Top Banner
Journal of Personal Selling & Sales Management, vol. XXIX, no. 2 (spring 2009), pp. 167–178. © 2009 PSE National Educational Foundation. All rights reserved. ISSN 0885-3134 / 2009 $9.50 + 0.00. DOI 10.2753/PSS0885-3134290205 Due to their impact on performance, customer orientation (Saxe and Weitz 1982) and adaptive selling (Spiro and Weitz 1990) have been the focus of “two prominent research streams in sales force research” (Franke and Park 2006, p. 693). Sales- people with high levels of customer orientation truly care about customers, and thus engage in actions that customers value, such as listening to customer feedback and solving customer problems. Salespeople with high levels of adaptive selling adjust their sales strategies in ways that better fit cus- tomers’ needs and preferences (Hunter and Perreault 2006). Together, then, customer orientation and adaptive selling lead to win-win outcomes because customers are served in better, more individually relevant ways, producing higher levels of success for the sales force and the organization. This suggests that organizations can use customer orien- tation and adaptive selling to predict whether a salesperson is a top or a bottom performer. Franke and Park’s (2006) meta-analysis showed that adaptive selling had a significant and direct effect on sales performance, regardless of whether performance was measured using self-reported, managerial ratings, or objective sales figures. Yet this meta-analysis also found that customer orientation had a significant effect on performance only when performance was measured with self- reported measures. Contrary to expectations, the meta-analysis found that the mean correlations linking customer orienta- tion with both managerial ratings of performance (r = 0.01, p > 0.1) and objective sales performance (r = 0.02, p > 0.1) were nonsignificant. One explanation for these findings is that the short-term impact of customer orientation is not recognized by management or reflected in sales figures in the short run, but could manifest itself in the long run (Franke and Park 2006). If so, the long-term impact of customer orienta- tion may be related to its positive association with customer satisfaction (e.g., Homburg and Stock 2005), customer trust (e.g., Langerak 2001), and customer’s willingness to maintain his or her business relationship with the salesperson (e.g., Jones, Busch, and Dacin 2003). Another possibility is that the impact of customer orientation on short-term objective performance is mediated by adaptive selling wherein “a high level of concern for customers does not automatically translate into higher objective performance because adaptive selling Fernando Jaramillo (Ph.D., University of South Florida), Assistant Professor of Marketing, College of Business Administration, Uni- versity of Texas at Arlington, [email protected]. Douglas B. Grisaffe (Ph.D., Vanderbilt University), Assistant Profes- sor of Marketing, College of Business Administration, University of Texas at Arlington, [email protected]. The authors thank Larry Chonko, Dawn Iacobucci, Paul Spector, and three anonymous reviewers for their helpful comments and recommendations. RESEARCH NOTE DOES CUSTOMER ORIENTATION IMPACT OBJECTIVE SALES PERFORMANCE? INSIGHTS FROM A LONGITUDINAL MODEL IN DIRECT SELLING Fernando Jaramillo and Douglas B. Grisaffe Since the inception of the concept, researchers have hypothesized that customer orientation plays a fundamental role in explaining sales performance. However, Franke and Park’s (2006) meta-analysis challenged this notion with findings of a nonsignificant effect of customer orientation on objective sales performance. This counterintuitive result was explained by noting that the impact of customer orientation on objective sales measures may be present in the long run. In this research note, we evaluate that notion by testing a model in which customer orientation is used to predict individual rates-of-change in sales performance over time. Longitudinal salesperson performance in dollars, from the database of a direct selling organization, is merged with survey responses and modeled using an emerging method called latent growth modeling (LGM). Results confirm Franke and Park’s findings that customer orientation has a nonsignificant direct effect on the static initial-level aspect of objective sales performance. However, as postulated, customer orientation does show a significant direct effect on longitudinal sales performance trajectories. Our findings also suggest that customer-oriented selling’s nonsignificant direct effect on cross-sectional performance may be due to a fully mediated indirect effect through adaptive selling.
13
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: Does Customer Orientation Impact Objective Sales Performance

Journal of Personal Selling & Sales Management, vol. XXIX, no. 2 (spring 2009), pp. 167–178.© 2009 PSE National Educational Foundation. All rights reserved.

ISSN 0885-3134 / 2009 $9.50 + 0.00. DOI 10.2753/PSS0885-3134290205

Due to their impact on performance, customer orientation (Saxe and Weitz 1982) and adaptive selling (Spiro and Weitz 1990) have been the focus of “two prominent research streams in sales force research” (Franke and Park 2006, p. 693). Sales-people with high levels of customer orientation truly care about customers, and thus engage in actions that customers value, such as listening to customer feedback and solving customer problems. Salespeople with high levels of adaptive selling adjust their sales strategies in ways that better fi t cus-tomers’ needs and preferences (Hunter and Perreault 2006). Together, then, customer orientation and adaptive selling lead to win-win outcomes because customers are served in better, more individually relevant ways, producing higher levels of success for the sales force and the organization.

This suggests that organizations can use customer orien-tation and adaptive selling to predict whether a salesperson is a top or a bottom performer. Franke and Park’s (2006) meta-analysis showed that adaptive selling had a signifi cant and direct effect on sales performance, regardless of whether

performance was measured using self-reported, managerial ratings, or objective sales fi gures. Yet this meta-analysis also found that customer orientation had a signifi cant effect on performance only when performance was measured with self-reported measures. Contrary to expectations, the meta-analysis found that the mean correlations linking customer orienta-tion with both managerial ratings of performance (r = 0.01, p > 0.1) and objective sales performance (r = 0.02, p > 0.1) were nonsignifi cant. One explanation for these fi ndings is that the short-term impact of customer orientation is not recognized by management or refl ected in sales fi gures in the short run, but could manifest itself in the long run (Franke and Park 2006). If so, the long-term impact of customer orienta-tion may be related to its positive association with customer satisfaction (e.g., Homburg and Stock 2005), customer trust (e.g., Langerak 2001), and customer’s willingness to maintain his or her business relationship with the salesperson (e.g., Jones, Busch, and Dacin 2003). Another possibility is that the impact of customer orientation on short-term objective performance is mediated by adaptive selling wherein “a high level of concern for customers does not automatically translate into higher objective performance because adaptive selling Fernando Jaramillo (Ph.D., University of South Florida), Assistant

Professor of Marketing, College of Business Administration, Uni-versity of Texas at Arlington, [email protected].

Douglas B. Grisaffe (Ph.D., Vanderbilt University), Assistant Profes-sor of Marketing, College of Business Administration, University of Texas at Arlington, [email protected].

The authors thank Larry Chonko, Dawn Iacobucci, Paul Spector, and three anonymous reviewers for their helpful comments and recommendations.

