Saturday, July 16, 2011

Four months on the Market

It's been almost four months since I opened my developer's account on the Android market and published my first app. Much to my amazement that app is not only constantly downloaded by someone but a significant part of the downloaders find it useful (or harmless) enough to keep it. In fact the number of downloads just broke the 500 mark. So what did I learn from this experiment?


Testing is hard but important

Well, first and maybe most important of all, testing Android application is harder than testing desktop or web applications. There is an extensive body of knowledge on how to do test driven development (TDD) for the later with whole books and innumerable blog posts devoted to TDD. But most of the techniques I use for testing web applications are useless for Android apps. Even unit testing feels awkward with the Android test framework being based on JUnit 3. And the books on Android app development all treat testing as an afterthought at best or ignore it completely at worst.

So I relied on manual testing for this app and naturally missed some bugs which I could have easily avoided using some simple automated integration tests. Guess I'll have to do more research on automated testing on Android. With a little luck somebody else will have solved this problem by the time I start my next app...

Satisfied users rarely give feedback


It's kind of sad but as a developer you get precious little feedback and the ones who are most likely to give feedback are those who are dissatisfied with your app. Of the 500 people who installed the app only four took the time to rate it. My favorite comment from a user is "Doesn't work....". Needless to say that that kind  of feedback doesn't help to improve an app. Another rather frustrating aspect of the Android market is that I've found no way to provide answers to the comments the users enter. Unless they send me an email I have no chance to get in touch with them.

Don't panic and stay polite

At one time I actually got an email from a user. Too bad it was a rather angry one in which he accused my app of having deleted all his calendar entries. After a few moments of panic during which I'd already started to write a note of warning in my app's description I wrote back that I didn't think it was my app that deleted his calendar but apologized anyway just in case I was wrong. The response I got back was already less angry and it later turned out that Google had some trouble with calendars not synchronizing properly.

Luckily this user contacted me by mail instead of venting his frustration with a negative comment on the market. This way we could clear up the misunderstanding and as far as I know he is still using the app.

Download statistics are addictive

This is not the first time I've released some tool free of charge but it is the first time I get nice download statistics telling me how many people are using my app, which language most of them are using and which country they are coming from. I must say it is strangely satisfying to see my app going places I've never been...

Don't expect easy money

Actually I didn't ;-) I released the app free of charge after all. But developing an app, even a simple one, takes  a lot of time, especially if you want to do a good job testing it in different environments. You'll need a lot of downloads to pay for that work and so far people only pay once for an app. Paid upgrades are not part of the app store model so you need a constant stream of new customers and / or new apps. So I guess it's not yet time to quit the day job...

No comments:

Post a Comment