Bonds, relationship management, and sex-type

Buyer-seller relationships: Bonds, relationship management, and sex-type

Brock Smith


Although a key objective of relationship marketing is building strong bonds with customers, there is little empirical research into the antecedents and consequents of relational bonds. Women are increasingly assuming key boundary-spanning roles in organizations, and understanding the extent to which sex differences affect relationship processes and outcomes is an important management issue. This study develops hypotheses linking relationship quality, relational bonds, facets of relationship management, and biological sex, and tests them in the context of buyer-supplier relationships. Social bonds and the relationship-management facets of communication/cooperation and relationship investment were found to be key predictors of relationship quality. Communication/cooperation, relationship investment and relationalism were found to predict social bonds, while relationship investment and relationalism were found to predict functional and structural bonds. Relationship type (male-male, male-female, female– male, and female-female) was found to have some effect on relationship quality and relational bonds. Post hoc analysis found significant interaction effects between the type of relationship and facets of relationship management.

The management of buyer-seller relationships is recognized as being integral to business success (Wilson, 1995) and can provide a key source of competitive advantage (Day & Wensley, 1983). While there are several views about the nature and scope of relationship marketing (Nevin, 1995), one core objective is to build strong bonds with customers (Cravens, 1995).

Interestingly, the antecedents and consequents of relational bonds remain relatively unexplored (cf. Han, 1992; Mummalaneni & Wilson, 1991).

Recent work by Palmer and Bejou (1995) and Smith and Bejou (1995) suggests that biological sex and/or socialized gender role may have a bearing on working relationships, processes, and outcomes. Although equivocal, research on interpersonal and work relationships suggests that men and women differ, at least in degree, in their relationship management styles, approaches, and values (e.g., Keys, 1985; Riger & Gilligan, 1980; Statham, 1987). These are thought to place a strain on coworker interaction (Devine & Markiewicz, 1990). It is not clear, however, to what extent, if any, sex differences are manifest in the management of buyer-seller relationships. This issue is particularly important since women have increasingly been assuming key management and boundary-spanning roles in organizations over the past two decades (Foster & Orser, 1994), and men and women may have different training requirements to be effective relationship managers.

This study investigates the role of relational bonds in the development of quality working relationships. The effects of biological sex on relationship management, relational bonds, and relationship quality are also explored.


Relationship marketing has been such a dominant, yet undefined, paradigm over the past 10 years that it is difficult to distinguish from marketing, as classically defined (Iacobucci, 1994). This has led to the paradox where considerable research has been conducted on service, consumer, channel, partner, and business-to-business relationships (cf. the special issue of the Journal of the Academy of Marketing Science, 1995), yet relationship marketing is considered to be in the very early stages of its development (Sheth & Parvatiyar, 1995).

Reflective of its early development relationship marketing research has been plagued by a variety of conceptual and methodological issues. There is little agreement on the definition of concepts, how they should be operationalized, or what labels should be attached to them (Wilson, 1995). This has led to the situation where the meanings of many of the concepts deployed in the literature overlap. An unresolved methodological issue is who can or should inform the researcher on the nature, characteristics, and dynamics of the relationship, its parties, their interaction, and their environment. Furthermore, if multiple informants are used, it is not clear how divergent or incongruent perspectives should be handled (Weitz & Jap, 1995).

Even the domain of relationship marketing is in dispute (Nevin, 1995). In the buyer-seller relationship literature, relationship marketing has generally been defined as the development and maintenance of close, long-term, mutually beneficial, and satisfying relationships between individuals or between organizations that are based on trust and collaboration (e.g., Wilson, 1995).

Central to the stability of buyer-seller relationships is the psychological attachment produced by relational bonds (Han, 1992). Despite conceptual recognition of their importance (e.g., Wilson, 1995), there have been few empirical studies of the antecedents and consequents of relationship bonds. Han (1992) found that the importance of the relationship was related to structural bonds, which in turn were related to commitment to a relationship. Satisfaction, (but not trust), was found to be related to social bonds, but social bonds were not related to commitment. Somewhat different results were obtained by Mummalaneni and Wilson (1991), who found that buyers and sellers who have a strong personal relationship are more satisfied and committed to maintaining the relationship than less socially bonded partners.

Recent research by Palmer and Bejou (1995) suggests that biological sex may play a role in the development and maintenance of buyer-seller relationships. They examined customer perceptions of sales representative’s empathy, pressure, and ethics and found some ambiguous differences in same-sex and cross-sex dyads. Smith and Bejou (1995) found that salespeople’s sex was not a predictor of relationship quality in a financial services context, but found evidence of sex differences (but also many similarities) in determinants of customers’ trust and satisfaction. This study, however, did not consider the sex of the customer.

Men and women have previously been thought to have different strengths in managing customer relationships (Swan, Rink, Kiser, & Martin, 1984). Comer and Jolson (1991) found that women are still often subjected to biased treatment based on sales managers’ beliefs in gender stereotypes. Other studies have found few or no sex differences in work related attitudes and behaviours (e.g., Schul, Remington, & Berl, 1990; Schul & Wren, 1992). This suggests sex differences are not ubiquitous, and, if they do exist, may be mitigated by other factors. Equivocal results in previous research indicates a need for further investigation.

This study begins to address two gaps in the literature, a lack of understanding of: the antecedents and consequents of relational bonds; and of the effects of biological sex on relationship management, bonds, and quality. These issues are particularly important given the strategic role of effective relationship management and the increasing number of women in key management and boundary-spanning positions (Foster & Orser, 1994). The study develops hypotheses linking relationship quality, relational bonds, relationship management, and biological sex, and tests them in the context of buyer-supplier relationships. Facets of relationship management are included in the study because they have received considerable research attention as direct and indirect determinants of effective work relationships (e.g., Anderson & Narus, 1990; Anderson & Weitz, 1989; Wilson, 1995).