RESEARCH NOTE

DOES CUSTOMER ORIENTATION IMPACT OBJECTIVE SALES PERFORMANCE? INSIGHTS FROM A LONGITUDINAL MODEL IN DIRECT SELLING

Fernando Jaramillo and Douglas B. Grisaffe

Since the inception of the concept, researchers have hypothesized that customer orientation plays a fundamental role in explaining sales performance. However, Franke and Park’s (2006) meta-analysis challenged this notion with fi ndings of a nonsignifi cant effect of customer orientation on objective sales performance. This counterintuitive result was explained by noting that the impact of customer orientation on objective sales measures may be present in the long run. In this research note, we evaluate that notion by testing a model in which customer orientation is used to predict individual rates-of-change in sales performance over time. Longitudinal salesperson performance in dollars, from the database of a direct selling organization, is merged with survey responses and modeled using an emerging method called latent growth modeling (LGM). Results confi rm Franke and Park’s fi ndings that customer orientation has a nonsignifi cant direct effect on the static initial-level aspect of objective sales performance. However, as postulated, customer orientation does show a signifi cant direct effect on longitudinal sales performance trajectories. Our fi ndings also suggest that customer-oriented selling’s nonsignifi cant direct effect on cross-sectional performance may be due to a fully mediated indirect effect through adaptive selling.

Page 2: Does Customer Orientation Impact Objective Sales Performance

168 Journal of Personal Selling & Sales Management

behaviors must occur fi rst” (Jaramillo et al. 2007, p. 69). Thus, the impact of customer orientation on objective performance could be mediated through adaptive selling in the short term, while having a direct infl uence on longer-term, longitudinal performance.

Although prior research has investigated the impact of customer orientation on the level aspect of salesperson’s perfor-mance, our current understanding of the impact of customer orientation on the longitudinal trajectory aspect of salesperson’s performance is limited. This limitation occurs because em-pirical studies have mostly relied on static, one-shot observa-tions that cannot capture dynamic longitudinal performance (Williams and Plouffe 2007). This is a serious shortcoming because sales leaders are not just interested in understanding the factors that can explain a salesperson’s static performance. As Ployhart and Hakel assert, understanding the dynamic nature of performance is critical because “individuals who are perhaps high-performing initially may be low-performing later” (1998, p. 860). In fact, longitudinal performance refl ects the salesperson’s ability to sustain, decrease, or increase initial sales levels—a vital dimension of performance beyond single static snapshots (Ash 1995).

The aim of this research note is to make a contribution beyond purely cross-sectional studies by investigating the impact of customer orientation on both level and trajectory

aspects of performance. Specifi cally, this paper tests a model in which customer orientation has (1) an indirect effect on the level aspect of objective performance through adaptive selling, and (2) a direct effect on objective performance trajectory (Fig-ure 1). As such, this research note serves as an initial response to Franke and Park’s (2006) call for longitudinal studies investi-gating the relationships among customer orientation, adaptive selling, and salesperson’s performance. The data structure is tested using latent growth modeling (LGM), an application of structural equation modeling for the analysis of longitudinal data (Bollen and Curran 2006; Duncan, Duncan, and Strycker 2006). Linear LGM is specifi cally designed to simultaneously test questions concerning individual’s static starting levels and subsequent longitudinal trajectories (i.e., intercepts and slopes, respectively) using longitudinal data. LGM also has explicit provision for explanatory latent covariates that have been purifi ed of measurement error (Bollen and Curran 2006; Williams, Edwards, and Vandenberg 2006). While not focused on customer orientation, but among the rare exceptions to cross-sectional studies, Ployhart and Hakel’s (1998) study of securities salespeople demonstrated that LGM can provide valuable insights in longitudinal sales contexts. LGM is thus well suited to our substantive research questions about cus-tomer orientation’s impact on short-term performance levels and long-run performance trajectories.

Figure 1Impact of Customer Orientation and Adaptive Selling on Sales Performance

Note: Dotted lines denote a nonsignifi cant relationship (at α = 0.05).

Page 3: Does Customer Orientation Impact Objective Sales Performance

Spring 2009 169

BACKGROUND AND RESEARCH HYPOTHESES

Franke and Park’s (2006) meta-analysis hypothesized that customer orientation and adaptive selling are two important characteristics of high-performing salespeople. Customer-oriented selling has been conceptualized as “the degree to which salespeople practice the marketing concept by trying to help their customers make purchase decisions that will satisfy customer needs” (Saxe and Weitz 1982, p. 344). Customer-oriented salespeople are solution providers who deliver value by assessing customers’ needs, then responsively helping customers identify alternatives, evaluate them, and select the best solution (Ehert 2004). In view of this, researchers have noted that customer orientation should have a positive effect on performance (e.g., Pettijohn, Pettijohn, and Taylor 2007; Saxe and Weitz 1982). Results from Franke and Park’s (2006) meta-analysis were thus quite counterintuitive in showing the relationship between customer orientation and objective sales to be statistically nonsignifi cant. In their interpretation of this fi nding, Franke and Park state that “customer-oriented selling does not consistently lead to sales” (2006, p. 700). This is in contrast, however, to some individual unaggregated studies, where in fact the expected relationship is found. For example, Joshi and Randall’s (2001) study shows that in a direct sell-ing environment, customer orientation had a positive effect on salesperson’s performance, measured in terms of self-rated achievement of sales objectives. All things considered, we start with the hypothesis that customer orientation does have a positive effect on job performance level:

Hypothesis 1: Customer orientation has a positive effect on sales performance level.

Psychologists have long argued that goal orientations may explain individual differences in the rate of change on performance (e.g., Hofmann, Jacobs, and Baratta 1993). We posit that the salesperson’s drive to serve the customer plays an important role in explaining performance improvement. Customer-oriented salespeople avoid short-sighted sales tactics that sacrifi ce customer interests, and engage in relationship-building behaviors directed at increasing long-term customer satisfaction (Saxe and Weitz 1982). It makes sense that a deeper focus on meeting customer needs would help to forge stronger relationships across time, paying off in the long term. Actually, researchers have used customer-oriented selling as a surrogate of salespersons’ long-term performance (see Joshi and Randall 2001). Ployhart and Hakel (1998) found that empathy had a signifi cant impact on sales performance growth rate. Empathic salespeople care about customer emotions and have a greater ability to understand and meet customer needs (Widmier 2002). This explains why empathy is a signifi cant driver of salesperson’s customer orientation (Widmier 2002). As also proposed by Franke and Park’s (2006) interpretation, the above

discussion suggests that customer orientation infl uences the rate of performance growth across time periods:

Hypothesis 2: Customer orientation has a positive effect on sales performance growth rate.

Adaptive selling refers to a salesperson’s ability to collect information from the customer, listen to customer input, and respond by altering sales behaviors during customer interac-tions (Jaramillo et al. 2007; Weitz, Sujan, and Sujan 1986). The salesperson modifi es technical, logistic, administrative, fi nancial, and organizational practices (Hagberg-Anderson 2006), thereby bringing a more customized, value-producing solution to the customer. The logical consequence of these aspects of adaptive selling is higher levels of sales. Numerous studies have found that adaptive selling is an important driver of salesperson’s performance (e.g., Giacobbe et al. 2006; Jara-millo et al. 2007). Therefore, we hypothesize that

Hypothesis 3: Adaptive selling has a positive effect on sales performance level.

