Forbes Firms continue to struggle with the implementation of sales force technology tools and the role they play in sales representative performance. This research expands previous literature in the area of sales force automation (SFA) and customer relationship management (CRM) by looking at the consequences after technology adoption by a sales force. Data were gathered from three sources to include 662 sales representatives, 60 sales managers, and ﬁrm archival data. Using structural equation modeling, our ﬁndings indicate that SFA usage has a direct impact on effort, thereby reducing number of hours worked, and CRM usage has a direct positive impact on adaptive selling behaviors. Moreover, experience moderates the relationship between CRM usage and adaptive selling. Discussion, limitations, and directions for future research are also discussed.
As competition increases and technology advances, organizations continue to seek ways to adjust to changing business environments. This is especially true in the personal selling context where salespeople are recognized as the boundary spanners and are expected to be relationship managers (Kotler 1984). Today’s salesperson is constrained to do more in less time, and technological advancements have become an integral part of the personal selling and sales management process.
Foreseeing this changing environment, Leigh and Tanner (2004) stressed the necessity for sales organizations to focus on technology-related strategies, business processes, and applications, and called on sales researchers to put forth theoretical models and empirical studies investigating these emerging issues. Notably, sales force technology usage has changed the methods of selling. Salespeople are no longer selling just a “product”; instead, they are providing a valuable “solution” to customer problems. Anderson and Dubinsky (2004) discussed the concept of consultative selling, where a salesperson acts as an expert and provides customized solutions. This role requires salespeople to develop a technological orientation to access, analyze, and communicate information in order to establish a strong relationship with customers (Hunter and Adam Rapp (Ph.D., University of Connecticut), Assistant Professor of Marketing, College of Business Administration (Marketing), Kent State University, [email protected] Raj Agnihotri (MBA, Oklahoma City University), Ph.D. candidate, College of Business Administration (Marketing), Kent State University, [email protected] Lukas P. Forbes (Ph.D., University of Kentucky), Assistant Professor of Marketing, Ford College of Business, Western Kentucky University, [email protected]
Perreault 2007). Sales technology enables salespeople answering the queries of customers to effectively provide competent solutions. This can lead to strong relationships between a salesperson and a customer. Thus, technology tools are used not only for smoothing the work process but they also have strategic utilizations. To this point, numerous models investigating technology acceptance have been postulated in the literature (Compeau, Higgins, and Huff 1999; Davis, Bagozzi, and Warshaw 1989; Venkatesh and Davis 2000; Venkatesh et al. 2003). These studies focus mainly on ﬁnding and examining the variables inﬂuencing salespeople’s motivation, or attitudes to adopt technology (Avlonitis and Panagopoulos 2005; Jones, Sundaram, and Chin 2002; Keillor, Barshaw, and Pettijohn 1997; Morgan and Inks 2001; Pullig, Maxham, and Hair 2002; Schillewaert et al. 2005; Speier and Venkatesh 2002). Notably, most existing research has focused on technology adoption with a few notable exceptions.
For example, Ahearne et al. (2008) and Hunter and Perreault (2007) investigated the mediating effects of relationship-forging tasks, and Ahearne, Jelinek, and Rapp (2005) proposed moderating effects of training and support on links between different types of sales technology use (adoption) and sales performance. However, there is still a need to investigate the links between technology use and performance (Sundaram et al. 2007). Within this study, we make two extensions to the prior research. First, we shift the focus from technology adoption to technology usage and consequence (Hunter and Perreault 2007; Sundaram et al. 2007). The rationale for this diversion is that the success of technology acceptance resides “not simply in whether or not salespeople adopt technology, but whether or not adoption (i.e. use) actually improves performance” (Ahearne, Jelinek, and Rapp 2005, p. 380). For this purpose, we ground our arguments in the technology-to-performance chain (TPC) model, which explores the link between technology and an individual’s performance and postulates that “performance impacts will depend increasingly upon task–technology ﬁt rather than utilization” (Goodhue and Thompson 1995, p. 216).
