Products & Services

Balancing Acquisition and Retention Resources to Maximize Customer Profitability

Balancing Acquisition and Retention Resources to Maximize Customer Profitability
of 18
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  Acquisition and Retention Resources / 63 Journal of Marketing  Vol.69 (January 2005),63–79 Werner Reinartz, Jacquelyn S. Thomas, & V. Kumar Balancing Acquisition and RetentionResources to Maximize CustomerProfitability In this research, the authors present a modeling framework for balancing resources between customer acquisitionefforts and customer retention efforts. The key question that the framework addresses is, “What is the customerprofitability maximizing balance?” In addition, they answer questions about how much marketing spending to allo-cate to customer acquisition and retention and how to distribute those allocations across communication channels. Werner J.Reinartz is Associate Professor of Marketing, INSEAD ( S.Thomas is Associate Profes-sor of Integrated Marketing Communications, Northwestern University ( is ING Chair Professor andExecutive Director, ING Center for Financial Services, School of Busi-ness, University of Connecticut ( authorsthank the JM  reviewers and the participants of the 2003 Marketing Sci-ence Conference for their valuable comments on a previous version of thearticle.They also thank the high-tech firm for providing access to the data,without which this study would not have been possible.All authors con-tributed equally. M easuring, managing, and maximizing customerprofitability is not an easy task. It requires that inresource allocation decisions, both the benefits andthe costs of marketing, sales, and customer interactions areconsidered. In this research, we conceptualize the market-ing resource allocation problem in terms of determininghow much to spend on customer acquisition and customerretention and how those expenditures are allocated. Giventhis conceptualization, the fundamental marketing resourceallocation question is, “What is the right balance of resources that optimizes customer profitability?”Prior research has examined parts of these issues, but todate, there has not been a comprehensive examination of marketing resource allocation that focuses on all three fol-lowing questions: How much? How? and What is the profit-optimizing balance? For example, Blattberg and Deighton(1996) address the question of how much to spend on cus-tomer acquisition and customer retention. However, theystop short of simultaneously considering acquisition andretention spending, which is critical to address the issue of balancing resources.Using Blattberg and Deighton’s (1996) framework,Berger and Nasr-Bechwati (2001) assume a budget amountand then suggest a model to address how that budget shouldbe allocated between acquisition and retention. However,their model is not tested empirically. In contrast, in thisresearch, we propose an integrated approach that sheds sta-tistical insight on this issue and thus goes beyond the deter-ministic approach that Berger and Nasr-Bechwati provide.Mantrala (2002) points out that Blattberg andDeighton’s (1996) approach is only a first step and thatthere is great scope for more research into customerprofitability–based decision modeling of the marketingresource allocation problem. However, this problem is notstraightforward, as Hanssens (2003, p. 16) highlights: The more challenging task is to assess long run marketingeffectiveness and to allocate the overall marketing budgetacross the key activities that generate customer equity....For any given set of business and customer response para-meters, there is an optimal level of customer acquisitionand retention which translates into optimal acquisition andretention spending levels. Prior models have begun to investigate these issues. Forexample, Blattberg, Getz, and Thomas (2001) incorporateacquisition, retention, and cross-buying into a model of cus-tomer lifetime value and customer equity but do not identifythe specific impact of marketing expenditures on customerprofitability. Thomas (2001) examines the link betweencustomer acquisition and customer duration. Reinartz andKumar (2000, 2003) examine the link between customerduration and customer profitability. Rust, Lemon, and Zeit-haml (2004) address both acquisition and retention aspects,but their model does not provide for separate or distinctinvestments in the acquisition of new customers and theretention of existing customers. Although Rust, Lemon, andZeithaml’s approach enables a trade-off analysis betweendifferent aspects of the marketing mix, it does not providefor an understanding of how to trade off specific invest-ments at different points in the customer–firm relationship.Bolton, Lemon, and Verhoef (2004) provide a conceptualmodel for linking marketing actions and expenditures tocustomer retention and profitability but do not provideempirical results. Focusing only on existing customers,Venkatesan and Kumar (2004) develop