THE CUSTOMER PYRAMID: CREATING AND SERVING PROFITABLE CUSTOMERS
Zeithaml, Valarie A.; Rust, Roland T.; Lemon, Katherine N.
California Management Review Summer2001, Vol. 43 Issue 4, p118
Innovative service companies today recognize that they can supercharge profits by acknowledging that different groups of customers vary widely in their behavior, desires, and responsiveness to marketing. Federal Express Corporation, for example, has revolutionized its marketing philosophy by categorizing its business customers internally as the good, the bad, and the ugly--based on their profitability. Rather than marketing to all customers in a similar manner, the company now puts its efforts into the good, tries to move the bad to the good, and discourages the ugly.(n1) Similarly, the customer service center at First Union, the sixth-largest bank in the U.S., codes customers by color squares on computer screens using a database technology known as "Einstein." Green customers are profitable and receive extra customer service support while red customers lose money for the bank and are not granted special privileges such as waivers for bounced checks. Providing different service to customers depending on their profitability is becoming an effective and profitable service strategy for firms like FedEx, U.S. West, First Union, Hallmark, GE Capital, Bank of America, and The Limited.
These firms have discovered that they need not serve all customers equally well--many customers are too costly to do business with and have little potential to become profitable, even in the long term. While companies may want to treat all customers with superior service, they find it is neither practical nor profitable to meet (and certainly not to exceed) all customers' expectations. Further--and probably more objectionable to quality zealots--in most cases it is desirable for a firm to alienate or even "fire" at least some of its customers. While quality advocates may be offended by the notion of serving any customer in less than the best possible way, in many situations both the company and its customers obtain better value.
Understanding the needs of customers at different levels of profitability, and adjusting service based on those differences, is more critical to the enterprise than has been previously held. Specifically, in examining customers by profitability--and understanding the key elements of the costs and revenues aspects of the profit equation--it is possible to actually increase the current and future profitability of all customers in the firm's customer portfolio. The Customer Pyramid is a tool that enables the firm to utilize differences in customer profitability to manage for increased customer profitability. Firms can utilize this tool to strengthen the link between service quality and profitability as well as determine optimal allocation of scarce resources. Companies can develop customized products and services that are more closely aligned with individual customer's underlying utility functions, thereby enabling the firm to capture more value from levels of customers, resulting in higher overall customer profitability.
Beyond a General Relationship between Service Quality and Profitability
Prior to the 1990s, the general link between service quality and profitability was still being questioned, but since the early 1990s, it has been persuasively established.(n2) The evidence to support the linkage came from a variety of sources and is now convincing enough to lead executives to believe that a positive relationship does exist. The link was first established through industry-wide, cross-industry, or cross-facility studies such as the PIMS (Profit Impact of Market Strategy) project, which demonstrated a correlation between quality and profits across both manufacturing companies and service companies.(n3) In more recent studies, quality improvement and customer satisfaction have been linked to stock price shifts, the market value of the firm, and overall corporate performance.(n4)
Because firms are managed at the individual level and not the industry level, executives still clamored for evidence that improved service quality resulted in increased firm profitability. A growing number of studies bear this out, showing that:
• service improvement efforts produce increased levels of customer satisfaction at the process or attribute level,(n5)
• increased customer satisfaction at the process or attribute level leads to increased overall customer satisfaction,(n6)
• higher overall service quality or customer satisfaction leads to increased behavioral intentions, such as greater repurchase intention,(n7)
• increased behavioral intentions lead to behavioral impact, including repurchase or customer retention, positive word-of-mouth and increased usage,(n8) and
• behavioral impact then leads to improved profitability and other financial outcomes.(n9)
What is still missing in this research evidence is the recognition that the link between service quality and profitability can be stronger if it is recognized that some customers are more profitable than others. Service investments across all customer groups will not yield similar returns and are not equally advantageous to the firm. Different profitability segments are likely to be sensitive to different service emphases and are likely to deserve different levels of resources. As a small number of progressive companies have discovered, they can become more profitable by acknowledging the difference in profit potential among customer segments, then developing tailored approaches to serving them.
The Limits of Traditional Segmentation
The idea of identifying homogenous groups of customers, assessing these segments for size and responsiveness, and then more precisely creating offerings and marketing mixes to satisfy them is not new. Traditional segmentation is most effective when it leads to more precise targeting that results in higher revenues or responsiveness to marketing programs. However, traditional segmentation is not typically grounded in knowledge of the different profitability of segments.
To build and improve upon traditional segmentation, businesses have been trying to identify segments--or, more appropriately, profitability tiers of customers--that differ in current and/or future profitability to a firm. This approach goes beyond usage segmentation because it tracks costs and revenues for groups of customers, thereby capturing their financial worth to companies. After identifying profitability tiers, the firm offers products, services, and service levels in line with the identified tiers. The approach has to date been effectively used predominantly in financial services, retail firms, and business-to-business firms because of both the amounts of data existing in those firms and the ability to associate data with individual customers.
One example of an innovator in the field is Bank One, which recognized that financial institutions were grossly overcharging their best customers to subsidize others who were not paying their keep. Determined to grow its top-profit customers, who were vulnerable because they were being under-served, the Bank implemented a set of measures to focus resources on their most productive use. The company used the data resulting from the measures to identify the profit drivers in this top segment and stabilized their relationships with key customers.(n10)
In another example, First Commerce Corporation knew that customer segmentation could improve the effectiveness of all of its operations. After dividing clients into mutually exclusive groups of individuals based on demographics, the company then identified the reasons for profitability swings (including balances, product mix, and transaction behavior). The firm then defined three unique segments: the smart money segment, the small business segment, and the convenience segment. Tailoring its marketing efforts differentially to those segments made the company's programs far more effective.(n11)
Conditions Necessary for Customer Tiers: An Empirical Example
In our view, four conditions are necessary for customer tiers to be used in a company.
