Picking the Perfect App Icon

I used this distinction because it works pretty well for describing the two major styles of human representation. It also maps pretty well to whether the target audience for a game is male (realistic) or female (cartoonish). This is especially true when women are represented. Very “realistic” women front grindy PvP games with high fantasy themes, which, based on historical app data from Facebook web games, I believe do best with boys. Cartoonish women front games with realistic themes (paradoxically enough) like Diner Dash or Campus Life. Again, this repeats a pattern seen in Facebook web games.

The realistic/cartoonish distinction works worst for describing animals, where many icons (including Gaia’s Monster Galaxy franchise) sit in the middle. I think this is because monster training games are less discretely gendered and are more likely to be popular with kids.

Cartoonish icons also appear more common in games with the largest audiences (Clash of Clans, Jetpack Joyride, Temple Run). This makes sense: an accessible icon that doesn’t divide by gender has a larger addressable audience (than, say, an anime babe in a metal bikini). The key, of course, is to put such an icon on a game that is also addressed to a similarly large audience.

An aside: there are a lot of Minecraft-related apps in the top grossing charts. Minecraft isn’t a game, it’s a platform.

The iconic/abstract category was originally two categories--iconic and abstract--attempting to parse the difference between representational (iconic) and non-representational (abstract). I broke down under the weight of slots apps, which are both iconic and abstract. (I would say that as icons and games they represent their own abstraction: they are just random number generators with pictures). I cannot understand how there are still so many slots apps, but there are. Slots aside, I was surprised how many games use iconic or abstract strategies when it seems to me like a bad way to represent a game. This is something I learned.

In top grossing, strongly gendered characters split evenly. I take this to mean that if you are making a gender play, either a male or a female character has equal chances of standing out to your audience.

If you look at subcategories, there are striking differences from overall top grossing. RPG is a good example of the extreme deviation that happens in a subcategory, where app icons use human much more frequently than in overall top grossing. There is also a marked change in gendering, with women (and most often women represented through a male gaze) more than doubling the number seen in overall. This could either mean that using a babe is a strategy with consistent results or that it is oversaturated.

Adventure seems to use realistic inanimate objects more than the other two charts. I don’t take this to mean that a realistic icon is s better or worse, but that you might change your icon to match the goals of a marketing campaign targeting a particular chart.

Finally, I looked at the very top of the overall charts (screenshot below), since the top nine or ten apps get the lion’s share of exposure and revenue.

The sample size is by definition very low, which is why I wanted to look at the top 100 of three charts in the majority of my study. The rates of representation are about what we see in the top 100. Kingdoms of Camelot and Marvel War of Heroes, for example, are apps I would class as “realistic,” and they have about as much representation in the top 10 as in the top 100 grossing. One thing of note: in the top 10 of grossing, paid, and free, there are no women except for She-Hulk tucked in the corner of the Marvel icon. The same is not true of the subcategories, where two apps in the top 10 of RPG have female characters.