Second, this research focuses on the multidimensionality of the technology usage construct. Hunter and Perreault (2006; 2007) made a distinction between sales force automation (SFA) and customer relationship management (CRM) tools and reinvigorated the issue of sales technology and its effectiveness. We extend that distinction. They suggested that SFA and CRM technologies should be considered as two different sales technology tools, and that “different use of technology have differential effects on various aspects of performance . . . thus, how a sales representative uses technology and on which behavioral tasks (work processes) matters” (Hunter and Perreault 2007, p. 30). Aligning with this logic, we perceive this new research direction as a valuable addition to an already established and rich literature of sales technology. The purpose of this research study, therefore, is to expand research with regard to the different dimensions of technology usage by investigating their impact on sales representative’s behavior that inﬂuences performance. We also investigate the role of salesperson experience within this model.
As mentioned previously, examining the relationship between technology acceptance and salesperson performance has only recently gained mainstream attention; however, studies investigating this link report positive ﬁndings. For example, researchers have documented that the growing use of technology tools inﬂuences salesperson performance positively (Ahearne, Srinivasan, and Weinstein 2004) by enhancing sales efﬁciency and sales effectiveness (Ahearne, Jelinek, and Rapp 2005). It has been argued that increasing the use of technology encourages salesperson knowledge attainment, which further improves his or her performance (Ko and Dennis 2004). More recently, Hunter and Perreault (2006) suggest that salespeople’s technology orientation inﬂuences their internal role performance.
In another study, Sundaram et al. (2007) theorize that technology use and technology impact on performance are directly proportional to each other. They provide empirical evidence suggesting that the extent to which salespeople use technology may improve overall sales performance. Bringing new insights into this subject, Hunter and Perreault (2007) propose new behavioral mechanisms that relate to sales representative performance. Speciﬁcally, they suggest that through relationship-forging tasks, salespeople are able to exploit different dimensions of technology utilization (i.e., accessing, analyzing, and communicating information), which in turn, affect different facets of sales performance.
Our research builds on the logic presented by previous researchers regarding the consideration of different dimensions of technology use and their differential effects on salespeople’s behavior. To provide theoretical grounding, we base our conceptual framework on the TPC model proposed by Goodhue and Thompson (1995). The TPC model emphasizes that in order to see a positive link between technology and performance, “the technology must be utilized, and the technology must be a good ﬁt with the tasks it supports” (Goodhue and Thompson 1995, p. 213, emphasis in original). Notably, tasks are viewed as activities performed by individuals to achieve outputs and technologies are tools that help them to perform these tasks. The use of certain applications of technology depends on the speciﬁc characteristics of the assigned task. Within the sales context, salespeople carry out operational (e.g., learning about existing and new products, generating automated reports) as well as strategic (e.g., identifying most important customers, preparing sales presentations based on customers’ speciﬁc needs) activities and need different tools to help perform these activities.
Moreover, the TPC model highlights the importance of an individual’s characteristics (e.g., training or experience), suggesting their impact on how “easily and well” one will use the technology tools (Goodhue and Thompson 1995, p. 216). The current research contributes to this idea by suggesting that the effect of technology use on salespeople’s behavior will depend upon whether they are using the technology for operational purpose (i.e., SFA) or for strategic purpose (i.e., CRM). Also, our framework incorporates salesperson experience to assess the inﬂuence of individual characteristics. Dimensions of Sales Technology Usage In a broad sense, technology is deﬁned as “an ability to act, a competence to perform, translating materials, energy and information in one set of states into another, more highly valued set of states” (Metcalfe 1995, p. 34). In a sales research domain, sales technology describes information technologies aiding or enabling the sales task performance (Hunter and Perreault 2007).
In the past, scholars from different research streams have raised the issue concerning the different dimensions and aspects of technology use and proposed several frameworks that support this concept (Goodhue and Thompson 1995; Orlikowski 1992). Although previous researchers build their arguments on different concepts, in essence, they all agree there are different aspects and dimensions of technology use. Considering the fact that different dimensions of technology use should be employed for different purposes, sales managers must develop and support an environment where salespeople use technology in accordance with the required behavior.
