Customer Churn – Go Beyond the ModelThe issue of customer churn is one of the most prevalent in business today. Attacking this problem typically begins with some analysis of the causes or indicators of churn. That is usually followed by the creation of a predictive model(s) to help determine which customers are at high risk of churning. These two outputs are then put into use as the basis of a larger customer retention strategy.

For some lines of business, macro approaches may be sufficient but most customer retention strategies have evolved to emphasize retention of high value customers, given the customer’s predicted Lifetime Value (LTV). Focusing on high value customers requires either an individualized approach (desirable) or the use of customer segments. Making use of the predictive churn model and customer LTV, a variety of retention strategies can be employed (targeted email, discounts, courtesy call, etc.) to an individual or segment. However a big opportunity to refine and improve these tasks is often overlooked – feedback into the customer retention strategy.

Which retention or “customer save” action has the greatest impact on a segment of customers? Which retention action not only “saves” the customer but leads to re-establishing loyal behaviour in the future? Is there an action which lowers a customer churn risk or raises their predicted LTV? Using cohort analysis, which is the optimal of a variety of “customer save” programs? Which indicators of churn will respond best to a certain offer?

Once the predictive model is put into use, focus tends to shift away from the data science tasks to the marketing and CRM focused acts of customer retention. The role of data science in the customer retention strategy shouldn’t end with the creation of a churn prediction model. The business value of applying data science increases after the predictive model is in place – the heavy lifting has already been done. The insights gained at this stage of the retention program can directly impact revenue and are a valuable resource to optimize the overall customer retention program.

Related Post

Leave a Comments