Predicting Wine Club Member Behavior

(Wines&Vines) - If you have a good idea of who your average wine club member is (age, gender and buying habits), then you’ll have a better chance of convincing him not to leave your club. That was the key message from direct-to-consumer sales experts at a forum Thursday evening at Sonoma State University. Understanding more about your existing customers can also help you target new wine buyers as well. Dr. Steven Cuellar, an economics professor who studies consumer behavior in the wine industry, said he’s developed an analytical model that he thinks the industry could use to understand club behavior and eventually optimize member lifetime and value. He said trying to get a statistical handle on wine club attrition is exceptionally challenging because of the nature of the data. One issue is that new members and existing members have to be “censored” from the data, leaving only those who drop out of clubs. “Think of it as if you’re a doctor, and the only people you observe are those who have died,” he said. Using ‘survival analytics’ for wine clubs Consequently, Cuellar said he turned to very specific type of statistical analysis used by the medical field and known as survival analytics. The approach is used to evaluate the effectiveness of treatment by tracking the survival rates of a certain group of people all diagnosed with the same disease. Using a similar approach for wine clubs, Cuellar monitored new club members over time for a handful of small wineries. He then pegged the probability of these members remaining in the club versus the length of the club membership in months.