Redefining performance of direct sales people

The purpose of this study is to broaden the definition of performance to include extra-role and in-role aspects in the conceptualisation of performance in direct selling. Using exploratory and confirmatory factor analysis, the authors first report the development of the extra-role performance scale. A model of extra-role performance consisting of five dimensions including individual initiative, helping behaviour, self-development, organisational loyalty and organisational compliance was identified. Nomological validity of the newly developed scale was established by relating the dimensions of extra-role performance to the in-role performance measure. The authors discuss the implications of their findings and suggest avenues for further research.


Introduction
With the emergence of the global marketplace, marketers have had to adapt their distribution strategies to capitalise on international markets.Direct selling, a non-store approach to retailing (Kotler, 1997) has become an attractive distribution channel due to a number of reasons: Firstly, direct selling organisations engage direct sales people who purchase products from the supplying organisation at a discount and resell these products to the consumers at a profit.As such, the financial risk of keeping stock is shared between the supplying organisation and direct sales people.
Secondly, the direct selling distribution channel provides firms increased ability to secure distribution and sales at low level of fixed cost (Granfield & Nicols, 1975).The burden of employing a large administrative and sales force is reduced because direct sales people are encouraged to recruit other direct sales people.The new recruits are trained and developed using a unique coaching and training system called sponsoring.In this system of sponsoring, the direct sales person, sometimes referred to as a distributor, shares his/her knowledge and expertise with new entrants learning the business for the first time.In return for this expertise the direct sales person earns commission based on a percentage of the sales from those recruited.Thirdly, direct selling as a channel of distribution makes it possible for firms to gain entrance to a market while avoiding excessive promotional and advertising expenses as well as potential price wars (Clothier, 1994).
While on one hand factors such as globalisation, deregulation and technological convergence have had profound impact on the way business is conducted, customers on the other hand are fundamentally changing the dynamics of the marketplace.A change in consumer demographics, lifestyles and work patterns coupled with the speed with which marketing information is spread has created highly discerning customers.As Prahalad, Ramaswamy and Venkatram (2000) point out, customers are no longer happy to be passive in the marketing exchange, they want to be co-creators of the content of their experiences.Direct selling fits this marketing environment very well in that direct sales people offer customers a customised and personalised service at a location that fits the customers' convenience.Distributors are usually the only interface between direct selling firms and the customers.They are not only responsible for meeting the needs of their customers, they are also responsible for generating sales and profits for the direct selling firms that they are associated with.Some authors (Raymond & Tanner, 1994) view direct sales people as the means of achieving a competitive advantage.
If distributors play such a significant role in generating business for direct selling firms, how can their performance be evaluated?What are the extra-role and in-role behaviours that they need to engage in to improve sales productivity?Are there industry specific extra-role behaviours that affect sales productivity?This paper seeks to address these questions by integrating extra-role and in-role behaviours in the conceptualisation of direct sales people performance.Inasmuch as the performance measured in many organisations is often in-role performance, extant research acknowledges the importance of distinguishing between inrole and extra-role performance in evaluating salespeople (MacKenzie, Podsakoff & Fetter, 1993;MacKenzie, Podsakoff & Ahearne, 1998).In-role performance is defined as performance that is within the duty of the organisation, which is seen as an expected part of the job and directly affects the organisation in terms of sales volume and managerial evaluations of sales effectiveness (Morrison, 1994;Kohli, Shervani, & Challagalla, 1998;MacKenzie, et al., 1998).Extra-role performance on the other hand is performance that is above and beyond the call of duty within the organisation, but make a contribution to organisational effectiveness (MacKenzie et al 1998).In their metaanalytical study Podsakoff, MacKenzie, Paine and Bachrach (2000) provided a number or reasons why extra-role performance behaviours might influence organisational effectiveness, and these include the fact that they: 1. enhance co-worker and managerial productivity 2. free resources up for more productive purposes, and 3. enhance the organisation's ability to attract and retain the best people by making it a more attractive place to work There has been a lack of consensus in literature about the dimensionality of extra-role performance and it is the purpose of this research is to identify the dimensions of extra-role performance in the direct selling context, and to relate these dimensions to in-role performance.In this paper we thus report the development and validation of the extra role performance measure.We then present a model that relates the extra-role performance dimensions to in-role performance.We believe that an understanding of the relationship between extra-role performance and in-role performance should facilitate efforts by direct selling practitioners and researchers to understand how and why performance of sales people improves or declines.Furthermore, organisational wellbeing can be enhanced when a more holistic approach in quantifying behaviours that improve sales performance is adopted.

