One of my (many) ambitions for SnoreLab is to have the app appear as one of the top 3 hits when someone searches for ‘snoring’ in the App Store.

The importance of appearing as a top hit in the App Store has increased dramatically since the release of iOS6. In iOS5, when a user searched for a keyword they were returned a list of 6 apps immediately visible on the page, with the top-25 quickly available with the swipe of the finger. It was possible to quickly and easily see the icons, names and ratings of 25 relevant apps, and your attention was quickly drawn towards whatever app had the most attractive icon and the best rating. One of the factors that encouraged me to create SnoreLab was the poor quality of the existing snoring apps, and I’m confident that SnoreLab would have performed well in that particular beauty parade.

However, in iOS6 App Store navigation was flipped on its head (or more precisely, tipped onto its side…). Instead of seeing 6 apps in response to a keyword search, in iOS6 only one app is returned from the keyword search. To view more apps, the user must side-swipe, with one new app revealed with each flick of the finger. App Store navigation has become much slower and much harder work as a result, and most users won’t make it beyond the first 5 apps or so.

Now, this is great if you are one of the top 5 apps. Where it all goes wrong, however, is in how the App Store is ranking the apps…

From my analysis, the results returned from App Store keyword searches are heavily skewed towards the historic number of downloads an app has had. This in turn, creates a heavy skew towards free apps (which typically get 5x-100x more downloads than paid apps), and also a skew towards apps which have been available for a long time. What barely seems to enter the equation, however, is the average rating that an app has.

And so you get the bizarre situation when Apple customers, whose choice of Apple implies to me they appreciate quality, search the app store they are first returned a list of junk apps before they get to something of quality. And so, in the UK, when someone searches for ‘snoring’ they are returned the following list…

1) Sleep Sounds Recorder LITE (Free) – 2011 – average UK rating 3 stars from 17 ratings

2) Sleep Recorder (Free) – released 2009 – average rating 1.5 stars from 6 ratings – described as “poor in the extreme” by one reviewer

3) Anti Snoring FREE – released 2010 – average rating 2.5 stars from 28 ratings – essentially a google advert that whistles at you – generously described by one reviewer as “absolute rubbish”

4) Sleeping U (Free) – 2011 – hasn’t even attracted 5 ratings in more than a year on the UK market

5) Stop Snoring ! (Free) – – 2011 – Has the grand total of one 1 star rating in the UK after more than a year

6) SnoreLab – (Paid) – Average rating 4.5 stars from 20 people, Top 5 medical app in the UK

Now, I’ll admit to having a bit of bias in the matter, but I’d have thought that most people looking for a snoring app would rather see a 4.5 star rated paid app rather than a 1.5 star rated free app. But that isn’t what they get. As a programmer who is obsessive about producing something high quality, it boggles my mind that… Apple… APPLE!… the most valuable company in the world that has produced some of greatest and most world-changing products ever…. could have such a poor search algorithm in their App Store. I mean, come on guys, sort it out!!! Hire someone from Google already!

In the US – SnoreLab is currently appearing around 13th in that list…. despite being amongst the best-rated apps in its category…. and partly as a result of that the app is getting fewer downloads there than in the UK – despite the fact that the US is 5x larger!

This current algorithm also has the frustrating consequence that if you want you app to score highly in a keyword search, it helps massively to be free. I witnessed the impact of this first hand when I experimented with making SnoreLab free to download… and downloads went through the roof.

What really annoys me about this situation is the implicit assumption within the search algorithm that people would rather have an inferior app for free than a superior app for a small sum of money – yet these are the same people who shelled out hundreds of dollars for a shiny iDevice. Personally, I would much rather pay a couple of bucks for something high quality then get a ‘free’ piece of junk.

I hope that one day Apple will get round to improving its app store search algorithms – surely, it would’t be too difficult to return the top free app and the top paid app for a keyword? Or to integrate ‘rank results by rating’ as an option? In the mean time, my faith in almighty Apple has taken another dent.

But SnoreLab is going to get up there one way or another, mark my words 😉

*** Update  7th Jan 2013 *** In the few days since writing this post SnoreLab has climbed from 13th to 10th in the US keyword search and from 6th to 3rd in the UK with no change in pricing strategy. This suggests that the App Store search algorithm is indeed learning more about the app and ranking it more highly, perhaps because it is attracting more clicks from the search term.

*** Update  20th Jan 2013 *** I decided to give the app away free for a limited time to boost its prominence and it was an amazing success. I gave the app away free through AppAdvice.com’s ‘Apps Gone Free’ and it got an incredible response. SnoreLab was the most downloaded medical app in 50 countries and within a few days it is approaching 100k downloads! Although this has yet to generate any cash, the positive reviews are flooding in (now 35 reviews in the US, average 4.5 stars) and will surely give it a massive boost when it is again a paid app. In terms of search rankings, it is now appearing 6th in the US. An improvement, but there are still a clutch of rubbish apps above it, including ‘iSnoring’ a 2-year old free app with 4 ratings ever.

*** Update 22nd Jan 2013 ***

As soon as I put SnoreLab back to being a paid app it dropped from 6th to 9th in the US keyword search, with 8 crappy free apps above it. This is madness! Apple, what is your crazy bias against paid apps?!

*** Update March 2013 ***

Despite excellent user reviews (4.7 average from more than 280 reviews) and a consistently high rank in the paid medical charts, the app still only appears 6th in a US keyword search for snoring, with 5 poorly reviewed. low-quality free apps above it, which I am sure gross virtually nothing. I find it a bewildering situation how the search algorithm would put those apps above SnoreLab, and have come to the conclusion that there is a systematic bias in the App Store search results towards free apps/pure download numbers vs any other measure of app quality. Hopefully, one day soon, Apple will improve this.

*** Update May 2013 ***

Success!! SnoreLab now appears as the top hit for ‘snoring’ in the US app store! Clearly, the algorithm now ranks the app more highly based on its success criteria. My suspicion is that the keyword search results are only updated periodically, perhaps even every few months. I was dismayed by the lack of movement in ranking for months, but then overnight it all changed and SnoreLab was in its rightful place as the king of the snoring apps!

Available on the App Store