2012 has been an exciting year for me, as I released my first commercial OS X App under Fingertip Tech, INC this year. Overall, the experience has been great and I am really enjoying all the great user feedback and am happy to report that a lot of users truly enjoy using the app. However, there have of course been a few missteps along the way.
Marketing: This is the biggest one by far. It is truly amazing what a positive effect a tweet about your product can have on your sales. Unfortunately, I didn’t realise this and failed to allocate enough capital for a proper marketing campaign. I am not talking about purchasing reviews here of course, but it is a reality of the current app marketplace that it is very hard to get a lot of traction without a significant marketing push.
Service Dependency: This is a closed second to my marketing misstep. Like Tweetro and so many developers before us, I have found that developing and maintaining an app that is built upon a third parties service is more than a little challenging to say the least. I won’t beat a dead horse here, since you’ve probably heard this from a number of other developers, but do not do this. Just a small example of one the most minor issues you will face: yesterday Github.com went down and stayed down for an extended period of time and I recieved a deluge of e-mails from frustrated users who could not interact with the service via the app because of this. The unfortunate truth is that if all or even a significant number of those users had decided to one-star the app instead of seeking support from me, sales would have been significantly impacted.
Like so many before me, I am unlikely to develop an app based off of someone else’s service again.
Distribution: Code Journal is currently sold on both the Mac App Store and its own site. Sales have been about equal in both places — some weeks favoring the site and other weeks the App Store. Overall, sales have been good, but since the sales are split about equally, the App Store ranking is significantly lower than it would be if all of the sales had come from there.
On the other hand, it has taken at least twenty (20) days for each update to be reviewed and approved in the Mac App Store. That’s simply too slow, since I have been trying to iterate quickly. Also, the App Store does not give me direct contact with my users and that’s something I greatly value
Focus: In many ways, Code Journal has been a success and I very happy with how it is performing. However, I’ve lost a significant amount of time and wasted a significant amount of energy attempting ports of the app to a number platforms using a number of tools. Code Radio listeners will know that there was a Mono port and a QT prototype already. What they don’t know is that there have been others — even a Java version. All of them were not satisfactory and were in many ways the lowest common denominator for whatever platform they targeted. I still hold some hope for QT in 2013 to bring truly powerful cross platform development, but, at least for now, the best apps will always be the ones written with the native toolchain.