Introduction

The determinants of sales performance may be some of the most studied yet most elusive variables in the marketing literature. The importance of sales performance to researchers is apparent given that the topic has been researched since 1918 and has yielded several meta-analytical studies (Brown and Peterson 1994; Churchill et al. 1985; Kohli 1985; Verbeke et al. 2011). From a practical standpoint, the Bureau of Labor Statistics’ (2010) projected 37.8% increase in the 2.8 million sales representatives in the United States by 2018 has huge implications for the recruiting, selection, and retention of well-qualified and high-performing candidates.

As the demand for well-qualified salespeople intensifies, the question of whether good salespeople are “born” or “made” will inevitably arise. With the underlying assumption that, to some extent, good salespeople are born, sales force hiring practices often target individuals with competitive backgrounds (Plotkin 1987) because competitiveness is thought to contribute to a person’s drive to succeed (Brown et al. 1998). Competitiveness has also been identified as a key personality trait of successful salespeople (Brewer 1994) and has been empirically shown to be positively related to sales performance (Brown and Peterson 1994; Carsrud and Olm 1986; Plotkin 1987). However, on its own, trait competitiveness has been able to account for only a small amount of the variance in sales performance. In fact, neither salesperson trait competitiveness nor any other personal characteristic has been able to explain much of the variation in sales performance (Brown et al. 1998; Churchill et al. 1985; Ford et al. 1985; Robie et al. 2005; Verbeke et al. 2011; Wang and Netemeyer 2002). What is more, despite agreement in the literature that supervisory behaviors positively influence sales performance, “the effects of organizational and environmental variables are inherently inconsistent in the direction of their influence on sales performance” (Verbeke et al. 2011, p. 412).

Since Churchill et al.’s (1985) seminal meta-analysis of sales performance, which spanned decades and included more than 1,500 correlations, concluded that “no one of the predictors themselves account for a great amount of variation in [sales] performance—less than 10% on average” (p. 113), they suggested a focus on determinants of sales performance that are “influenceable” in terms of “selecting, motivating, coaching and training by sales managers” (p. 117). More recently, Verbeke et al.’s (2011) meta-analysis of 268 sales performance determinant studies conducted between 1982 and 2008 revealed results not unlike those of Churchill et al. (1985). However, the analysis by Verbeke et al. (2011) was based on more “finely grained” category subgroups to better reflect our current economic and academic climate. They identified five sub-category sales performance determinants within four categories: selling-related knowledge and degree of adaptiveness from the skill-level category; role ambiguity from the role perceptions category; cognitive ability from the aptitude category; and work engagement from the motivation category. Since selling-related knowledge can be taught, work engagement can be motivated, and adaptability and cognitive ability can be recruited, Verbeke et al. (2011) echoed the importance of influenceable sales performance determinants. One influenceable aptitude factor yet to be examined in a personal selling context is salesperson coachability, which is an individual personality difference that relies on both the competitive context of personal selling and the salesperson–sales manager exchange to be elicited and manifested.

This two-part study contributes to the literature by adapting a measure of athletic coachability and applying it to salespeople in a business-to-business sales context and by exploring the relationship between salesperson coachability and sales manager leadership style. We conceptualize and test relationships among salesperson coachability, salesperson trait competitiveness, sales manager leadership style, and sales performance. We demonstrate that salesperson coachability interacts with sales manager transformational leadership style and salesperson trait competitiveness to enhance sales performance. We also identify the mediating effects of salesperson coachability on the trait competitiveness–sales performance relationship and on the transformational leadership–sales performance relationship. The theoretical and managerial implications of the results are also discussed.

The analogy of competitive sport and business-to-business sales

The analogy of competitive sport has often been used to describe the world of business-to-business sales (Rich 1998), and salespeople have been compared to athletes (Mosca et al. 2010). It has also been argued that the job of a sales manager is comparable to that of an athletic coach in that both are concerned with the learning curves and ultimate performance of those for whom they are responsible (Mosca et al. 2010). Since personal selling, like competitive sport, is an achievement situation in which individuals may receive guidance and support in an effort to enhance performance outcomes for themselves and for their organizations, the concept of athletic coachability (Giacobbi 2000) may be particularly relevant when applied to salespeople in a sales setting.

Salesperson coachability

Drawing on Fridhandler’s (1986) distinction between personality traits and states and on the tenants of interactional psychology, generally speaking, athletic coachability is the result of the interaction of the athlete’s personality traits of achievement motivation, agreeableness, and conscientiousness with such situational influences as a motivating coach and adequate support in the context of competitive sport (Giacobbi 2000). More specifically, coachability is characterized by the “motivation to improve one’s sport skill, inquisitiveness, openness to learning, and trust and respect for the coach and his or her training process” (p. 42). Consistent with Bem’s (1983) triple typology and call to “convert observations of particular persons behaving in particular ways in particular situations into assertions that certain kinds of persons behave in certain kinds of ways in certain kinds of situations” (p. 566), coachability’s multidimensional measure was designed to provide a summary statement of an athlete’s most likely and public behaviors in the context of competitive sport.

