Customer Lifetime and Value Analytics

Citation metadata

Date: 2013
Document Type: Article
Pages: 9
Content Level: (Level 5)

Document controls

Main content

Full Text: 
Page [23]

Customer Lifetime and Value Analytics

The Value of the Customer

Much of the conventional wisdom and guidance for an organization’s relationship with its customers is largely linked to a general recommendation that all customers are of equal value and importance. We are bombarded with aphorisms about customer relationships and interactions: “the customer is always right,” “the customer is king,” or “it’s about the customer, always.” The common themes seem to convey the idea that the goal of any business is to consistently and continuously ensure that every single customer is completely satisfied. And to some extent, there is some wisdom in suggesting that customers be treated well, since a business cannot survive without customers.

In reality, though, not every customer is the same, nor is each customer equally valued. A quick web search about the distribution of profitability at banks yielded some (unscientific) results that in general the Pareto principle (aka the “80/20” rule) generally held: 20% of the customer base accounts for 80% of the profit. In one case study, the top 16% of the customer base accounted for 105% of the profit and that the bottom 28% of the customer base actually accounted for –22% or an effective loss. 1 One can interpret this factoid to draw two interesting conclusions. First, approximately one-sixth of the customer base accounted for all of the bank’s profit; and second, more than one out of every five customers accounted for a loss to the bank.

This relative inequity of profitability can be seen as being somewhat troubling. At the worst, it suggests that a significant corporate effort is expended on serving low-profitable or unprofitable customers, and that certainly raises a bunch of questions about customer relationships, customer engagement, and ultimately, the relationship between customer value and customer centricity. In this chapter, we explore some of these issues and questions in greater detail, as well as the corresponding ramifications. These considerations lend some weight to the use of data, thoughtfulness, analytics, and insight to drive a cohesive customer centricity model.

1 See Hughes, “How Banks Use Profitability Analysis” http://www.dbmarketing.com/articles/Art195.htm.

Page 24  |  Top of Article

Defining Customer Value

What is the value of a customer? Clearly, different types of customers have different kinds of value; and the bank case study shows different types of customers account for different levels of profitability, and consequently, value. In addition, different customers have different values over both the lifetime of the company as well as the lifetime of the customer. Attempting to quantify the value of a customer is somewhat complex, especially when organizations struggle to even define what a customer is. Fortunately, there are some strategies for analyzing customer characteristics and behaviors in relation to key dimensions of value that help in evolving a model for customer value:

  • Customer categorization: It is a process of organizing the customer community in terms of specific groups. In our bank case study, there is an implicit stratification of customers that includes specific levels like “best customers,” “ good customers,” and “unprofitable customers.” These are not arbitrary categories; rather they are defined with respect to some set of performance measures. What are ways to stratify or organize the different types of customers, and how are those levels defined? Are they keyed to revenue? Costs associated with customer support? Risk of customer attrition? Sometimes the first step in understanding ways to interact with your customers is to look at how they can be organized in relation to the value drivers for your business.
  • Differentiation: Presuming that there are discrete categories for customers, how do you classify each customer within those categories? This process blends the definition of customer categories with weighted measurement thresholds that can be used to distinguish customers at different levels. What are the business demographic characteristics that are common among your “best customers?” Examples of business demographics might include number of products purchased, monetary value of subscriptions, or the length of the business relationship.
  • Classification: As opposed to what we referred to as differentiation, which focused purely on the business value criteria for assigning a customer’s category, one might instead seek to determine what are the qualitative demographic characteristics that are common among customers within the same category, such as annual income, whether they own their own home, or educational attainment, to name a few. A combination of differentiation and classification helps in developing predictive models about the future of a customer relationship.
Page 25  |  Top of Article

These three processes together can be used to assess the customer community and to provide the basis for a customer value analysis.

Additional Aspects of Customer Value

The preceding activities shed light on some quantitative aspects of developing customer value models. In addition, there are some additional quantitative as well as qualitative aspects of managing a relationship with a customer.

One example is the cost of doing business with the customer. Many companies track the level of effort expended in customer acquisition, but do not continue to follow through in tracking the ongoing maintenance costs of the relationship. The answer to these questions factor into assessing the value of a customer:

  • What is the relative cost of operations associated with customer relationship management?
  • What are the average costs for customer acquisition, retention, as well as ongoing service and maintenance among all customers?
  • Can these costs be estimated or calculated on a customer-by-customer basis?

These can be factored with qualitative perceptions of the investments in engagement. In other words, what is the proper level of engagement and interaction with a customer? And at what point does the level of engagement overwhelm the value of customer retention? There is a good example of setting limits on customer engagement using a story involving Southwest Airlines former CEO Herb Kelleher’s response to a perennially complaining customer. This customer was responsible for a never-ending series of letters complaining about various aspects of the Southwest Airlines business model (such as the absence of first class, not assigning seats, and casual uniforms). At one point, this customer’s latest rant had been escalated to the top level of the company. Herb’s reaction was to immediately draw the line between customer satisfaction and retention: he wrote back: “We’ll miss you.”

