A Measure of Customer Experience

Peter Kriss is a senior research scientist at Medallia and the Director of Research for Vision Prize has recorded some quantifiable evidence that the customer experience provides significant performance improvements for companies. He focused the research on two different companies with different revenue models.  One company was involved in a transactional-based business while the other was a subscription-based business.

The transactional-based business was focused primarily on how frequently the customers returned and how much was spent each time the customer returned. On the other hand, the subscription-based business focused on retention.  In his research his team controlled for other factors that might affect repeat purchasers to remove any bias by those factors.

His results were truly amazing as noted below:

1.    For the transaction-based business, customers who had the best past experiences spent 140% more than those who had the poorest past experience

2.    For the subscription-based business, members who gave one of the top two experience scores had a 74% chance of remaining a member for at least another year, whereas those who had the poorest experience had only 43% chance of being a member a year later.  The real difference was that members who gave the lowest scores were likely to remain a member for a little over year, compared to the members who gave the highest scores who were likely to remain a member for another six years.

 The bottom line is obvious because the numbers tell the story. Customers that rated their experience high were much more profitable and loyal.  While these two companies provide only anecdotal experiences, each presents a great case for considering the metric of customer experience as a way of measuring customer loyalty while providing increased profit.

Share of wallet may be the best loyalty yardstick

Tim Keiningham, global chief strategy officer for Ipsos Loyalty has developed an interesting statistics regarding loyalty.  Many companies rely on the NPS measurement to define their customer loyalty. However, Mr. Cunningham has some statistics that seem to provide some serious questions about the NPS measurement of loyalty. In effect, a report for a client with customers right in a five as loyal with the NPS criterion. However, more than 90% of the time the same customers had an NPS loyal score for another brand.

There is an interesting new book titled the wallet. How location rule which was co-written by Prof. Lerzan Aksoy and Mr. Luke Williams also of Ipsos.  In this new book the authors give an example regarding a retail grocery chain.  The authors note that while 53% of the customers gave high marks when asked would they would promote the store, only 43% listed the store as their primary shopping location. The authors point out that 57% of the customers with a solid satisfaction score still prefer competitors. It was further pointed out that customers often chose to shop elsewhere due to lower prices and one of the other stores had a location that was more convenient.

Rather than using the NPS score estimator of loyalty. The authors offer three performance indicators that may provide a better measure of loyalty. The authors suggest the following three metrics:

1.       1.  Measure the percentage of customers that list them as their first choice

2.       2   The average number of brands used by the customer

3.       3.   The share of the customer’s wallet captured by the company.

Dr. Keiningham references another study of a financial services firm that found 90% of the customers who would indicate they would promote a brand might not list it as their primary brand in the field. 

The bottom line is that customer loyalty identified as a 9 or 10 on the NPS scale may not accurately reflect loyalty according to the following definition.   Many companies believe loyalty means that their company will be considered first when an engagement opportunity occurs (the customer will spend his money with the company because they are loyal).  Based on these statistics, that may not be true.