For example, salespeople involved in operational activities such as exchanging information with colleagues and managers, taking or tracking inventory levels, or learning about existing, new, and competitive products will employ different technology tools as compared to situations where they execute strategic activities such as identifying potential customers, identifying the most important customers from the list of current customers, or working on improvement of sales presentation skills. Thus, it will be beneﬁcial for sales managers, as well as for salespeople, to understand how employing different technology tools will inﬂuence performance-enhancing behaviors (Hunter and Perreault 2007). Accordingly, we view the use of SFA and the use of CRM as two dimensions of sales technology based on their level of speciﬁcity for inﬂuencing different salespersons’ behaviors. SFA usage, with an operational orientation, includes the utilization of technological tools supporting the routine sales functions (Jacobs 2006). CRM usage, with a strategic orientation, includes methods and employing technology tools that help salespeople develop sales strategies (Rigby and Ledingham 2004).
Importantly, both the routine sales functions and strategic sales activities that a salesperson engages in can develop, sustain, and strengthen customer relationships. Use of SFA Technology Hunter and Perreault suggest that SFA technology includes “tools that are intended to make repetitive (administrative) tasks more efﬁcient” (2007, p. 17). Highlighting its potential beneﬁts, previous research views SFA use as a competitive equivalence (Morgan and Inks 2001) and suggests that it enhances the “precision” of salespeople’s activities (Honeycutt et al. 2005) by providing fast and reliable information ﬂow among customers, salespeople, and ﬁrms (Speier and Venkatesh 2002). Sales managers and salespeople alike identify SFA as a tool to enhance efﬁciency (Erffmeyer and Johnson 2001) and to improve productivity (Engle and Barnes 2000; Pullig, Maxham, and Hair 2002). SFA tools assist with routine tasks, allow salespeople to focus on relationship-oriented activities, and free up time for more customer-centric tasks (Cotteleer, Inderrieden, and Lee 2006).
To attain the advantages of SFA, salespeople need to understand the speciﬁc purpose of using SFA. Keeping this in mind, we adapt the Rivers and Dart’s conceptualization of SFA that describes it as transforming “manual sales activities to electronic processes through the use of various combinations of hardware and software applications” (1999, p. 59). We view SFA as a tool that converts repetitive and routine manual processes to automated processes, and assists salespeople trying to operate in an efﬁcient and timely manner. Examples of SFA tools could include programs such as quarterly automated sales reports and calendaring tools. The domain for SFA applications includes the attainment and storage of information. However, the information being utilized, analyzed, and obtained with the help of SFA tools is unlike that from CRM tools. Use of CRM Technology Unlike the routine purpose of SFA applications, CRM technology usage focuses on relationship and strategy building (Rigby, Reichheld, and Schefter 2002). Day views CRM as “a cross-functional process for achieving a continuing dialogue with customers” (2001, p. 1).
CRM is also described as a “business strategy that includes information technology to provide an enterprise with a comprehensive, reliable, and integrated view of its customer base” (Zikmund, McLeod, and Gilbert 2003, p. 3). In essence, salespeople use CRM technology tools for developing and managing customer relationships (Yin, Anderson, and Swaminathan 2004). This characterization is aligned with the analysis aspect of sales technology use suggested by Hunter and Perreault (2007). They deﬁned it as the degree to which salespeople depend on technology “to study and synthesize data and understand the implications of data relevant to the demands of their sales jobs” (Hunter and Perreault 2007, p. 21). Outlining the functionality of sales technology, Widmier, Jackson, and McCabe (2002) postulate different sales functions (e.g., presentations, informing, communications, sales reporting) that can be supported by sales technology.
Importantly, these functions of sales technology can be separated on the basis of whether their strategic orientation is “customer” centric or “back-ofﬁce” centric (Geiger and Turley 2006). We characterize the use of CRM as utilizing customer-centric technology tools that help salespeople formulate strategies to achieve effectiveness in their selling methods. Therefore, the optimal utilization of CRM tools will depend on how well salespeople assimilate the information obtained through data patterns in their job-speciﬁc behaviors. We believe that the use of CRM technology tools not only accelerates the regular sales operation, but also aids salespeople in developing and managing long-term customer relationships. CONCEPTUAL MODEL DEVELOPMENT In light of the above-mentioned arguments, we propose a model (Figure 1) examining the effects of SFA and CRM on salespeople’s behaviors after technology adoption and how these behaviors can inﬂuence salesperson performance.