Franke and Park (2006) tested the premise that customer orientation also might play an indirect role in performance through adaptive selling. This mediated formulation was speci-fi ed in their meta-analysis and was used to explore signifi cant shared variance between the two marketing constructs. The specifi ed causal direction is consistent with research indicat-ing that salespeople who display a high level of concern for customers are more likely to adapt their behavior to effectively satisfy individual customer needs and preferences (Saxe and Weitz 1982). Jaramillo et al.’s (2007) study found that adap-tive selling mediates the impact of customer orientation on objective sales performance. That logic, combined with our earlier hypothesis about customer orientation’s direct effect on sales level, leads us to propose the following partially mediating hypothesis in our model:

Hypothesis 4: Adaptive selling partially mediates the impact of customer orientation on sales performance level.1

Control Variable

Experienced salespeople generally are assumed to be more capable of identifying customer needs, more willing to en-gage in adaptive selling, and eager to provide customers with solutions, which ultimately helps them achieve higher overall performance levels (Franke and Park 2006). Also, experienced salespeople tend to have larger, more established installed customer bases, logically resulting—all else equal—in more potential repeat business (Franke and Park 2006). Further, if customer reference and word-of-mouth effects are in play, as is often the case in direct selling, these larger established customer bases could “compound” at incrementally faster

Page 4: Does Customer Orientation Impact Objective Sales Performance

170 Journal of Personal Selling & Sales Management

rates across time. In their study of insurance salespeople, Hof-mann, Jacobs, and Baratta (1993) found that experience in the company had a signifi cant impact on performance growth rates. In their study of sewing machine operators, Deadrick, Bennett, and Russell (1997) also found that experience was related to the rate of performance improvement over time. Together, these effects may cause experience to relate to higher performance levels and growth rates. Thus, we treat sales experience as a control variable affecting customer orienta-tion, adaptive selling, performance level, and performance growth trajectories.

METHODOLOGY

Data

We drew our data from a sample of independent sales con-sultants working for a large direct selling organization. Direct selling is an important context for sales research, as it has become a sizable and rapidly growing part of the sales domain and the economy. The Annual Review of the Direct Selling Association (2006) highlights the importance of this sector in the United States, noting total annual sales of $2.6 billion, 80 percent growth in the past 10 years, and a workforce of over 13.6 million people. Outside the United States, total sales are greater than $105 billion, with 100 percent growth in the last decade, and employment of over 57 million people.

Direct sellers are formal representatives of a company but are also considered to be independent agents. They are not viewed as employees and are not subject to the same kind of control systems found in traditional sales departments. They are required to follow the legal and ethical guidelines set forth by the parent organization (Pratt 2000), albeit through a less formal structure. Direct salespeople typically work with a sales director who is responsible for ensuring adherence to the overall policies of the fi rm. Beyond that guidance, direct salespeople often organize in more informal collaborative working groups of several independent representatives to ex-change information and selling tips, hold sales workshops to create shared opportunities, and assist each other by loaning inventory (Sparks and Schenk 2001).

We started with a sample of 1,266 independent sales repre-sentatives of a single large direct selling organization, randomly selected from across the United States. These individuals received online access to a questionnaire that we designed. Our executive contact posted this survey on a company Web site and sent an e-mail invitation to the sample, stressing the importance of the project to the fi rm. The invitation was sent from corporate headquarters, worded with contextually ap-propriate terminology, and contained typical company logo headings and other familiar icons of the fi rm. It also contained a password granting access to the survey for each potential

respondent. Data were accumulated through the Web site and electronically transmitted to the researchers.

A total of 608 respondents completed the survey, 455 of which had matched sales data from company records, render-ing an effective response rate of 36 percent. Because the data were collected quickly by electronic means over a two-week period, assessment of early versus late respondents is not war-ranted. Descriptively, respondents came from 49 states and Puerto Rico. Respondents were all female due to the nature of the products sold by this particular organization. Participant ages ranged from 19 to 78 years, with an average age of 39.86 years (standard deviation [SD] = 11.1). The average tenure with the organization was 3.86 years (SD = 3.67), ranging from 0.1 to 26.8 years of experience. Among respondents reporting their ethnicity, the majority were Caucasians (60 percent) followed by Hispanics (34 percent) and African Americans (6 percent).

Longitudinal Dependent Variable Measurement

In addition to the survey data, the company provided us with 12 months of sales data matched at the individual sales representative level. These sales fi gures came from a separate and independent internal company database. At the time of the data capture, six months of postsurvey sales data were available. We also had access to sales data for the six months immediately prior to the survey, creating an equal number of pre- and postobservations exactly balanced around the survey administration. To smooth some of the month-to-month variability in sales fi gures, we aggregated the 12 months of sales data into quarterly groupings (e.g., Ployhart and Hakel 1998). This produced two quarterly presurvey observations (Time 1 and Time 2) and two quarterly postsurvey observa-tions (Time 3 and Time 4). Of the 608 respondents, 455 had matched sales data from all four time periods (74.8 percent of respondents). We used these individuals with complete matches to the four-period sales data to perform our analysis. We also applied a logarithmic transformation to correct posi-tive skewness inherent in the raw dollar measures (Bollen and Curran 2006; Sriram, Balachander, and Kalwani 2007).

Measures

The constructs depicted in Figure 1 were operationalized with items taken from established scales. Table 1 presents the scale items and relevant summary statistics for each of the measures. An initial confi rmatory factor analysis (CFA) was used to test the measurement properties of these scales in our sample (Gerbing and Anderson 1988). Resulting indices suggest adequate fi t (χ2 = 19.78, degrees of freedom [df ] = 8, p < 0.01; root mean square error of approximation [RMSEA] = 0.071, CI

90% 0.031 to 0.011; comparative fi t index

Page 5: Does Customer Orientation Impact Objective Sales Performance

Spring 2009 171

[CFI] = 0.99; goodness-of-fi t index [GFI] = 0.94, normed fi t index [NFI] = 0.98). All intraconstruct indicator loadings (λ) were signifi cant at α = 0.01, thus providing evidence of convergent validity (cf. Speier and Venkatesh 2002). Evidence of discriminant validity is indicated by a comparison of a two-factor confi rmatory model (χ2 = 19.78) that treats customer orientation and adaptive selling as separate constructs with a one-factor confi rmatory model (χ2 = 336.17) that treats them as a single construct. The difference in χ2

df = 1 was signifi cant

at α = 0.01, in support for the two-factor model (Gerbing and Anderson 1988).