The choice of relationship quality as the dependent variable in the study is consistent with previous studies of relationship management (e.g., Crosby, Evans, & Cowles, 1990; Palmer & Bejou, 1995). This subjective indicator of effectiveness reflects the interactive nature of a relationship better than more objective measures, which can be confounded by competitive and environmental factors (Gladstein, 1984). Briefly, relationship quality is hypothesized to be determined in part by the level of social, functional, and structural bonds developed in a working relationship. Facets of relationship management, such as open communication and functional conflict resolution, are hypothesized to affect both relationship quality and the level of relational bonds. Finally, relationship type, a categorical variable reflecting same- and cross-sex working relationships, is hypothesized to affect relationship management, relational bonds, and relationship quality. These hypotheses are illustrated in Figure 1 and further developed and justified below.

Relationship Quality

Relationship quality is an overall assessment of the strength of a relationship and the extent to which it meets the needs and expectations of the parties based on a history of successful or unsuccessful encounters or events (Crosby et al., 1990). Although there is no consensus on what constructs make up relationship quality (Kumar, Scheer, & Steenkamp, 1995), it is generally conceptualized as being concerned with the extent to which relators trust each other, are satisfied with the relationship, and are committed to its long-term maintenance. Trust, the confident belief that the salesperson can be relied upon to behave in such a manner that the long-term interest of the customer will be served (e.g., Anderson & Narus, 1990), is thought to be a building block or foundation for satisfactory interaction. Satisfaction, the emotional state that occurs in response to an evaluation of interaction experiences in relation to alternatives (Westbrook, 1981), serves to strengthen bonds of trust. Commitment is an enduring desire to maintain a valued relationship (Moorman, Zaltman, & Deshpande, 1992). Trust, satisfaction, and commitment are thus intimately interconnected in the conceptualization of relationship quality.


Bonds are the psychological, emotional, economic, or physical attachments in a relationship that are fostered by association and interaction and serve to bind parties together under relational exchange (McCall, 1970; Turner, 1970). While previous researchers conceptualized two types of bonds, structural and social (Han, 1992; Wilson, 1995), we propose that functional bonds also serve to bind parties to a relationship.

Social bonds are personal ties or linkages forged during interaction at work (Turner, 1970). They include the degree of personal friendship and liking shared by a buyer and seller (Wilson, 1995), as well as linking of personal selves or identities through self disclosure; closeness; providing support or advice; being empathetic and responsive; feelings of affiliation, attachment, or connectedness; and shared experiences (Turner, 1970).

Functional bonds are the multiplicity of economic, performance, or instrumental ties or linkages that serve to promote continuity in a relationship. Described as task bonds by Turner (1970), functional bonds are created by the economic, strategic, technological (knowledge or information), and instrumental (product- or service– related) benefits derived by the exchange parties. These benefits are evaluated in comparison with alternative relationships, either experienced or suspected.

Functional bonds thus draw on the exchange theory concept of comparison-level given alternatives (Thibaut & Kelly, 1959). Han (1992) and Wilson (1995) described these bonds as structural bonds in the sense that parties are tied by relationship benefits at the organizational level. However, ties of organizational benefits are conceptually quite distinct from contractual or physical linkages between organizations, which can also bind parties to a relationship. We view these latter linkages as being more structural than functional.

Structural bonds are ties relating to the structure, governance, and institutionalization of norms in a relationship. The rules, policies, procedures, or agreements that provide formal structure to a relationship; the norms or routines that informally govern interaction; and the organizational systems and technologies, such as electronic mail or electronic data interchange, that enable or facilitate interaction can provide psychological, legal, and physical ties that bind parties to a relationship and make it difficult to consider other exchange partners.

Social exchange theory (e.g., Homans, 1961) suggests that relational bonds are developed through a series of successive interactions. These interactions, successful or unsuccessful, become threads in the woven fabric of a relationship and provide a context or history that draws and keeps parties together and shapes their interaction. As such, social, functional, and structural bonds provide the context or bases from which relational outcomes such as trust, satisfaction, and commitment are evaluated. In addition, bonds reduce the risk inherent in voluntary exchange relations and provide a foundation for the trust needed to risk greater commitment and feel satisfied with the overall relationship. Thus it is hypothesized that:

H1a-c: Relational bonds (social, functional and structural) will positively affect relationship quality.

Existing empirical support for this hypothesis is limited. Han (1992) found a strong relationship between structural bonds (which included functional bonds) and commitment, but not between social bonds and commitment. Han also found a relationship between social bonds and satisfaction but not between social bonds and trust. In that study, satisfaction and trust were specified as antecedents of social bonds. We believe the opposite specification of causality is more appropriate in a crosssectional study, since bonds embody the history of a relationship, and evaluations of trust and satisfaction are based on an assessment of that history. We recognize, however, that in ongoing relationships bonds influence relational outcomes, which in turn influence the further development of bonds.

Relationship Management

Relationship management is the extent to which relators have the orientation or behavioural tendency to actively cultivate and maintain close working relationships (Crosby et al., 1990). It is a higher-order construct representing the mix of behaviours, approaches, and styles used to effectively manage relationships.