Effort A salesperson’s effort, in general, can be characterized as “persistence—in terms of the length of time devoted to work and continuing to try in the face of failure” (Sujan, Weitz, and Kumar 1994, p. 40), and it can be assessed via a litany of measures. Speciﬁcally, “the duration of time spent working and the intensity of work activities” are viewed as components of effort (Brown and Peterson 1994, p. 71); other research studies measure effort by the number of hours invested by salespeople to accomplish their sales goals or the number of sales calls made (e.g., Rapp et al. 2006). Past scholars have conceptualized that the utilization of technology tools improves salesperson efﬁciency (Keillor, Barshaw, and Pettijohn 1997; Pullig, Maxham, and Hair 2002) and that technology assists salespeople in formulating a professional sales encounter (Marshall, Moncrief, and Lassk 1999). Salespeople can maintain direct contact, even with remote customers, through e-mails and cell phones, thus saving travel hours.
They can receive and manage orders from customers in an easy, timely fashion. Various SFA applications (e.g., calendaring; routing tables) inject activeness in salespeople’s typical sales routines and reduce downtime. Salespeople, in today’s competitive environment, face a great deal of data that include information about distributors, dealers, retailers, and ultimately, the end customer. In addition to this, salespeople need to keep track of competitor’s activities as well as product market situations. Notably, SFA tools provide answers to salespeople in such complex data utilization and management scenarios. Different application tools, spreadsheets, Web browsers, inventory management software, and other database software enable salespeople to manage the records of products, competitors, and customers in timely manner.
Hence, salespeople using SFA tools will be more organized and able to complete their schedules on time (Ahearne, Jelinek, and Rapp 2005). One key representation of salespeople’s efforts to realize their job objectives is the activity through which they complete their tasks (Brown and Peterson 1994). The use of SFA reduces “the amount of time spent on such activities as managing contacts, scheduling sales calls, developing sales plans, and planning sales routes” (Widmier, Jackson, and McCabe 2002, p. 190). Also, salespeople using SFA tools can cut down work hours doing paperwork and other reporting-related tasks (Colombo 1994; Parathasarathy and Sohi 1997). Importantly, these administrative tasks (e.g., sales reporting) are non-customer-centric activities (Geiger and Turley 2006); however, salespeople spend many hours completing them. Thus, reductions in such activities, with the help of SFA, will provide salespeople with an opportunity to reduce their working hours. Formally stated, Hypothesis 1: Relative to salespeople reporting lower use of SFA, salespeople reporting higher use of SFA will report fewer hours of work.
Underlining the importance of CRM usage, Ahearne, Jelinek, and Rapp (2005) argued that such technology tools aid salespeople by managing information about a larger number of customers. Salespeople equipped with such valuable information are able to relate to customers without as much difﬁculty and can be more responsive to critical issues, thereby shortening the duration of each sales encounter. They will also complete tasks with less effort (Ahearne, Jelinek, and Rapp 2005). Mostly, CRM tools make customers’ cross-referencing possible among different departments within an organization that can generate more sales potential and reduce efforts by evading multiple attempts on the same prospective clients (Widmier, Jackson, and McCabe 2002).
Moreover, the use of CRM tools will ease the processes of presale planning activities and improve the accuracy of sales forecasts, speeding up the overall sales process (Hunter and Perreault 2006). Parallel to this thought, it is pragmatic to think that salespeople using CRM tools will not ﬁnd examining customer data to be an overly complex and time-consuming process. Moreover, they can promptly focus on vital information, which, in turn, enables them to develop winning strategies in shorter time. We believe that salespeople equipped with CRM technology will be able to conserve their efforts by speeding the process of strategy development.