a resource allocationmodel that provides guidance on how much to invest in dis-tinct communication channels. By estimating the frequencyof buying and the change in the contribution margin fromone period to the next, they compute and seek to maximizethe future value of the firm’s existing customer base. Interms of the data similarities and the discussion of spending  64/ Journal of Marketing,January 2005 across communication channels, substantively our researchis similar to that of Venkatesan and Kumar. However, ourresearch takes a more longitudinal perspective and exam-ines resource allocations more comprehensively because itbegins before the successful acquisition of a customer. Wedepict a conceptual view of our perspective in Figure 1.A natural extension to the contributions made by theprevious studies is a conceptual framework and model thatcan be used to balance resource allocations between cus-tomer acquisition and retention and whose objective is tomaximize the firm’s long-term profitability. It is importantthat the model be flexible enough to address simultaneouslyhow much and how to invest. This study is that extension.Specifically, Figure 1 conceptually depicts the key pro-cesses in the evolution of the customer–firm relationship. Inaddition, our study presents a comprehensive system of equations that links the acquisition and retention processesto customer profitability and can be used for resource allo-cation decisions. Because of the linkage, the system can beused to assess the trade-offs that occur in resource alloca-tion decisions. It is important to acknowledge that ourapproach necessitates that resources can be split betweencustomer acquisition and retention, which is the case inmany business-to-business (B-to-B) or direct marketingcontexts.Based on the existing research, our specific objectivesin this article are to 1. Present a resource allocation model that addresses the ques-tions of how much to invest in customer relationships and 1 We use the terms “customer profitability” and “customer value(to the firm)” interchangeably in this research. Both expressionsrepresent a multiperiod measure of the economic value of a cus-how to invest at different points of the customer–firmrelationship;2. Illustrate the application of the statistical model with anempirical example; and3. Show by a simulation how varying different inputs to themodel (e.g., expenditures, number of communication con-tacts) affects acquisition rates, retention rates, customerprofitability, and the magnitude of the firm’s return oninvestment. The context in which we address the issue of balancingresource allocations is customer contact strategies. How-ever, the framework that we provide can extend beyond thecustomer contact strategy. With the rise of the Internet andelectronic technology, the question of how firms shouldinteract with their customers is gaining in importance, espe-cially as firms consider the cost differences between tradi-tional communications media, such as television and salesforces, and electronic media, such as the Web and e-mail. Allocating Resources to CustomerContact Channels To answer the questions of how much to spend on cus-tomers and how to allocate expenditures, we must under-stand the key drivers of customer profitability (for a review,see Berger et al. 2002). 1 Given our data, the specific drivers FIGURE 1Linking Customer Acquisition,Relationship Duration,and Customer Profitability Prospects Acquisition Process Retention Process Acquired customers Nonacquired customers Customer profitability Relationship duration •Firm actions •Customer actions •Competitor actions •Customer characteristics  Acquisition and Retention Resources / 65 tomer to the firm, expressed in contribution margin terms.Although the term “customer lifetime value” has been used abun-dantly in that context, we refrain from doing so. Conceptually,there may be reservations about using “customer lifetime value”because it implies complete knowledge (i.e., past and future) abouta customer’s value to the firm. We do not take such a viewpoint. that we focus on relate to customer contact channels. Theallocation of a budget to customers and across differentcontact channels is a classic problem that has gained height-ened attention in today’s multichannel environment. How-ever, typical media planning investigations have been con-ducted at the firm level (see, e.g., Aaker 1975; Rust 1986).According to Tellis (2003, p. 45), the use of the most disag-gregate measure—the individual customer—is probablymost appropriate for allocating media expenditures becausepersuasion is created at the individual level and becausemedia increasingly can be targeted at the individual level.Therefore, an individual-level investigation is a contributionof our study.At the most simple level, different contact channels mayhave independent effects on a specific dependent variable.Contact channels (e.g., personal selling, telephone, directmail, e-mail) have been characterized as