Prognosis : Lessons learnt from developing a top 10 iPhone app

http://blog.medicaljoyworks.com/prognosis-lessons-learnt-from-developing-a-to

by npsomaratna • 16 years ago

62 19 16 years ago

blog.medicaljoyworks.com

zackattack• 16 years ago

you can't afford $50/month even with a top 10 iPhone app? does not compute. if you're waiting to get your monies from apple, just email tim and i'm sure he'll let you pay on net 60 terms (or whatever apple's terms are)...

npsomaratna5• 16 years ago

Our app is free. We do have plans to add advertising later on - but not for several months at least.

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.

bjonathan2• 16 years ago

I'm sorry if it's a dumb question but why do you plan to wait to add advertising? You are on the TOP10 right now, you are having good review (by users AND blogs). Its seems to me like the perfect moment to add ads to your product.

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)

dmix• 16 years ago

I'd recommend figuring out monetization earlier. It's never as easy as flipping on a switch. It takes a while to nail down your margins, pricing, ad networks, etc.

Sooner the better.

arn• 16 years ago

So, first thing to note is the difference between Top 10 in the free Medical category vs. Top 10 of all free apps. Not to talk down about their achievements, but the #1 Top Free Medical App doesn't even break the Top 200 Overall Top Free Apps.

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.

npsomaratna5• 16 years ago

>> So, first thing to note is the difference between Top 10 in the free Medical category vs. Top 10 of all free apps.

Fixed on blog. Can't edit post.

npsomaratna5• 16 years ago

You do have a good point about adding advertisements later on.

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"

BornInTheUSSR• 16 years ago

Regarding the ios testing issues, would setting the iPhone emulator that comes with apple developer tools to the various versions have worked here?

xsmasher• 16 years ago

The emulator doesn't contain older versions of iOS, and it should not be relied on for real testing. It links against different libraries, runs different code, and won't find issues with iPod music, sleep/wake, memory usage, multi-touch, case sensitivity, and more.

It's great for development and testing functionality, not good for testing compatibility.

callmeed• 16 years ago

Re: #2 ... I'm about to launch my first app and it will talk to a Rails app on Heroku.

Should I also keep a slicehost/linode vps running the app?

whyleym• 16 years ago

I have recently released a paid app - which has turned up in the top 10 sports section in the UK for both iPad and iPhone. I have designed this for the very latest version of the iOS - i'm guessing i'm losing sales doing this - have you or anyone else for that matter found any statistics around the numbers of users who download the app from previous versions of the iOS ?

npsomaratna5• 16 years ago

Unfortunately Flurry's new SDK does not collect iOS version info, so we're as much in the dark as you are.

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.

gyardley• 16 years ago

The 'iOS4 devices' listing in Flurry means the operating system is running iOS 4 or greater. It doesn't mean the device is an iPhone 4.

Prognosis : Lessons learnt from developing a top 10 iPhone app

Related Stories