CRM use will help salespeople conﬁgure product offerings per customer stipulations without showing extra efforts (Widmier, Jackson, and McCabe 2002). Under these circumstances, salespeople will be able to decrease their efforts by investing less time in the formulation of customer relationship strategies, reducing backorders, and lessening the number of calls required to ﬁnalize a sale (Columbo 1994; Thetgyi 2000). Based on this, we hypothesize: Hypothesis 2: Relative to salespeople reporting lower use of CRM, salespeople reporting higher use of CRM will report fewer hours of work. Adaptive Selling Adaptive selling is deﬁned as “engaging in planning to determine the suitability of sales behaviors and activities that will be undertaken, the capacity to engage in a wide range of selling behaviors and activities, and the alteration of sales behaviors and activities in keeping with situational considerations” (Sujan, Weitz, and Kumar 1994, p. 40). In more general terms, adaptive selling can be deﬁned as an approach to personal selling in which selling behaviors are altered during the sales interaction or across customer interactions, based on information about a customer and nature of the selling situation.
Acquisition, analysis, and use of customer information are particularly important for salespeople in demonstrating adaptive selling behaviors (Weitz, Sujan, and Sujan 1986). Moreover, if salespeople have precise customer information, they will be more capable of practicing such behaviors (Hunter and Perreault 2006). CRM tools can also aid salespeople in tracking customer purchase patterns and enabling them to recognize potentially viable customers. Salespeople, with the help of CRM technology, can obtain critical customer information to successfully plan an effective sales encounter (Ahearne et al. 2008). CRM tools will be useful for keeping salespeople informed as well as for developing, implementing, or revising sales planning. Such customer database systems provide opportunities to meticulously research customers and design their sales presentations according to particular customer needs and wants (Marshall, Moncrief, and Lassk 1999).
Equipped with sound customer information, salespeople will better anticipate customer responses, prepare appropriate ways to meet customer needs, and overcome customer objections. We propose that CRM tools provide access to customer information that enables salespeople to improve or enhance their adaptive selling skills. Based on this argument, we hypothesize: Hypothesis 3: Salespeople’s use of the CRM technology will be positively related with their adaptive selling behaviors. Experience as a Moderator Salesperson’s experience has been deﬁned as a composite of three different areas: the employee’s general sales experience, the amount of time spent working with their current company, and the time spent in their territory (Rapp et al. 2006). Previous studies document the positive relationship of experience with different individual outcomes. For example, individual’s performance adaptability has been associated positively with greater amounts of relevant work experience (Pulakos et al. 2000). It has been argued that individuals seeking knowledge usually carry dissimilar wants and expectations (Markus 2001).
This idea is especially applicable in a personal selling context, where no single formula or approach can guarantee success of every salesperson. Salespeople with different breadth and depth of experience will have different abilities and expectations. Within this research study, we suggest that less-experienced salespeople, even if they use sales technology tools (i.e., SFA and CRM), will be less likely to exploit such tools in a better way, relative to more experienced salespeople. Importantly, our research differs from the previous work of Ko and Dennis (2004) in that we examine different dimensions of technology use as well as behavioral outcomes of the technology/experience interaction rather than outcome-based performance. Thus, our hypotheses differ according to our proposed arguments. Sales researchers agree that the uses of CRM technology tools are essential for making customer alliances; however, individual characteristics can affect this process (Jones, Sundaram, and Chin 2002).
Because CRM is used in crafting customer relationship strategies, salespeople’s experience will play a critical role in the relationship between CRM utilization and adaptive selling behaviors. CRM will provide valuable customer information; however, to be successful in utilizing such information, salespeople need to have a “deep base of organizational, contextual, and domain knowledge” (Ko and Dennis 2004, p. 313) and be well versed in handling difﬁcult objections. Salespeople with relatively less experience will have had less exposure to the capabilities of CRM tools, and a lower level of understanding about adaptive selling. With the lack of knowledge regarding various tasks and selling situations, less-experienced salespeople will be less capable of exploiting the rich data available in a CRM repository. Experienced salespeople are more likely to maintain focus on the task-related activities, identify and realize the link between CRM tools utilization and adaptive selling behaviors, and smartly engage in activities relevant to task completion. To sum, we argue that more-experienced salespeople will employ information toward formulating plans in a better way that helps them to practice adaptive selling than those salespeople with less experience.