Applying the concept of athletic coachability to the personal selling context, we conceptualize salesperson coachability as the degree to which salespeople are open to seeking, receiving, and using external resources to increase their sales performance in a personal selling context. While salesperson coachability is an individual difference, it requires the right situation to be triggered and to manifest. Situational factors that can trigger, or draw out, salesperson coachability include a motivating manager or coach; a sales manager’s emphasis, priorities, values, communication, and leadership style; adequate resources; and peer group support. When elicited, observable coachable behaviors include a salesperson’s willingness to learn from the sales manager or sales coach, the provision of information to the sales manager, a display of trust and respect toward the sales manager or sales coach, the demonstration of flexibility and adaptability to changes in routine, and the desire to seek and deal with feedback and information from other sources so as to improve his or her own sales performance. The potential outcomes of such observable coachable behaviors include enhanced interpersonal relationships and/or sales performance.

In terms of Verbeke et al.’s (2011) classification categories of sales performance determinants, salesperson coachability falls into the aptitude sub-category of personal concerns. Unlike a decontextualized and enduring personality trait, such as competitiveness, salesperson coachability relies on the context of personal selling and on the salesperson–sales manager exchange to be elicited. Coachability is therefore neither strictly a personality nor a situational determinant of sales performance. Rather, coachability is an interactional variable whose impact on sales performance depends not only on the extent to which salespeople are coachable but also on the extent to which sales managers exhibit coaching behaviors.

Sales coaching and sales manager leadership style

Sales coaching is “a sequence of conversations and activities that provides ongoing feedback and encouragement to a salesperson or sales team member with the goal of improving that person’s performance” (Corcoran et al. 1995, p. 118). The sales coach plays a crucial role in empowering, guiding, counseling, and mentoring salespeople to achieve desired performance levels (Mosca et al. 2010). As such, sales coaching has been referred to as one of the most significant opportunities available to an organization to influence salesperson performance (Corcoran et al. 1995).

Although a variety of coaching methods exist, Grant’s (1981) developmental coaching, which is concerned with salesperson skill and knowledge enhancement, hinges on the evolution of the salesperson’s abilities under the direction of the sales manager. Given the interactive nature of developmental coaching (Grant 1981), one can appreciate that the method is more closely related to some leadership styles than to others. Armstrong’s (2001) application of the transformational leadership model (Bass 1985) to the role of athletic coaching revealed that the practices of many successful athletic coaches echo those associated with the transformational leadership style, which in turn results in unparalleled performance. Transformational leadership is often contrasted with supervisory feedback (i.e., Kohli 1985; Teas and Horrell 1981; Teas et al. 1979), which influences sales performance through the manager’s provision of feedback to salespeople contingent on their effort or performance. Because of the give-and-take exchange process associated with supervisory feedback, it is often referred to as transactional leadership (Bryman 1992).

The interaction of personality and situational factors

Sales managers create the context in which salespeople’s personal characteristics manifest and affect sales performance. This underlines the importance of examining the interaction of personality and situational variables when attempting to better explain variation in sales performance. Just as there has been much debate in the sales literature regarding the effects of personal versus situational factors on sales performance, there has long been a debate in the psychology literature concerning the relative role of traits or dispositions (Allport 1937; Cattell 1965; Eysenck 1952) versus the impact of the social context, the situation, or the environment (Skinner 1972; Watson 1924) in determining a person’s behavior (Kendrick and Funder 1988).

Personality traits are enduring and consistent patterns of thoughts, feelings, and behaviors that help distinguish people from one another (Johnson 1997). Personality states are more fleeting behaviors that manifest in specific situations (Fridhandler 1986). Interactionism is the compromise position between the trait and state views that personality and situational factors combine to produce personality states and behaviors (Vealey 1992). Summarized by Giacobbi (2000), the interactionist view of personality is as follows: behavior is a function of the person and the environment/situation in that the person and environment interact to produce behavior (McAdams 1997); individuals select environments that allow their personal characteristics to thrive (Bowers 1973; Snyder 1983); personal characteristics can reciprocally alter the situation a person encounters (Wachtel 1973); and a person’s pattern of responses to situations in a particular domain are idiosyncratic (Endler and Magnusson 1976).

In the sales literature, interactionism has been used to examine the joint effects of personality-by-situation interactions and to demonstrate that the combination of personality traits and organizational characteristics can explain incremental variance beyond the main effects of either type of variable on its own (i.e., Brown et al. 1998; James et al. 1990; Wang and Netemeyer 2002). These results support past research in the organizational behavior and applied psychology literature (i.e., Caldwell and O’Reilly 1990; Holland 1985) that points to the interaction of employees’ personalities with situational factors as a key determinant of such variables as motivation and performance. The impact that the interaction of personality and situational variables can have on sales performance is highlighted by the findings that the combination of trait competitiveness and management practices fostering a competitive organizational climate can result in more effective goal setting and better performance (Brown et al. 1998) and that trait competitiveness has a significant effect on performance when combined with learning and self-efficacy (Wang and Netemeyer 2002).

Conceptual development and hypotheses

Goal perspective theories of achievement motivation posit that task-involvement and ego-involvement goals or purposes exist in achievement situations (Nicholls 1989). Individuals in achievement situations display varying degrees of task and ego orientations depending on their perceptions of their own competence and definitions of success. Those who are more task-oriented tend to focus on task involvement, personal improvement, working hard, and being immersed in the activity. They gauge their success in terms of improvement relative to their own past performance. Those who are more ego-oriented are highly concerned with displaying superior competence in relation to others and, accordingly, define success and competence as defeating others in a competitive situation.

