HackFwd Build 0.10 - Berlin June 2012
Stephanie Kaiser from Wooga and Torsten Reil from NaturalMotion are HackFwd referrers and games industry experts. In this high-rated talk, they combine their expertise in KPIs and optimizing them, especially for games.
Torsten and Stephanie outline the 'magic numbers' that they manage their games with. In this talk, they concentrate on retention and virality, measured in terms of Daily Active Users (DAU).
They start off by analyzing the first day retention in games. By analyzing the tutorial funnel, step by step, you can increase the retention on the first day, which makes a large overall impact on your business as you keep bringing new users in every day and keeping them.
Particularly for games that a skill-based, it is important for people to feel they are good already on the first day. In Torsten's example, NaturalMotion's game Backbreaker introduced rather incompetent opposing team players in the very beginning so that the player feels they can play the game well even in the very beginning.
Day-to-day retention is another important KPI to track and optimize. Daily tasks in Monster World and foreshadowing of upcoming events and challenges are rigged to make people come back often. In NaturalMotion's My Horse, push notifications remind the players that their horse needs attention and they should go back to the game.
K-factor is the virality metric to track. It is the number of new people that users on average bring into the game. A K-factor of 0.2 means that for each five players, one new users is introduced into the game. A K-factor of 1 would mean that every player in the game brings in one new person. This is cumulative, so the new people brought into the game will also bring in more people. Optimizing the K-factor is a major leverage in a game's growth, as the examples in the talk show.
But remember, you have to fix retention first, Torsten says. Otherwise you are carrying water with a leaky bucket.
A/B testing and the performance landscape
The performance landscape for games is set of hills, and it depends entirely on the design what hill you land on, or close to. A/B testing the UX and the game play will then help you climb that hill as high as possible. It may not help you in finding new higher hilltops in how your game performs, but it will help you find a 'local maximum' of game performance. The idea of performance landscape and A/B testing is directly comparable to fitness landscape in evolutionary biology, where evolution by natural selection (equivalent to A/B testing in this example) drives the fitness of a species or a group towards a local maximum, but it may not help them change to a higher maximum, as that would require crossing a fitness valley.
After the talk, Torsten and Stephanie shared the KPI management model they used in the talk with the HackFwd startups. To see more great talks for startup founders, head to Passion Meets Momentum. If you are interested in applying to HackFwd, read the HackFwd Experience.