Calculating the Worth of Your Best Customers

With all the time, money and effort spent on identifying best customers, surely such an animal must exist. Or is it time to step back and take another look at how we define our quarry? Maybe our “best customer” search is just a unicorn hunt.

Most often, the best customer concept is articulated in terms of a cross-sectional comparison; i.e., if we looked at all customers today and ranked them by revenue, the most valuable customers might be defined as “best.” Of course, we might be more sophisticated in our value calculation by considering not only revenue, but product ownership, service costs and other factors.

For the moment, let’s stick with a simple value calculation based on revenue. How do we handle the following?

Customer A has been a customer for two years, with monthly revenue of $50 and total revenue of $1,200. Customer B has tenure of 12 months and has spent $100/month for a total of $1,200. Are they equally valuable? Are they both qualified to be best customers?

Based on this information, it would be safe to select Customer B over Customer A for the Best Customer award. For example, if we normalize the data to account for average spending over a period of time (say, the past twelve months), then B is twice as valuable.

But, what if we have an attrition model that forecasts a low risk of attrition for Customer A, but a much higher one for Customer B? This suggests that B’s value may go to zero in the coming months, while A will continue to produce revenue. What if Customer B uses the call center more often, and therefore has higher servicing costs?

There are many other attributes and assumptions that we could add, each which may favor one customer over another. This is an area where any simple definition of “best” is usually inadequate. A cross-sectional analysis ignores important time-based components of value, but adding additional dimensions, such as time, estimates of future value, services costs, and others may quickly introduce so much complexity that no clear definition emerges.

One reason companies often invest so much in identifying best customers is that they wish to “clone” such customers. Finding and attracting more of them may be even more daunting than creating the original definition. Once we try to recruit more of them, we may find that there is great variation between such customers and persuading them to switch to our brand may be expensive.

The real answer may be that we need to view the customer base more as a portfolio, much as we might view a financial portfolio. In this case, Customer A might be equivalent to a bond—providing lower, but dependable, yields over time—and Customer B might be viewed as a stock—more valuable recently, but also more volatile. When it comes to acquisition, we need to make sure that our portfolio is properly weighted, and we should avoid getting too concentrated.

Of course, that’s a less sexy story to present to executive management. But that same management—looking to profile their customers predictably and profitably—is growing dissatisfied with the hunt for that mythical beast, the best customer.

David King is chief executive officer of Fulcrum.