Personal selling is an achievement situation in which we conceptualize salesperson coachability as a display of task orientation in that it is the manifestation of proactively working to get better at selling. We conceptualize salesperson trait competitiveness as an ego orientation since it is concerned with the desire to win or keep a customer’s business from competitors in a competitive market environment. Spence and Helmreich (1983) described trait competitiveness as the “enjoyment of interpersonal competition and the desire to win and be better than others” (p. 41). This definition is consistent with Kohn’s (1992) concept of intentional competitiveness, which is internal and “concerns the desire on the part of the individual to be number one” (p. 4). Further, personal selling is an achievement situation that can be greatly influenced by the sales manager since the manager creates the context in which salespeople are managed, in which they perform, and in which their coachable characteristics are brought to light. While salesperson coachability can be elicited by a sales coach or manager’s emphasis, priorities, values, and means of communication, it is the coach or manager’s leadership style and associated behaviors that are the key contextual features eliciting or inhibiting the level of coachability of a particular subordinate (Giacobbi 2000). Thus individual salespeople vary not only in the degree to which they are coachable but also in terms of how a particular contextual feature will influence the manifestation of their coachable characteristics.

Just as salespeople vary in terms of their coachability, sales managers vary in terms of the coaching behaviors they exhibit. In the relationship between sales manager leadership style and sales performance, the transformational leadership style has been shown to lead to higher sales performance (Humphreys 2002; MacKenzie et al. 2001; Russ et al. 1996; Shoemaker 1999). Since the transformational leadership style exhibits characteristics most similar to successful athletic coaching in practice and research (Armstrong 2001), we expect that transformational leadership creates the most favorable situation for a salesperson’s coachability characteristic to manifest. As sales manager coaching behaviors increase, salesperson coachability characteristics should be triggered to a greater extent, allowing them to have a greater impact on sales performance.

Business-to-business sales is based on overt competition, either between firms for the business of a particular customer or between the selling firm and the prospect itself in a make-or-buy situation. Trait competitiveness has been shown to be positively related to performance (Brown and Peterson 1994; Carsrud and Olm 1986; Plotkin 1987). In other words, as levels of trait competitiveness increase, so do performance levels. In the achievement situation of personal selling, it is expected that the interaction of salesperson coachability, transformational leadership, and trait competitiveness will be such that sales performance will be highest when the level of each is high. Referred to as effect congruence in the interactional psychology literature (Joyce et al. 1982), this more-is-better perspective assumes that the variance explained will continue to improve as additional independent variables that reflect attributes of both the individual and the situation occur at increasing levels (see Fig. 1). We therefore propose the following:

Fig. 1
figure 1

Conceptual model of salesperson coachability, trait competitiveness, transformational leadership style, and sales performance. Note: The large outer box (dashed lines) depicts the hypothesized effect congruence pattern, which suggests that sales performance will be more positive when higher levels of salesperson coachability, salesperson trait competitiveness, and sales manager transformational leadership interact than when the three interact at lower levels

Full size image
  1. H1:

    The interaction among salesperson coachability, trait competitiveness, and sales manager transformational leadership will be such that sales performance will be more positive when salesperson coachability, salesperson trait competitiveness, and sales manager transformational leadership occur at higher levels than at lower levels.

Mediating effect of salesperson coachability on the leadership style–sales performance relationship

The development of Hypotheses 2 and 3 is also built around the model shown in Fig. 1. We begin by discussing the hypothesized mediating effects of salesperson coachability on the transformational leadership–sales performance relationship and then move to the hypothesized mediating effect of salesperson coachability on the trait competitiveness–sales performance relationship.

Sales managers have been called the necessary link between company strategy and its execution in the field (Good 1993). As such, in personal selling and sales management research, the sales manager is often a focal variable of interest that influences sales performance (Brown 1968). Through their leadership styles and accompanying behaviors, sales managers create the context in which salespeople’s personalities and personal characteristics manifest and affect sales performance. Influencing sales performance through the sales force means that salespeople are the necessary link between the sales manager and sales performance.

Managers practicing transformational leadership impact sales performance through a developmental process and their coach-like behaviors (Grant 1981; MacKenzie et al. 2001). We suggest that the extent to which developmental coaching influences sales performance depends on the degree to which a manager triggers the coachable characteristics of his or her subordinates. Since salesperson coachability can impact sales performance only when it is elicited, we expect salesperson coachability to fully mediate the transformational leadership–sales performance relationship. We therefore propose the following:

  1. H2:

    Salesperson coachability will fully mediate the relationship between transformational leadership and sales performance.

Mediating effect of salesperson coachability on the salesperson trait competitiveness–sales performance relationship

Those who are highly competitive constantly monitor their performance in relation to others to make sure that they are surpassing their peers. They attach a great deal of importance to outperforming others and hence exert extra effort in preparation and execution (Kohn 1992). Bartkus et al. (1989) suggest that a Type A behavior pattern, of which competitiveness is a component, affects sales performance through the mediation of work effort. Plouffe et al. (2010) reinforce the argument that “the relative presence, or absence, of trait competitiveness drives some of the salesperson's most basic behaviors” (p. 540). A key contribution of their research was to show that trait competitiveness can result in mediating behaviors—in their case exploratory navigation—that then go on to impact performance.