Many dimensions of relationship management stand out in social exchange theory and the extant literature. Preliminary interviews with purchasing professionals suggested five that could be particularly important in the buyer-supplier context: relationship investments, open communication, cooperation, functional conflict resolution, and relationalism. Relationship investments are the resources, effort, and attention devoted to a relationship that do not have outside value (Mummalaneni & Wilson, 1991). Open communication is the extent to which buyers and sellers communicate openly, sincerely, and substantively with customers, either formally or informally (Anderson & Weitz, 1989; Crosby & Stephens, 1987). Cooperation is the extent to which relators undertake voluntary coordinated action to achieve mutual or reciprocated outcomes (Anderson & Narus, 1990). Functional conflict resolution is the degree to which disagreements between relators are resolved productively to clear the air of tensions and ill-will (Anderson & Narus, 1990). Relationalism is the extent to which relators actively and purposefully manage their relationship and promote behaviours to maintain or improve the relationship (Noordewier, John, & Nevin, 1990).

Exchange theory suggests that these facets of relationship management facilitate exchange by removing barriers of risk and uncertainty, signalling commitment to making a relationship work, and enhancing norms of reciprocity and fair exchange, which facilitate positive relational outcomes (e.g., Homans, 1961). Empirically, the importance of relationship management for quality relationships is well documented. Crosby et al. (1990), for example, found the selling behaviours of mutual disclosure, cooperative behaviour, and contact intensity to be a key determinant of relationship quality with life insurance customers. Other buyer-seller and channel relationship studies have linked trust and satisfaction to aspects of relationship management such as constructive conflict resolution (Lau, 1990), open communication (Anderson & Narus 1990), and functional conflict resolution (Smith, 1993). Thus it is hypothesized that:

H2a-c: Relationship management (relationship investments, open communication, cooperation, functional conflict resolution, and relationalism) will positively affect relationship quality.

Relationship management behaviours should also facilitate the development of social, functional, and structural bonds. Behaviours aimed at cultivating and maintaining close working relationships should influence the success of each transaction and, as such, provide the catalyst for bond development over successive transactions. Buyers and sellers are more likely to develop social ties of closeness, personal disclosure, concern, and liking when there is open communication, cooperation, functional conflict resolution, and time and effort invested in the relationship. Greater functional benefits are likely derived from relationships where there is better flow of information, some give-and-take, and some effort expended. The frequency of interaction inherent in greater relationship management activities might also be expected to result in the development of relational norms, the formalization of the relationship, or other structural and physical linkages to facilitate exchange. While empirical investigation of bonding has been scant, Han (1992) found that adaptation (a construct similar to our relationship investments) was a significant determinant of structural bonding (our functional and structural bonds). Thus there is primarily conceptual support to hypothesize that:

H3a-e: Relationship management (relationship investment, open communication, cooperation, functional conflict resolution, and relationalism) will positively affect social bonds.

H4a-e: Relationship management (relationship investment, open communication, cooperation, functional conflict resolution, and relationalism) will positively affect functional bonds.

H5a-e: Relationship management (relationship investment, open communication, cooperation, functional conflict resolution, and relationalism) will positively affect structural bonds.

Relationship Type

Biological sex rather than gender role identity was chosen as a focal construct. Marketing researchers have found sex to be as relevant, and sometimes more relevant, a construct than gender role (Schmidt, LeClarc, & Dube-Rioux, 1988). Sex is also an immediately apparent classification category that elicits well developed sex– based schema for exchange (Deaux & Major, 1987).

Biological sex is an individual-level construct. Since our unit of analysis is the buyer-seller relationship, a more appropriate conceptualization is relationship type: a categorical variable reflecting male-male, male-female, female-male, and female-female buyer-seller dyads.

Theoretical rationale for why researchers might expect sex differences in relationship management can be found in two basic perspectives: biological and sociological (Riger & Gilligan, 1980). The biological perspective argues that men and women have physiological differences that result in women having greater innate relational abilities. The sociological perspective suggests that men and women are socialized into different gender roles (e.g., Bem, 1981) and, consequently, develop different traits, values, and characteristics that differentially shape their work-related interests, decisions, and behaviours.

As a result of biological or cultural differences, women are thought to be more concerned with relationships, develop closer relationships, and more concerned than men with helping others achieve their goals (Gilligan, 1982). Empirically, men and women have been found to differ, at least in degree, in relational orientation (e.g., Riger & Gilligan, 1980), expectations (Sherrod, 1989), interaction styles (Carli, 1989), and communication styles (Henley, 1977; Tavris, 1992). In work relationships, sex differences have been found in management styles (Statham, 1987), network development (Ibarra, 1992), work values and interests (Keys, 1985), and ethical orientation (Betz, O’Connell, & Shepard, 1989).

Together, these theoretical and empirical underpinnings suggest that women may manage buyer-seller relationships differently than men and may have greater relationship management skills. However, work role socialization, training, or natural selection may reduce any biological or culturally socialized sex differences. In addition, observed differences in behaviours may simply reflect differences in environment, as discrimination, sexual stereotyping, and external role relationships are thought to differentially affect men and woman (Riger & Gilligan, 1980). Recent empirical studies have found more sex similarities than differences in salespeople’s attitudes, motivation, and outcomes of and evidence against stereotypical weaknesses of saleswomen (Schul & Wren, 1992). Thus, while support is mixed, it is hypothesized that:

H6a-e: Relationship type affects relationship management in terms of relationship investments, open communication, cooperation, functional conflict resolution, and relationalism.

