RWRetailWire
Back to Discussions
DISCUSSION

Can customer lifetime value scores work against retailers?

Written by Tom Ryan

Photo: RetailWire

A recent Wall Street Journal article — “On Hold for 45 Minutes? It Might Be Your Secret Customer Score” — pointed to some of the benefits and downsides that consumers may receive from the use of customer lifetime value (CLV) scoring by retailers. While the headline focused on the negatives for consumers associated with the use of the metric, retailers may pay an even bigger price if CLV is not put to its proper use.

Among the challenges: 

  • Transparency: Unlike credit scores, CLV scores aren’t accessible by consumers and aren’t monitored by any government agency. Many propriety formulas are not disclosed for competitive reasons.
  • Accuracy: Some CLV measurements are based only on transactional data. Others take in online behavior, customer-service interactions, social-media profiles, geolocation data, demographics and other factors. The myriad of inputs being used by various firms can lead to inaccurate scores.
  • Inequality: A customer scored as “high-value” may not receive a discount as quickly as the average customer because they’re expected to remain loyal shoppers, but they generally receive better perks and service than perceived “low-value” customers.

In an interview with RetailWire, Peter Fader, a marketing professor at the University of Pennsylvania’s Wharton School and a CLV pioneer, agreed with many of the concerns in the article and called for greater transparency.

As with credit scores, giving consumers knowledge of CLV scores could incentivize better shopper behavior. Moreover, similar to how an actuary can estimate when a person is going to die based on an average individual with similar characteristics, marketers might lessen the “creepiness” of CLV by disclosing their calculations were disclosed, Prof. Fader believes.

Standards should be established and be based on transactions or RFM segmentation (Recency, Frequency, Monetary Value), according to Prof Fader. Demographics and other behavioral data “really hurts the accuracy” of CLV scores, but can be used in a second step to support profiling efforts.

Prof. Fader doesn’t believe, however, that CLV should be used to “treat anyone badly.” A baseline of “reasonable standards” should be established for low to mid-tier customers. Then, CLV can be tapped to “treat the better customers better.”

Said Prof. Fader, “It’s playing defense with the rank and file and then playing offense with those at the high end.”

Discussion Thread0