Since “competitive salespeople are known to be proactive” (Plouffe et al. 2010, p. 546) and since coachability is characterized as being proactive in working toward improving sales performance, we suggest that in the achievement situation of personal selling, competitive individuals recognize that performance gains can be made through being open to the coaching process. This recognition triggers one’s coachable characteristics so that they may enhance performance. We therefore expect salesperson coachability to have a mediating effect on the salesperson trait competitiveness–sales performance relationship and propose the following:

  1. H3:

    Salesperson coachability will have a mediating effect on the relationship between salesperson trait competitiveness and sales performance.

Empirical study

The empirical study is reported in two parts. In Part 1, we adapt the Athletic Coachability Scale (Giacobbi 2000) to assess salespeople in a personal selling context. Here, the focus is on determining the reliability of the adapted measure and its appropriateness for the business-to-business sales context. An additional aim of Part 1 is to explore the relationship between the adapted salesperson coachability construct and constructs already established in sales force research (i.e., sales manager leadership style, trait competitiveness, and sales performance) via correlation and regression analyses. In Part 2, we test our hypotheses to determine whether and to what extent salesperson coachability interacts with sales manager leadership style and salesperson trait competitiveness to enhance sales performance. Existing measures were used and adapted only when necessary to maintain original integrity to facilitate comparisons and generalizations with other sales management research endeavors (Hensel and Bruner 1992).

Part 1: Assessment of the adapted athletic coachability scale

Through industry contacts, four food and beverage consumer packaged goods companies in the U.S. mid-south were solicited to participate in the study, three of which agreed. Of the 135 surveys distributed to this convenience sample of sales representatives, 94 completed surveys were returned, yielding a response rate of 69.6%. Data were collected by self-administered paper and pencil questionnaire, which sales representatives returned to the researcher via self-addressed envelope. Twenty-nine (31%) of the respondents were female and 65 (69%) were male. The average respondent age was 31, and the average number of years in present job was six. The average number of years in selling or in a selling-related profession was nine.

Measures

Salesperson coachability

Giacobbi’s (2000) Athletic Coachability Scale was adapted to assess salespeople in a personal selling context. References to sport were replaced with references to selling, references to the athlete were replaced with references to the salesperson, and references to the coach were replaced with references to the sales manager. The scale’s 24 items were measured on a seven-point scale (see Appendix 1). To increase the validity of the instrument and to guard against introducing systematic error to the scale, the items that were negatively worded in the original measure were positively worded in the adapted scale (Hinkin 1995; Jackson et al. 1993; Schriesheim and Hill 1981).

Sales manager leadership style

The Multifactor Leadership Questionnaire (MLQ) (Bass 1985) was adapted to assess the transformational and transactional leadership behavior of sales managers. Following Bycio et al. (1995) and Martin (2001), we did not assess laissez-faire leadership, making the adapted version a 28-item scale. On a five-point frequency scale, transformational leadership was assessed by using 20 items from the MLQ, and transactional leadership was assessed by using the remaining eight items. The items measuring each leadership style were combined into single composite indicators to allow for the analysis of the primary influence of each leadership style as its own general construct (Dubinsky et al. 1995).

Salesperson trait competitiveness

Trait competitiveness was measured by using Helmreich and Spence’s (1978) four-item, seven-point scale.

Sales performance

Following Yammarino and Dubinsky (1990) and Dubinsky et al. (1995), we used an adapted for self-report five-point, ten-item measure of subjective sales performance to assesses overall work attitude, activity reporting, product knowledge, and problem-solving effectiveness. Although the scale is multidimensional, the items were combined into one composite indicator to allow for the analysis of subjective salesperson performance as a general construct.

Extant research suggests that since subjective and objective salesperson performance measures assess different things, they are not interchangeable (Martin 2001; Rich et al. 1999). Because some variables show consistent relationships across performance measures and others do not, Franke and Park (2006) reinforce Churchill et al.’s (1985) recommendation of using multiple indicators of different types to assess sales performance. Therefore, our survey for Part 2 of the study included two self-report questions pertaining to objective sales performance: percentage of quota achieved for the year (Dalrymple and Strahle 1990; Dubinsky et al. 1995; Martin 2001; Martin and Bush 2006; Weitz 1978) and total sales volume for the most recent full year of employment with the respondent’s present organization (Dubinsky and Hartley 1986; Martin 2001; Martin and Bush 2006; Weeks and Kahle 1990). Though objective sales performance data are most desirably obtained from company sales records or payroll departments, the difficulty of actually obtaining such data has long been established and continues to persist (Churchill et al. 1985; Hunter and Perreault 2007).

Discussion of measure assessment

The primary purpose of Part 1 was to assess the 24-item coachability measure in a personal selling context. Although the other measures intended for use in Part 2 of the study were used in previous academic research studies, Part 1 of this study was undertaken to empirically examine the reliability of the scales by using respondents similar to those who would provide responses for the second part of the study. Table 1 presents the results of the analysis. The coefficient alphas of all measures (salesperson coachability α = .95; transformational leadership α = .94; transactional leadership α = .93; trait competitiveness α = .92; and sales performance α = .87) exceed the minimum levels acceptable by social science research standards (Fornell and Larcker 1981), indicate highly reliable measurement scales, and provide initial support for the reliability of the adapted measure of athletic coachability in a personal selling context.

