I have written several times about Moneyball and many times about customer lifetime value (LTV), so I wanted to bring the two together. Moneyball was the Michael Lewis book turned into a successful film about Billy Beane and how he made the Oakland A’s competitive by relying on analytics over intuition (for more detail, please see Lessons from Moneyball for the Social Game Industry and Moneyball Strikes Again). The same principles that help the Oakland A’s compete effectively could help social game companies compete, even against better financed firms. The same phenomenon holds with LTV, in which many of the metrics people focus on do not have maximum impact on long-term success.
Runs = LTV
LTV serves the same role in your business as runs do in baseball. Beane and his analysts realized that the success of a baseball team comes down to scoring more runs than your opponent. They thus reverse-engineered the game and its players into what contributed to scoring runs and what contributed to preventing runs. They then used their resources that maximized the delta between runs scored by the A’s and runs that they allowed.
With games, you win when your LTV is greater than your cost per install (CPI); the larger the difference, the bigger the win. Thus, to be successful with limited resources (or even with virtually unlimited resources), LTV shows you where to dedicate your resources.
Retention is the social gaming equivalent of on-base percentage
What Beane and his Moneyball team determined was that a lot of the traditional ways players were measured did not contribute to the success of the A’s and may have even hindered it. For example, a player who hits many home runs would usually be given a very high value (and salary). Many home run hitters, however, do not get on base very often. Beane’s analysts discovered that these home run hitters were thus much less valuable than players who would walk frequently (for my non US friends, this means the player gets to take first base because the pitcher did not throw enough strikes). A player must get on base before he can score a run. Thus, on-base percentage is a much more important metric.
In the game space, retention serves the role of on-base percentage. Your customer has to be in the game for them to have value to you. They can only spend money (monetize) when they are in the game. They can only reach out to other people (virality) when they are in your game. Without retention, you cannot achieve a positive LTV, just as you cannot score runs without getting on base.
Session length is like stolen bases
Beane also discovered that the stolen base was an over-rated metric. The negative impact of occasionally being out trying to steal a base greatly reduced the expected value of a base-stealer.
Session length (which is sometimes referred to as engagement, though engagement can also be defined differently and be a useful retention metric) is a statistic many developers like to point to as an indication of the quality of their games. However, a free-to-play game may have users playing initially for 30-45 minutes because it is not well-balanced, not overly challenging, not sufficiently gated, or a myriad of other reasons … and then the player will not monetize or ever come back. Thus, in many cases it is a misleading statistic that diverts resources from areas that would improve performance.
Act like Billy Beane
To succeed, you must devote your resources to where they will have the biggest impact on increasing LTV and decreasing CPI. If you focus on these metrics, you can compete with bigger companies that may get enticed by the sexiness of metrics that are not as important.
The Business of Social Games and Casino 