This discussion also suggests that men and women may develop different types of relational bonds. This is also supported by the interaction theory notion that characteristics of relators can impact the nature and quality of the interaction. From a role theory perspective, buyers may have different role expectations for male and female sales representatives that could influence their desire for ties that bind and their evaluation of the relationship. Hall (1993), for example, found that the sex of the service provider influenced customers’ perceptions of good service in restaurants. Thus it is hypothesized that:

H7a-c: Relationship type affects social, functional, and structural bonds.

Being more oriented toward, and sensitive to, relationships, women might be expected to develop higher levels of trust, satisfaction, and commitment in their relationships than men, or to develop stronger relationships with other women than with men. This has been evident in network relationships (e.g., Ibarra, 1992), but mixed evidence has been found in other types of work relationships (for a review, see O’Leary, 1988). Salespeople, however, are trained or socialized to be adaptive. Saleswomen may adopt more typically masculine traits such aggressiveness and goal orientation when required, while salesmen may adopt more typically feminine traits such as a relational orientation when required. While few studies have examined the politically sensitive issue of performance, Kanuk (1978) found men and women to be equally effective in performing the selling role. Whether women develop higher quality relationships than men is an empirical issue to be investigated with the hypothesis:

H8: Relationship type affects relationship quality.

The literature suggests that relational processes and outcomes are affected to a large degree by the duration of working relationships and differences in the experience levels and expertise of the parties. Anderson and Weitz (1989), for example, found that trust and expectations of continuity increase as relationships mature. Crosby et al. (1990) found that expertise strongly affected relational selling behaviours and to a lesser extent, relationship quality and sales effectiveness. Expertise has also been found to be a key mitigating factor in sex difference research (e.g., Deaux, 1984). To account for potential differences in relationship duration and supplier representative expertise, these constructs were included as covariates in the analysis.


The study was conducted in two stages. First, to gain a better understanding of buyer-seller relationship impediments and issues, perspectives on sex differences, and field terminology, preliminary interviews were conducted with a convenience sample of 9 purchasing professionals and 9 supplier representatives in a mediumsized Canadian city. These interviews informed the design and wording of a national mail survey and assisted in interpreting the results. Hypothesized relationships were tested using this cross-sectional, self-report survey data and hierarchical regression. Post hoc analysis using MANOVA and ANOVA was conducted to better understand the regression results. LISREL was used to refine the measures, but was considered inappropriate for testing hypotheses at the early stage of theory and measure development (Pedhazur, 1982).


A self-report survey was mailed to a random sample of 220 male and 200 female members of the Purchasing Management Association of Canada. Half of these men and women were asked to consider their working relationship with the last male supplier representative (rep) with whom they had talked and whom they had known for at least three months. The other half were asked to consider a female supplier rep meeting those criteria. The recipients were then asked a series of questions pertaining to that working relationship and reported on it as key informants.

Of the 420 surveys mailed, 21 were undeliverable and 185 useable surveys were returned (46% useable response rate). Of these, 105 were returned by male purchasing professionals reporting on relationships with 45 male supplier reps and 60 female supplier reps, and 80 were returned by female purchasing professionals reporting on relationships with 45 male reps and 35 female reps. There were no significant differences in response rates (p

Nonresponse bias was assessed with a short telephone survey of 19 (about 10%) of the nonrespondents selected on a random basis. Of these, 31% provided a valid response for not responding (21% did not receive the survey and 10% were no longer in a purchasing role). The demographic profile of the nonrespondents in terms of age, sex, and purchasing experience was found to be very similar to that of the respondents. For example, the mean purchasing experience was 15.9 years for the respondents and 16.1 for the nonrespondents. The proportions of respondents aged 19-35, 3645, and over 46 were 17.4%, 39.1%, and 43.5%, respectively, while for nonrespondents they were 23.1%, 31.8%, and 46.2%, respectively. No sex bias was evident in the profiles of the nonrespondents. Consequently, we concluded the respondent sample was at least demographically representative of the sampling frame.

As might be expected given traditional male dominance in both the purchasing and sales professions, our sample of male purchasing professionals reported, on average, relationships of greater duration than did female purchasing professionals (t = p


Multiple items summed into scales were used to measure modelled constructs (Appendix A), with the exception of sex and relationship duration, which were measured with single items. Seven-point Likert-type scales ranging from strongly agree to strongly disagree were used to measure the relationship quality facets of trust, satisfaction, and commitment. These items were adapted to this context from published scales (e.g., Anderson & Narus, 1984; Anderson & Weitz, 1992; Crosby et al., 1990) and were found to have reliabilities (Cronbach’s alpha) of .83, .83, and .76, respectively.

Social, functional, and structural bonds were measured with original items based on the conceptual works of Han (1992) and Turner (1970). Respondents were asked to assess on a 5-point Likert-type scale the extent to which their relationship was characterized by listed attributes. Thirteen social bond items, eight functional bond items, and five structural bond items were found to have reliabilities of .92, .79, and .52, respectively. The low reliability for the structural bond items is not too surprising as they represented distinct sources of structure, not all of which would be expected in any given relationship, but when summed would indicate the extent of structural bonding.

The relationship management facets of relationship investment, open communication, cooperation, functional conflict resolution, and relationalism were measured with multiple items adapted from published scales (e.g., Anderson & Narus, 1990; Heide & John,1992; Mummalaneni & Wilson, 1991). The measures were found to have reliabilities of .75, .64, .68, .83, and .76, respectively.

Finally, the mediating factor supplier rep expertise was measured with 2 items adapted from Crosby et al. (1990), which had a reliability of .84 in this context.