Table 1 Part 1: Scale reliabilities and correlations of study variables
Full size table

Discussion of study variables

Consistent with our conceptualization that salesperson coachability more positively influences sales performance when a salesperson’s coachable characteristics interact with a sales manager who uses a coaching-based leadership style, salesperson coachability was significantly positively correlated with the transformational leadership style (r = .666). An additional result reinforcing this conceptualization was that salesperson coachability was negatively correlated with the transactional leadership style (r = −.516). These correlations support the notion that the salesperson must encounter the right context to elicit the salesperson’s coachable characteristics so they can positively influence performance (see Table 1).

Table 1 also shows the significant positive correlations between salesperson coachability and trait competitiveness (r = .806) and between salesperson coachability and sales performance (r = .628). Regression analysis confirmed that salesperson coachability R2 = .374 (β = .245, p = .000), trait competitiveness R2 = .781 (β = .450, p = .000), and transformational leadership R2 = .479 (β = .485, p = .000) were all significant positive predictors of sales performance. These results suggest that the relationships proposed for examination in Part 2 of the study are theoretically sound.

Part 2: Hypothesis testing

Four hundred eighty-five surveys were distributed via registration packets to registrants at a healthcare sales and marketing professionals’ conference that was held over three days. Of the 351 surveys that were returned, 80 were deemed unusable because significant portions of the survey were not completed, which yielded a response rate of 56%. The convenience sample comprised salespeople from a cross-section of companies with varying management philosophies and sales cycles and a wide variety of product lines (pharmaceuticals, 43%; animal health, 23.7%; medical equipment, 21.1%; medical devices, 9.3%; surgical supplies, 2.2%; and other, .70%). The study sample is representative of the sales and sales-related occupations as reported by the Bureau of Labor Statistics (2010). With respect to gender, 55.6% of the sample was male and 44.4% female, matching the national averages of males (52.9%) and females (47.1%). The sample’s ethnic composition was also comparable to national averages: Caucasian, 88.5% versus 73.4%; African American, 2.6% versus 9.8%; Asian, 3.0% versus 5.0%; and Hispanic, 3.3% versus 11.8%. Further, 84.1% of the study’s respondents had attained at least a bachelor’s degree. The national average for people in sales and sales-related occupations is 76.7%.

Measure reliability and validity assessment

Our first step in the empirical assessment of the reliability and validity of each measure was the use of exploratory factor analysis (EFA). Before conducting EFA, inter-item correlations were examined via the corrected item to total correlations. Items below .30 (COA3, COA20, PERF8, TRNS2, TRNS5, TRNS6, TRNS7, LEAD13) were eliminated and dropped from further analysis (Bearden et al. 2001; Robinson et al. 1991). We then conducted EFA on the remaining 58 items, employing maximum likelihood extraction with oblique rotation. As per Bartlett’s test (Snedecor and Cochran 1989) of sphericity (χ 2 (1596) = 11984.30, p < .000) and KMO’s measure of sampling adequacy (.933), the sample correlation matrix revealed that it was suitable for factor-analytic procedures. Using Rummell’s (1967) .40 cutoff as the minimum acceptable factor loading, the two items that did not meet this criterion (PERF2, PERF9) were removed and dropped from further analysis.

Although we were expecting a ten-factor solution from the EFA, only nine factors emerged. The coaching feedback and coping with criticism sub-dimensions of the coachability scale had substantial cross loadings. This is consistent with Giacobbi (2000), who suggested that the coaching feedback sub-dimension warranted further analysis because of its potential to cross load with other factors and that further work on the scale might reveal that the component may be better if combined, modified, or dropped. Since criticism is a form of feedback and since it crossed loaded highly with the coaching feedback sub-dimension, the two were combined into a new sub-dimension we called “coping with feedback” and further assessed for reliability and convergent and discriminant validity.

As shown in Table 2, the internal consistency reliability (coefficient alpha) of each measure was conducted and was found to be above the commonly accepted threshold of .70 (Nunnally 1978). Also, the composite reliability of the salesperson coachability and sales manager leadership style scales was conducted as per Fornell and Larcker (1981). The results of the analysis indicate that both scales (salesperson coachability α = .91 and sales manager leadership style α = .90) are reliable. Harmon’s single-factor test suggests that common method bias does not likely pose a problem for this analysis since the variance explained from the first factor (32.9%) was less than 50% (Podsakoff et al. 2003).

Table 2 Part 2: Confirmatory factor analysis of scale items and their measurement properties
Full size table

The measurement items were then subjected to confirmatory factor analysis by using LISREL 8.51 (Joreskog and Sorbom 1993) with the maximum-likelihood estimation to examine construct validity. For the two multidimensional scales, salesperson coachability and sales manager leadership style, composite scores of each dimension’s set of items (sums of scores divided by the number of items) were created, resulting in five composite indicators of salesperson coachability, four composite indicators of transformational leadership, and one composite indicator of transactional leadership. The use of composite scores to represent a partially aggregated model acknowledges the constructs’ multidimensional natures (Bagozzi and Heatherton 1994) while enhancing parsimony and facilitating model estimation for each construct (Pechmann et al. 2003).

