iTunes AppStore & The Tyranny of Choice

When I’m looking to buy a new video game, I have 2 primary sources: The first is Video Game blogs (I generally read Joystiq & the IGN Xbox feeds), the second is my friends – whom I mostly to use to ask about games I’ve first heard about via one of the above.

But my practice for learning about new apps is different – it’s nearly 99% from my friends, 1% from blogs (if I’m being honest, mostly from Daring Fireball, which is my (and many, many other people’s) go-to place for new Apple/iOS-related opinion. Since about 1 month after it opened, the iTunes App Store itself has been more or less useful useless (thanks, Evan!). Cream doesn’t rise to the top in that store. Look at the top 5 apps in any category – you’re as likely to see utter garbage as you are a beautifully designed app or brilliant, original game (based on how network TV works, I’d go as far to say as the most original NEVER get the most attention). This because the metric used for the top charts is based on downloads (Top) and sales (Top Grossing). There’s not really a mechanism for “most interesting” and if there were, unless it was weighted by people’s whose opinions I like, would still not be useful.

It ends up that I now more or less completely ignore the App Store as a source of recommendations for apps. Even the app-related blogs aren’t so hot, because, for them to be useful filters, Its needs to be focused on particular categories of apps – or it’s just too general – similar to how sites that review all forms of music pale compared to sites that mine just a few genres. Whatever the issue is, the App Store is failing for me as a source of new apps – I’m paralyzed by the Tyranny of Choice: Which of the 18283 education apps do I want? Hell, even deciding between the 120 “Hot” Education apps is too much choice.

What’s needed is a sort of “social ranking” mechanism – sort of like the vote up/down of Reddit, Digg, etc. But actually, what I’d like is something even more refined. I’ll state now that I really don’t care what the vast majority of the world thinks. I care about what people that I respect think. These vote up/down algorithms should be measured in concentric circles: First & foremost, what my friends like. If, say David likes an app, there’s a pretty good chance that I’ll like it too – at least based on historic precedent. & I don’t care about 5-star ratings. All I care about is thumbs up/thumbs down. The next circle out from that should be what friends of David, but who aren’t my friends, think about an app – one of those birds-of-a-feather ideas – if David likes this people, maybe I will to. Beyond that, should be people who’re influential “generally” think. Leveraging something like Klout could be useful for this level of recommendation. I might not know the person from Adam, but if more often than not apps that they recommend are subsequently recommended by people in their circle of friends, that’s a good sign.

This seems like a perfect area to start exploring interest data-mining/app possibilities. While you could simply shoe-horn on an up/down voting method over the existing, more-or-less useless 5-star rating system, I don’t think even this is necessary. Here’s what I’m imagining:

  • An iTunes plug-in that allows me to share what Apps I own (I’m equating ownership as a “thumbs-up” – this might be overly simplistic, but it’s a place to start). This sharing could be anonymous or not. Maybe I need to actually “rate” an app to share it, so I can hide apps I don’t want to rate or share.
  • Leverage my existing social networks to see which friends I care about. This is a well-established method: Let me “follow” my twitter, facebook, linked in, google/yahoo/hotmail contacts, etc. I strongly believe that this system needs to asynchronous – more like twitter than like facebook.
  • As I like apps that someone else has already liked, the weighting the system gives to that person’s likes, relative to me, should be weighted higher – because it means that I’m more likely to agree with them in the future – a bayesian weighting system.
  • Over time, as my circle of followers grows & shrinks, as people in my circle add/remove/rate up/rate down apps, I’ll have an ever-changing list of suggested apps. Which makes app developers more money, makes apple more money, makes my devices more useful to me.

So there’s some hitches that I see in all of this as I currently have it down:

  1. People have to download a plugin to iTunes (or, they have to sign into a website then manually find their apps). This manual start-up process is a rather large barrier to entry.
  2. People actually have to rate apps & do it regularly as their app-library changes. A code-snippet that could be added into apps, similar to the existing “rate us in the app store” would be nice & helpful – but would require this get big enough.

Thoughts, people? Is there already something like this out there? If not, and you’ve got some money, want to fund me to make it? Or, want to make it yourself? Let me know – I’ll be an eager tester of it!