LISREL was used to assess the validity of the conceptualized facets of relational management, relationship bonds, and relationship quality (see Appendix B). For the relationship-management facets, confirmatory factor analysis found support for a 3-, not 5-factor model: a joint communication/cooperation factor (3 items, alpha = .85), relationship investments (2 items, alpha = .85), and relationalism (3 items, alpha = .82) [see Table 1, chi^sup 2^ = 18, df = 17, p = .386, GF = .98, AGF = .95].

Evidence also supported distinct social bond (5 items, a = .84) and functional bond (5 items, alpha = .72) constructs [see Table 1, chi^sup 2^ = 40, df = 34, p = .235, GF = .96, AGF = .93]. Structural bond items were not included in the LISREL analysis because they constitute an index, and LISREL assumes reflective indicators. Structural bonds were measured with the original 5-item scale plus the item information exchange, which was found in a preliminary principle components factor analysis to load highly on this construct and not on functional bonds.

Support was found for a 6-item, unidimensional, relationship quality scale that captured commitment, satisfaction, and trust elements (alpha = .88) [see Table 1, chi^sup 2^ = 27, df = 9, p = .001, GF = .95, AGF = .89]. The fit of the whole model (excluding structural bonds and covariates) was modest [chi^sup 2^ = 446, df = 237, p = .000, GF = .84, AGF = .80]. However, these results were considered acceptable given the early stage of measure development and that the items would be summed into scales.

Analysis and Results

Hierarchical regression analysis was used to test the hypothesized relationships. Standardized beta coefficients and the change in multiple correlation (R2) were used to assess the relative importance of the effects. MANOVA and ANOVA analyses were conducted post hoc to further understand and interpret the results.

Predictors of relationship quality explained 67% of the variance in the construct (Table 2, Equation 4). An examination of the effects of four blocks of variables (covariates, relational bonds, relationship management, and relationship type) indicates that individually (Equations Ib-d), social bonds, functional bonds, communication/cooperation, relationship investment, and relationship type have a significant effect on relationship quality. These effects were still significant after accounting for the effects of the covariates (of which expertise of the supplier representative, but not relationship duration, were significant). However, relationship type did not significantly improve the variance explained in relationship quality beyond that of expertise (Equation 2c).

Interestingly, when all predictors were considered (Equation 4), expertise (beta = .38), social bonds (beta = .34), communication/cooperation ( = .19), and relationship investment (beta = .15) were the only significant predictors of relationship quality. Functional bonds and relationship type appear to be mitigated by relationship management factors (Equations 3b and 3c). Post hoc ANOVA1 found a significant interaction between communication/cooperation and relationship type (p = .039).

Predictors of social bonds explained 33% of the variance in the construct (Table 3, Equation 6). As hypothesized (H3), communication/cooperation (beta = .33), relationship investment (beta = .18) and relationalism (beta = .19) were all significant predictors of social bonds. These were found to mitigate the effect of expertise (Equation 4a and 5b). Relationship type was found to have no effect on social bonds.

Predictors of functional bonds explained 19% of the variance in the construct (Table 3, Equation 9). These were expertise (beta = .24), relationalism (beta = .18), and relationship type (beta = .16). Interestingly, relationship investment had a significant effect when individual blocks were considered (Equation 7b), but this effect was mitigated partially by expertise (Equation 8a) and relationship type (Equation 9). Post hoc ANOVA found no direct effects for the predictors of functional bonds, but found significant two-way interactions between relationship investment and relationship type, communication/cooperation and relationship type, and relationship investment and relationalism.

Predictors of structural bonds also explained 19% of the variance in the construct (Table 3, Equation 12). When individual blocks were considered, relationship investment (beta = .21) and relationalism (beta = .19) were found to have a significant effect (Equation 10b), as were the covariates expertise (beta = .30) and, to a lesser extent, relationship duration (beta = .13, p

The interaction effects relating to relationship type are intriguing. Although these findings are post hoc and information is lost when categorizing the independent variables, they do suggest a need to more fully understand these relationships. To begin this process, we conducted further post hoc analysis to understand how the male-male, male-female, female-male, and female– female- buyer-supplier working relationships differed.

MANOVA (Table 4) found that at least one of the relationship pairs differed from the others in terms of relationship quality (p = .058), functional bonds (p = .019), communication/cooperation (p = .033), relationship investment (p = .005), and relationalism (p = .012). One-way ANOVA found that male-female buyer-supplier relationships were significantly different from malemale, female-male, and female-female buyer-seller relationships (Table 4). The mean levels of all the dependent constructs were lower for male-male relationships than for female-female relationships, but none of these differences was significant.

Male purchasing professionals reported greater relationalism, relationship investment, and functional bonds with male supplier reps than with female supplier reps. Female purchasing professionals reported greater relationship quality, functional bonds, communication/cooperation, and relationship investment with female supplier reps than male purchasing professionals did with female supplier reps. Finally, female purchasing professionals reported greater relationship quality, functional bonds, relationalism, relationship investment, and communication/cooperation with male supplier reps than male purchasing professionals reported with female supplier reps.


These results provide mixed support for the hypothesized relationships (see Table 5). However, they are interesting in a number of ways. First, they support the traditional marketing thought on the importance of developing social bonds and personal rapport in managing working relationships. Sales training programs usually stress the need to develop close relationships and demonstrate empathy, concern and other ties with customers on a personal level. These activities, however, are often the first to be abandoned in the name of professionalism, when confronted with increasing time constraints on the part of both buyers and sellers, and with the adoption of impersonalizing technologies such as fax machines, voice mail, and electronic mail. As one sales representative we talked to in the preliminary interviews put it: “I don’t need to do a lot of the social thing … I prefer to be in a more sterile environment, [it’s] more professional … Three martini lunches just aren’t done anymore.” We don’t advocate three martini lunches, but study results do question trends in both purchasing and sales to minimize the social dimensions of work relationships, particularly since the effects of functional bonds are mitigated by other factors, and structural bonds do not effect relationship quality.