Results of the confirmatory factor analysis using the sample covariance matrix as input to LISREL 8.51 (Joreskog and Sorbom 1993) show that each indicator loaded significantly on its respective underlying concept. A variety of fit indices were examined since the chi-square test statistic (χ 2) is not reliable for samples larger than 200 (Cudeck and Henly 1991). With the exception of the χ 2, the results indicate an acceptable fit of the measurement model (χ 2 = 753.93, df = 183, p = .00, Root Mean Square Error of Approximation [RMSEA] = .07; Goodness-of-Fit Index [GFI] = .97; Normed Fit Index [NFI] = .97; Non-normed Fit Index [NNFI] = .98; Comparative Fit Index [CFI] = .98) (Bentler and Bonnett 1980; Hu and Bentler 1999; Joreskog and Sorbom 1993).

As shown in Table 2, the magnitudes of the standardized loading estimates ranged from .60 to .92 and all loadings were significant (i.e., all t-values were larger than 2.00). In addition to these loadings, average variance extracted (AVE) was used to demonstrate convergent validity since all values were greater than .50 (Fornell and Larcker 1981). Discriminant validity is also present since the largest value for shared variance between all pairs of constructs was .36, which is less than the lowest value for AVE (.51) (Fornell and Larcker 1981). Taken together, the measurement model shows that all the measures exhibit reliability and convergent and discriminant validity.

Non-response bias

Over a three-day period, 134 respondents failed to return completed surveys. To assess the potential impact of non-response bias, a time-trend extrapolation test was conducted to compare the characteristics of early and late respondents (Armstrong and Overton 1977). Each survey was dated when received, which resulted in three pools of responses. The first and third pools of responses were considered early and late groups, respectively. Independent sample t-tests were conducted to detect significant differences between early respondents and late respondents in terms of demographic characteristics, salesperson coachability, salesperson trait competitiveness, sales performance, and respondents’ managers’ transformational and transactional leadership styles. Since all t-tests were found to be non-significant, non-response bias was not considered to be a concern for this study.

Results

H1 was tested by using hierarchical regression. Following Brown et al. (1998), we examined the regression weights of the salesperson trait competitiveness, salesperson coachability, and sales manager transformational leadership style constructs at high and low levels to interpret the form of the interaction. To test whether high levels of each construct have a more positive effect on sales performance than do low levels, the regression weights of each construct were examined at one standard deviation above and one standard deviation below each of the variable levels. Examining these “simple slopes” is analogous to analyzing simple main effects in ANOVA (Aiken and West 1992). The results of this analysis were consistent with H1. Salesperson trait competitiveness was positively and significantly related to sales performance at high levels (β = .287, p < .001) but not at low levels (β = − .131, p = .482). Similarly, salesperson coachability was positively and significantly related to sales performance at high levels (β = .450, p = .012) but not at low levels (β = .031, p = .474). Finally, sales manager transformational leadership style was positively and significantly related to sales performance at high levels (β = .351, p = .026) but not at low levels (β = − .237, p = .126). Further, all three constructs were positively and significantly related to sales performance. To provide further support to our focus on transformational leadership, transactional leadership was tested by using the procedures explained above. The regression results indicate that sales manager transactional leadership style is not significantly related to sales performance at either high (β = .081, p = .537) or low levels (β = − .239, p = .187).

A further analysis of the three-way interaction was conducted by using multiple regression (Avolio et al. 1999; Krishnan et al. 2006). The sample was partitioned into groups by using the same procedures as above to examine salesperson coachability, transformational leadership style, and salesperson trait competitiveness at high and low levels. Hierarchical regression was used to test for a three-way interaction and to confirm whether the high-level group produced higher levels of sales performance than the low-level group. The results indicate that the three-way interaction of the variables from the high-level group is significantly higher and more positively related to sales performance (β = .553, p < .01) than it is from the low-level group (β = − .061, p = .811). H1 is therefore supported, which suggests that sales performance is more positive when salespeople are high in trait competitiveness and high in coachability and when the sales manager’s level of transformational leadership is high.

To more closely examine the relationship between salesperson coachability and the other study variables, additional regression analyses were performed. Our results confirmed that salesperson coachability R2 = .242 (β = .365, p < .000), trait competitiveness R2 = .179 (β = .226, p < .000), and transformational leadership style R2 = .077 (β = .195, p < .000) were each significant predictors of sales performance (see Table 3). A multiple regression model was also created by entering trait competitiveness into the equation first, since it is an enduring personality trait, followed by transformational leadership, since the sales manager creates the context in which salespeople are managed and in which they perform, and finally salesperson coachability, since only after being elicited by the situation can salesperson coachability be revealed. The pattern of results is consistent with effect congruence in that the variance explained in sales performance continued to improve as additional independent variables occurred at increasing levels (see Table 4).

Table 3 Part 2: Summary of regression statistics for each variable as an independent predictor of sales performance
Full size table
Table 4 Part 2: Multiple regression statistics for incremental change in variance explained in sales performance
Full size table

Mediating effect of salesperson coachability

The mediating effect of salesperson coachability on the relationship between transformational leadership style and sales performance was tested by using Baron and Kenny’s (1986) method of a series of regression equations. First, salesperson coachability was regressed on sales manager transformational leadership style, which was significant at p = < .00 with an unstandardized beta of .493. Second, salesperson performance was regressed on sales manager transformational leadership style, which was significant at p = <.00 with an unstandardized beta of .195. Finally, salesperson performance was regressed on both sales manager transformational leadership style and salesperson coachability, with salesperson coachability being significant at p = <.00 with an unstandardized beta of .354 and sales manager transformational leadership style being not significant at p = <.607 with an unstandardized beta of .004. That all the conditions held in the predicted direction, coupled with the finding that the effect of sales manager transformational leadership style on salesperson performance was not significant in the third equation when salesperson coachability was controlled, establishes full mediation (see Table 5). The Sobel (1982) significance test for the indirect effect of the independent variable on the dependent variable through the mediator also was significant at p = < .00. This provides support for H2 and suggests that salesperson coachability fully mediates the relationship between transformational leadership style and sales performance.