Results also evidence adequate reliability (Bagozzi and Yi 1988). As shown in Table 1, all Cronbach alphas and structural equation modeling (SEM) composite reliability indices are above 0.7. In addition, all average variance extracted indices (ρ

v ) were above 0.5. Finally, we note that the longitudinal

dependent variable—actual sales—originated from sources that were independently measured and differently scaled. This enhances the validity of our fi ndings because it minimizes the possibility of empirically observed relationships occur-ring simply as a result of common method bias (Podsakoff et al. 2003).

MODEL DEVELOPMENT AND RESULTS

Latent Growth Models

LGM was used to test all hypotheses. LGM is an advanced application of SEM used to analyze longitudinal data. LGM allows researchers to explicitly address questions about pat-

terns in measures observed across multiple points in time (Bollen and Curran 2006; Duncan, Duncan, and Strycker 2006). We leveraged LGM’s capability to model (1) the ini-tial status of a construct (e.g., level of sales performance at the fi rst time period), (2) the longitudinal trajectory of the construct (e.g., rate of sales performance growth across all time periods), and (3) a system of latent explanatory covariates that can be modeled as infl uencers of the initial status and trajectory variables.

LGM estimation typically follows a two-step procedure (Bollen and Curran 2006). An unconditional latent growth model (ULGM) is tested fi rst. Thus, we tested an uncon-ditional linear model (Figure 2) that fi t only unobserved intercept and slope “constructs” to the repeatedly measured dependent variables. The initial level is an “intercept” and the rate of growth is a “slope.” Because these are modeled as “random effects” (i.e., a person-specifi c intercept and slope is estimated for each respondent), also assessed in this step is the degree of between-person variability in these estimates (Bollen and Curran 2006). Statistically signifi cant variability (i.e., nonzero) implies the presence of variance that may be explainable in subsequent modeling steps.

Given that the unconditional model demonstrates evidence of adequate model fi t and statistically signifi cant individual-level variability, a second step is undertaken called the con-ditional latent growth model (CLGM) (Bollen and Curran 2006). In the CLGM, not only are intercepts and slopes modeled, but a system of explanatory covariates can be intro-duced following familiar SEM principles (see Figure 1). These covariates are theoretically justifi ed predictors of the previously

Table 1Scale Items and Scale Statistics

Mean Average (Standard Composite Variance FactorConstruct Name and Items Deviation) α Reliability Extracted Loading

Customer Orientation (Saxe and Weitz 1982) 6.44 0.94 0.95 0.86 (0.82) I offer the product that is best suited to the 0.90 customer’s need. I answer a customer’s question about products 0.95 as correctly as I can. I try to bring a customer with a need together 0.94 with a product that satisfi es that need.Adaptive Selling (Robinson et al. 2002) 5.08 0.87 0.89 0.74 (1.38) When I feel my selling approach is not working 0.82 in a selling situation, I tend to change to another approach. I experiment with different sales approaches. 0.98 I tend to use a wide variety of selling approaches. 0.76

Note: Factor loadings are standardized and signifi cant at α = 0.01.

Page 6: Does Customer Orientation Impact Objective Sales Performance

172 Journal of Personal Selling & Sales Management

modeled ULGM intercept-slope structure. The covariates help to explain “which cases will start high versus low and which will increase steeply and which will not” (Bollen and Curran 2006, p. 9). The entire CLGM system is estimated simulta-neously using the maximum likelihood method. Just as in traditional SEM applications, the causal effects are estimated wherein the theoretical constructs have been disattenuated for the infl uence of measurement error. The usual variety of standard structural equation model fi t indices are produced as outputs. A brief description is given in the Appendix of the functional forms of the ULGM and CLGM.

Unconditional Latent Growth Model Results

Per Bollen and Curran (2006), we estimated the ULGM us-ing LISREL 8.80 (Figure 2). This model fi t well (χ2 = 7.39, df = 5, p > 0.05; RMSEA = 0.034 (CI

90% 0.00 to 0.079);

CFI = 0.993; GFI = 0.995, NFI = 0.980), with the mean intercept and slope estimates signifi cantly differing from 0 (μα = 6.12, t = 78.5, μβ = –0.34, t = –8.86), and with statisti-cally signifi cant individual-level variation suffi cient for further modeling in both the intercepts (σ2

α = 1.53, t = 8.69) and slopes (σ2

β = 0.24, t = 4.96).Results of the ULGM show that (1) the sales performance

level (intercept in natural log dollar units) is positive and different from 0; (2) the sales performance slope trajectory is also different from 0, with the negative sign indicating that sales are decreasing with the passage of time (consistent with the observed quarterly means in Table 2); and (3) there is suffi cient variation in individual-level random intercept and

slope effects to warrant introduction of explanatory covariates. Thus, we move to the CLGM.

Conditional Latent Growth Model Results

CLGMs include a set of variables used to predict the inter-cepts and slopes of the ULGM (Bollen and Curran 2006). Results for our model of Figure 1 indicate that the CLGM has an adequate fi t: χ2 = 69.6, df = 42, RMSEA = 0.038 (CI

90% 0.021 to 0.054), CFI = 0.989, GFI = 0.974, and

NFI = 0.973 (Bagozzi and Yi 1988). Also, as suggested by Bollen and Curran (2006), model fi t was assessed by examin-ing the magnitude of parameter estimates, squared multiple correlations, size of modifi cation indices, and differences between the observed and predicted variance–covariance ele-ments. These inspections further supported the presence of an adequate model. Finally, the model explains 6.3 percent of the variance in performance level2 and 15.3 percent of the variance in performance growth. Given the adequate model diagnostics, the next section describes results for our specifi c hypothesis tests, as summarized in Table 3.

Model Hypotheses

H1 predicted a positive effect of customer orientation (CO) on objective sales level (SL). Contrary to our expectations, this relationship was nonsignifi cant (H1: β = 0.11, p > 0.10). It is, however, consistent with the nonsignifi cant CO → SL fi nding in Franke and Park’s (2006) meta-analysis. Results of additional hypothesis tests help to clarify and extend the

Figure 2Unconditional Latent Growth Model (ULGM)

Page 7: Does Customer Orientation Impact Objective Sales Performance

Spring 2009 173

interpretation of this replicated nonsignifi cant path. In H3, we predicted a positive direct effect of adaptive selling (AS) on sales level. We found support for this hypothesis (H3: β = 0.18, p < 0.05). Thus, the fi nding of a nonsignifi cant CO → SL relationship and a signifi cant AS → SL relationship in our data reaffi rm, and are directly in line with, the fi ndings of Franke and Park’s (2006) meta-analysis.