Second, the results provide further empirical evidence to support the relationship management paradigm in marketing and suggest continued emphasis be placed on relational skills in sales training programs. Two key determinants of relationship quality were communication/cooperation, a factor concerned with being candid, open, and cooperative in decision making; and relationship investments, a factor concerned with the investment of energy and resources. These were also important predictors of social bonds. This is consistent with previous work that has identified communication openness, cooperation, and relationship investments as critical factors for effective exchange relations (e.g., Anderson & Narus, 1990; Anderson & Weitz, 1992). Interestingly, relationalism, the act of purposefully managing and seeking to improve a relationship, had an impact on social, functional, and structural bonds but had no direct effect on relationship quality. The latter finding may be a result of multicollinearity, where the explanatory power of relationalism was assumed by other relational management factors. This finding, our finding of 3 factors underlying 5 hypothesized dimensions of relationship management, and post hoc findings of interaction effects support the need for further measurement of relationship management constructs.

Finally, relationship type was found to have an effect on relationship quality, but this was mitigated entirely by relationship management constructs. Relationship type was also found to have a direct effect on functional bonds. Post hoc analysis suggests that relationship type may interact with relationship investment, communication/cooperation, and relationalism to produce complex effects that need further understanding. Post hoc MANOVA results found that the working relationships between male purchasing professionals and female supplier representatives tend to be poorer (in terms of relational management, bonds, and sometimes quality) than those of any other sex and role combination. Relationships involving female purchasing professionals tend to be stronger than those involving male purchasing professionals (except where men deal with men).

While the conclusions that can be drawn from the results of one study and post hoc analysis are clearly limited, these results are interesting and somewhat discouraging. They suggest that even with heightened awareness of discrimination issues and an increase in the number of women in professional selling roles in the 1990s, women still face greater relationship barriers than their male counterparts in dealing with male purchasing professionals. While there are probably many exceptions, on average, male purchasing professionals put more time and effort into managing working relationships and develop stronger economic, technological, and strategic bonds with salesmen than with saleswomen. Our results suggest, however, that saleswomen are able to overcome these challenges, as no differences were observed in the quality of male-female and male-male relationships. Interestingly, female purchasing professionals tend to treat supplier reps more equally than their male counterparts and develop high quality relationships with men and women. This is consistent with previous studies of interpersonal and work relationships that suggest that, for biological or sociological reasons, women are predisposed to the development, nurturing, and maintenance of relationships (e.g., Ibarra 1992; Riger & Gilligan, 1980). However, no significant differences were found between male-male and female-female relationships, which is inconsistent with this theory. The results of other studies of the quality of same-sex relationships have also been inconsistent (O’Leary, 1988). One explanation is that some studies examined the quantity of relationship attributes, processes, and outcomes, while others examined their nature, quality, or development.

The rationale for why some male buyers treat female supplier reps differentially or why male-female buyer-supplier relationships are more challenging than others was not examined in this study. However, our preliminary interviews identified a number of factors that could create friction or pose barriers to the development of effective male-female buyer-seller relationships.

Age was one factor noted by some of the younger women interviewed. Melanie, a young financial services supplier, reported: “As a young woman, older men don’t trust me.” Melanie also noted that her marital status sometimes hindered her choice of an appropriate social outing with clients: “Being married, I can’t stay out all night or go out of town fishing as would [my] male colleagues.” Previous studies have found women, and to a lesser extent men, to be highly sensitive to the spectre of sexuality in work relationships (Devine & Markiewicz, 1990). This was expressed by Cathy, a management consultant who felt uncomfortable in meeting male buyers outside of the office. She said: “[I] don’t want anything that [I] do or say to be taken the wrong way-[I’m] not sure whether this [male client] is looking for something.” Sensitivity to workplace sexuality, discrimination, and harassment was reported by some male buyers interviewed as a source of friction in their interactions with women. Bob, a public sector education buyer, reported: “I might not tell a [certain] joke to a woman, there’s a fear [among males] of being misperceived … I’m very careful when I meet a woman, I do a lot of dancing until I know their boundaries.” Role expectations were also acknowledged as barriers by some of the women interviewed, who observed greater difficulty interacting with men of some cultural backgrounds than others. While potential relationship barriers were identified in the interviews, most of the men and women observed few sex differences in relationship management practices and reported that any difficulties they had experienced were the exception rather than the norm.

Summary and Conclusions

Although a key objective of relationship marketing is building strong bonds with customers, little empirical research has examined the antecedents and consequents of relational bonds (Han, 1992; Wilson, 1995). This study found that social bonds, as well as the communication/cooperation and relationship investment facets of relationship management were key predictors of relationship quality. The relationship management facets of communication/cooperation, relationship investment, and relationalism were also found to predict social bonds, while the latter two facets also predicted functional and structural bonds. Relationship type (malemale, male-female, female-male, and female-female) was found to have some effect on relationship quality and relational bonds, but its effect appears to be primarily in interaction with facets of relationship management.

In addition to the tests of hypotheses, this study contributes to the literature by enhancing Han’s (1992) and Mummalaneni and Wilson’s (1991) conceptualizations and measures of relational bonds. While only social and structural bonds had been previously identified, we conceptualized distinct social, functional, and structural bonds that bind parties together under relational exchange. As relational bonds are central concepts in relationship marketing research (Wilson, 1995), further development of these constructs is clearly warranted.