Extra-role performance
According to Smith, Organ and Near (1983) there are three basic types of behaviours that are essential for a functioning organisation: (1) people must be encouraged or persuaded to enter and remain in a system; (2) they must carry out specific role requirements in a dependable fashion; (3) people must engage in innovative and spontaneous activities that go beyond role prescription.It is the latter activities that managers consider in addition to objective performance.These innovative and spontaneous activities have been referred to as 'organisational citizenship behaviours' by Bateman and Organ (1983), 'prosocial organisational behaviours' by Brief and Motowidlo (1986), 'organisational spontaneity' by George and Brief (1992) and 'contextual behaviours' by Motowidlo and Van Scotter (1994).
Recently, literature has defined these behaviours as extrarole behaviours, that is, positive, discretionary behaviours that are above and beyond what is expected of the job prescription in a way that is organisationally functional (MacKenzie et al., 1998;Morrison & Phelps, 1999;Van Dyne & LePine, 1998).This study is based on this conceptual definition of extra-role performance.Podsakoff et al. (2000) identified several overlaps in the conceptualisation of extra-role performance and expressed the need to explore the differences and similarities between the various types of extra-role performance.In their metaanalytical study Podsakoff et al. (2000) reorganised the different types of extra-role performances into seven dimensions: (1) helping behaviour, (2) sportsmanship, (3) organisational loyalty, (4) organisational compliance (5) individual initiative, (6) self development, and (7) civic virtue.During the same year Coleman and Borman (2000) published a study that delineates the conceptual organisation of citizenship behaviours which they called citizenship performances.In their study they proposed a framework that represents three broad categories of behaviours (1) including interpersonal citizenship performance, (2) organisational citizenship performance, and (3) job/task citizenship performance.The three broad categories are further divided into subcategories that capture helping behaviours, altruism, civic virtues, organisational loyalty and organisational compliance similar to findings of several authors (Smith et al, 1983;Organ, 1988;Williams & Anderson, 1991;Podsakoff et al., 2000).
Helping behaviour, which involves the voluntary helping of others with work-related problems has attracted the attention of a number of researchers.For example, Coleman and Borman (2000) conceptualised helping behaviour as a behaviour that assists and supports the performance of organisational members through cooperation and facilitation efforts that go beyond expectations.Coleman and Borman (2000) included interpersonal conscientiousness in their conceptualisation of helping behaviour.They define conscientiousness as a helping behaviour directed at others within the organisation, but with a notion of contributing to the organisation's well-being.In an earlier study, Podsadoff and MacKenzie (1994) conceptualised helping behaviour as a second-order construct consisting of altruism, courtesy, peacekeeping and cheerleading sub-dimensions.Altruism and courtesy have also appeared in a number of studies prior to Podsakoff and MacKenzie's (1994) study (see Smith et al., 1983;Organ, 1988;Van Scotter & Motowidlo, 1996).
Literature defines the sportsmanship behaviour as willingness to tolerate the inevitable inconveniences and impositions of work without complaining (Podsakoff et al., 2000).This definition was based on an earlier work by Mackenzie, Podsakoff and Paine (1999), who captured this dimensions with 5 items including (1) consuming a lot of time complaining about trivial matters; (2) finding fault with what the company is doing; (3) making problems bigger than they are; (4) focusing on what's wrong with the situation instead of the positive side of things.The contention that sportsmanship contributes to the effective functioning of an organisation was empirically supported (MacKenzie et al., 1999).Their findings showed that salespeople who engage in helping behaviour, civic virtue and sportsmanship are seen as contributing more to the effective functioning of the organisation than those who do not.
In a number of studies (Podsakoff & MacKenzie, 1994;Podsakoff, Ahearne & Fetter, 1997MacKenzie, Podsakoff, & Fetter, 1993) civic virtue is conceptualised as commitment to the organisation shown by the willingness (1) to participate in its governance; (2) to monitor its environment from threats and opportunities; and (3) to look out for its best interest.On the other hand Coleman and Borman contend that organisation commitment is demonstrated through allegiance, loyalty and through compliance with organisational rules, policies and procedures.This kind of commitment has been captured by other authors as two separate constructs -(1) organisational loyalty and organisational compliance (refer to George & Brief, 1992;Bormon & Motowidlo, 1997;Smith et al., 1983).

