یادداشتی در تمایل به صرف و ارزش مادام العمر مشتری برای شرکت های با ظرفیت محدود

The paper draws a distinction between customer lifetime value (CLV) and willingness to spend (WTS). By WTS we mean the maximum amount the firm should be willing to spend to acquire (retain) the customer relationship. In order to avoid the double counting of cash flows when summing the CLVs of customers, we suggest including only direct cash flows in the formulation of CLV. This convention means that CLV will equal WTS if (and, for the most part, only if) the firm's relationships with customers are independent. By independent we mean that the acquisition (retention) of Jane Doe has no effect on the cash flows of any other current or future customers. In contrast to well-understood demand-side dependencies among customer relationships (such as referrals), this paper highlights a particular kind of supply-side dependency—that created when the firm is limited in the number of customers it can serve. Using an extended version of the model of Blattberg and Deighton (“Manage Marketing by the Customer Equity Test,” Harvard Business Review, July–August 1996, 136–144) of customer equity, we demonstrate that, for a firm at capacity (in this model), CLV is no longer relevant to marketing spending decisions and the firm can prefer a lower-CLV customer.

To the best of our knowledge, it was Bursk (1966) who introduced the concept now commonly referred to as customer lifetime value (CLV) with his suggestion that firms use the “investment value” of a customer to guide marketing spending decisions. Attention directed toward CLV helps shift focus from transactions (finding more buyers for the firm's products) to relationships (finding more ways to serve the firm's customers). Using CLV to guide marketing decisions also encourages firms to recognize differences among customers and begin to create value though differential treatment. For these and other reasons, the concept of CLV receives much attention from marketing practitioners and academics (e.g., Rust et al., 2000, Blattberg et al., 2001, Gupta et al., 2006 and Blattberg et al., 2009). For our purposes, CLV is defined as the present value of the future cash flows attributed to the customer relationship (Pfeifer, Haskins, and Conroy 2005). By design, this definition is flexible; it can be applied at any point in the firm's relationship with a customer. Thus, it makes sense to talk about the (remaining) lifetime value (or CLV) of an existing customer (and attempts by the firm to maximize that value) as well as the value of a newly acquired customer. Although the definition of CLV means it applies to both new and existing customers, to keep things simple and avoid confusion, we adopt the default assumption that (unless otherwise noted) CLV refers to the present value of customer cash flows if and when acquired. Notice also that the definition is silent with respect to what cash flows should be attributed to the customer relationship. There appears to be agreement, however, that the sum of the CLVs of the firm's current and future customers (net of acquisition costs) is a measure of the value of the firm (see, for example, Bayon et al., 2002, Berger et al., 2006, Gupta et al., 2006 and Rust et al., 2000). In order for CLVs to sum to something meaningful, there should be no double counting of cash flows. One simple way to avoid double counting is to attribute only direct cash flows in the formulation of CLV. We will adopt that convention throughout this paper. So, for example, if Jane Doe encourages her friends to spend more, the value of that extra spending should be included in the CLVs of the friends and not in the CLV of Jane Doe. Under this convention, the decision to try to acquire Jane Doe requires that the firm recognize that acquiring (retaining) Jane Doe will increase the CLVs of her friends. This is consistent with the approach taken by Kumar, Petersen, and Leone (2007) who treat “referral value” as distinct from CLV. It is also consistent with the Kumar et al. (2010) treatment of CLV as but one of four components of the overall value (referred to as “Customer Engagement Value”) of the customer to the firm.