Based on these arguments, we propose that Hypothesis 4: The relationship between use of CRM and adaptive selling will be more positive for employees who report higher levels of experience, as compared to those who report lower levels of experience. In the case of technology use, it has been argued that the inﬂuence of technology is moderated by contextual variables (Orlikowski 1992). It seems especially true in a situation where technology is being used as a tool to formulate strategies or as a medium to support routine tasks. Experienced salespeople are more likely to have created an optimal schedule (i.e., necessary efforts required to accomplish maximum output), and given the strategic utilization of CRM, they can further cut down their efforts to achieve sales objectives.
Consistent with the arguments of Hunter and Perreault (2006), we argue that more-experienced salespeople have learned the necessary skills to execute different activities. We also suggest that moreexperienced salespeople have discovered ways to reduce their levels of effort while maintaining their higher levels of performance, relative to those with less experience. Importantly, for those who have already adopted technology, more-experienced sales representatives will feel the greatest inﬂuence on their behavioral outcomes. Formally stated, Hypothesis 5: The relationship between use of CRM and effort will be more negative for employees who report higher levels of experience, as compared to those who report lower levels of experience. Salesperson Performance In a general sense, job performance is an outcome of effort and strategy (Bandura 2002). Sales literature has recognized the signiﬁcance of salesperson efforts in different theoretical frameworks of performance (Walker, Churchill, and Ford 1977) and proposed a signiﬁcant positive relationship between effort and adaptive selling behaviors and salesperson’s productivity (Sujan, Weitz, and Sujan 1988).
Previous literature enjoys a relatively wide consensus about the critical role of effort and adaptability in achieving high performance objectives. To this point, numerous researchers have examined the links between performance and adaptive selling and effort (Anglin, Stohlman, and Gentry 1990; Brown and Peterson 1994; Goolsby, Lagace, and Boolrom 1992; Holmes and Srivastava 2002; Predmore and Bonnice 1994; Sujan, Weitz, and Kumar 1994). Within this research study, we revisit these links and offer that, parallel to previous ﬁndings, both salesperson behaviors of adaptive selling and effort will demonstrate unique positive relationships with their performance. Hypothesis 6: Salesperson effort will be positively related with salesperson performance. Hypothesis 7: Salesperson adaptive selling behaviors will be positively related with salesperson performance. RESEARCH METHOD Sample Our sample was drawn from the human health-care segment of a medium-sized pharmaceutical company.
Data were collected from three separate sources in the form of salesperson surveys, manager surveys, and archival job performance data from company records. Sales representatives in this ﬁrm were responsible for marketing directly to physicians within a speciﬁc geographical area. All sales representatives were responsible for a particular portfolio of products and completed training for each product line. In sum, 900 sales representatives of the human health-care division of the company were contacted for this study. Usable survey responses were obtained from 662 (74 percent) of the representatives and from 60 different sales managers. There was an average of 11 sales representatives per manager. Respondents completed and returned a questionnaire mailed directly to them by the researchers. Management’s strong endorsement of questionnaire completion via e-mail and telephone, coupled with two waves of mailings, led to the high response rate. The sample was 40 percent male, the average age was 34.9 (standard deviation [SD] = 4.6), and 91 percent reported their ethnicity as Caucasian. Measures Use of SFA SFA usage was assessed by the sales representative using a fouritem scale.
The scale asked sales representatives the amount of usage on four speciﬁc SFA applications. All four items were representative of tasks that helped salespeople streamline or automate some of the basic processes and functions of the sales tasks. Item responses were anchored by (1) “I do not use this technology at all” and (7) “I use this technology to a great extent.” The scale demonstrated acceptable reliability (α = 0.72). See the Appendix for a complete list of scale items. Use of CRM Similar to the above, CRM usage was assessed by asking the salesperson four questions regarding his or her use of technologies that helped manage customer relationships. These questions were speciﬁc to the software and database applications that the ﬁrm had in place. Again, item responses were anchored by (1) “I do not use this technology at all” and (7) “I use this technology to a great extent.” The scale demonstrated acceptable reliability (α = 0.75). Effort Salesperson effort was measured as a self-report item assessing average number of hours worked per week.