Page 26  |  Top of Article

All of these factor into a different take on customer centricity that goes beyond the conventional expectation of the always-right customer. Instead, a model for customer centricity considers the value of the customer over the duration of the customer’s relationship and enables continuous refinement of the level of engagement so as to maximize “profitability” while minimizing costs and risks. And customer value analytics will prove to be a key variable in this refinement.

Evaluating the Value of a Customer

There are different kinds of customers who provide different kinds of value at different times during the lifetime of the customer relationship. But from an objective perspective, it is valuable to develop criteria for customer valuation that are relevant within the corporate business context. These criteria can provide quantifiable measures for characterizing customer types, enable the development of customer analytics models, and help in crafting a set of customer engagement strategies that maximize the different aspects of profitability.

In a perfect model, this could be characterized as an optimization problem by enumerating a collection of variables whose positive values we would seek to maximize or whose negative values we’d seek to minimize. We can attempt to work out this enumeration from the perspective of the most desirable optimized outcome by looking at the key dimensions of customer value and profitability. This would specify different variables to be applied to each customer, or potentially to a collection of customers, including:

  • Tangible aspects such as increased revenue and decreased costs.
  • Material, yet somewhat fuzzy aspects such as minimizing corporate risk.
  • Practical aspects such as minimizing effort of engagement or elongating the duration of the customer relationship.
  • Somewhat intangible beneficial aspects such as increased goodwill and word-of-mouth positive publicity.
Page 27  |  Top of Article

Another way of saying this is that the value of a customer is a function of variables such as the examples given in Table 4.1 .

The general wisdom is to seek to tweak your corporate business processes for engaging the customer so as to optimize across the full palette of these value variables. Of course, attempting to only concentrate on one or a few of these aspects may impact maximizing the overall cumulative benefit.

For example, increasing the duration of the customer relationship (i.e., increasing the “customer lifetime”) may be beneficial if the customer type is one that has a predictable revenue stream across the full duration of that time horizon. But if the cost of elongating the relationship increases the level of effort and costs associated with retention (i.e., increased “retention costs”), the anticipated benefit of the revenue stream may be offset by the increased costs. As another example, decreasing the investment in maintaining a customer may result in lowered overall costs, but may increase the probability of attrition, thereby reducing (or really eliminating) the predictability of the future revenue stream.


Table 4.1 Sample Variables for Customer Valuation

Table 4.1

Table 4.1 Sample Variables for Customer Valuation
Variable Measure Description
Revenue stream Revenue The net present value of the customer’s revenue stream over various time horizons
Customer lifetime Time The duration of the customer relationship
Maintenance cost Cost The costs necessary to ensure the realization of the customer’s future revenue stream
Risk cost Cost The quantified cost of risk attributed to the customer
Acquisition cost Cost The costs and the level of effort associated with acquiring a new customer
Maintenance cost Cost The costs and the level of effort associated with maintaining the customer
Retention cost Cost The costs and the level of effort associated with retention
Endorsement value Revenue A valuation of the contribution to the revenue streams resulting from the customer’s endorsement across the customer’s sphere of influence
Model refinement Revenue A valuation of the contribution to the revenue streams resulting from the customer’s contributions to customer profile models
Page 28  |  Top of Article

Developing a Customer Valuation Model

From a practical perspective, the customer valuation model should incorporate any variable that contributes to some measurable aspect of corporate value (as discussed in Chapter 2). Some of those given in Table 4.1 seem obvious, such as the net present value of the customer’s future revenue stream. On the other hand, some have correlations that are more complex to envision, such as the revenue value of improving or enhancing the customer profile model. It is worth allocating a brainstorming session to identify those variables that can possibly contribute to corporate value and populate a table like the one provided here by taking these steps:

  1. Begin with an enumeration of the high-level dimensions of value.
  2. Suggest variables associated with customer engagement and touch points that potentially contribute to any of those value dimensions.
  3. Specify a measure of value for each of the variables.

Once that table is populated, the next step is to consider factors for weighting the measures as they contribute to the calculation of customer value. For example, if there are costs relating to customer risks, the analysts executing this refinement task would seek to ascertain which customer characteristics are relevant in identifying the key customer risks and then assessing the costs that relate to those customer risks. Some examples include investigating the relationship between the customer’s credit score or the number of times the customer has a missed or late payment and the customer’s overall propensity to pay on time. These identified customer characteristics coupled with the measure calculation become a part of the valuation model.

This highlights the fact that it would be a challenge to presume that a customer valuation model can be created in the absence of any history of the results of the various customer relationships and experiences. The implication is that the developed model is going to rely on analysis of customer engagement histories involving transactions and interactions at any touch point, as well as relationships between customers and products/services and between customers and other individuals. On the other hand, it also suggests that there are some key variables that contribute more significantly to the valuation than others. This implies that you can suggest a baseline model whose variables and weightings can be refined over time.

Page 29  |  Top of Article

Using the Customer Valuation Model for Customer Centricity

A customer lifetime valuation model not only provides a means of quantifying the value of the relationship with the customer, but also enables a number of ways of implementing customer centricity, both at the different engagement points along the customer engagement life cycle and at touch points within specific business processes. Operationally, this means embedding decision processes within business processes that are driven by the integration of customer profiling and customer lifetime value models. The objective is to positively influence customer behavior in ways whose outcomes are optimized across the collection of value metrics.