We predicted in H4 that adaptive selling would partially mediate the impact of customer orientation on sales level (i.e., CO → SL, and CO → AS, AS → SL). Because the impact of customer orientation on sales level was nonsignifi cant in H1, our data do not support this partially mediated structure. However, results show that customer orientation affects sales level through a fully mediated process that involves adaptive

Table 2Pairwise Correlation Coeffi cients

Variables

Measures 1 2 3 4 5 6 7

1. Sales Time 12. Sales Time 2 0.293. Sales Time 3 0.36 0.384. Sales Time 4 0.33 0.39 0.435. Customer Orientation 0.12 0.15 0.23 0.206. Adaptive Selling 0.08 0.19 0.06 0.05 0.357. Experience 0.01 –0.03 0.15 0.13 –0.03 –0.07

Mean 1,000.1 827.0 763.1 687.5 6.44 5.08 3.86Standard Deviation 1,041.8 840.5 687.5 868.0 0.82 1.38 3.97Number of Scale Items — — — — 3 3 —Cronbach’s α — — — — 0.89 0.83 —Composite Reliability — — — — 0.89 0.83 —Average Variance Extracted — — — — 0.74 0.62 —

Note: Correlations coeffi cients with magnitudes greater than 0.11 are signifi cant at α = 0.05.

Table 3Results of the Conditional Latent Growth Model (CLGM)

SP1, UP2 Signifi canceHypothesized Relationships (t-value) (α = 0.05, two-tailed)

H1: Customer orientation → sales level 0.11, 0.19 (1.57) Nonsignifi cantH2: Customer orientation → sales growth rate 0.32, 0.21 (3.60) Signifi cantH3: Adaptive selling → sales level 0.18, 0.20 (2.42) Signifi cantH4: Adaptive selling partially mediates the impact of Partial mediation iscustomer orientation on sales performance level. nonsignifi cant Customer orientation → adaptive selling 0.42, 0.65 (8.71) Mediation is signifi cant3

Control Variable Experience → sales level 0.008, –0.000 (0.007) Nonsignifi cant Experience → sales growth rate 0.263, 0.003 (3.40) Signifi cant Experience → adaptive selling –0.053, –0.001 (–1.14) Nonsignifi cant Experience → customer orientation –0.042, –0.001 (–0.88) Nonsignifi cant Adaptive selling → sales growth rate –0.140, –0.06 (–1.49) Nonsignifi cantPercentage of Variance Explained (R 2) Sales level (intercept) 0.063 — Sales growth rate (slope) 0.153 — Adaptive selling 0.180 — Customer orientation 0.002 —

Notes: 1 SP = standardized path. 2 UP = unstandardized path. 3 Mediation requires: (1) signifi cance of the CO → AS relationship, (2) signifi cance of the AS → SL relationship (after controlling for the CO → SL relationship), and (3) a nonsignifi cant CO → SL relationship (Kenny 2008).

Page 8: Does Customer Orientation Impact Objective Sales Performance

174 Journal of Personal Selling & Sales Management

selling. As indicated by Kenny (2008), full mediation can be concluded under conditions of (1) signifi cance of the CO → AS relationship (β = 0.42, p < 0.01), (2) signifi cance of the AS → SL relationship (β = 0.18, p < 0.05, after control-ling for the CO → SL relationship), and (3) a nonsignifi cant CO → SL relationship (β = 0.11, p > 0.10). This is also in line with results from Jaramillo et al.’s (2007) study.

Perhaps of central interest to the core purpose of our paper, and an explicit benefi t of our modeling approach, is the test of H2—that customer orientation affects longitudinal sales trajectory. We found clear support for this hypothesis in our data (H2: β = 0.32, p < 0.01). The experience control variable was the only other signifi cant infl uence on sales trajectory (β = 0.26, p < 0.01).3

Taken together, the results of our hypothesis tests clarify and extend an understanding of how customer orientation affects sales performance. In our data, customer orientation did not impact sales performance level directly, consistent with Franke and Park (2006). It did affect sales performance in two ways: (1) by its indirect effect on sales level through adaptive selling, and (2) by its direct effect on sales growth trajectories over time. Regarding our test of this second means of infl u-ence, as called for by Franke and Park (2006), we know of no other study that has yet brought empirical evidence to bear on this clarifi cation/extension of the positive role of customer orientation on longitudinal sales performance.

DISCUSSION

A well-known quip has long been recognized regarding sales: “nothing happens until somebody sells something” (Gordon 1965, p. 25). This acknowledged connection between sales and the fi rm as a whole magnifi es the importance of salesperson performance and helps to explain why, despite meta-analyses of hundreds of previous studies (e.g., Brown and Peterson 1993; Churchill et al. 1985; Franke and Park 2006; Vinchur et al. 1998), new attention to sales performance continues to emerge in leading marketing literature (e.g., Evans et al. 2007; Hunter and Perreault 2007; Piercy et al. 2006). However, with few exceptions (e.g., Hofmann, Jacobs, and Baratta 1993; Ployhart and Hakel 1998), extant sales research has mostly relied on static data to explain the level aspect of performance, to the neglect of its longitudinal trajectory side.

Our paper addresses this limitation by testing a model that explains the effect of customer orientation, adaptive selling, and experience simultaneously on the level and trajectory as-pects of sales performance. Firms are rightly concerned about variables that can predict whether a salesperson is a top or a bottom performer (a “level” question). But, in addition, fi rms must be concerned about what drives a salesperson to sustain, decrease, or increase their initial level of performance (a “trajec-tory” question). As stated by one prominent leader in direct

selling, “The world is rife with fl ashes in the pan—people who perform well for short periods but fi zzle out over time . . . a strong beginning is a good thing only when coupled with a strong fi nish” (Ash 1995, p. 52).

We modeled longitudinally observed sales performance in dollars using LGM. Our study confi rms that adaptive selling has a direct effect on performance level but does not show an impact on sales growth. Conversely, our study also shows that customer orientation explains individual differences in perfor-mance growth but does not have a direct effect on performance level. This is an important fi nding because it evidences that performance level and performance growth may have a distinct set of predictors, in terms of signifi cance and magnitude. As Deadrick, Bennett, and Russell assert, “the determinants of ini-tial performance differ from the determinants of performance improvement” (1997, p. 755). Our fi ndings also bring further support to Ployhart and Hakel’s (1998) claim that studying sales performance trends over time is vital because it avoids the possible erroneous conclusions that can result from studies that rely on performance data collected at one time period. The enhanced understanding offered by the type of model-ing in our research is critical to the broad overarching goal of explaining sales performance more comprehensively.

Understanding the longitudinal aspect of performance is critical because performance is a dynamic process (see Walker, Churchill, and Ford 1979). Researchers have claimed a longer-term payoff of customer-oriented approaches through the development of mutually reinforcing exchanges and relation-ships that tend to deepen and strengthen over time (Franke and Park 2006). Yet again, most studies examining the effects of those constructs on sales performance also rely on static data (Saxe and Weitz 1982; Siguaw, Brown, and Widing 1994; Weitz, Sujan, and Sujan 1986). In fact, static data are often recognized as a serious limitation of extant empirical research (e.g., Franke and Park 2006; Siguaw, Brown, and Widing 1994). For instance, Williams and Plouffe’s (2007) review of 1,012 sales articles published in 15 leading marketing journals found that less than 1 percent of all articles include longitudinal data.