In-role performance
In-role behaviours are the required or expected behaviours that are within the duty of the organisation.Much of in-role performance is subjective as it varies from industry to industry and firm to firm, depending on specific tasks that need to be performed.For example O'Reilly and Chatman (1986) used three items to measure in-role performance of university employees: (1) I work a full eight-hour day, (2) I complete my assigned duties on time and (3) I comply with the rules and regulations of this organisation.Williams and Anderson (1991) on the other hand captured the in-role performance variable with 6 items: (1) adequately complete assigned duties; (2) fulfil responsibilities specified in the job prescription; (3) perform tasks that are expected of me; (4) meet formal performance requirements of job; (5) engage in activities that directly affect my performance evaluation; (6) fail to perform essential duties.The authors argue that an objective sales measure may not be a good indicator of in-role performance, especially given that there are other factors such as salesperson skill and effort that impact on sales performance.Churchill, Ford, Hartley, and Walker (1985) identified six factors that affect sales person performance and these include role, skill, motivation, personal, aptitude and organisational factors.

Research method -development of extra-role performance scale
Development of the extra-role performance scale followed the standard psychometric procedures as suggested by Nunnaly (1978).The first step in the scale development process, as illustrated in Figure 1, was to establish the domain of the extra-role performance construct.As mentioned earlier review and synthesis of past literature identified seven dimensions of extra-role performance.Past literature also provided the definitions required in specifying the domain of the construct as well as items that capture it.Thirty seven items (see Table 1), representing the seven dimensions of extra-role performance were included in a pre-test survey presented to 14 direct selling industry experts.The purpose of the pre-test was two-fold: (1) to tap into the insights of direct selling experts and to identify items that are specific to direct selling that might not have been captured in literature; (2) to determine if the respondents felt that the items were relevant and clear in meaning.The suggestions and comments of the respondents were incorporated in the final survey instrument outlined in Table 1.Our review and examination of relevant literature and existing scales resulted in selecting the seven items illustrated in Table 1: (1) willingly give of time to help member with business related problems; (2) willingness to help and encourage co-members with personal problems; (3) helping other members even though it does not benefit me directly; (4) acting as a peacekeeper when others in the team have disagreements, (5) touching base with others before initiating actions that might affect them; (6) taking steps to try to prevent problems with other members in the team, (7) being a stabilising influence in my organisation when dissension occurs (Brief & Motowidlo, 1986;Van Scotter & Motowidlo, 1996;Netermeyer, Boles, MacKee & McMurrian, 1997) .

Individual initiative
This dimension of extra-role performance involves the performance of task-related behaviours at a level that is far beyond minimally required or generally expected levels.It involves acts such as taking on extra responsibilities beyond the call of duty, with the aim of improving organisation performance.Individual initiative is a behaviour that also captures the enthusiasm and extra-effort displayed by an employee in accomplishing a task.This dimension is Domain of extra-role performance specification Item generation using past literature and data from expert survey Scale refinement using two analytical procedures: (1) exploratory factor analysis, and (2) confirmatory factor analysis

Assessment of reliability and validity
Assessment of nomological validity with in-role performance as an endogenous variable and extra-role behaviours as exogenous variables captured by seven items adapted from Van Scotter and Motowidlo's (1996) job dedication scale (see Table 1).

Sportsmanship
The third dimension of extra-role performance is captured by three items presented in Table 1.This dimension has been defined as willingness to accept the expected inconveniences and obligations of work without complaining (Organ, 1990).Podsakoff et al. (2000) broadens this definition as he points out that in addition to the willingness to accept the inconveniences of work without complaining, sportsmanship includes the ability to maintain a positive attitude at all times, and not to take rejection of one's ideas personally.

Self development
The fourth dimension of extra-role performance has been defined by George & Brief (1992) as a behaviour that employees engage in to improve their knowledge, skills, and abilities.Self-development has received very little attention in the extra-role performance literature.As such the list of items was bolstered through discussion with industry experts to derive the five items shown in Table 1.

Organisational loyalty
This dimension is captured by six items that represent behaviours that promote the organisation to outsiders, as well as protecting and defending it against external threats (Borman & Motowidlo, 1997).According to Podsakoff et al. (2000) additional work is warranted given the conflict in literature regarding the confirmatory factor solution of this dimension.

Organisational compliance
While the focus of organisational loyalty is on spreading goodwill of the organisation, organisation compliance on the other hand captures a person's internalisation and acceptance of organisation rules, regulations and procedures.Although all employees are expected to obey organisational rules, regulations and procedures, they do not always do so.As such those who faithfully obey the rules even when no one is watching are regarded as engaging in an extra-role behaviour.The organisation compliance items that appear on Table 1 have been adapted from Coleman and Borman (2000); and Smith et al. (1983).