This study contributes to the literature by beginning to examine the extent to which sex differences observed in interpersonal relationship management are also manifest in the workplace, particularly in buyer-seller relationships. Our findings suggest that male-male, malefemale-, female-male, and female-female relationships do differ in at least some aspects of relationship management, relational bonds, and in some cases relationship quality. It appears that female supplier representatives still face greater obstacles in managing working relationships with male purchasing professionals than their male counterparts do with male or female purchasing professionals. This challenge is further aggravated by the fact that women still face barriers in gaining entry to some selling jobs and have been subjected to biased treatment based on sales managers’ beliefs of gender stereotypes (Comer & Jolson, 1991) . Awareness is a first step in addressing these inequalities, and we hope our study is a first step towards this end.

The conclusions that can be drawn from a single study are clearly limited, and we heed Deaux’s (1984) caution of overinterpreting the effects of sex as an explanatory variable. Self-reports about sex-appropriate behaviour are notoriously subject to reporting bias, and there may be confounding variables that we did not consider. For example, it is not known what proportion of purchasing professional relationships were with male or female supplier reps. Frame of reference may very well play a role in setting relationship expectations. There may also be industry effects that were not considered. The finding of some sex differences in relationship management and quality, however, is interesting and may serve to encourage further research on sex differences and relational barriers.

This study was limited by its use of key informant data (e.g., Phillips, 1981). Future research needs to consider the perspectives of both purchaser and supplier reps. A critical incidence approach may be more appropriate in future research for uncovering and understanding barriers to effective buyer-seller relationships. To overcome the limitations of self-reports, future research might also incorporate projective techniques or measure nonverbal responses or overt behaviour.

More work is clearly needed to enable us to better understand the effects of sex on the management of buyer-seller relationships and interaction effects with age, culture, role expectations, and other potential relationship barriers. The paucity of such research is probably a reflection of the political sensitivity of the topic. We received written and oral comments by some of our respondents and nonrespondents who questioned the motivation for, and appropriateness of, the study. Given the heterogeneity and multicultural nature of the Canadian workplace and the increasing importance of relationship management as a competitive strategy, it is critical for researchers to stop ignoring sex, culture, and other sensitive variables that could impede the development of effective relationships.


1. For this analysis, independent variables were recoded into low, medium, and high categories with about 33% of the observations in each category.


Anderson, E., & Weitz, B. (1989). Determinants of continuity in conventional industrial channel dyads. Marketing Science, 8, 310-323.

Anderson, E., & Weitz, B. (1992). The use of pledges to build and sustain commitment in distribution channels. Journal of Marketing Research, 29, 18-34.

Anderson, J. C., & Narus, J. A. (1990). A model of distributor firm and manufacturer firm working partnerships. Journal of Marketing, 54, 42-58.

Bem, S. L. (1981). Gender schema theory: A cognitive account of sex typing. Psychological Review, 88, 354-364. Betz, M., O’Connell, L., & Shepard, J. (1989). Gender differences in proclivity for unethical behavior. Journal of Business Ethics, 8, 321-324.

Carli, L. (1989). Gender differences in interaction style and influence. Journal of Personality and Social Psychology, 57 (6), 565-576.

Comer, L., cec Jolson, M. (1991). Perceptions of gender stereotypic behavior: An exploratory study of women in selling.

Journal of Personal Selling and Sales Management, 11 (1), 43-59.

Cravens, D. W. (1995). Introduction to the special issue.

Journal of the Academy of Marketing Science, 23, 235. Crosby, L., Evans K., & Cowles, D. (1990). Relationship quality in services selling: An interpersonal influence perspective. Journal of Marketing, 54, 68-81. Crosby, L., & Stephens, N. (1987). Effects of relationship marketing on satisfaction, retention and prices in the life insurance industry. Journal of Marketing Research, 24, 404-411.

Day, G., & Wensley, R. (1983). Marketing theory with a strate

gic orientation. Journal of Marketing, 47, 79-89. Deaux, K. (1984). From individual differences to social categories: An analysis of a decade’s research on gender. American Psychologist, 39, 105-116.

Deaux, K., & Major, B. (1987). Putting gender into context: An interactive model of gender-related behavior. Psychological Review, 94 (3), 369-389.

Devine, I., & Markiewicz, D. (1990). Cross-sex relationships at work and the impact of gender stereotypes. Journal of Business Ethics, 9, 333-338.

Foster, B., & Orser, B. (1994). A marketing perspective on women in management: An exploratory study. Canadian Journal of Administrative Sciences,11, 339-345. Gilligan, C. (1982). In a different voice: Psychological theory and women’s development. Cambridge, MA: Harvard University Press.

Gladstein, D. (1984). Groups in context: A model of task group effectiveness. Administrative Science Quarterly, 29, 499517.

Hall, E. (1993). Smiling, deferring, and flirting: Doing gender by giving `good service: Work and Occupations, 20 (4), 452471.

Han, S. (1992). Antecedents of buyer-seller long-term relationships: An exploratory model of structural bonding and social bonding. (Working Paper No. 6-1992). University Park, PA: Institute for the Study of Business Markets, Pennsylvania State University.

Heide, J. B., & John, G. (1992). Do norms matter in marketing

relationships? Journal of Marketing, 56, 32-44. Henley, N. M. (1977). Body politics: Power, gender, and nonverbal communication. Englewood Cliffs, NJ: Prentice Hall.