Table 5 Part 2: Means, standard deviations (SD), and intercorrelations
Full size table

H3 was also supported, which suggests that salesperson coachability does have a mediating effect on the relationship between salesperson trait competitiveness and sales performance. Salesperson coachability’s partially mediating effect on the relationship between salesperson trait competitiveness and sales performance was confirmed by using Baron and Kenny’s (1986) method of a series of regression equations. First, salesperson coachability was regressed on salesperson trait competitiveness, which was significant at p = < .00 with an unstandardized beta of .296. Second, salesperson performance was regressed on salesperson trait competitiveness, which was significant at p = < .00 with an unstandardized beta of .226. Finally, salesperson performance was regressed on both salesperson trait competitiveness and salesperson coachability, and both were significant at p = < .00 with unstandardized betas of .142 and .284, respectively. That all the conditions held in the predicted direction, coupled with the finding that the effect of salesperson trait competitiveness was less in the third equation than in the second equation, establishes partial mediation (see Table 5). The Sobel (1982) significance test for the indirect effect of the independent variable on the dependent variable through the mediator was also significant at p = <.00.

Social desirability bias test

Strahan and Gerbasi’s (1972) social desirability bias scale was included in the survey to determine whether social desirability had any impact on the results of the study, and a confirmatory factor analysis produced an acceptable fit of the underlying social desirability construct (χ 2 = 3.51, df = 2, p = .17; CFI = .99; IFI = .99; NFI = .99; NNFI = .98; RMSEA = .04) (Bentler and Bonnett 1980; Joreskog and Sorbom 1993; Hu and Bentler 1999). The scale’s items were then summed to obtain a composite score, and this variable was given a path to each variable in the original structural model. Since the difference between the original and the modified structural models was Δ χ 2 = 4.83, Δ df = 11, p > .05, social desirability bias appears to have little effect on the standardized parameter estimates, the fit indices, and the study’s hypothesized relationships.

Discussion

The measurement of salesperson coachability and the exploration of its interaction with variables known to be linked to sales performance may help to fill some of the gaps in our understanding of the elusive outcome variable sales performance. Salesperson coachability was explored in the context of transformational leadership (Bass 1985) to demonstrate the importance that context has in eliciting certain personality characteristics. Since transformational leadership style is the style that most resembles coaching (Armstrong 2001) through its use of developmental coaching methods (Grant 1981), it was expected, and our results suggest, that this approach and these methods draw out and allow salesperson coachability to positively affect salesperson performance.

The general pattern of this study’s findings provides support for the overall purpose of this research. With respect to the hypothesized three-way interaction, salesperson coachability, salesperson trait competitiveness, and sales manager transformational leadership style were all positively and significantly related to sales performance when they occurred at high levels but not at low levels. The results of the mediation analyses indicate that salesperson coachability fully mediates the transformational leadership–sales performance relationship and partially mediates the salesperson trait competitiveness–sales performance relationship.

As for the question of whether good salespeople are “born” or “made,” our results suggest that good salespeople, much like athletes, are willing to take a proactive approach to performance improvement. However, this study’s findings suggest that salesperson success is a function of the combination of the salesperson’s personal characteristics with the situation created by the sales manager. While good salespeople may naturally be highly coachable and highly competitive, the sales manager can play a critical role in creating a situation in which these personal characteristics can be elicited and can thrive.

Theoretical contributions

Salesperson coachability is an important “influenceable” cognitive aptitude sales performance determinant that has not to date been studied in sales force research. Supporting the examination of personality-by-situation interactions when undertaking sales research (i.e., Bagozzi 1978, 1980; Brown et al. 1998; Flaherty et al. 2009; Licata et al. 2003; Weitz 1981), this study’s findings serve as another example of the value of adopting an interactionist perspective. In addition to finding support for our hypotheses, our multiple regression analysis highlights how the variance explained in sales performance increases when the interactional, situational personality variable salesperson coachability combines with the personality variable trait competitiveness and the situational variable sales manager leadership style.

MacKenzie et al. (2001) suggest that sales managers who use developmental-type coaching methods have a more positive impact on sales performance and that sales managers may be able to improve their effectiveness substantially by paying more attention to their transformational leadership behavior. Our results extend MacKenzie et al.’s (2001) by demonstrating that transformational leadership behavior influences sales performance through salesperson coachability. As a mediator variable, salesperson coachability appears to represent the process or “generative mechanism” (Baron and Kenny 1986, p. 1173) through which sales manager transformational leadership style influences sales performance. Further, the partially mediating effects of salesperson coachability on the trait competitiveness–sales performance relationship add to the trait competitiveness literature in general and build on the findings of Plouffe et al. (2010) in particular, who noted that “although the trait competitiveness-performance relationship has received a moderate amount of attention in the literature (e.g., Brown et al. 1998; Harris et al. 2005), the relationship has typically been proposed as a direct effect” (p. 546). We provide another example of how trait competitiveness can drive important performance-enhancing behaviors.