This paper draws a distinction between CLV and WTS (the maximum amount the firm should be willing to spend to acquire [retain] the customer). Adopting the reasonable convention that only direct cash flows are included in the formulation of CLV (so as to avoid double counting when summing the CLVs of the firm's customers), then CLV will equal WTS if (and, for the most part, only if) customer relationships are independent. By independent, we mean that the acquisition (retention) of Jane Doe does not affect the cash flows from any other customer relationship. Note that this proposed convention for incorporating only direct cash flows in CLV is consistent with the Kumar, Petersen, and Leone (2007) treatment of referral value as distinct from CLV. But it is not consistent with the suggestion by Noone et al., 2003 and Blattberg et al., 2008 to include the value of referrals in the calculation of the referring customer's lifetime value. Referrals are a good example of a demand-side dependency. Acquiring Jane Doe affects the cash flows of other relationships if referrals from Jane Doe make it less costly for the firm to acquire (retain) the referred customers. In such a situation, the value of acquiring Jane Doe (the firm's WTS to acquire Jane Doe) is greater than her CLV. In addition to demand-side dependencies, there are supply-side dependencies. This paper identifies and takes initial steps in examining a particular and common supply-side dependency: the limited capacity to serve customers. If firms are limited in the number of customers they can serve, the decision to serve Jane Doe runs the risk of displacing some other customer in the future. In this situation, the firm's WTS to acquire (retain) Jane Doe will be something less than her CLV (in recognition of the fact that Jane Doe occupies a scarce and valuable unit of capacity). Conversely, the usual interpretation of CLV as the limit on acquisition (retention) spending rests on the (usually unstated) assumption that the firm has unlimited capacity to serve customers. To examine what CLV means for a firm with limited capacity, we combine several convenient assumptions about customer relationships to create the multi-period BD model. Although this model is rather stylized, it seems appropriate for an initial examination of marketing spending and CLV for a firm with limited capacity. Within the context of this model, we first show that if the capacity limit is infinite, then CLV is, indeed, the firm's WTS to acquire (retain) the customer. At the other extreme we examine the firm operating optimally at capacity and show that CLV is not directly relevant to acquisition and spending decisions. This happens because for the firm at capacity that will remain at capacity, the per-period margins received from customers are not at stake. What is at stake are the acquisition and retention dollars required to maintain the firm at capacity. Consequently, the firm at capacity in the multi-period BD model should balance acquisition and retention spending so that the marginal cost to acquire a new customer equals the marginal cost to retain an existing customer—regardless of the value of the customers. The multi-period BD model makes many convenient assumptions. In particular, all customers and prospect pools are identical. As a first step toward dealing with customer heterogeneity, this paper considers customer selection for a firm operating optimally at capacity. For that decision, it is the total PV from the unit of capacity that is the appropriate metric to use. Total PV includes both the CLV of the selected customer and the expected value of the unit of capacity freed up when the customer churns. At equal CLVs, the firm prefers the customer who frees up capacity sooner. It follows that the firm at capacity can prefer the lower CLV candidate. Before discussing the limitations of our model, let us position this work relative to existing research on the integration of customer relationship management (CRM) with revenue management (RM). At the risk of oversimplifying things, the two-by-two matrix in Table 6 distinguishes among CRM, RM, and more traditional (consumer) marketing management based on whether the firm is relationship or transaction focused and whether capacity is limited or unlimited. RM is (usually) practiced by transaction-focused firms changing prices and controlling inventory to make the most revenue (profit) from a limited supply of capacity (airline seats, hotel rooms, etc.). CRM is (usually) practiced by relationship-focused firms interested in acquiring customers with the intention of retaining them and profiting from the relationships. As pointed out throughout the paper, CRM is usually practiced in an environment where the capacity to serve customers is unlimited. The lower-right cell in the matrix corresponds to firms that are transaction focused with unlimited capacity. We label this as traditional (consumer) marketing management in recognition that most of traditional marketing is about positioning and pricing and promoting products (transaction-focused) under the assumption that production can keep up with the demand that marketing generates. The unlabeled, upper-left cell represents somewhat uncharted territory. As noted by Noone, Kimes, and Renaghan (2003), “the importance of integrating CRM and RM strategies has been noted by a number of authors (Belobaba, 2002, Dickinson, 2001, Jonas, 2001 and Lieberman, 2002); however, the implications of integration of the two strategies for hotel organizations have received little attention.” They go on to note that “the RM model is oblivious to the specific characteristics or value of the customer (Karadjov and Hornick 2000). While this approach maximizes the room revenue generated from a single transaction, it may not lead to optimal long-term gains.” Cross, Higbie, and Cross (2009) might refer to the upper-left cell as “customer-centric revenue management,” something they suggest will be part of the future of revenue management. Von Martens and Hilbert (2009) offer suggestions about how RM might account for long-term customer value when deciding which booking requests to accept. So there is agreement there are unexplored opportunities in the upper-left cell. In contrast to the papers mentioned above that approach the upper-left cell from below (asking how RM can be improved by accounting for and managing individual customer relationships), this paper approaches from the direction of CRM (see arrow). What we do is start with a well-established CRM model (the BD model) and ask how CRM principles change if the firm has limited capacity. We ask that question not only because of its relevance to managers, but also to explore the boundaries of the upper-left cell. The main insight is that the usual interpretation of CLV as the limit on acquisition (retention) spending is appropriate in situations where the firm has the capacity to serve all customers. So if our apartment building has several vacancies and will always have several vacancies, then customers and prospects are the scarce resource and the usual CRM advice applies. The firm should be willing to spend up to the candidate customer's CLV to acquire (retain) the customer. If our apartment building is at capacity and will always be at capacity, however, then prospects and customers are no longer the scarce resource—the units in the apartment building are. The firm should spend so as to get the most value from each apartment unit, which can sometimes mean preferring a lower CLV customer and never means being willing to spend up to CLV to acquire a new customer. It also means cutting retention spending (or raising rent) if it is less costly (at the margin) to acquire a new customer than to retain the existing one. The list of convenient assumptions of the multiple-period BD model provides a road map for future research designed to deliver more specific guidance for firms with limited capacity managing customer relationships. A companion paper (Ovchinnikov, Boulu, and Pfeifer 2011) takes some important steps in that direction by considering a dynamic programming model of a firm with limited capacity making sequential acquisition and retention spending decisions with heterogeneous customers in a customer retention situation in which customer relationships would otherwise be independent. The main finding is that optimal spending increases in capacity and decreases in number of customers. That paper goes on to consider a one-time decision to expand capacity. Its fundamental difference from this paper is that Ovchinnikov, Boulu, and Pfeifer 2011 considers the dynamics of spending and derives the optimal response of the forward-looking firm to stochastic realizations of acquisition and retention outcomes. As such that paper is more a decision-support tool, while this one offers conceptual insights. Both this and the companion paper consider customer retention situations. (Customers who are not retained are considered lost for good.) While this assumption is reasonable for apartments, season tickets, and college enrollment, there is a wider variety of situations (called customer migration situations) in which customers come and go and come back again. This paper also only considers marketing spending and customer selection (briefly) for the simplest of “hard” capacity constraints (on the total number of customers served). Alternative forms of “softer” constraints are of interest for future research.