By contrast, our study used truly longitudinal performance data to address questions about factors that account for in-dividual differences in rates of change in sales performance over time. We found that customer orientation and sales experience relate directly to sales performance growth. Thus, salespeople who are more likely to increase their initial level of performance (or decrease at a lower rate) are those who truly care about customer needs and who have been in the company for a long time. Other factors not included in this study may also affect performance growth rates. For instance, researchers have argued that differences in cognitive ability (Sturman, Cheramie, and Cashen 2005), learning orienta-tion (Hofmann, Jacobs, and Baratta 1993), and age (Avolio,

Page 9: Does Customer Orientation Impact Objective Sales Performance

Spring 2009 175

Waldman, and McDaniel 1990) may also affect the trajectory of job performance. Future research in sales that investigates the impact of these factors on the dynamics of performance is warranted. Dynamic models are available and appropriate for such explorations, including the LGM approach we used, or the hierarchical linear modeling approach used by others in the past (e.g., Mathieu, Ahearne, and Taylor 2007).

While clarification of customer orientation’s role was central to our study following the call from Franke and Park (2006), we also note that the effect of experience on trajec-tory is explainable in light of theories of relationship selling. Salespeople who remain loyal to the fi rm are better able to build and maintain long-term relationships with customers (Palmatier et al. 2006). Positive consequences of building long-term relationships include cumulative fi nancial effects that grow across time (Reichheld 1996). Again, all else equal, we clearly would expect better long-term performance from well-established relationships with customers, as opposed to relationships that are right-censored by employee departures, or those that have not had suffi cient time to develop and mature in broader and deeper ways (Bolton, Lemon, and Verhoef 2004).

Taken together, these are important fi ndings with clear managerial implications. They stress the need to hire and retain customer-oriented salespeople who, in the short run, can adaptively sell in transactions, and who, in the long run, can help to build lasting customer-focused relationships. There also may be sales training implications. Our fi ndings suggest that a customer-centric point of view helps salespeople to bet-ter recognize cues from customers and prospects that indicate when an approach is not working, and to implement adaptive selling techniques to get back in alignment with whatever is valued by a particular customer or prospect.

Limitations and Future Research

Whereas our sales performance data were truly longitudinal, the self-reported construct measures were captured at a single point in time (time-invariant latent covariates). LGM can also accommodate longitudinal latent covariates and outcomes. There may be “virtuous circles” of mutually reinforcing effects between the covariates or sales performance that longitudinal modeling approaches such as LGM can be used to explore. We were unable to test such notions with our data. The con-straints of accessing a large organization’s sales force made reasonable and feasible for us only a single administration of the survey.

An additional limitation is the placement of our time-invariant covariates in the center of the longitudinal sales per-formance sequence. This means two presurvey observations are predicted by a later measurement, posing a technical problem for causal interpretation. However, given the reliability of the

well-established constructs used, we would expect extremely high test–retest reliability correlations with measures obtained at the beginning of the dependent variable sequence—that is, six months earlier. According to standards in psychological test theory, six months is a reasonable minimum for computing such a coeffi cient of stability (Crocker and Algina 1986). Thus, we would expect nearly identical responses had our measures been obtained six months prior. Further, it is not any pre- or postmeasure in isolation being predicted in the longitudinal portion of our analysis. Rather, LGM estimates the slope values for the entire sequence taken as a whole. It is those whole-sequence estimates that are actually predicted in the model. Finally, we note that LGM’s static tests of the intercept explicitly model the fi rst time point in the sequence, making those tests similar to many cross-sectional sales stud-ies that use survey data to model prior objective performance measures (e.g., Jaramillo et al. 2007).

Other limitations include the fact that our data were drawn from a single company, in a direct selling context, with an exclusively female sales force. This specifi city raises questions about generalizability. The concern may be assuaged partly by clear replications found in the intercept component of our modeling. Confi rmation of previous research fi ndings from other diverse contexts provides some reassurance that our data are not especially anomalous. Finally, our longitudinal study relied on a one-year period, which could be deemed too short to capture the dynamic aspect of performance. However, di-rect selling is characterized by a short selling cycle, and thus, this period was viewed as suffi cient to explore the impact of customer orientation and experience on performance growth (Deadrick, Bennett, and Russell 1997).

Given the gaping dearth of longitudinal studies as revealed in Williams and Plouffe (2007), our research minimally adds one empirical study to an otherwise underpopulated body of work. Further, since we know of no other published longitudinal sales research which has yet applied LGM, this study also offers an introductory application of an important methodological approach available for other studies beyond the particular limitations of our data. Regarding such future research, extensions to other types of samples, conditions, and sales contexts are clearly in order.

With these limitations and needed extensions noted, our research still offers contributions to the sales literature by assessing salesperson performance longitudinally, and by implementing a powerful emergent approach to modeling longitudinal trajectories, to clarify two ways that customer orientation may infl uence sales performance. Future research should acknowledge that sales performance has both level and trajectory components, and that application of techniques such as LGM to model both simultaneously may uncover different sets of forces and dynamics driving these distinct components.

Page 10: Does Customer Orientation Impact Objective Sales Performance

176 Journal of Personal Selling & Sales Management

NOTES

1. Following Franke and Park (2006), we also tested a competing model with adaptive selling infl uencing customer orientation. In the competing model, this path was signifi cant, with model fi t, diagnos-tics, and all other estimates and signifi cance levels extremely similar to those for the model of Figure 1. We prefer as more plausible the notion that the “mind-set” of customer orientation infl uences the “behaviors” of adaptive selling.

2. Results are comparable with Franke and Park’s (2006) meta-analysis that found that the combined effect of gender, experience, job satisfaction, customer orientation, and adaptive selling predicted only 11.2 percent of the variance in objective performance.

3. In testing our key hypothesis of the effect of customer orienta-tion on sales trajectory, we controlled for the potential infl uence of adaptive selling. Adaptive selling and sales trajectory showed a statistically nonsignifi cant relationship.

REFERENCES

Ash, Mary Kay (1995), Mary Kay: You Can Have It All, Rocklin, CA: Prima.

Avolio, Bruce J., David A. Waldman, and Michael A. McDaniel (1990), “Age and Work Performance in Nonmanagerial Jobs: The Effects of Experience and Occupational Type,” Academy of Management Journal, 33 (2), 407–422.

Bagozzi, Richard P., and Youjae Yi (1988), “On the Evaluation of Structural Equation Models,” Journal of the Academy of Marketing Science, 16 (Winter), 74–94.

Bollen, Kenneth A., and Patrick J. Curran (2006), Latent Curve Models: A Structural Equation Perspective, Hoboken, NJ: John Wiley & Sons.

