[drewcrawford]

I think this is nice in theory, but it's naive. From a developer's point of view, an icon is an ad that causes you to click and read the description in the app store. I don't like it, but those are the facts.

The author argues that a movie app icon can be designed in two ways:

1. To stand out against other movie apps in the app store

2. To stand out against Mail and Safari on a device home screen.

The fact is, developers are incentivized to do #1 and incentivized against #2. It would be nice to live in a world where that was reversed, but as long as developers are competing for eyeballs in an app store against other apps in a search query that will simply not be the world in which we live.

I've run A/B testing on dozens of app icons. #1 becomes optimal as your app becomes more niche, #2 becomes more optimal as the niche becomes increasingly competitive. Compare and contrast the icons for "cinema" as for "sign language" to watch this battle play out. Cinema you get brand names, abstract art. Sign language you get a hand sign, or little kids (in the case of baby sign).

[lookatpete]

The premise of my post is that your first assertion ("an icon is an ad") is not a fact, but a choice. If you search for "email" in the App Store, would Sparrow jump out at you? No, but they've chosen not to pursue that "advertising" strategy. Their growth is based on fully delighting their users with the product, and an icon which, IMO, looks like it always belonged on my phone.

Sparrow's growth comes from happy customers (and some traditional advertising), not an icon that looks like an ad.

I'm curious how you A/B test app icons...Do you switch from one version to the next and compare downloads?

Also, can you clarify what you mean by "developers are...incentivized against #2"?