Finally, our additional regression analyses of the data drawn from two different sales contexts revealed that salesperson coachability on its own accounted for 37.4% (Part 1) and 24.2% (Part 2) of the variance in sales performance. These results are consistent with Verbeke et al. (2011), who identified cognitive aptitude as a significant sales performance driver. Even more, when combined with salesperson trait competitiveness and transformational leadership, it explained 30.1% of the variance in sales performance. The variance explained is comparable to the 32% predicted by Verbeke et al.’s (2011) multivariate causal model, though their model contained five sales performance antecedents (role ambiguity, cognitive aptitude, degree of adaptiveness, selling-related knowledge, work engagement). Thus, while our mediation analyses confirm that salesperson coachability is an important “influenceable” cognitive aptitude sales performance determinant, our regression analyses suggest that salesperson coachability is an important predictor of sales performance.

Managerial implications

Given the anticipated rise in demand for good salespeople, our study’s findings suggest the importance of recruiting and retaining coachable salespeople while fostering the kind of environment that allows a salesperson’s coachable characteristics to thrive. We conceptualize, formalize, and empirically validate what many successful firms may have already suspected and have been practicing—that highly coachable salespeople may be less effective if their sales managers do not create an environment that facilitates coachability. Alternatively, manager leadership style and accompanying coaching behaviors may be irrelevant if salespeople are low in coachability.

The salesperson coachability measure could be used by sales managers to assess their sales force and discern which dimensions are lacking in poorly performing salespeople, allowing sales managers to focus sales training on overcoming any coachability weaknesses (i.e., working with others, becoming more open to feedback and criticism). Even more, with many firms spending on average four months and more than $7,000 in training for every new salesperson hired (Johnston and Marshall 2011), salesperson coachability may be a useful criterion in the selection process. Identifying in candidates such coachable qualities as inquisitiveness and openness to criticism and such coachable behaviors as reading sales-related books in their spare time and working well with others in the past can easily be incorporated into the interview process. The candidate’s level of coachability could also be more formally and comprehensively assessed by modifying the scale based on the candidate’s most recent subordinate–superior experience. Screening for coachability would not be unlike many sales recruiting and selection procedures that include cognitive tests, integrity-oriented assessments (Hunter and Schmidt 1998), and tests of emotional intelligence (Kidwell et al. 2011).

Limitations

This study’s findings are subject to several limitations. First, while our study shows the importance of coachability as a mediator of sales performance, our results do not indicate that causal relationships exist among the study variables. Further, the salesperson coachability measure was used in only two sales settings with convenience samples. The choice of the healthcare sales industry may be an additional limitation since pharmaceutical salespeople composed the largest proportion (43%) of the main study sample. Pharmaceutical salespeople may differ from salespeople in other industries since they engage in ethical promotion as they inform healthcare professionals of the benefits, safety, and side effects of drugs so that the healthcare professional has the correct information on which to base their choices to prescribe medication to their patients (Wright and Lundstrom 2004). However, pharmaceutical firms are similar to many other service organizations in terms of organizational structure and frequent fellow salesperson interactions (Lam et al. 2010).

Additional study limitations include the use of self-report measures and the reliance on a subjective measure of sales performance. Since self-report measures have the potential to introduce common methods bias, respondents were assured of the anonymity of their responses and were not informed of the specific objectives of the study. Further, while there may be a tendency to discount research findings because of a perceived upward bias produced by subjective measures, they may not be problematic when methodological checks are put in place (Churchill et al. 1985; Franke and Park 2006). Even more, “academic literature does not support a consistently low correlation between self-report and objective measures of performance” (Sharma et al. 2004, p. 136). As per Podsakoff et al. (2003) and given the results of our test for common method bias, the measures did not likely inflate the association between the study’s determinants and sales performance.

A further limitation is that the focus of the data collection was on outside salespeople. Future consideration should be given to how the research variables affect inside salespeople (Jones et al. 2005). The study also failed to consider whether the respondents were working in a team-selling or individual-selling environment. The increase in team-selling in practice has attracted the attention of sales researchers, who propose that this concept may help explain factors that contribute to sales performance (Jones et al. 2005; Verbeke et al. 2011).

Future research directions

A number of future research directions arise given the results of the study. First, the original coachability scale was based on six sub-dimensions. However, our results suggest that an alternative five sub-dimension scale may be more appropriate. Further, in this study, transformational leadership style was the situation variable of interest because of its resemblance to coaching in general. Future research should examine other leadership styles and how salesperson coachability may be elicited or inhibited under each. Additionally, since sales manager leadership style was the only situational variable examined in the current study, others, such as sales team environment, should be examined in future studies. Future studies should also examine the reciprocal effects of the sales manager leadership style–salesperson coachability relationship.

While our focus was the interaction and mediating effects of salesperson coachability on sales performance, there is ample opportunity to examine the potentially moderating effects of gender, age, and years of experience on the relationship between salesperson coachability and sales performance. Furthermore, this study focused on outside salespeople. Differences exist, however, between outside and inside salespeople with respect to salary, training, and environmental aspects (Johnston and Marshall 2005). Future studies should therefore include inside salespeople in the sample. In addition to a difference in work environment, there may be marked individual differences between inside and outside salespeople and inside and outside sales managers.