Civic virtue
The final dimension of extra-role performance is reflected by a willingness on the part of the employee to participate actively in the organisation's governance, to monitor the organisation's environment for threats and opportunities and to look out for the company's best interest (Podsakoff et al., 2000).The three items shown in Table 1 reflect the willingness to keep up with developments in the company; the willingness to attend functions not as a company requirement but as a commitment to the company; and the willingness to risk disapproval in order to express personal beliefs about what is best for the company.These items were adapted from the work of Netemeyer, Boles, MacKee and McMurrian (1997).

In-role performance
The in-role performance measure is adapted from earlier work conducted by Msweli-Mbanga (2001)

Sample and data collection
Three participating organisations, one in South Africa, one in Uganda and one in Mauritius provided names, telephone numbers and addresses of salespeople.All three organisations from the three countries source their products from House of Health, South Africa.A total of 900 questionnaires were distributed via the supplying organisation in South Africa and only 80 were returned from South Africa, 44 from Mauritius and 52 from Uganda.A total of 176 responses were generated resulting in a 20% response rate.Fifty eight percent of the respondents were female, their median age was 31-40 years, and their education ranged from high school to a graduate degree.
ANOVA was conducted to determine the appropriateness of pooling the data across three firms in three different countries.The Levene Test in the One-way ANOVA procedure was used to test the null hypothesis that the three firms come from populations with the same variance.The procedure was repeated for each of the seven dimensions.
All the test statistics were nonsignificant and the null hypothesis was thus accepted.As such, the remainder of the analyses was based on pooled data.

Exploratory factor analysis
To assess the structure of the extra-role performance scale, all the 37 items were factor analysed, using principal component analysis.Principal component analysis was chosen instead of common factor analysis, as it is more appropriate when the primary concern is to summarise data in a minimum number of factors (Hair, Anderson, Tatham & Black, 1998).Principal component analysis was followed by varimax rotation.According to Hair et al. (1998) orthogonal rotational approaches are used more frequently because the analytical procedures for performing oblique rotations are not as well developed and are still subject to 'controversy' (p.109).Kaiser's criterion was used for deciding which factors to eliminate (Bryman & Cramer, 1994).The initial factor solution resulted in six factors with eigenvalues greater than 1.The six-factor solution accounted for 74,2 % of the variance, and individual initiative was explained by the greatest amount of variance (33,5%).McMurrian (1997) In order to purify the list, items with loadings of 0,3 and less were eliminated.Hair et al. (1998) defines factor loading as 'a correlation between each variable and the factor' (p.106).
Items that correlated high with more than one factor were also eliminated to ensure that true discriminant validity has been established among the factors.This resulted in removing three items from helping behaviour (#1, #3, and # 6); three items from organisational loyalty (#2, #4, and #5); three items from organisational compliance (#4, #5, and #6); and two items from self-development (#4 and #5).The final items (see Table 2) were tested for internal consistency using Cronbach's alpha.As Table 2 illustrates, the constructs are within the benchmark of ,70 as suggested by Nunnaly and Bernstein (1994), with the exception of sportsmanship (α = ,60).

Confirmatory factor analysis
The 23 items selected from factor analysis were used as indicators of the 6 latent variables in a confirmatory factor analysis model.Structural equation model played a confirmatory role as it allows for a statistical test of the goodness-of-fit for the proposed confirmatory factor solution (Hair et al., 1998).Hair et al. (1998) also point out that confirmatory factor analysis is particularly useful in the validation of scales for the measurement of specific constructs.The first step in the model testing was to estimate the path coefficients relating observable variables to latent constructs using AMOS and the correlation matrix presented in Table 3. Next, the variance extracted by each dimension was compared to the variance due to measurement error.The sportsmanship dimension was found to be below the ,50 cut, and was thus eliminated.The explained variance for the rest of the dimensions were significantly higher than the variance due to measurement error, indicating adequate convergent validity for the 5 dimensions illustrated in Figure 2 (Fornell & Larcker, 1981).Elimination of insignificant paths resulted in a total of 15 observable variables as shown in Figure 1.The fit indices of the revised model were within the traditionally accepted levels (GFI = ,92; CFI = ,90; RMSEA = ,08).

