I have asked NimbleBit about getting acquired before. The brothers seem like prime candidates for a larger company to come along and buy them out, and indeed Ian Marsh says during this whole affair that Zynga has offered to purchase NimbleBit and its games before. But as they told me, these guys aren’t in it for the money. Certainly their games are very lucrative, but the Marsh brothers have said that they just like making great games on their own and will continue to do that for as long as they can.
Zynga is the leader in the Facebook game market with about 221 million users according to AppData. Now Wooga (36 million users) is looking to edge out Zynga at their own game: using metrics as the primary design method for their games.
The basic idea at Zynga and Wooga (amongst others) is to use a data-driven analysis of game design. Everything the player does is tracked and put through some basic analysis to see how to make the games more sticky for the players. The end goal for these companies is to drive sales and most design decisions lead back to this core goal.
Users are often subjected to A/B testing to see what works best – this happens almost daily. The best design (most clicked or other similar measurement) is then put into the game until another approach is found and that is A/B tested.
The result is a rigorous process that practically automates the creation of a social game, and maximises each title’s chance of success. “We A/B test everything, we optimise everything,” says Stephanie Kaiser, a lead game designer. “In the product department, it’s very simple,” says Thorbjörn Warin, a former employee. “They have all of their KPIs [key performance indicators] and metrics. It’s really, ‘This week, we focus on nothing but retention, let’s identify ten activities that can increase that.’ In the first 60 seconds of Monster World, there are 13 to 15 tracking points. For a new user, when they start playing, every three or four seconds, Stephanie and Jens can see what is happening. Usually something has to be improved, and that’s when creativity comes in.” Wooga’s users don’t just play a game; they design it.
… Monster World launched in April 2010. But it soon stalled, with only 300,000 daily active users by August (wooga considers one million users the minimum mark of success). “It was not growing virally and it was not a success at all,” says Kaiser. The team focused on three topics: engagement, virality and monetisation, and went to work on the first. Kaiser began with the “user funnel”; she studied 38,863 users who began the game tutorial one week, to see where they dropped off. “A 1.3 percent drop is unacceptable and the game is optimised accordingly,” she says. When such a loss was identified, Kaiser’s team would develop two solutions, put them both live as an A/B test, and find out which performed better. And so on across every part of the game. It worked. On November 16, Monster World reached a million daily active users. “What we learned was that you could really turn a game around post-launch,” says Begemann. “It had always been my belief, but that’s the first time we really proved it, by doing nothing else other than A/B testing, and of course being creative. Four months after launch, that’s when some companies would have given up.”
With this larger user base established, wooga “switched on the monetisation”. Monster World offers several ways for a user to pay to customise their game. But two-thirds of wooga’s revenues come not from these adornments, which account for around only two percent of total sales, but from the items that turbo-charge a player’s progress in the game: magic wands, which harvest crops instantly (240 wands cost 480 Facebook credits, worth roughly £76), and “woogoo”, which produces several other items of value. These may sound whimsical, but “a feature has to be driven by metrics, it can’t just be cool,” says Kaiser.
Is this a bad thing? Well, before Zynga and Wooga made it big on Facebook Jonathan Blow was critiquing their core design philosophy of retention (and later monetization) as the primary goal:
Is metrics-driven design inherently a good or bad approach?