• Tiers have different and identifiable profiles. Profitability differences in customer tiers are most useful when other variables can identify the tiers. As with customer segmentation, it is necessary to find ways in which customers vary across tiers, especially in terms of demographic characteristics. These descriptions can help understand the tier's customers and identify appropriate marketing activities.
• Customers in different tiers view service quality differently. Customers in different tiers can also have different needs, wants, perceptions, and experiences. Understanding the factors that affect the customer's decision to purchase a new product or service from an existing provider as well as the factors that affect the decision to increase the volume of purchases from an existing provider are crucial for managing customers for profitability. If customers in different tiers have different expectations or perceptions of service quality, these differences will allow the company to offer different groups of attributes to the tiers.
• Different factors drive incidence and volume of new business across tiers. Differences in characteristics, needs, wants, and definition of service quality are likely to result in different drivers for the incidence and volume of new business. If this condition is met, a company can target customers that are likely to end up in higher tiers.
• The profitability impact of improving service quality varies greatly in different customer tiers. Just as direct marketers routinely qualify lists to test for potential profitability, companies need to qualify their customer tiers for potential profitability. If customer tiers are appropriate, the way customers respond to service and marketing should differ among tiers. Higher tiers should produce a much higher response to improvements in service quality that will be evident in increases in new business, volume of business and average profit per customer. Taken together, the disproportionately greater response to changes in service quality in each of these areas will result in an overall greater return on service quality improvements for the higher tiers of customers.
An Empirical Test of the Conditions in a Two-Tier Situation
Virtually all firms are aware at some level that their customers differ in profitability, in particular that a minority of their customers accounts for the highest proportion of sales or profit. This has often been called the "80/20 rule"--twenty percent of customers produce eighty percent of sales or value to the company. We recently conducted an empirical study to examine this simple "80-20" scheme.
A major U.S. bank provided profitability information about retail products and customer information files with descriptive information including average account balance, average profit from account, and average age, gender, and income. Data on a random sample of 796 of these customers were merged with responses to a service quality survey from the same set of customers. Eight months later, information regarding the amount of new business, including both the incidence and volume of new business (revenue from new accounts), was added to the data file by examining behavior following the survey. In this way, service quality measures could be used to predict future behavior using a cross-sectional, time-series approach.
Demographic Differences
We examined differences in customer descriptive statistics, service quality perceptions, drivers of incidence of new business, and drivers of volume of new business across tiers using various statistical analyses.(n12) We also projected both the increase in the percentage of customers who would open a new account and the increase in the average account balance. Multiplying the projected average account balance by the average profit per account balance for each tier yielded an estimate for the projected increase in average profit per account. Multiplying that by the number of accounts yielded the total projected new profits from each tier.
We then divided the customer base into two customer tiers: the most profitable 20% (top 20%) and the least profitable 80% (lowest 20%). The results met all the conditions described above. First, customers in different profitability tiers had different customer characteristics. The top tier had a higher percentage of women than the lower tier, an average account balance about five times as big, and average profit about 18 times as much. The top 20% was also older than the lowest 20%, had more upper-income customers, and had far fewer lower-income customers. The top 20% produced more profit per volume of business, with an average profit per account balance of 2.53%, versus 0.71% for the lowest 20%. Finally, the top 20% produced 82% of the bank's retail profits, an almost perfect confirmation of the 80/20 rule in this profit setting.
Views of Service Quality
Second, customers in different tiers viewed quality differently. The top 20% viewed service quality in terms of three factors: attitude, reliability, and speed. By contrast, the lower 20% had a less sophisticated view of service quality, viewing service as only two factors, attitude and speed, with slightly different interpretations of the factors. The reliability factor was not a driver for the lowest 20%. A particularly compelling finding emerged from these data. When we combined all customers into a single group, all appear to want the same factors and the factors meant the same thing to both groups. The important insight here is that blending customer tiers resulted in an imprecise view of what service quality meant to the customer base.
Drivers of Incidence and Volume of New Business
Third, we found that different tiers had different drivers of incidence and volume of new business. Since we measured what customers did after they reported what was important to them, we captured what actually drove customers to make purchases, rather than what they thought would make them do so. For the top 20%, speed was key to driving incidence of new business whereas attitude was the key driver for the lower tier. As before, analyzing the entire customer base as a single group would have been misleading. Both the combined attitude/ reliability factor and the speed factor were key drivers for the group as a whole, but the combined analysis would not reveal the fact that different strategies should be used for different profitability levels.
Profitability Impact
Finally, the profitability impact of improving service quality varied greatly in different customer tiers. An across-the-board service quality improvement of the key drivers (approximated by a 0.1 increase in average satisfaction with each driver in each tier) resulted in a projected 3.65% increase in incidence of new accounts in the top 20%, but only a 2.00% increase in the Lower tier. This result suggests that the Top 20% was almost twice as responsive to the changes in service quality than the lowest 20%. When examining the projected increase in average account balance, the results were even more encouraging. The projected increase in average account balance was $6.19 in the top 20%, but a meager $0.69 in the Lower tier. Here, the Top 20% appeared to be almost 10 times as responsive to changes in service quality. Finally, the projected increase in average profit per customer was 15.7 cents in the top 20%, but 0.5 cents in the lowest 20%. Again, the top 20% provided a substantially greater return on the service quality improvement. Of particular interest was that simultaneous improvement of the key drivers for both tiers produced a projected 89% of the new profits in the top 20%, while only 11% of the new profits could be attributed to the lower 20%. This was an even higher percentage than the current percentage.