Some Apps Do Sell: Developer Makes $750K in 3 Weeks on Android Market

This tells it all. Looks like they were developing it for at least 1 year, I'm guessing they had something to the tune of 4-6 people (if not more) actually working on the development.

Still, if they keep up this revenue for a year, it could end up being some very nice returns. Will they really be selling ~$1m a month, or will it taper off soon?

well, in the android market apps tend to be on the cheaper side. You rarely see app for more than $5, so compared to the rest of the market, the price might seem high.

I think it depends on the kind of startup. If your startup is trying to grow an audience around a service with a free app, Android is a good place to start, if you are dependent on sales of the app for revenue, I'd go with iOS.

Are you sure it's a good idea to develop iPhone or Windows Phone apps either? On average, those don't sell either... and they can be deleted by the Overlords for no reason whatsoever.

Despite the extremely high risk, people do it anyway. I'm guessing the risk vs. reward is about the same with Android, but since there's no Kool-Aid to drink, there are more people wasting time on iOS apps. Just a guess, though.

(As an Android user, I do think it's strange to pay $15 for the home screen when the whole OS is free.)

There are a few advantages to writing apps for iPhone / Android, one lots of people are looking for consultants, when people hire consultants they want to see a portfolio, so you can increase your skills and possibly make some money building a portfolio while there is the chance of making money. It's much better than doing spec work.

So what you do is to write trivial apps, when you get consulting gigs you can say "30 apps in app store", and have a portfolio of apps that look good, also, you write apps around features so you teach yourself the feature, etc, GPS, Gyro, Camera, BUMP API,FB Connect, etc. This is also an important part, spend time making your app look good. Customers can't see code quality, but they do see graphics. To customers good graphics == good code quality.

Now if you have 30 apps and one happens to hit a niche you've got a great way to promote your other apps and don't have to do consulting BS anymore, and because you have lots of apps in the app store and lots of experience writing against the various features you can justify a much higher hourly because you'll be delivering their app faster.

I definitely agree with you that writing a single app to try and strike it rich in the app store is foolhardy, unless you already have people signed up to buy it.

Yes, particularly when it comes to Android. Exactly the reason why Angry Birds is free and ad-supported on Android, as opposed to iOS.

Unfortunately the main problem is that Android devices allow arbitrary packages to be installed. There will always be someone willing and able to crack and distribute any given app that relies solely on client-side security. Your best bet is to roll your own obscure anti-piracy and hope you get most of your sales in the first few days, or look at a solution based on IAP receipts.

The same problem exists on iOS, but jailbreaking is an extra inconvenience that only 10% or so of users will accept.

Also I doubt many users think twice about downloading a pirated app nowadays (and they should, due to the very real threat of trojans). It's been over a decade since "don't copy that floppy" was replaced by the music/film industry's crusade.

The following is a good talk given at this year's Google I/O on how to use secure Android applications:

http://www.google.com/events/io/2011/sessions/evading-pirate...

If you foster a positive community relationship, you don't get hurt as bad by it. I don't think its really worth the struggle. (See Swype for example, people would just like for them to take their money, I still don't know if you can buy it yet.)