more or less inter-personal. Personal selling, at one extreme of the communi-cations continuum, is dyadic in nature, offers the ability formessage customization, enables rich interaction, and allowsfor personal relationship building (Moriarty and Spekman1984; Stewart and Kamins 2002). Venkatesan and Kumar(2004) find that higher-level bidirectional communication isassociated with higher purchase frequencies. Prior researchalso asserts that if the buying environment can be describedas high involvement decision making (such as a B-to-B pur-chase), a more involving and interpersonal contact channel,such as a personal sales call, will have a much higher con-version rate on average than will a less involving contactchannel, such as e-mail or telesales (Anderson and Narus1999, p. 302). Therefore, the ability to customize the mes-sage easily and build personal bonds with customers willeventually lead to greater retention through personal selling,especially in B-to-B settings. In addition, research hasshown that buyers and sellers that have strong personal rela-tionships are more committed to maintaining their relation-ships than are less socially bonded partners (Mummalaneniand Wilson 1991).Extrapolating prior findings to this context suggests thatmore interpersonal channels will have a greater positiveimpact on customer acquisition (Mohr and Nevin 1990;Moriarty and Spekman 1984). Similarly, more interpersonalcontact channels will be associated with greater customerretention compared with less interpersonal contact chan-nels. However, there is little theory or rigorous testing tofocus our understanding about the efficacy of contact chan-nels with regard to customer profitability. Thus, a uniqueaspect of this research is that it addresses the issue of themarginal efficiency of contact channels with respect to cus-tomer acquisition and two longitudinal performance mea-sures: customer retention and long-term customerprofitability.An important consideration when investigating theimpact of communication modes is the potential interactioneffect between contact channels. The investigation of inter-action effects between different promotional vehicles iscomplex and rarely addressed by researchers (Sethuramanand Tellis 1991). According to Farris (2003), there is a needto develop resource allocation models that reflect mediasynergies and interactions. For example, Jagpal (1981) stud-ied radio and print advertising for a commercial bank andwas the first to present empirical evidence of synergy inmultimedia advertising. More recently, Naik and Raman(2003) find empirical evidence for the existence of syner-gistic effects between television and print media. In a hypo-thetical scenario, Berger and Nasr-Bechwati (2001) accountfor the possibility of media interaction effects in their deter-ministic model of customer equity. Yet so far, no customerprofitability approach has modeled empirically the interac-tion between the marketing-mix variables. In addition, theempirical tests have all been conducted at the aggregatelevel.In addition to the marginal effects, this study empiri-cally tests for the synergistic effect of multiple communica-tions media on individual consumers’acquisition, retention,and profitability. Investigating the impact of media interac-tion on the allocation decision for each of the dimensions iscritical because it demonstrates whether it is necessary tochange the communications strategy at different stages of the customer life cycle. For example, customer acquisitionmight be optimized by means of more (highly involving)personal sales calls, but when customers have beenacquired, the retention strategy may be most effectivelymanaged by less obtrusive or less interpersonal communi-cation, such as e-mail or Internet-based interactions. Theidea that different types of communication channels playvarying roles in the acquisition and retention processes hasonly been discussed conceptually so far (Dwyer, Schurr,and Oh 1987). We investigate this assertion empirically in aB-to-B setting. Thus, the testing of the effect of media syn-ergies on individual customers’behavior is another uniqueaspect of this research. Data Data for the study come from a large, multinational, B-to-Bhigh-tech manufacturer. The company’s database includesfirms that function in B-to-B and business-to-consumer (B-to-C) markets. The product categories in the database repre-sent different spectra among high-technology products.Even though the products are durable goods, they requireconstant maintenance and frequent upgrades; this character-istic provides the variance required to model the customerresponse. The choice of vendors for the products is nor-mally made after much deliberation by the buyer firm. Forthe product categories, the buyer and seller choose whetherto develop their relationships, and there are significant ben-efits to maintaining a long-standing relationship for bothbuyers and sellers.  