Developer Brian Robbins from Riptide Games gave a fascinating talk today about the apparent paradox of trying to make money off of a free app on Apple's App Store. Originally the App Store had only the two choices of a free or paid release. More recently, developers have a bevy of options for bringing in revenue even on free apps, including in-app ads, in-app purchases and freemium models, or even using "lite" versions of games to promote the paid versions. Robbins walked through a general overview of the models available and filled in the blanks with a lot of actual, clear numbers from his own company and its releases.
Robbins first mentioned making free apps for portfolios and lite versions. Some developers will just put apps together to try to build up their brand or show off work for hire abilities, then release those as free apps for everyone to try. Lite apps are also a possibility, where devs can release free apps to promote their paid versions (or even promote upgrades via in-app purchase). But Robbins was unimpressed with the lite strategy; he said conversions to paid apps were basically a numbers game and not a really workable strategy. Many devs have been releasing free apps to try reaching a larger audience, then hoping that a larger audience would go for paid apps. "I don't personally believe that the lite version is a really good model going forward," Robbins said.
Robbins was much more bullish on ad-supported apps, a strategy that his company has used to solid effect. There are a few different large ad networks on the iOS platform, the most high profile of which is Apple's own iAds service. But while Robbins said ads were a pretty stable way to make money (sometimes even earning up to $2 or $3 per CPM -- basically the number of clicks or views on a thousand ad shows), the "fill rate" is still a problem. iOS developers request ads from these ad networks all the time, but ad networks don't always have ads to show.
"Fill rate" is the rate at which advertisers actually insert ads in an iOS app, for a user to see and a developer to make money off of. These fill rates are surprisingly low, no matter what ad network you're talking about. Even an established network like AdMob (owned by Google) only fills ads at a rate of about 80 percent. And that's high in the market; Robbins said other advertisers only filled at about 50 percent. Apple's own iAd service only filled at about 25-30 ercent. Outside of the US, said Robbins, the fill rate for iAd on apps was practically zero, and another developer in the audience said that when he tried to use full-screen iAds on the iPad, the fill rate turned out to be something like 0.01 percent. In other words, Apple is apparently having a tough time selling ads (something that we've heard before).
However, Robbins also said that when Apple does sell ads, they are profitable. Last holiday season, Apple's ads were paying as much as $40 per CPM (though they've since been down to $0.90), and Robbins's set of ad-supported "iLook" apps take their biggest chunk of revenue from Apple's own service. The iLook apps have made over $53,000 for Robbins and his company so far, of which $26,000 came from iAds. Most of that came from a big Christmas boost last year. Robbins said the other networks paid less for shown ads, and he was surprised that networks have started to pay much less for games and a little more for non-game apps. iAds for Dove, for example, never showed up in his games, but he saw them all the time in his non-game photo apps.
Finally, Robbins talked about the freemium model, which he and his company have since moved to using in a big way. The first game Riptide released under this model was Gravity Slide (a game actually prototyped at a previous 360iDev game jam). Since its release the app has pulled in 450,000 downloads, making up $14,000 in revenue from in-app purchases of level and content packs. The app has seen a 1 percent conversion rate (so 1 percent of people using the app have purchased something), but Robbins noted that among OpenFeint users specifically, that conversion rate more than doubled to 2.5 percent.
Finally, Robbins walked through the stats on My Pet Zombie, his company's latest game, designed completely for the freemium model. This game uses "bones" as an in-app currency and has pulled in $53,000 in revenue though it's only been out on the App Store for a little while. In-app purchases made up $19,000 of that revenue, while ad campaigns and "incentivized promotions" (where advertisers pay for things like other app installs, or even getting users to watch a short video) made up most of the rest. Robbins also stressed the importance of metrics in freemium development, or the ability of developers to track and monitor just what their users are doing in the games and apps and how they're doing them.
When My Pet Zombie first launched, said Robbins, it was only keeping users around for the first full week at a rate of 7 percent -- that is, after seven days post-download, only 7 percent of users were still playing the game. That kind of stat also led to a low conversion rate, which meant low revenue. But through some testing, Robbins has managed to get the seven-day retention figure up to 13 percent, with consequent revenue growth as well. He credited that growth to using metrics to figure out the audience, and he said he plans to grow the game even bigger in the future.
Robbins's talk was really interesting, and those numbers he shared offer up some real food for thought for developers in this space. The freemium model definitely shows promise (which the "lite" app model is apparently lacking), and ad-supported apps look like a potential revenue boon for developers of free apps as well.