to say, Uber can break many laws, screw over drivers, flood the market with way too many cars, all while losing money, but none of that means that the dispatch model isn't inherently more efficient.
  - chipotle_coyote 8y ago
    The dispatch model is also relatively easily copyable, though. I don't want to downplay the amount of tech work that went it, but it's not an insurmountable obstacle -- taxi companies could sign up with a service like Curb or Easy, or even create their own like Yellow Cab Houston's Hail a Cab.
  - dbt00 8y ago
    Uber wins through three things. one, the convenience of the app model. two, subsidizing rides to flood the market with drivers. three, offloading risks like employment law and fleet management onto drivers. None of the stuff you said about traffic routing is currently true nor is it likely to be relevant. Taxi and Uber drivers do the same thing -- show up where and when there are predictable flows of people. They do this using human intelligence. It is unlikely that all the crazy big data analytics is going to do significantly better than this. Occasionally they work off hours and pick up a few people that need rides. That's it.
    - hakfoo 8y ago
      1 and 3 are undefensible territory. 2 is selling dollar notes for 90 cents; eventually you're going to have to convert to something more cash-flow positive. Their only realistic "exit as a world champion" play was to get to self-driving and own a fleet themselves (being able to create economies of scale on vehicle purchase and maintenance that would beat local taxis). That dream ended when they ran over a woman in Tempe. Their best hope now is to be one of several large commodity players-- modest margins, some limited ability to spread fixed costs like software engineering over large volume.
  - jacques_chester 8y ago
    This brilliantly advanced dispatching still makes cars drive around New York City blocks at rush hour instead of letting me walk 20 metres to the car that is not far ahead of me on the avenue.
  - AlexandrB 8y ago
    > At its most efficient, an Uber model can automatically and predictively allocate rides based on realtime rider and driver current and predicted locations with realtime traffic/travel data. [emphasis mine] This seems like a bizarre claim. Precisely when a specific rider will call for an Uber is inherently unpredictable (at least without a lot of surveillance). Meanwhile Uber drivers are not paid to drive from place to place without a rider - so Uber has limited capability to make them do something like head to a high-volume area in anticipation of demand. This is especially true since almost all drivers I talk to now drive for Uber, Lyft, and whoever else at the same time. The micro-optimizations of dispatching drivers based on what side of the street I'm on or how many left vs. right turns are on the route are nice but only have a meaningful impact on wait time when there are already numerous drivers in the area to choose from (i.e. when the wait time is already short). Edit: On second glance I realized you may mean that the dispatch is predictive in terms of anticipating traffic conditions and factoring this into the matchmaking between riders and drivers. While this is valuable, the key innovation here is the same kind of predictive routing algorithms that both Google and Apple already provide. Building this yourself may be hard, but is far from novel.
    - rconti 8y ago
      Yes, perhaps the best example I can think of was actually a failure -- my Uber that was supposed to arrive in "5 minutes" was actually driving away from me at the time, turning into FB HQ a minute later, and then spending easily 10+ minutes getting lost driving from building to building on the campus. But their algorithm can predict (usually!) which car will be nearest me the soonest; not which car is currently closest to me (regardless of occupancy) or which car is empty and nearby.
- icebraining 8y ago
  Uber has also managed to skirt a bunch of labor laws by claiming their employees are contractors. I guess that's a kind of "efficiency". In the US, taxi drivers were always considered contractors, Uber just followed the existing model. In fact, they used to pay to work[1], starting each day in the hole. Only after Uber et all appeared have drivers started unionizing nationally and suing taxi companies to change that. [1] https://www.nytimes.com/1995/04/09/nyregion/driving-a-taxi-difficult-in-best-of-times-gets-tougher.html https://www.nytimes.com/1995/04/09/nyregion/driving-a-taxi-d...

by rconti 8y ago

AlexandrB 8y ago

> Tech allows a company like Uber to allocate rides vastly more efficiently It's more accurate to say that venture capital allows Uber to subsidize rides more aggressively. I see no evidence that Uber is substantially more efficient (when you include the driver's costs) than traditional taxi companies. Edit: Uber has also managed to skirt a bunch of labor laws by claiming their employees are contractors. I guess that's a kind of "efficiency".

