We told you so before, we’re obsessed with ratings, rankings, so much that we’ve created our own all encompassing App Score; the reason we care so much is partly because we have a very successful iOS app out there, used by millions of users, with near religious recurrence - and we keep tabs on general and specific signals, by tracking essentially everything and also engaging in deep CRM with our users. With that sort of loyalty, our users also await our updates with great anticipation and they typically upgrade in full confidence without blinking…
Yet we’re humans, and therefore we err… Version 2.4.1 of Appsfire Deals was pushed two months ago with a nasty bug in it; it crashed right away when launched on iOS 4.0. The intent here is not to put the blame on the faulty 3rd party library that caused it. The fact is that we failed to check that minor update on older devices running iOS 4, not because we don’t check our code on iOS 4 in general (we do), but because we wrongly assumed that this 3rd party vendor had done their own homework (they hadn’t), and because that update was theoretically really minor for us…
Well, we learnt our lesson well: it’s wrong to assume anything with so many moving parts. Even if iOS doesn’t suffer much from fragmentation, the permutations well exceed 100 unique configurations (vs. many thousands on Android). In the end, if end-users matter at all, only 3 words: QA, QA and QA…
The takeaway was much bigger than that in fact, and is two fold:
- assuming that an app can be updated “quickly” in case of trouble is wrong; in fact, our quick fix took over 6 weeks to be delivered… to simplify, let’s blame it on the increased review time around WWDC2012 and the impending iOS6/iPhone5 launch. And who knows, an update could potentially NEVER be approved; think about it, that version that you have published COULD be the last one!
- it will cost your app precious stars! yes, afflicted users will punish you with a one-star review, for sure!
On that last point, we were hurt! We pride ourselves on having a near perfect score, the elusive 5-star. Typically, on any app update prior to that one, within a couple of weeks, we were accustomed to receiving an average rating of 5-stars from over 1000 users for that given version, in the US App Store alone. In the case of our buggy update, our rating dropped to 4.5, and our own App Score dropped 2 points to 90. Ouch! We’re climbing back up slowly and surely now. Version 2.5 of our app fixes everything, and disgruntled users hopefully forgiving us and revising the grade they gave us!
And we're back in style with version Appsfire Deals v2.5!!!
So for you developers out there, you’ve been warned! Publish a crashing app on the now under-represented iOS 4; lose half a star, and few points on your App Score (and of course, expect a lot more damage if your app crashes on a more recent, more represented OS!).
As a side note, Apple, if you’re listening: please provide a roll-back mechanism, actionable for 24 hours post release… Not sure yet if it’s a good idea, but in some cases, it would save the day, the pain, the disappointment, the headache… and half a star 