Homans, G. (1961). Social behavior: Its elementary forms. New York, NY: Harcourt, Brace and World. Iacobucci, D. (1994). Some thoughts toward defining relationship marketing. In J. Sheth & A. Parvatiyar (Eds.), Relationship marketing: Theory, methods, and applications-1994 research conference proceedings (section 2.1.1, pp. 1-10). Atlanta, GA: Center for Relationship Marketing, Emory University.

Journal of the Academy of Marketing Science (1995), fall spe

cial issue.

Ibarra, H. (1992). Homophily and differential returns: Sex differences in network structure and access in an advertising firm. Administrative Science Quarterly, 37, 422447.

Kanuk, L. (1978). Women in industrial selling: How great are

the career opportunities? How well are they performing? Journal of Marketing, 1, 87-91.

Keys, D. (1985). Gender, sex role, and career decision making of certified management accountants. Sex Roles, 13, 3346.

Kumar, N., Scheer, L. K., & Steenkamp, J. B. (1995). The effects of supplier fairness on vulnerable resellers. Journal of Marketing Research, 32, 54-65. Lau, G. T. (1990). The effects of buying task characteristics on buyer-seller conflict management behavior. Unpublished doctoral dissertation, University of Western Ontario, London, ON.

McCall, G. J. (1970). The social organization of relationships. In G. J. McCall et al. (Eds.), Social Relationships (pp. 334). Chicago, IL: Aldine Publishing Company. Moorman, C., Zaltman, G., & Deshpande, R. (1992). Relationships between providers and users of market research: The dynamics of trust within and between organizations. Journal of Marketing Research, 29 (3), 314-28.

Mummalaneni, V., & Wilson, D. (1991). Modelling the influence of a close personal relationship on buyer commitment to a supplier. (ISBM Working Paper). University Park, PA: Pennsylvania State University. Nevin, J. R. (1995). Relationship marketing and distribution channels: Exploring fundamental issues. Journal of the Academy of Marketing Science, 23, 327-334. Noordewier, T. G., John, G., & Nevin, J. R., (1990). Performance outcomes of purchasing arrangements in industrial buyer-vendor relationships. Journal of Marketing, 54, 80-93.

O’Leary, V. E. (1988). Women’s relationships with women in the workplace. In B. Gutek, A. H. Stromberg, & L. Larwood (Eds.), Women and work: An annual review (Vol. 3). Newbury Park, CA: Sage.

Palmer, A., & Bejou, D. (1995). The effects of gender on the development of relationships between clients and financial advisors. International Journal of Bank Marketing, 13 (3), 18-27.

Pedhazur, E. J. (1982). Multiple regression in behavioral

research (2nd ed.). New York, NY: CBS College. Phillips, L. W. (1981). Assessing measurement error in key informant reports: A methodological note on organizational analysis in marketing. Journal of Marketing Research, 18, 395-415.

Riger, S., & Gilligan, P. (1980). Women in management: An exploration of competing paradigms. American Psychologist, 35, 902-910.

Schmidt B., LeClarc, F., & Dube-Rioux, L. (1988). Sex typing and consumer behavior: A test of gender schema theory. Journal of Consumer Research, IS, 122-128. Schul, P., Remington, S., & Berl, R. (1990). Assessing gender

differences in relationships between supervisory behaviours and job-related outcomes in the industrial sales force. Journal of Personal Selling and Sales Management, 10, 1-16.

Schul, P. L., & Wren, B. M. (1992). The emerging role of women in industrial selling: A decade of change. Journal of Marketing, 56, 38-54.

Sherrod, D. (1989). The influence of gender on same-gender friendships. In C. Hendrick (Ed.), Review of personality and social psychology: Close relationships (pp. 164-186). Newbury Park, CA: Sage.

Sheth, J. N., & Parvatiyar, A. (1995). Relationship marketing in consumer markets: Antecedents and consequents. Journal of the Academy of Marketing Science, 23, 255-271. Smith, J. B. (1993). Horizontal selling alliances: Distance and mutual trust. Unpublished doctoral dissertation, University of Western Ontario, London, ON. Smith J. B., & Bejou, D. (1995). Sex differences in relationship management: An exploratory analysis. In Ken Grant & Ian Walker (Eds.), Proceedings of the World Marketing Congress (Vol. 7, pp. 125-132). Melbourne, Australia: Academy of Marketing Science.

Statham, A. (1987). The gender model revisited: Differences in the management styles of men and women. Sex Roles, 16, 409-429.

Swan, J. E., Rink, D. R., Kiser, G. E., & Martin, W. S. (1984). Industrial buyer image of the saleswoman. Journal of Marketing, 48, 110-116.

Tavris, C. (1992). The mismeasure of woman. New York, NY: Simon & Schuster.

Thibaut, J. W., & Kelley, H. H. (1959). The social psychology

of groups. New York, NY: John Wiley. Turner, R. H. (1970). Family interaction. New York, NY: John Wiley.

Weitz, B. A., & Jap, S. D. (1995). Relationship marketing and distribution channels. Journal of the Academy of Marketing Science, 23, 305-320.

Westbrook, R. (1981). Sources of consumer satisfaction with

retail outlets. Journal of Retailing, 57, 68-85. Wilson, D. T. (1995). An integrated model of buyer-seller relationships. Journal of the Academy of Marketing Science, 23, 335-345.

Brock Smith

University of Victoria

Funding for this research by the SSHRC and the University of Victoria and research assistance by Maria Barnes is gratefully acknowledged. Address all correspondence to Brock Smith, Faculty of Business, University of Victoria, PO Box 1700, Victoria, BC, Canada, V8W 2Y2.

Copyright Administrative Sciences Association of Canada Mar 1998

Provided by ProQuest Information and Learning Company. All rights Reserved