Hacker News Reader

> you need a great iOS app

> you need a great Android app

Neither of these is rocket surgery. Good devs can create an app very quickly, especially if your directions are "clone this other app."

> you need a backed that is both fast and has rock solid 24/7 reliability. it can't be like Twitter in early days

So you need something that isn't node + redis + mongo + riak running heroku through cloudflare. This is also quite easy.

> you need to become a payment processor that interjects itself between users and drivers, which I'm sure comes with its own set of legal and logistic challenges * and you need to do it internationally

> and you need a massive operation team whose job is to onboard new drivers (recruit, train, supervise)

> you need a legal team because every state can have different rules

> and you need them also internationally

The beauty of all of these things is that Uber has already done many of them for you. Unless they have all of their drivers under NDA (possible, but unlikely) and haven't published any of that informaton on their website, you're going to profit from hundreds of thoussands if not millions of dollars of legal work that Uber has already done for you. It's not as simple as a sed script on a legal document, but it's way (way way) easier than starting from scratch.

The best part is that your drivers can also be Uber drivers, so you don't need the volume that Uber has today - you can slowly build up both sides of your market as long as the friction for drivers is minimal and the ease for users is equal or better.

Is it easy? Fuck no. Is it about 1000% easier for you now that Uber has done all of the hard work? Hell yes.

And you could start in your small college town, work out the bugs in your process and then go national, then international. @debacle I agree. Micro-Ubers could pop up all over.

Nice theory you have there but Uber uses Node.js.

Drivers often list themselves on both Uber and e.g. Lyft in the same way riders use both apps. So there are no network effects in play. All the drivers need to do is switch the signs on their windows.

It's the reason Uber is taking such extreme measures to scare "it's" drivers into not simultaneously listing themselves on competing apps. See: http://money.cnn.com/2014/08/04/technology/uber-lyft/

Note this is not an example of "no network effects". This is the opposite. Both services benefit from the shared network effects. The real question is whether the future market for ridesharing is zero-sum, which seems unlikely looking at analogous markets.

Building a few pretty apps is hardly a barrier to entry. It might be beyond the reach of a single founder bootstrapping but it's quite reasonable to build with a small VC investment.

Note that autowale.in (an Uber competitor, currently profitable) was started on an investment of about 30 lac - roughly speaking 1 year runway for 3-5 engineers. Various tricks: start in 1 city, do the onboarding manually, and ignore the legalities until you are big enough to be noticed.

Autowale.in dealt with the bootstrap issue directly buying inventory from drivers. I.e. autowale.in pays the driver for 20km each day, and then tries to find customers to use that inventory.

The major barrier to entry here is the fact that Uber is really good. But if Uber decides to exploit their market condition and rip people off, that will change.