Hacker News Reader

Comment by cody3222 | Hacker News Reader

The network effects are pretty obvious actually: as the number of people on the platform increases, the amount of time a driver has to drive to pick you up becomes increasingly shorter, thus enabling the drivers to spend more of their time with a passenger in their car (which is when they are earning money). If drivers are making more money, Uber can pay them less (reduce the subsidy).

Old taxi cab companies can't compete because they have to drive much farther on average for each pickup

View on HN

Uber built its network city by city and a competitor could do likewise. As others have pointed out, most rides are local - if you live in Dallas, TX, you probably require rides mostly in Dallas, TX. If a hypothetically better (in some way) competitor came along that only served Dallas, you might be interested in trying it out.

So a competitor could just focus on Dallas to start out, raising enough capital to hire enough initial drivers for acceptable wait times in Dallas only. Once Dallas was successful, they could then go on to other cities.

This is unlike a website such as Ebay, which truly becomes stronger and stronger due to national and international network effects, not just local ones. If a user selling the beanie babies you want is in Atlanta, having them shipped to Dallas is hardly more difficult than having them shipped within Dallas itself. So that Atlanta user has increased the utility of Ebay to you.

  Uber built its network city by city and a competitor could do likewise.

A smart competitor would cherry-pick the most profitable markets and ignore the rest.

Yet there is a minimum viable issue to contend with. The average car will have an average value of dollars to go a unit distance and it cannot go lower than that very simply. The US Highway Safety board has computed this number for decades. Today it stands at ~$0.52/mile driven. This takes into account many many factors like gas prices, car maintenance, driving style, etc. Uber must pay drivers, on average, something more than 52 cents per mile. Yes, they may pay much less than that, but then you have to monkey about with driver mentality, average automobile efficiencies, road conditions, etc. Those are very difficult problems to tackle and are mostly outside of the 'wheelhouse' of Uber as a software company.

Yes, but it's easy to calculate out the results for everyone (drivers, Uber) if vehicles are x% more utilized, and at no point does it become sustainable for anyone - let alone profitable. As the Naked Capitalism article points out the only way this works is if Uber finds ways to reduce the fixed costs involved in their service, which taxis around the world have thus far been unable to do and Uber shows no evidence of achieving.

But if Uber is banking on autonomous vehicles none of this matters, because once there are autonomous vehicles providing sufficiently supply is purely a capital problem, and capital problems are relatively easy to solve if there are sufficient profits to be made. And if there aren't sufficient profits to be made, then Uber isn't a very attractive business.

Yes but how much are you ready to pay to divide your waiting time from 3min to 1min?

Reply on news.ycombinator.com