rconti 8y ago
I think it's unquestionable that tech like Uber's can/does allocate cars and rides more efficiently. The taxi model relies on drivers circling looking for a fare and phone dispatch with, at best, a dispatcher seeing GPS cab location on a screen and trying to eyeball-match it to the caller's location. (or, at worst, caller-reported location, and radioing a bunch of cabs asking them "hey, where you at?") At its most efficient, an Uber model can automatically and predictively allocate rides based on realtime rider and driver current and predicted locations with realtime traffic/travel data. There's simply no comparison. The extent to which rides are cheaper due to skirting of regulations, extracting money from VCs, pushing costs onto drivers, is effectively a separate issue. That is to say, Uber can break many laws, screw over drivers, flood the market with way too many cars, all while losing money, but none of that means that the dispatch model isn't inherently more efficient.
- chipotle_coyote 8y ago
  The dispatch model is also relatively easily copyable, though. I don't want to downplay the amount of tech work that went it, but it's not an insurmountable obstacle -- taxi companies could sign up with a service like Curb or Easy, or even create their own like Yellow Cab Houston's Hail a Cab.
- dbt00 8y ago
  Uber wins through three things. one, the convenience of the app model. two, subsidizing rides to flood the market with drivers. three, offloading risks like employment law and fleet management onto drivers. None of the stuff you said about traffic routing is currently true nor is it likely to be relevant. Taxi and Uber drivers do the same thing -- show up where and when there are predictable flows of people. They do this using human intelligence. It is unlikely that all the crazy big data analytics is going to do significantly better than this. Occasionally they work off hours and pick up a few people that need rides. That's it.
  - hakfoo 8y ago
    1 and 3 are undefensible territory. 2 is selling dollar notes for 90 cents; eventually you're going to have to convert to something more cash-flow positive. Their only realistic "exit as a world champion" play was to get to self-driving and own a fleet themselves (being able to create economies of scale on vehicle purchase and maintenance that would beat local taxis). That dream ended when they ran over a woman in Tempe. Their best hope now is to be one of several large commodity players-- modest margins, some limited ability to spread fixed costs like software engineering over large volume.
- jacques_chester 8y ago
  This brilliantly advanced dispatching still makes cars drive around New York City blocks at rush hour instead of letting me walk 20 metres to the car that is not far ahead of me on the avenue.
- AlexandrB 8y ago
  > At its most efficient, an Uber model can automatically and predictively allocate rides based on realtime rider and driver current and predicted locations with realtime traffic/travel data. [emphasis mine] This seems like a bizarre claim. Precisely when a specific rider will call for an Uber is inherently unpredictable (at least without a lot of surveillance). Meanwhile Uber drivers are not paid to drive from place to place without a rider - so Uber has limited capability to make them do something like head to a high-volume area in anticipation of demand. This is especially true since almost all drivers I talk to now drive for Uber, Lyft, and whoever else at the same time. The micro-optimizations of dispatching drivers based on what side of the street I'm on or how many left vs. right turns are on the route are nice but only have a meaningful impact on wait time when there are already numerous drivers in the area to choose from (i.e. when the wait time is already short). Edit: On second glance I realized you may mean that the dispatch is predictive in terms of anticipating traffic conditions and factoring this into the matchmaking between riders and drivers. While this is valuable, the key innovation here is the same kind of predictive routing algorithms that both Google and Apple already provide. Building this yourself may be hard, but is far from novel.
  - rconti 8y ago
    Yes, perhaps the best example I can think of was actually a failure -- my Uber that was supposed to arrive in "5 minutes" was actually driving away from me at the time, turning into FB HQ a minute later, and then spending easily 10+ minutes getting lost driving from building to building on the campus. But their algorithm can predict (usually!) which car will be nearest me the soonest; not which car is currently closest to me (regardless of occupancy) or which car is empty and nearby.
icebraining 8y ago
Uber has also managed to skirt a bunch of labor laws by claiming their employees are contractors. I guess that's a kind of "efficiency". In the US, taxi drivers were always considered contractors, Uber just followed the existing model. In fact, they used to pay to work[1], starting each day in the hole. Only after Uber et all appeared have drivers started unionizing nationally and suing taxi companies to change that. [1] https://www.nytimes.com/1995/04/09/nyregion/driving-a-taxi-difficult-in-best-of-times-gets-tougher.html https://www.nytimes.com/1995/04/09/nyregion/driving-a-taxi-d...