Here is a simple example: leveraging offers to increase airline loyalty, revenue, and customer experience. Frequent travelers may opt to allocate their airline travel “spend” across multiple carriers. At the same time, airlines offer an effective “future rebate” in the form of a “restricted currency”—airline frequent flyer miles. These frequent flyer miles are units of value that are allocated by the flight carrier to the traveler as their future rebate. These frequent flyer miles can be accumulated and exchanged for free travel, as well as premier statuses that provide specific benefits such as greater flexibility in booking frequent flyer trips, premier-level customer service, more comfortable seating, or upgrades to higher level of travel accommodations.

A specific airline may analyze their customer profiles to assess the airline’s “wallet share,” or the percentage of the individual’s total travel spend that has been allocated to the airline. This type of analysis requires the modeling of the different customer flyer categories (such as “vacation traveler,”“ intermittent business traveler,” all the way to “very frequent business traveler,” for example), followed by developing methods or ways of determining which individuals fall into the named categories. Third, the collection of customers must be classified according to those differentiation rules into the defined categories.

This categorization and classification can then be used to identify those customers that are likely to be traveling a lot but are not using the airline as much as potentially possible. This is the determination of the airline’s wallet share. The determination that there is an opportunity to capture a greater share of wallet creates a scenario for influencing customer behavior to decide to purchase from that airline instead of any other alternative. Theoretically, influencing purchase decisions has cumulative impacts to customer lifetime value:

Page 30  |  Top of Article
  • Improved loyalty benefits improve the customer experience.
  • Improved customer experience strengthens the customer relationship.
  • A strong customer relationship leads to increased purchases.
  • Increased purchases continue to contribute to increased loyalty benefits for the customer.
  • Increased loyalty leads to an elongated customer lifetime.

In other words, triggering an improvement in the customer’s loyalty benefits leads to improved customer experience, increased revenues, increased wallet share, and longer lifetime relationships. To that end, when it becomes apparent that a target customer is a candidate for this type of influence, the airline might trigger the cycle by offering incentives for increasing the rate or volume of accumulating frequent flyer miles that ratchet the traveler into a premier status level.

Considerations: Influencing Customer Behavior

Customer-centric relationship management processes can blend information from customer profiles with the lifetime value model to influence behaviors in a number of ways:

  • Migration into premier status levels: We can generalize the example above to any scenario in which loyalty benefits are used as leverage for extending the customer relationship. The decision point is influenced by offers that ease the process of increased benefits. Aside from loyalty programs, other examples include various tiered special “member discounts” provided to shopping club members willing to pay for increased membership levels or ratcheting up a level when using an affinity credit card.
  • Pumping the customers relationship network: Providing incentives to a customer to encourage recommending your company’s products or services to his/her friends.
  • VIP service: Using the customer’s lifetime value and profile to assign level of effort applied to customer touch point activities. For example, best customers have minimal call center wait times, while undesirable customers are pushed further back on the wait queue. Page 31  |  Top of Article
  • Leveling-upin profile: Identifying “good” customers who have the potential to be turned into “great” customers and providing incentives for the behaviors that would ratchet the customer into a more profitable segment from a lifetime value perspective.
  • Proactive retention management: Identify scenarios in which providing incentives for continuous reengagement of good customers as a way of elongating the relationship, as well as reducing the benefits provided to undesirable customer to encourage their attrition.

Each of these opportunities for influencing customer behavior must be weighed in terms of the overall improvement in all relevant aspects of value over the anticipated customer lifetime. It means recognizing that sometimes what appears to be optimal for one performance measure in the short term may not necessarily be the best decision for other performance measures for the long term. For example, lowering the price of one product to increase overall number of products sold may reduce the risk of attrition, lead to a reduction in customer credit risk, or even increase product profitability by clearing out the inventory more quickly. Providing a full refund for a returned item may reduce revenue as well as incur an immediate cost of processing, but the goodwill generated may increase positive word-of-mouth, thereby increasing the network value of the customer.

In all of these cases, decisions that may seem to decrease immediate value may lead to increased lifetime values for a broad range of customers. Managing the decision processes with this perspective should increase overall profitability over the long term.

Using Information to Develop a Culture of Customer Centricity. DOI: http://dx.doi.org/10.1016/B978-0-12-410543-0.00004-4
Copyright © 2013 Elsevier Inc. All rights reserved.

Source Citation

Source Citation   (MLA 8th Edition)
Loshin, David, and Abie Reifer. "Customer Lifetime and Value Analytics." Using Information to Develop a Culture of Customer Centricity: Customer Centricity, Analytics, and Information Utilization, Morgan Kaufmann, 2013, pp. [23]-31. Gale Ebooks, https%3A%2F%2Flink.gale.com%2Fapps%2Fdoc%2FCX6994800011%2FGVRL%3Fu%3Dmnarasmuss%26sid%3DGVRL%26xid%3De86fe100. Accessed 23 Sept. 2019.

Gale Document Number: GALE|CX6994800011