Bolton, Ruth N., Katherine N. Lemon, and Peter C. Verhoef (2004), “The Theoretical Underpinnings of Customer Asset Management: A Framework and Propositions for Future Research,” Journal of the Academy of Marketing Science, 32 (3), 271–292.

Brown, Steven P., and Robert A. Peterson (1993), “Anteced-ents and Consequences of Salesperson Job Satisfaction: Meta-Analysis and Assessment of Casual Effects,” Journal of Marketing Research, 30 (February), 63–77.

Churchill, Gilbert A., Neil M. Ford, Steven W. Hartley, and Orville C. Walker, Jr. (1985), “The Determinants of Sales-person Performance: A Meta-Analysis,” Journal of Marketing Research, 22 (May), 103–118.

Crocker, Linda, and James Algina (1986), Introduction to Clas-sical and Modern Test Theory, New York: Holt, Rinehart and Winston.

Deadrick, Dianna L., Nathan Bennett, and Craig J. Russell (1997), “Using Hierarchical Linear Modeling to Examine Dynamic Performance Criteria Over Time,” Journal of Management, 23 (6), 745–757.

Direct Selling News (2006), Annual Review, Lake Dallas, TX: Direct Selling Association.

Duncan, Terry E., Susan C. Duncan, and Lisa A. Strycker (2006), An Introduction to Latent Variable Growth Curve Modeling, 2d ed., Mahwah, NJ: Lawrence Erlbaum.

Ehert, Michael (2004), “Managing the Trade-Off Between Re-lationships and Value Networks: Towards a Value-Based Approach for Customer Relationship Management in Business-to-Business Markets,” Industrial Marketing Manage-ment, 33 (August), 465–473.

Evans, Kenneth R., Timothy D. Landry, Po-Chien Li, and Shaoming Zou (2007), “How Sales Controls Affect Job-Related Outcomes: The Role of Organizational Sales-Related Psychological Climate Perceptions,” Journal of the Academy of Marketing Science, 35 (September), 445–459.

Franke, George R., and Jeong-Eun Park (2006), “Salesperson Adaptive Selling Behavior and Customer Orientation: A Meta-Analysis,” Journal of Marketing Research, 43 (Novem-ber), 693–702.

Gerbing, David W., and James C. Anderson (1988), “An Updated Paradigm for Scale Development Incorporating Unidimen-sionality and Its Assessment,” Journal of Marketing Research, 25 (May), 186–192.

Giacobbe, Ralph W., Donald W. Jackson, Jr., Lawrence A. Crosby, and Claudia M. Bridges (2006), “A Contingency Approach to Adaptive Selling Behavior and Sales Performance: Selling Situations and Salesperson Characteristics,” Journal of Person-al Selling & Sales Management, 26, 2 (Spring), 114–142.

Gordon, Howard L. (1965), “What Is the Next Breakthrough in Research?” Journal of Marketing, 29 (January), 25–27.

Hagberg-Anderson, Asa (2006), “Does Adaptation Pay Off?” In-dustrial Marketing Management, 35 (February), 202–209.

Hofmann, David A., Rick Jacobs, and Joseph E. Baratta (1993), “Dynamic Criteria and the Measurement of Change,” Journal of Applied Psychology, 78 (2), 194–204.

Homburg, Christian, and Ruth M. Stock (2005), “Exploring the Conditions Under Which Salesperson Work Satisfaction Can Lead to Customer Satisfaction,” Psychology & Marketing, 22 (5), 393–420.

Hunter, K. Gary, and William D. Perreault (2006), “Sales Tech-nology Orientation, Information Effectiveness, and Sales Performance,” Journal of Personal Selling & Sales Management, 26, 2 (Spring), 95–113.

———, and ——— (2007), “Making Sales Technology Effec-tive,” Journal of Marketing, 71 (January), 16–34.

Jaramillo, Fernando, William B. Locander, Paul E. Spector, and Eric G. Harris (2007), “Getting the Job Done: The Moderat-ing Role of Initiative on the Relationship Between Intrinsic Motivation and Adaptive Selling,” Journal of Personal Selling & Sales Management, 27, 1 (Winter), 59–74.

Jones, Eli, Paul Bush, and Peter Dacin (2003), “Firm Market Orientation and Salesperson Customer Orientation: Inter-personal and Intrapersonal Infl uences on Customer Service and Retention in Business-to-Business Buyer–Seller Relation-ships,” Journal of Business Research, 56, 4 (April), 323–340.

Joshi, Ashwin W., and Sheila Randall (2001), “The Indirect Effects of Organizational Controls on Salesperson Performance and Customer Orientation,” Journal of Business Research, 54, 1 (October), 1–9.

Kenny, David A. (2008), “Mediation: Baron and Kenny Steps,” David A. Kenny’s Webpages (May 20) (available at http://davidakenny.net/cm/mediate.htm).

Page 11: Does Customer Orientation Impact Objective Sales Performance

Spring 2009 177

Langerak, Fred (2001), “Effects of Market Orientation on the Behaviors of Salespersons and Purchasers, Channel Rela-tionships, and Performance of Manufacturers,” Interna-tional Journal of Research in Marketing, 18, 3 (September), 221–234.

Mathieu, John, Michael Ahearne, and Scott R. Taylor (2007), “A Longitudinal Cross-Level Model of Leader and Salesperson Infl uences on Sales Force Technology Use and Performance,” Journal of Applied Psychology, 92, 2 (March), 528–537.

Palmatier, Robert W., Rajiv P. Dant, Dhruv Grewal, and Ken-neth R. Evans (2006), “Factors Infl uencing the Effectiveness of Relationship Marketing: A Meta-Analysis,” Journal of Marketing, 70 (October), 136–153.

Pettijohn, Charles E., Linda S. Pettijohn, and A.J. Taylor (2007), “Does Salesperson Perception of the Importance of Sales Skills Improve Sales Performance, Customer Orientation, Job Satisfaction, and Organizational Commitment, and Reduce Turnover?” Journal of Personal Selling & Sales Man-agement, 27, 1 (Winter), 75–88.

Piercy, Nigel F., David W. Cravens, Nikala Lane, and Douglas W. Vorhies (2006), “Driving Organizational Citizenship Behaviors and Salesperson In-Role Behavior Performance: The Role of Management Control and Perceived Orga-nizational Support,” Journal of the Academy of Marketing Science, 34 (Spring), 244–262.

Ployhart, Robert E., and Milton D. Hakel (1998), “The Sub-stantive Nature of Performance Variability: Predicting In-terindividual Differences in Intraindividual Performance,” Personnel Psychology, 51 (4), 859–901.

Podsakoff, Philip M., Scott B. MacKenzie, Jeong-Yeon Lee, and Nathan P. Podsakoff (2003), “Common Method Bias in Behavioral Research: A Critical Review of the Literature and Recommended Remedies,” Journal of Applied Psychology, 88 (October), 879–903.