Although not an ideal evaluation tool, this approach is similar to other research that has demonstrated that self-report evaluations are often representative of objective measures of evaluations (Sharma, Rich, and Levy 2004). Adaptive Selling Adaptive selling was measured using a shortened four-item scale stemming from the adaptive selling scale originally developed by Spiro and Weitz (1990). Items were adapted slightly to ﬁt the speciﬁc selling context. This measure was assessed by the sales manager and exhibited strong reliability (α = 0.90). In this setting, sales managers have frequent contact with their salespeople. By meeting with sales representatives, conducting customer follow-up visits, and participating in ride-alongs, we argue that the sales manager can observe the behavior of the salesperson, in this circumstance, adaptive selling tendencies.
Experience Experience was a composite measure consisting of three separate measures of sales experience. Sales representatives were questioned about the length of time they had worked in their territory, for their company, and in a sales ﬁeld. These scores were each z-scored and then averaged to form an overall experience index. Salesperson Performance We operationalized salesperson performance as the outcomebased measure of percentage of quota. Percentage of quota achieved is deﬁned as the total sales brought to a close by a salesperson relative to the sales organization’s sales targets for that individual. Percent of quota, or total sales divided by expected sales target, is a strong measure of sales representative performance because it controls for some potential contaminating factors such as territory size (Churchill et al. 1985). Sales representatives’ quotas are annually set by a consulting company, in conjunction with corporate sales management, and are based on market information and company records.
Quotas are discussed with sales representatives to ensure that the representatives understand the methods used to set their annual quotas. Analytical Strategy We analyzed our data using a covariance-based structural equation modeling package, AMOS 5.0 (Arbuckle 1997). In evaluating this model, we followed the procedures recommended by Anderson and Gerbing (1988). First, we conducted a conﬁrmatory factor analysis (CFA) to examine the adequacy of the measurement component of the proposed model and evaluate discriminant validity. After ensuring an appropriate ﬁt, we then derived the full structural model from our hypotheses. To gauge model ﬁt, we report the comparative ﬁt index (CFI) (Bentler 1990) and the standardized root mean square residual (SRMR) (Hu and Bentler 1999).
The CFI has been identiﬁed as a strong approximation of the population value for a single model, with values ≥ 0.90 considered indicative of good ﬁt. SRMR is a measure of the standardized difference between the observed and unobserved covariance and predicted covariance, with values ≤ 0.08 considered a “relatively good ﬁt for the model,” and values ≤ 0.10 considered “fair” (Hu and Bentler 1999). Based on an exploratory and follow-up CFA, we determined that all items loaded signiﬁcantly on their respective factors and no cross-loadings were present. Each indicator exhibited a highly signiﬁcant estimate (p < 0.001), which suggests high convergent validity (Gerbing and Anderson 1988). Also, discriminant validity was assessed according to the Fornell and Larcker (1981) suggested approach.
By examining the amount of variance extracted for each of the latent constructs and comparing this to the squared correlations among the constructs, we found that the shared variance among any two constructs was always less than the average variance explained by the construct, which suggests that discriminant validity has been achieved. Finally, because four of the variables were collected from the same source, we conducted checks for common method variance, which could inﬂate any observed correlations between the dependent and independent variables. As suggested by Grifﬁth and Lusch (2007), we used a CFA approach to assess Harman’s one-factor test. To do this, one would create a single latent factor for all same-source indicators as an alternative explanation to our results. Based on our analysis, our measurement model ﬁt yielded a χ2 of 295.61 (88), p < 0.01; CFI = 0.93; SRMR = 0.04.