Nomological validity -relating extra-role behaviours to in-role performance
Nomological validity represents how scores on one instrument relate to scores on other constructs.If the expected relationships between constructs are empirically supported, it is assumed that the measures of those constructs have a certain degree of nomological validity (Churchill, 1979).In this study, nomological validity was tested by relating the scores on each of the 5 dimensions of extra-role performance (individual initiative, helping behaviour, self development, organisational loyalty, organisational compliance) to in-role performance using the structural equation methodology.To achieve model identification the covariances were fixed to a unity.Cronbach alphas for extra-role behaviour are depicted in Table 2, and that of in-role performance is ,7, indicating unidimensionality of the constructs.Table 4 presents the standardised parameter estimates for the extra-role/in-role performance model and the overall goodness-of-fit indices.As shown in the table the relationship between the extra-role performance dimensions and in-role performance are statistically significant.These findings are encouraging for the nomological validity of extra-role performance measure.The findings are also consistent with the findings in Podsakoff and MacKenzie's (1994) study where they examined the impact of three forms of extra-role performance (helping behaviour, sportsmanship, and civic virtue) on the performance of life insurance agencies.The authors found that all three forms of extra-role performance had significant effects on performance.As can be observed in Figure 1, individual initiative has more impact on in-role performance than other behaviours (β=,76), followed by self development (β= ,48).

Discussion, implications and avenues for further research
The objective of this study was to enhance the understanding of performance of direct sales people.To achieve this objective, we first identified and evaluated the validity of a seven dimensional measure of extra-role performance.Confirmatory factor analysis supported a five dimensional scale consisting of helping behaviour, individual initiative, self development, organisational loyalty and organisational compliance.Through confirmatory factor analysis, the discriminant validity of the five extra-role dimensions was also established.As literature points out, extra-role behaviours vary according to the beneficiary of the behaviours (Coleman & Borman, 2000), and in our case the direct selling organisations are the beneficiaries.It was not surprising to find civic virtue not forming the underlying dimensions that explain extra-role performance in the direct selling context.This could be because direct sales people are independent contractors and their involvement is limited in the running of a direct selling company.One would therefore not expect them to participate directly in the governance issues, and in the monitoring of the organisation environment from threats and opportunities.It was somewhat surprising to find a lack of value for sportsmanship within direct sales people.The lack of this form of extra-role behaviour suggests that direct salespeople are intolerant of the inconveniences of their jobs.Lack of value for sportsmanship also indicates that direct sales people are more inclined to put their own interest above the interest of the collective.To overcome this problem, direct sales managers should ensure that a complaint handling procedure that is effective is put in place.It might also be beneficial for direct selling organisations to monitor and control this behaviour through training and development.Further research is required to understand the antecedents of sportsmanship and to articulate its effect on organisational effectiveness.
Perhaps the most important finding is that individual initiative accounted for a substantial proportion of variance (33,5%) and had the highest factor loading in both the exploratory and confirmatory factor solution.Even more interesting, is the fact that when the dimensions of extra-role performance were related to in-role performance, individual initiative had the greatest effect on in-role performance compared to other extra-role performance dimensions.The strength of the impact of individual performance in relation to in-role performance suggests that direct sales people regard this dimension as the most important behaviour for achieving increased performance.Given the fact that direct sales people are commission salespeople, it makes sense to expect them to engage in individual initiative before reaping the rewards of increased sales productivity.
Direct sales managers need to consider what they can do to foster the five extra-role behaviours proposed to impact on in-role performance.One way of increasing extra-role behaviours could be by designing incentives that reward direct sales people who display the behaviours identified in this study.The measure of extra-role performance should also prove useful in the recruiting and training of direct sales people.For example, direct sales people could focus their recruiting efforts on people who have a higher propensity to display the extra-role behaviours.This study is not without limitations.First, extra-role performance research is conducive to socially desirable responses.To reduce this impact, direct salespeople were assured confidentiality of their responses.Although there are 36 direct selling firms operating in and around South Africa, our sample is limited to three direct selling organisations.Although the respondents are from three different African countries, they have a typical direct salesperson profile consistent with findings in other studies conducted in Africa (Sargeant & Msweli, 1999;Msweli & Sargeant, 2001).As much as we do not believe that the results apply directly to the universe of direct selling organisation in Africa, we do believe that our results are likely to apply to a considerable number of direct selling firms.Nevertheless, it will be useful to test the extra-role/inrole performance relationship in a broader group of direct selling organisations.
Further research is required to identify the enablers or determinants of extra-role performance.This would further equip direct sales managers to create an organisational environment that increases extra-role behaviours.It is also necessary to test the effects of cultural differences on the relationship between extra-role performance and in-role performance, as it likely that salespeople from different cultures may relate differently to extra-role behaviours.

Figure 1 :
Figure 1: Summary of procedure followed to developing extra-role performance measure Measures Helping behaviour

Figure 2 :
Figure 2: Revised confirmatory factor analysis model for the dimensions of extra-role performance