Pratt, Michael G. (2000), “The Good, the Bad, and the Am-bivalent: Managing Identifi cation Among Amway Dis-tributors,” Administrative Science Quarterly, 45 (September), 456–493.

Reichheld, Frederick F. (1996), The Loyalty Effect, Boston: Har-vard Business School Press.

Robinson, Leroy, Greg W. Marshall, William C. Moncrief, and Felicia G. Lassk (2002), “Toward a Shortened Measure of Adaptive Selling,” Journal of Personal Selling & Sales Man-agement, 22, 2 (Spring), 111–119.

Saxe, Robert, and Barton A. Weitz (1982), “The SOCO Scale: A Measure of the Customer Orientation of Salespeople,” Journal of Marketing Research, 19 (August), 343–351.

Siguaw, Judy A., Gene Brown, and Robert E. Widing (1994), “The Infl uence of Market Orientation of the Firm on

Sales Force Behavior and Attitudes,” Journal of Marketing Research, 31 (February), 106–116.

Sparks, John R., and Joseph A. Schenk (2001), “Explaining the Effects of Transformational Leadership: An Investigation of the Effects of Higher-Order Motives in Multilevel Market-ing Organizations,” Journal of Organizational Behavior, 22 (December), 849–869.

Speier, Cheri, and Viswanath Venkatesh (2002), “The Hidden Minefi elds in the Adoption of Sales Force Automation Technologies,” Journal of Marketing, 66 (July), 98–111.

Spiro, Rosann L., and Barton A. Weitz (1990), “Adaptive Selling: Conceptualization, Measurement, and Nomological Valid-ity,” Journal of Marketing Research, 27 (February), 61–69.

Sriram, S., Subramanian Balachander, and Manohar U. Kalwani (2007), “Monitoring the Dynamics of Brand Equity Us-ing Store-Level Data,” Journal of Marketing, 71 (April), 61–78.

Sturman, Michael C., Robin A. Cheramie, and Luke H. Cashen (2005), “The Impact of Job Complexity and Performance Measurement on the Temporal Consistency, Stability, and Test-Retest Reliability of Employee Job Performance Rat-ings,” Journal of Applied Psychology, 90 (2), 269–283.

Vinchur, Andrew J., Jeffrey Shippmann, Fred S. Switzer, and Philip L. Roth (1998), “A Meta-Analytic Review of Predic-tors of Job Performance for Salespeople,” Journal of Applied Psychology, 83 (August), 586–597.

Walker, Orville C., Gilbert A. Churchill, and Neil M. Ford (1979), “Where Do We Go From Here? Some Selected Conceptual and Empirical Issues Concerning the Motiva-tion and Performance of Industrial Salespeople,” in State of the Art and Future Research Needs, Gerald Albaum and Gilbert A. Churchill, eds., Eugene, OR: University of Or-egon Press, 10–75.

Weitz, Barton A., Harish Sujan, and Mita Sujan (1986), “Knowl-edge, Motivation, and Adaptive Behavior: A Framework for Improving Sales Effectiveness,” Journal of Marketing, 50 (October), 174–191.

Widmier, Scott (2002), “The Effects of Incentives and Personality on Salesperson’s Customer Orientation,” Industrial Market-ing Management, 31, 7 (October), 609–615.

Williams, Brian C., and Christopher R. Plouffe (2007), “Assess-ing the Evolution of Sales Knowledge: A 20-Year Content Analysis,” Industrial Marketing Management, 36 (May), 408–419.

Williams, Larry J., Jeffrey R. Edwards, and Robert J. Vandenberg (2006), “Recent Advances in Causal Modeling Methods for Organizational and Management Research,” Journal of Management, 29 (6), 903–936.

Page 12: Does Customer Orientation Impact Objective Sales Performance

178 Journal of Personal Selling & Sales Management

APPENDIXLATENT GROWTH MODEL

We are necessarily limited here and provide only a short de-scription of latent growth modeling (LGM) features relevant to our application. More complete technical details and ap-plication possibilities may be found in Bollen and Curran (2006) and Duncan, Duncan, and Strycker (2006).

Unconditional and Conditional Latent Growth Models

The assumption of a linear unconditional model is that the trajectory of the observed longitudinally measured dependent variable (y) is explained by Equation (1):

yit = α

i + λ

i + ε

it, (1)

where yit is the dependent variable for the ith person at time

t, αi is a latent random intercept for case i, and β

i is a latent

random slope for case i. The λt parameter is a constant, which

by convention in linear trajectories takes on values = t – 1. Thus, with four observations, λ

1 = 0, λ

2 = 1, λ

3 = 2, and

λ4 = 3. Finally, ε

it is a disturbance term with E (ε

it ) = 0 for

all i and t.Also estimated across all individual growth curves are

aggregate parameters for the mean and variance of the ar-ray of individual intercepts (μα and σ2

α), and the mean and variance of the array of individual slopes (μβ and σ2

β). Each individual’s intercept and slope may be expressed in relation to the aggregate-level parameters according to the intercept and slope equations for unconditional latent growth modeling (ULGM), where

αi = μα + δαi

(2)

βi = μβ + δβi

. (3)

In the conditional model, covariates are introduced to ex-plain variation in α

i and β

i. Now, the random intercepts and

slopes are a function not only of the overall mean parameters but also of a set of covariates X

1 ... X

k. Thus, the intercept

and slope equations for the conditional latent growth model (CLGM) are expanded to

αi = μα + γα1

X1i + γα2

X2i ... + γαk

Xki + δαi

(4)

βi = μβ + γβ1

X1i + γβ2

X2i ... + γβk

Xki + δβi

. (5)

The X1 ... X

k covariates may be observed variables, or they

may be latent constructs indicated by a set of observed vari-ables, or a combination thereof. We also note that some Xs may be specifi ed as exogenous (ksi “ξ” in LISREL notation) while some Xs may be specifi ed as endogenous (eta “η” in LISREL notation), the latter being infl uenced in the model by other covariate Xs (ksis or other etas). This capability, along with others (e.g., model testing, modifi cation indices, latent constructs with multiple indicators, estimation of measure-ment error, fl exibility in specifying a variety of types of effects, etc.), distinguish LGM from other possible approaches to longitudinal estimation (e.g., multilevel models) (see Bollen and Curran 2006 for contrasting approaches).

Estimating the Latent Growth Models

The ULGM in our study is depicted in Figure 2. The yit of

Equation (1) are the four quarterly sales performance values (for the ith salesperson at time t). The pattern of individual sales performance over those four time periods is explained by two individual-level random effect parameters—the in-tercept (α

i ) and the slope (β

i ). Also shown are the aggregate-

level means and variances across individuals on the intercept estimates (μα and σ2

α ), and on the slope estimates (μβ and σ2

β ). Then, as depicted in Figure 1, our CLGM models that structure as a function of a causal system of explanatory latent covariates.

Page 13: Does Customer Orientation Impact Objective Sales Performance