One of the biggest complaints I hear about customer-centric strategy is that it’s too complicated. Sure, companies acknowledge that their customer base is heterogeneous. They understand that they have “good” customers (willing to pay premium prices, renew contracts annually and require low-service costs) and “bad” customers (tie up firm resources while giving little back). But it seems daunting to figure out who’s who and to use this information in day-to-day operations.
Take the issue of calculating Customer Lifetime Value, or CLV. I define CLV as the present value of the future net cash flows associated with a particular customer. Most companies’ CLV calculations lump together their entire customer base and hope that differences will come out in the wash.
It’s easiest to understand the principle in a contractual setting—say, a mobile-phone plan. Traditional CLV paradigm assumes that each customer “flips a coin” to make a renewal decision, and will renew with some fixed probability equal to the average retention rate. Say 100 people join a mobile-phone plan with the overall retention rate of 85 percent. Using this logic, at the start of the second year 85 people will renew, then 72 in year three and 62 a year later.
However, this gently declining survival curve in no way resembles reality. In the real world, there is a steeper drop-off of customers in the first year or two followed by a flattening out.
Nearly every company explains this pattern by asserting that loyalty increases over time. And while for many companies this is true, it’s rarely correct. A more valid (less intuitive) explanation is that these “churn dynamics” are merely an artifact of heterogeneity—with absolutely no changes in loyalty over time for any single customer.
In contrast, imagine different coins, each associated with a different customer segment. Some customers flip a coin that is very renewal-prone, and some flip a coin that is heavily weighted toward nonrenewal. Over time, more of the customers with the churn-weighted coins would leave, leaving us with a much more homogeneous customer base composed predominantly of renewal-oriented customers.
It has everything to do with weeding out the bad customers.
The implications for companies are huge. Rather than trying to bribe departure-prone customers to stick around with discounts and other goodies, firms should spend their resources finding new loyalty-prone customers.
This doesn’t mean that companies need to drill down to the individual customer level when applying CLV calculations. On the contrary, even breaking the customer base into just two or three different segments can dramatically improve the accuracy of CLV measures. However granular you get with customer data, the lesson is clear: firms ignore customer heterogeneity at their own peril. When companies fail to properly apply segmentation to their CLV calculations, they vastly and systematically undervalue the collective CLV of their customer base.