I am NOT a baseball fan, which is sort of a lie. Rather I’m not a fan of watching baseball, but I’m a fan of the statistics. I’ve been that way since I was a kid. I could not wait to get my hands on the Sunday paper and analyze all the stats they were willing to publish.
The statistics, and more importantly the predictability of them, have been greatly improved since Bill James and others founded the Society for Baseball Research (SABR) in 1971. James and others sabermetricians changed the game of baseball by collecting, summarizing and analyzing in-game data to address specific questions. In other words, they applied the scientific principle to baseball and it payed big dividends.
Without going into great detail, there are plenty of resources out there for that, it’s important to note that this measurement and analysis strategy has completely changed the way that baseball organizations are managed and the way the game is played.
Sabermetrics is big business and there are more measures added seemingly every year. Yet no matter how complex the analyses get, they still follow the same principals, which can be applied to your business.
1) You need to create and test hypotheses, not just mine data and hope to find something. Be specific about what you are trying to predict and be bold about using different measures to predict that outcome.
2) Not all events are created equal. For instance baseball used to only track if a player got a base hit. It didn’t matter if the hit was a single or a home run. It makes perfect sense now, but statisticians kept equal weighting to compute batting average for well over a 100 years. You should not value all actions of customers the same. Some are truly more important than others. You should talk to customer to identify what matters most to them to weigh those experiences accordingly.
3) In many cases, like in sabermetrics, you may find that you already had been collecting the data to give you enough insight on the questions you have. However, you may need to transform some of this information to give you the ability to draw great insights.
4) If you don’t have good data, you will need to invest in a quality measurement system. However, before jumping out and grabbing more data, take a learned approach and determine what matters most.
5) Challenge the old guard. It’s likely that your company has already established the metrics that matter. These are likely outcome measures such as revenue, profitability, customer retention, etc. However, you need to find those leading and highly predictive indicators to really make a difference in customer's experiences.
6) Analysis is not good enough, to truly get results you need to implement your findings. It took Major League Baseball a long time (in some cases 30 plus years) to get organizations to accept the importance of their findings. Don’t be guilty of knowing what to do, but choosing to do nothing.