Prognosis : Lessons learnt from developing a top 10 iPhone app
blog.medicaljoyworks.com
blog.medicaljoyworks.com
My thoughts are if you are planning on adding ads eventually, you should ad them with version 1. Otherwise, people will complain when they come later.
We launched a free App (AppShopper) a couple of months ago, and we did hit the top 10 overall free.
We planned ahead with a dedicated server with a caching system (Squid) and the ability to spin up EC2 instances to mirror further. Fortunately, the one server (and the db server behind it) was able to handle the single traffic peak (which was over 100,000 downloads in a day).
I can identify with the testing thing (or lack thereof), as you can't fix a problem quickly, since you have to go through App approval. Which can be frustrating when there's a crashing bug.
Fixed on blog. Can't edit post.
We thought that having advertisements at the very start might turn off some users - but there is also the possibility of a backlash if we add them later.
Note that we were not planning on adding a generic ad network, but rather partnering with a pharma company and deploying ads customized for the iPhone.
The other possibility is to have 'sponsored cases', similar to what they have on Medscape - there would just a be a single line of text, i.e 'Sponsored by XYZ"
However, 90.9% of our users are listed as having iOS4 devices - I am not sure whether this indicates the iPhone 4, or whether iOS 4 prevents flurry from detecting the device type (in which case this would be an accurate estimate of the number of devices running iOS 4)
I guess that the percentage of users with iOS 3 devices should therefore be at most 9.1% - probably less than that.
Another statistic is that 7.43% of the users had jailbroken devices - I suppose that a significant portion of them would be using iOS 3, as it is so painful to upgrade to iOS 4.
Should I also keep a slicehost/linode vps running the app?
It's great for development and testing functionality, not good for testing compatibility.
All costs are out of my pocket (and Sandaruwan's and Parinda's)
Note that $50 is a LOT when bootstrapping from Sri Lanka - my salary as a Doctor (which is much higher than that of the general population) was only $400 / month - with most of that getting burnt up in living costs.
Arent you afraid that you loose your momentum without made any dollar from your app? You can find clean way to do advertising on an iphone app (look at Twitterific for example)
Yes, we have several clean advertising plans. i.e - sponsored cases.
If you are not confortable with that maybe you could contact them offer them that to have an idea of what they are ready to pay to be in your app. Maybe the amount will change your mind :)
I dont mean to be disrespectful with my comment but I made that mistake to wait to long before monetising my startup because I always wanted more and more data before selling ads on it. I lost a lot of time (and money) for nothing.
Now that I think about it, another point that should have gone into the blog post is that "be prepared to monetize if your app goes viral in few days"
If you look at the top 10 apps in the medical section (or the top 25 for that matter), you'll see that they have been there for months (or years) on end.
My feeling is that the medical section of the app store is fairly different from the games section (where apps have a very limited shelf life).
Sooner the better.