By ﬁtting the same-source factor model, our ﬁt was signiﬁcantly worse with a χ2 of 789.53 (101); p < 0.01; CFI = 0.77; SRMR = 0.08. Second, we employed the partial correlation procedure of including a marker variable (i.e., a variable not theoretically related to at least one other variable in the study). By using a measure of sales ethics as the marker variable, we found no signiﬁcant relationships to other variables in the model. These analyses indicate that our structural equation analysis is not as susceptible to an inherent common method bias in the responses to the survey. Table 1 provides descriptive statistics and pairwise correlations for this study. As mentioned, model ﬁt for the measurement model was good (χ2 = 295.61 (88), p < 0.01; CFI = 0.93; SRMR = 0.04). Next, we ﬁt a linear effects model that amounts to the hypothesized model depicted in Figure 1 minus the two interactions (i.e., H4 and H5). This model was ﬁt in order to test the linear relationships.
This model also serves as a baseline model for tests of the interactions. Notably, the linear relationships between experience and both adaptive selling and effort, although not hypothesized, were included in this model so as to serve as a baseline for the hypothesized model. To test the interaction effects, CRM usage and experience were both mean-centered (by virtue of using z-scores) so as to reduce effects of multicollinearity. We then calculated a multiplicative interactive term between the two variables and ﬁt a second model that included this product as an antecedent of both effort and adaptive selling. Because the linear effects model is nested in the hypothesized model, a signiﬁcant Δχ2 between them indicates that one or both of the interactions are signiﬁcant (Cortina, Chen, and Dunlap 2001). To note, we speciﬁed the relationship between the observed scores and their respective latent variables by ﬁxing the measurement error terms for each construct at (1 – rxx) times the variance of each scale score. Following procedures advanced by Mathieu, Tannenbaum, and Salas (1992) and supported by Cortina, Chen, and Dunlap (2001), the reliability of the interaction term was estimated using the formula presented by Bohrnstedt and Marwell (1978). RESULTS We derived the full structural model from our hypotheses.
Structural model ﬁt was within acceptable limits (χ2 = 240.21 (95), p < 0.01; CFI = 0.95; SRMR = 0.04) (see Table 2). Although the χ2-statistic is signiﬁcant, it is not always the best indication of model ﬁt (e.g., Bagozzi and Yi 1988), because it has the drawback of being sensitive to sample size and the number of parameters in the model. Notably, our initial ﬁndings show that SFA usage is negatively related to effort (H1: β = –0.123, p < 0.05) and that CRM usage does not have a negative relationship with effort as originally hypothesized (H2: β = 0.091). As expected, the linear effect of CRM usage to adaptive selling was positive and signiﬁcant (H3: β = 0.122, p < 0.05). Finally, although not hypothesized, the linear effect of experience to effort (β = 0.166, p < 0.01) and adaptive selling (β = 0.106, p < 0.05) were both signiﬁcant. Next, we tested the hypothesized model. By adding the interaction terms, we found that the model demonstrated an excellent ﬁt (χ2 = 233.53 (93), p < 0.01; CFI = 0.95; SRMR = 0.04) and was a signiﬁcant improvement over the linear effects model (Δχ2 (2) = 6.68, p < 0.05).
The moderating effect of experience on CRM usage to effort was not present (H5: β = 0.083); however, the moderating inﬂuence of experience on the relationship between CRM and adaptive selling was evident (H4: β = 0.112, p < 0.05). The ﬁnal portion of our model examined both adaptive selling and effort as predictors of salesperson performance. We found that both effort (H6: β = 0.115, p < 0.05) and adaptive selling (H7: β = 0.086, p < 0.05) had signiﬁcant relationships with performance. As a post hoc analysis, we included experience as an additional predictor of performance and found that it exhibited a signiﬁcant relationship (β = 0.107, p < 0.05), while not changing the levels of signiﬁcance of the other two variables.
The proportions of variance of the endogenous variables accounted for were as follows: R²Effort = 0.047; R²Adaptive selling = 0.034; and R²Performance = 0.031. To interpret the nature of the interaction, we plotted it using standard practices (Aiken and West 1991). Speciﬁcally, using the information from the moderated regression analysis, we plotted the relationship between CRM usage that correspond to the average, low (one SD below the mean) and high (one SD above the mean) values of the experience moderator (see Figure 2). Corresponding to our expectations, we ﬁnd that CRM usage has a positive linear effect on adaptive selling and that increased levels of experience enhance this relationship as demonstrated by the steeper slope (more positive) for high-experience salespeople.