66/ Journal of Marketing,January 2005 2 The manager here refers to the manager at the firm who sup-plied the data. 3 A limitation of the data is that we cannot distinguish an inter-action between media from a media pulsing strategy. Therefore,we may underestimate the true level of media interactions in thedata. 4 In these data, there are no incidences of more than two contactmodes used in the same month. The data used in the study cover a four-year period fromthe beginning of 1998 to the end of 2001. All of the cus-tomers are new to the firms and made their first purchasefrom the manufacturer in the first quarter of 1998. A total of 12,024 prospects were contacted for potential acquisition,and of those, 2908 made at least one purchase in the firstquarter of 1998. The average interpurchase time for an indi-vidual customer ranged between 1.5 and 21 months.To help the manager make an allocation decision, he orshe has at his or her disposal information about eachprospect before acquisition and each customer after acquisi-tion, as follows 2 : date of each purchase, number of proac-tive manufacturer-initiated marketing campaigns before thatdate, type of campaign (face-to-face, telephone, e-mail),and the number of customer-initiated contacts with the sup-plier firm (through the Web). From this information, weconstructed the variables FACE-TO-FACE, TELEPHONE,EMAIL, and WEB to measure the number of contacts thatthe firm had with the customer through the specific contactmode. In the acquisition equation, the variables representthe total number of preacquisition contacts in each channelbefore the first purchase. In the duration equation, the vari-ables measure the total number of contacts in each channelafter the first purchase. In the profitability equation, thevariables are operationalized as every contact (pre- andpostacquisition) that the customer has with the firm. If anytwo modes of contact with a customer or prospect occurredin a given month, we formulated an interaction termbetween those two contacts. Specifically, we operational-ized the interaction terms as the number of times any twocommunication modes occurred in the same month, 3 whichhelped us assess whether the use of two different contactmodes (e.g., telephone and e-mail) in a given period pro-vides added effectiveness. 4 Additional decision variables under the firm’s controlare the amount of acquisition dollars spent for eachprospect (ACQUISITION DOLLARS) and the amount of retention dollars spent for each customer (RETENTIONDOLLARS). These dollar expenditures are allocated to thefour different communication channels. Thus, the expendi-ture amounts cannot change without adjustments in theallocation of effort to the communication channels. In addi-tion to linear terms for the expenditures, we also includedquadratic terms for acquisition dollars (ACQUISITIONDOLLARS 2 ) and retention dollars (RETENTION DOL-LARS 2 ) in the model. As firms increase their acquisitionand retention budget, the associated acquisition rate, reten-tion rate, and customer profitability will be less responsive(concavity). So far, this effect has not been demonstrated inempirical customer lifetime value literature, except for thedeterministic (not statistical) approach taken by Blattbergand Deighton (1996). The quadratic terms do not impose acurvature but help uncover nonlinear effects (e.g., diminish-ing marginal effects) of the relationship between the expen-diture and the dependent variables.We calculated customer profitability (PROFIT) by sub-tracting direct (product-related) cost, total retention costs,and acquisition cost from the total revenues the customergenerates for the firm during the observation period. Control Variables  We introduce several covariates to control for exchange andcustomer characteristics. Exchange characteristics that mayhave important bearings on the different dependent vari-ables include customer-initiated contacts, the degree of cross-buying, the frequency of transactions (e.g., Reinartzand Kumar 2003), the customer’s share-of-wallet with thefocal firm (e.g., Verhoef 2003), and the relationship dura-tion (Bolton 1998; Bolton and Lemon 1999; Reinartz andKumar 2000).In this context, the firm records the number of customer-initiated contacts that is executed through theInternet. From a utility perspective, customers who havegreater expected benefits and utility from an ongoing rela-tionship are more likely to commit to it. Customer-initiatedcontacts are a way to signal this commitment, and there isample evidence that frequency of communication is posi-tively associated with a partner’s commitment (Andersonand Narus 1990). We introduce this count measure as acovariate for all three dependent variables.Cross-buying, which is an indicator of stronger relation-ships (Kamakura et al. 2003), should have a potentialimpact on both relationship duration and customer prof-itability. We operationalize the variable CROSS-BUY as thenumber of different categories from which the customerbuys.Frequency of transactions is also a sign of the quality of a relationship (Anderson and Weitz 1992; Kalwani andNarayandas 1995) and therefore should have an impact onboth relationship duration and customer profitability. Weoperationalize the variable FREQUENCY as the number of purchase occasions for each customer.The firm’s share-of-wallet with a particular customercaptures the competitive aspect. As a customer allocates rel-atively more category purchases to a focal vendor, competi-tors have less access to the customer. Firms that own agreater share-of-wallet of their customers have a strategicadvantage over their competitors. A larger share-of-walletallows for (and requires) greater learning about customerrequirements, allows for (and requires) more communica-tion between the parties, and justifies greater relationship-specific investments (Anderson and Narus 2003). Thus, alarger share-of-wallet should have an impact on relationshipduration and customer profitability. We operationalize thevariable SOW as the percentage of the customer’s informa-tion technology budget that is spent with the focal firm.Finally, we introduce the LENGTH OF RELATION-SHIP as a covariate for modeling customer profitability.The expectation with respect to relationship duration andcustomer profitability is that as the length of the customer  Acquisition and Retention Resources / 67 tenure rises, it allows for more transactions (volume andfrequency). If the transactions are profitable, there shouldbe overall greater relationship profitability (Kamakura et al.2003; Reinartz and Kumar 2000).To control for observed heterogeneity across customers,we include additional determinants in the specification. Thethree available variables represent the following characteris-tics of the potential targets: type of industry, annual rev-enues, and number of employees. The variable INDUSTRYTYPE classifies customers as either B-to-B or B-to-C firms.In addition, we use ANNUAL SALES REVENUE ($ mil-lions) and SIZE OF FIRM (number of employees) in thisanalysis.Based on these data, the model underlying this researchand the potential drivers of acquisition, duration, and cus-tomer profits appears in Figure 2. Figure 2 is a conceptualrepresentation of the three equations that make up our sta-tistical model. In terms of the statistical specification, allthree equations are similar with respect to firm (e.g., expen-diture amounts, communication channels) and customer(e.g., Web-based communications) action variables. In addi-tion, customer characteristics enter the acquisition equationbecause the information is available for each prospect.Through the correlation structure, the duration and profitequations also capture the influence of these customer char-acteristics. Finally, the control variables for customerbehavior (e.g., FREQUENCY, SOW, CROSS-BUY) areapplicable only to acquired prospects and thus only appearin the duration and profit equations. Right Censoring  Because of the noncontractual nature of the relationship,customers are subject to silent attrition. Our model includesan estimate of the customers’relationship duration, and wetherefore must account for the possibility of right censoring.This possibility was established with the use of Allenby,Leone, and Jen’s (1999) approach to compute the expectedtime until the next purchase. If the expected time until thenext purchase exceeds the time elapsed since the last pur-chase, the account is considered active and the duration isconsidered right censored. If this is not the case, the rela-tionship is assumed to have been terminated at the lastpurchase.We provide descriptive statistics in Table 1. The uniquestrength of the data set lies in the availability of individual-level marketing-mix contacts/communications, costs associ-ated with the channel contacts, and profile data. These dataenable us to use individual-level models and derive optimalmarketing guidelines for each individual customer or at thesegment level. Research Methodology Statistical Model  To link customer acquisition, relationship duration, andprofitability, we use a system of equations known as a pro-bit two-stage least squares model. We provide mathematicalrepresentations of the model in Equations 1, 2, and 3. FIGURE 2Determinants of Focal Constructs Firm Actions •Retention expenditures•Contact mix•Contact mix interactions Customer Actions •Customer-initiated contacts Control Variables •Cross-buying•Frequency•Share-of-wallet Firm Actions •Acquisition expenditures•Contact mix•Contact mix interactions Customer Actions •Customer-initiated contacts Customer Characteristics •Industry type•Annual revenue•Firm size (employees)AcquisitionlikelihoodRelationship durationCustomer profitability Firm Actions •Acquisition expenditures•Retention expenditures•Contact mix•Contact mix interactions Customer Actions •Customer-initiated contacts Control Variables •Cross-buying•Frequency•Share-of-wallet•Relationship duration

Sigi Los

Apr 16, 2018

Powerpoint mg

Apr 16, 2018
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks