Abstract
Horizontal referrals—when suppliers recommend other suppliers—are a common phenomenon in complex B2B markets. For the referring supplier, giving the best possible horizontal referral may strengthen the relationship with its customer, yet it may also threaten the referring supplier's future revenues and cross-selling opportunities. Instead, the supplier could make an obligatory referral, one that fulfills the obligations of recommending another supplier while keeping the referring supplier's own interests paramount. The authors rely on role theory and its antecedents (mutual trust and referring supplier's dependence) to determine when a referring supplier adopts the role of a friend (vs. a businessperson) and gives the best possible referral (vs. an obligatory referral). Study 1, an experiment, supports the theoretical model. Study 2, a conjoint study, links the observable antecedents of the referring supplier–customer relationship to the choice of horizontal referrals. Study 3, another experiment, looks at the consequences of the horizontal referral on the referring supplier–customer relationship and shows that providing an obligatory referral can hurt the customer's intent to continue the relationship with the supplier. This effect is mediated by the customer's perceived alignment of interest with the supplier. For B2B marketing research and practice, the authors report that the supplier's dependence is critical in predicting the quality of horizontal referrals, even though an exploratory survey showed that customers overlook that dimension and focus on mutual trust when seeking referrals.
Following the conclusion of a marketing analytics consulting engagement, [the client organization] asked us whether we could further develop an advanced predictive model. That kind of modeling effort required specific resources our company did not have access to. Even if it had, developing this particular capability was not deemed strategic to our firm. Consequently, we replied that no, unfortunately, our company did not offer such services. The customer then asked: “If you cannot do it, then, do you know someone who can?” We did. Unfortunately, the best recommendation for the job was also one of our firm's biggest competitors. (CEO of an analytics consulting firm, interview for this research, 2019)
The marketing technology industry has seen the number of suppliers increase seventyfold in less than a decade, from 150 in 2011 to more than 9,930 in 2022 (Brinker 2022). In such complex business-to-business (B2B) industries, suppliers frequently know more about other suppliers than their customers do. Customers often attempt to tap into that knowledge and reduce their search costs by asking their suppliers to recommend another supplier (as the preceding example demonstrates). Hada, Grewal, and Lilien (2010) define such a recommendation—where a supplier recommends another supplier to their customer—as a horizontal referral. However, giving a horizontal referral can lead to a conundrum for the referring supplier (the supplier giving the recommendation).
On the one hand, an ideal recommendation for the customer, irrespective of the future competitive threat it provides for the referring supplier, could position the referring supplier as an advisor who always acts in its customer's best interests. The business press recognizes the value of building a strong relationship with one's customers by recommending other suppliers, even competitors (Green 2011): What? Why would you ever recommend a competitor? … If there is any sure sign that a salesperson is a trusted advisor, it is this: the salesperson's willingness to turn down a sales opportunity—even better, to recommend a competitor—if that were in the client's best interests. Given how tough it has become to win new work, you might question the logic of ever recommending a competitor for work that your firm could perform (or any kind of work). My boss certainly did. … [But] Recommending a competitor is an effective way to demonstrate that you can be trusted to act on the client's behalf.
On the other hand, if the best supplier for the customer is one that could take away future revenue (as our interview quote highlights), the referring supplier must choose between a referral in the customer's best interest (which we refer to as a best possible referral) or another that would not pose a future threat to the supplier, but still satisfy the supplier’s obligations to the customer (an obligatory referral).
Surveys and interviews we conducted underscore this problem. In a survey we conducted with 132 B2B customers, 72% reported actively seeking horizontal referrals at least once from their existing suppliers. In an interview, a supplier confirmed receiving such customer requests “at least once a week …, if not more.” For instance, a firm supplies its health care customer with customer relationship management (CRM) analytics services. The customer aims to acquire a CRM software solution. Seeking to benefit from the supplier's market knowledge, the customer asks the supplier to recommend a CRM software supplier (i.e., to make a horizontal referral). The dilemma the referring supplier faces is that such a horizontal referral might also provide CRM analytics services (competing directly with the referring supplier's current offering), may overlap with the referring supplier on adjacent markets such as cloud services (impeding the referring supplier on potential cross-selling opportunities with its customer), or both.
Thus, although the referring supplier might wish to act as a “trusted advisor” (as the Green [2011] article indicates) and give the best possible recommendation to the customer, such a recommendation could threaten its current revenue or future growth opportunities. In this research, we ask: How do referring suppliers behave in a situation of possibly misaligned interests, and what is the effect of their behavior on the referring supplier–customer relationship? To answer these questions and better understand the phenomenon of horizontal referrals in B2B markets, we take a multipronged approach (Figure 1).
Research Questions and Studies.
First, we conduct an exploratory survey with customers and interviews with referring suppliers and describe how the phenomenon of horizontal referrals works in practice. Our survey shows that customers extensively rely on horizontal referrals when they need to find new suppliers (81% of our sample). Customers report relying on suppliers they trust and expect the referring supplier to act in the customer's best interests. However, our interviews with suppliers show that they are conflicted about which supplier to recommend (as the opening quote demonstrates). A literature review on horizontal referrals and interfirm relationships in economics and marketing fails to provide a definitive answer about how referring suppliers will behave in such situations.
Second, we develop a theoretical model to address how referring suppliers act by integrating role theory with social exchange theory in interfirm relationships (Koch and Schultze 2011; Kumar, Heide, and Wathne 2011). We show that referring suppliers tend to act either as friends or as businesspersons (e.g., Heide and Wathne 2006) and determine that the roles referring suppliers are most likely to adopt depend on supplier–customer mutual trust and referring suppliers’ dependence on the customer. In turn, the roles referring suppliers adopt influence the horizontal referrals they make (i.e., best possible or obligatory). We validate our theoretical framework with an experiment in Study 1 (see Figure 1).
Third, we acknowledge that firms in business relationships can be incorrect about their perceptions regarding these relationships and seek to understand when that incorrectness occurs (Vosgerau, Anderson, and Ross 2008). We identify objective and measurable antecedents of mutual trust (relationship length and interaction frequency) and the referring supplier's dependence on the customer (economic dependence and product breadth) and hypothesize how each of these four antecedents will impact the likelihood of a referring supplier making a best possible referral (acting in the customer’s interest) versus an obligatory one (acting only to fulfill the referring supplier's obligations toward the customer). These predictions provide guidelines to managers seeking to obtain horizontal referrals. We test these predictions with a choice-based conjoint experiment in Study 2.
Fourth, building on Studies 1 and 2, we assess the effect of the referring supplier's choice of horizontal referral (best possible vs. obligatory) on the customer’s intentions of continuing with their relationship if the customer detects that the referring supplier did not make the best possible referral. We test our hypotheses with a between-subjects experiment in Study 3.
We make three contributions. First, we contribute to the literature on horizontal referrals in economics and marketing. The economics literature assumes that suppliers’ interests are misaligned with those of the customers when giving horizontal referrals (e.g., Arbatskaya and Konishi 2012; Garicano and Santos 2004). This literature concludes that referring suppliers will act in their own interests and choose to service the customers themselves and not refer other suppliers. In contrast, we rely on the social exchange perspective in interfirm relationships to argue that the referring suppliers’ behaviors depend on the referring supplier–customer relationships and the referring suppliers’ characteristics. Our perspective, counter to the economics literature, predicts that referring suppliers will often act in their customers’ interests.
Second, we rely on role theory to uncover why some referring suppliers act only out of obligation while others go above and beyond to act in the customers’ interest. These findings contribute to the literature on suppliers’ behavior, wherein suppliers stay within obligations in their relationships or go beyond them (e.g., Wuyts, Verhoef, and Prins 2009). We also contribute to this literature by studying a situation in which referring suppliers’ and customers’ interests can be misaligned and highlight the role that dependence in the relationship plays in influencing how each entity behaves.
Third, we provide managerial implications for both customers and suppliers. Our exploratory survey indicates that when customers seek horizontal referrals, they tend to rely on trusted suppliers. We show that, in fact, they should also seek the advice of suppliers that are not economically dependent on them.
Literature Review: Horizontal Referrals and Suppliers’ Behavior
Horizontal Referrals
Research on horizontal referrals has primarily been conducted in the economics domain, assuming that horizontal referrals almost always lead to misaligned interests between referring suppliers and customers (see Table WA1 in Web Appendix A). For example, Grassi and Ma (2016) note that when the supplier knows of another supplier that would be better for the customer (as opposed to itself), the referring supplier faces a trade-off between “honestly advising clients to build a good reputation [and giving a horizontal referral] and reaping a quick profit at the client's expense [by serving the client itself]” (p. 938).
Consequently, the economics literature, primarily via analytical models, finds that suppliers will act in their own interests and attempt to service customers themselves, as opposed to giving a horizontal referral (e.g., Bolton, Freixas, and Shapiro 2007; Garicano and Santos 2004). Interestingly, in a rare model of a repeated long-term B2B relationship, Park (2005) finds that the supplier will give a horizontal referral only when there is fear of the customer punishing the referring supplier for the deception of not recommending. Thus, the economics literature argues that suppliers will deceive customers to keep business for themselves, except when customers could discover the deception.
The marketing literature on horizontal referrals is sparse (Table WA1). Similar to the economics literature, Mayzlin and Yoganarasimhan (2012) build an analytical model and show that there are trade-offs in blogs’ referrals to other blogs (referrals promote readership, but referrals to better blogs can lead to loss of readers). In empirical research, Blanchard, Hada, and Carlson (2018) show that a salesperson giving a horizontal referral for a nonfocal product (e.g., bed frame) can increase sales of the focal product (e.g., mattress). As Table WA1 shows, horizontal referrals have not been studied extensively for existing relationships in B2B markets; hence we consider the literature that addresses suppliers’ behavior in B2B marketing relationships.
Suppliers’ Behavior
Horizontal referrals—when the customer relies on the supplier to reduce its search costs—occur because the customer–supplier relationship is built on socioeconomic exchanges. In such situations, the supplier's behaviors toward its customers have been shown to be driven by prosocial considerations as well as economic ones. These dual and often contradictory considerations lead the supplier to act either within the expected obligations of the relationship (in-role behaviors) or go beyond them (extra-role behaviors; e.g., Wuyts 2007).
Extra-role behaviors are efforts voluntarily expended beyond the call of duty (Kim and Mauborgne 1996). A distinguishing feature of extra-role behaviors is that they are voluntary (Kim, Hibbard, and Swain 2011) and are oriented toward helping the customer (Wuyts 2007). For example, a supplier might go beyond its obligations and absorb demand and pricing shocks for its customer. In such a case, the supplier has gone above and beyond the stated guidelines of the relationship. Not surprisingly, positive relationships characterized by a cooperative atmosphere (Wuyts 2007), benevolence (Li 2010), and extant gratitude (Mangus et al. 2017) lead to a firm exhibiting extra-role behaviors. At the same time, suppliers also exhibit extra-role behaviors when they face the risk of losing their business, such as when the customer has a multisourcing strategy (Wuyts 2007). Regardless of the motivations of the supplier, research has shown that going beyond obligations in a relationship leads to positive relational consequences such as relational performance (Li 2010) and profitability (Wuyts, Verhoef, and Prins 2009; Table WA2 in Web Appendix A).
However, this stream of research has also shown that extra-role behaviors are likely to have greater positive effects on early-stage relationships than on more mature ones. Mangus et al. (2017) argue that this relative difference is due to the rising expectations of customers about suppliers’ extra-role behavior in positive relationships. Similarly, Kim, Hibbard, and Swain (2011) argue that in-role behaviors are considered insufficient for maintaining the relationship when firms can switch between suppliers. If so, suppliers’ extra-role behaviors in positive relationships might not lead to the expected rewards.
Although this stream of literature considers behaviors that are both within and beyond obligations in a relationship, it has not looked at situations with misaligned interests between the customer and the supplier. Which relational factors govern the supplier's likelihood of putting the customer’s interests ahead of its own? And what are customers’ expectations in this regard? As a starting point, we rely on customer and supplier surveys and interviews to gain insights.
Exploratory Studies: Horizontal Referrals in Practice
To better understand horizontal referrals in B2B markets and to provide a foundation for our conceptual model, we conducted two exploratory studies: a survey of customers’ perspectives on asking for—and relying on—horizontal referrals, followed by in-depth interviews addressing referring suppliers’ perspectives on giving horizontal referrals.
Perspectives from B2B Customers (Survey)
Sample
We hired Dynata Inc., a market research company with specialized B2B panels, and surveyed 126 managers in supplier-facing roles. Our sample averaged 22 years of work experience, and 45% had roles related to purchasing (see Web Appendix B).
Findings
Among the surveyed customers, 81% reported horizontal referrals: 72% had asked their current suppliers to recommend other suppliers, 74% had received unsolicited horizontal referrals from their current suppliers, and 65% had experienced both. We asked these respondents to recall one specific horizontal referral and answer questions based on that instance (Table 1).
Examples of Horizontal Referrals from Exploratory Customer Survey.
Respondents reported using horizontal referrals because suppliers know their industry better than customers (e.g., “he had more experience than I did in that particular market” and “they are the ones in the industry, and I feel they would have the best information”; Table 1). Customers usually asked for a recommendation for a different product/solution in the supplier's industry than the one the referring supplier provides (e.g., asking the supplier for office information technology [IT] supplies to recommend a supplier for audiovisual IT support; Table 1), but there were instances in which customers asked for a horizontal referral to reduce their reliance on a single source for parts (e.g., “reduce backlog” and “reduce supply risk by developing a secondary supplier relationship”).
Respondents often mentioned the positive relationship they have with the referring supplier as a reason for asking for a horizontal referral (e.g., “good relationship,” “long-term relationship,” and “worked with for several years”) and that they trusted the referring supplier to act in the customer's interests (over 27% of the open-ended responses mentioned trust; Table 1 provides examples). Although three of the respondents indicated caution in relying on the referring supplier's recommendation (“only when their volume is larger than mine” and “I’m not always confident they’ll be unbiased in their responses”), such responses were rare.
We also asked respondents to rate the relationship they had at the time with the referring supplier (1 = “Strongly disagree,” and 7 = “Strongly agree”). They reported a relationship high on trust (M = 6.17, SD = .89) and low on conflict (M = 2.41, SD = 1.69), with frequent interactions (M = 5.59, SD = 1.54) and relationship lengths that varied from six months to 30 years. They also reported that, on average, referring suppliers were economically dependent on them, although with great variability in the sample (M = 4.72, SD = 1.71). Reflecting on their experience, respondents stated that they felt referring suppliers generally referred them to appropriate new vendors (M = 5.40, SD = 1.40).
Overall, the survey revealed that customers rely extensively on horizontal referrals to reduce their search costs for a new supplier (as the economics literature has emphasized) and that they largely trust the referring supplier to act in their (the customer's) interests.
Perspectives from B2B Referring Suppliers (Interviews)
To understand how referring suppliers perceive horizontal referrals, we conducted ten in-depth interviews with customer-facing managers in B2B supplier firms. These firms ranged from large companies selling industrial products to small firms offering technological and analytical services (see Web Appendix B). These managers also reported that horizontal referrals are common practice. For example, when asked how often he receives such a request, a vice president of a marketing analytics firm answered: “At least once a week. If not more.” A director at an education solutions firm concurred: “I can’t even count the number of times this has happened in my career.”
The interviews indicated that suppliers are keen to maintain a good relationship with their customers but are aware of the potential threats from certain suppliers they could recommend, suppliers that could be strong competitors for a customer's business. A vice president of sales at a consulting firm explained his thoughts when customers request horizontal referrals: We have to show good faith by giving a good referral. But we really don’t want to bring in a possible competitor; our model is to expand our share of wallet with the customer—everyone who is in this solution industry focuses on “land and expand.” We don’t want competitors to take that [future revenue]. Our approach is to enter into a relationship with a customer and then get a good chunk of their business. Whenever I have to refer another firm, typically because we lack capability in that country, I am very, very wary of who will walk in. It is not difficult to take away business with a cheaper person-per-hour quote.
The director of a digital publishing firm presented a similar perspective: We are often capacity constrained in add-on services, so we have recommended other firms—but smaller firms. For our key long-term customers, we try and outsource to these smaller firms ourselves and keep the client with us. It's a bit tricky, I want to do my best for these long-term customers, but these add-on services have higher margins and we do not want to lose that either! The customer did buy [the horizontal referral's] device, but long-term, I think I came out on top.
A CEO of a chip manufacturing start-up stated: We fill a niche; all my customers are annual repeat customers and we trust each other. I am quite confident that I am not going to lose a big chunk of future work. So, I [recommend] the best I can. I’m a one-person firm. My clients are with me because they trust me and my judgment. I will do the best I can. But I also need every single one of my clients. It's a trade-off.
The interviews indicate that referring suppliers are conflicted when giving horizontal referrals: they are aware of the long-term view that requires them to put their relationship first, but at the same time, they are concerned about losing revenue by recommending suppliers who offer lower prices or products that could compete with their offerings.
Study 1: Role Theory and Horizontal Referrals
Our exploratory studies have indicated that customers are likely to rely on referring suppliers in complex markets (81% of our survey sample) and that those referring suppliers are unlikely to recommend badly performing suppliers. However, even assuming a positive relationship between the supplier and the customer, it is unclear what a referring supplier will do when interests are misaligned. When presented with a situation requiring a horizontal referral, a referring supplier can follow four courses of action:
No referral: The referring supplier denies/dodges the request. Substandard referral: The referring supplier gives a blatantly substandard horizontal referral (e.g., low performing, overly expensive, incapable of meeting the customer's needs). Best possible referral: The referring supplier makes the best referral possible and does so with the customer's interests at heart. Obligatory referral: The referring supplier considers the suppliers it knows and chooses one that satisfies the customer's request (thus fulfilling the obligations toward the customer) but is purposely not the best for the customer, hence limiting the competitive threat to the referring supplier.
Our exploratory surveys, and a review of the literature on horizontal referrals, indicate that referring suppliers are unlikely to give no referral (they realize the importance of the request) or a substandard referral (they want to maintain their positive relationship with the customer). Therefore, in this research we address an important but unexplored research question: under which circumstances will a referring supplier give a best possible referral versus an obligatory referral?
Park (2005) finds that in a repeated long-term relationship, the supplier will give a horizontal referral only when there is a fear of the customer punishing the referring supplier for the deception of not recommending. However, in complex B2B markets, a customer may not easily discern whether the referring supplier suggests an obligatory or a best possible supplier (at least not ex ante). The referring supplier's choice is rather a reflection of how it sees itself behaving in specific situations (as obliged or beyond), irrespective of the likelihood of detection. Role theory focuses on how entities behave because they perceive themselves as following certain roles, depending on the situations in which they find themselves, even devoid of the chance of detection (March 1994). Therefore, role theory is well suited to the context of horizontal referrals.
Theoretical Development: Role Theory
Heide and Wathne (2006) define an organizational role as “an organizational ‘identity’ or ‘collective mind’ which provides the foundation for shared perceptions and decision-making” (p. 91). According to role theory, a decision maker in a firm asks the following questions: (1) What kind of situation is this? (2) What kind of organization is this? and (3) What does an organization such as this do in a situation such as this? Thus, the decision maker first considers the situation and then acts based on what the firm considers appropriate in such situations (Koch and Schultze 2011).
Montgomery (1998) identified two role archetypes that firms adopt: (1) a friend, whose decisions are guided by the norms and expectations in the relationship, and (2) a businessperson, whose decisions are guided by utility maximization considerations and not bound by expectations of behavior. Role theory has seen applications in marketing, especially in how frontline employees can feel role conflict when there is ambiguity (De Ruyter, De Jong, and Wetzels 2009) or how sales reps adopt a friend or a businessperson role depending on the situation (Grayson 2007).
Applying role theory to interfirm relationships, Heide and Wathne (2006) illustrate the differences in the friend and businessperson roles, taking a cooperative interfirm relationship as an example: A supplier and a customer may develop the norm of cooperation and act in each other's interests. A supplier that adopts the role of a friend will cooperate with the customer because the supplier believes such cooperation is expected, even if it is not in the supplier's interest. A supplier that adopts the role of a businessperson will cooperate or not with the customer based on the decision that maximizes the supplier's interest in that situation. Heide and Wathne (2006) argue that these roles can coexist, and firms do not always make decisions according to only one role. Depending on the situation, a firm can act as a friend or a businessperson.
Thus, we consider how the referring suppliers’ adoption of the role of friend or businessperson is likely to affect their choice of making a best possible or obligatory referral. The role the referring supplier adopts when making a horizontal referral depends on the situation, which, in turn, is defined by the customer–supplier relationship (Koch and Schultze 2011).
Hypotheses
As discussed previously, interfirm relationships are economic exchanges embedded in a social structure, and a horizontal referral takes place in such a social environment. Thus, to consider antecedents to the referring supplier's behavior when giving a horizontal referral, we rely on the social exchange theory, which leads to two perspectives in marketing: trust–commitment and dependence (Palmatier, Dant, and Grewal 2007).
The trust–commitment perspective argues that “commitment and trust, not power or dependence, are the keys to inter-organizational relational performance … help foster mutual goals and mitigate each channel member from acting entirely in their own self-interest” (Watson et al. 2015, p. 550). Trust in a relationship is defined as the belief that an exchange partner is benevolent and honest and will act with integrity toward the firm (Grayson, Johnson, and Chen 2008), whereas mutual trust is defined as the degree to which a partner believes that both entities trust each other and will act accordingly toward each other (e.g., Anderson, Lodish, and Weitz 1987). For suppliers and customers to rely on each other when considering other suppliers, we assume that bilateral reciprocation of trust is needed. Therefore, we study mutual trust (for brevity, we use the word “trust” hereinafter) between the referrer and the supplier as the key antecedent to the role the referring supplier might adopt. Trust involves the willingness to rely on a partner, which has as its basis an expectation that the motives, intentions, and behavior of the partner are trustworthy. With high mutual trust between the supplier and its customer, the supplier is likely to act based on altruistic/communal motivations and would be willing to go beyond obligations for the customer (e.g., Wuyts 2007). A supplier that is willing to go beyond obligations and genuinely wants to help the customer is likely to adopt the role of a friend. Therefore:
Conversely, the dependence perspective builds on the premise that dependence of one firm on another affects that firm's influence on the other, affecting the interfirm relationship (e.g., Antia, Zheng, and Frazier 2013). A referring supplier's dependence on a customer refers to the need to maintain the relationship with the customer to achieve the supplier’s current and future goals (Ganesan 1994). The higher the referring supplier's dependence on the customer, the more threatened the supplier is by losing current revenues. In such a situation, a supplier is likely to be driven by instrumental motivations rather than altruistic ones (e.g., Bolino 1999). Thus, the supplier desires to “generate a positive impression for competitive reasons (to have an edge over competing suppliers) or for cost reasons (to avoid the costs associated with relationship termination)” (Wuyts 2007, p. 303).
A decision maker adopts the role of a businessperson when acting according to the consequences of the decision and not according to the expectations established in the relationship (Heide and Wathne 2006; Montgomery 1998). Since dependence is likely to lead the supplier to act under instrumental motivations (Wuyts 2007), we hypothesize:
Grayson (2007) finds a similar effect in the context of sales reps in a network-selling scheme. When relationships are financially valuable, the rep is less likely to act as a friend and more likely to act as a businessperson. A firm that acts as a friend makes decisions based on what is appropriate and expected in a friend-based relationship. A referring supplier that acts as a friend would likely give a referral that is in the customer's best interests, even if it poses a risk to the supplier’s future business with the customer.
A referring supplier will want to appear as if it is acting with the customer's best interests at heart to generate a positive impression. Yet, doing so by giving a horizontal referral that could take away future revenue may jeopardize its current and future profit. A firm that adopts the role of a businessperson makes decisions based on the consequences of its choices, choosing what maximizes its own utility. Therefore:
Figure 2 summarizes our hypotheses.
Research Design
We summarize the conceptual model and hypotheses in Figure 2. We test H1–H3 with a 2 × 2 between-subjects experiment. We manipulate trust (high vs. low) and the referring supplier's dependence on the customer (high vs. low) and measure the propensity of the referring supplier to act as a friend (vs. a businessperson) and make a best possible (vs. obligatory) horizontal referral. For our empirical context, we require a situation in which the referring supplier and customer have an existing relationship and the customer requires a solution in the referring supplier's industry. This difference in industries between the referring supplier and the customer forms the basis of horizontal referrals, as the referring supplier knows more about its industry than the customer does. Our respondents play the role of a referring supplier in the analytics and technology industry. Their client is a health care insurance company building a data center and asking for a recommendation for a new technology supplier 1 (Web Appendix C provides details).
Conceptual Model for Study 1.
Manipulations and Measures
To manipulate high (low) trust, we adapt the following from Srivastava and Charkrvarti (2009): Referring supplier and customer have worked well together for more than three years. In past dealings, both firms have always (mostly) kept their word and looked out for each other in business transactions. There have been no (several) instances in which the firms have had to resort to legal action to enforce business agreements. Referring supplier is (not) highly dependent on the customer. If this relationship were to end, referring supplier, and you as the key account manager, would (not) see a significant financial and professional loss.
There is little research on the roles of friend and businessperson in interfirm relationships. We adapted items from Dong, Tse, and Hung (2010) to assess referring suppliers’ role adoption, modifying the items to correspond to firm actions. Our theoretical arguments rely on the referring supplier perceiving horizontal referral as a situation of misaligned interests with the customer. Therefore, we measure respondents’ perceived misalignment of interests and control for it. All measurement items presented in Web Appendix C.
For referral choice, respondents must choose between the following two options: A supplier that would be best for the customer from the available suppliers you know and would take away referring suppliers’ future revenue from customer. [best possible referral]
A supplier that would give the impression of being the best for the customer even if there are better available suppliers you know and would not take away referring suppliers’ future revenue from customer. [obligatory referral]
Model
We estimate a system of two equations. The first equation linearly estimates whether the supplier will adopt the role of a friend (vs. businessperson) measured on a continuous scale (1–7) as a function of customer–supplier mutual trust (manipulated, high/low), supplier's dependence (manipulated, high/low), and their two-way interaction (see Table 2 and Figure 3). The second equation uses a logit specification to estimate whether the supplier will make a best possible referral (coded as 1) or an obligatory one (coded as 0) as a function of the role the supplier adopts (i.e., friend vs. businessperson). It also controls for the direct effect of trust, supplier dependence, and their two-way interaction (see Table 2 and Figure 3). We estimate the two equations simultaneously and use bootstrapping to estimate the indirect and total effects of trust and dependence on referring suppliers’ choice of horizontal referral (for the estimation approach, see Hayes [2015]).
Results (Study 1): Supplier’s Choice to Make a Best Possible Referral.
Parameter Estimates (Study 1).
*p < .1. **p < .05. ***p < .01 (one-tailed tests).
Notes: N = 387. Trust and supplier dependence are significant antecedents of the roles the referring suppliers will adopt, and referring supplier is more (less) likely to make a best possible horizontal referral if it adopts the role of a friend (businessperson). See Figure 3 for illustration.
Results
Our sample consists of 387 respondents with at least two years of work experience recruited on Prolific Academic (details in Web Appendix C). We find an equal split (approximately 50%) between respondents who stated that they would make a best possible referral and an obligatory referral, indicating that referring suppliers often—but not always—act in their customer's best interests.
Hypothesis testing
All hypotheses are supported. We find that a referring supplier is more likely to act as a friend in the context of a high-trust relationship (.23; p < .05) but less likely to act as a friend in a high-dependence context (−.15; p < .05), confirming both H1 and H2. The more a supplier acts as a friend, the more likely the supplier is to make a best possible referral (.95; p < .01), in strong support of H3. We report the results in Table 2 and Figure 3.
Mediation analysis
We use bootstrapping (Hayes 2015) to also estimate the indirect and total effects of trust and dependence on the likelihood of making a best possible referral. Trust does not have a direct effect on best possible referral; however, as hypothesized, the total effect of trust through its mediator is significant and positive (.22; 95% CI: [.40, .05]). The total effect of dependence is negative but marginally significant (−.15; 90% CI: [.01, −.13]). Furthermore, we find that dependence has a significant, negative, direct effect on the likelihood of making a best possible referral (−.41; 95% CI: [−.08, −.70]).
Interactions
Figure 3 displays the marginal total effect of trust and dependence on the likelihood of selecting a best possible referral. An interesting pattern emerges. As dependence increases, the likelihood of a best possible referral decreases across the board. The effect of trust is more subtle. As trust increases, the likelihood of a best possible referral does, in fact, increase, but that effect is only significant in the low-dependence condition. In all cases, this effect is far from sufficient to compensate for the negative effect of a high-dependence relationship. Thus trust matters, but dependence matters far more.
Study 2: Observable Antecedents and Choices of Horizontal Referrals
While Study 1 deciphered the underlying mechanisms and theoretical underpinnings of the phenomenon of interest, it relied on perceptual measures. Methodologically, perceptual measures limit external validity, and results could be affected by respondents’ social desirability. Theoretically, firms can incorrectly read the perceptual aspects of their relationship (Vosgerau, Anderson, and Ross 2008). It is also unclear how an obligatory referral—a referral that balances the referring supplier's obligations to the customer and future competitive threat from another supplier—would manifest itself. Study 2 addresses these two limitations by studying (1) the observable antecedents of trust and referring supplier's dependence and (2) referring suppliers’ observable choice of either a best possible or obligatory referral. We summarize Study 2 in Figure 4.
Conceptual Model for Study 2.
We first explore the observable antecedents of trust and dependence. We then identify the characteristics of best possible versus obligatory referrals.
Theoretical Development: Antecedents of Trust and Dependence
Antecedents of trust
In Study 1, we considered mutual trust between the referring supplier and the customer. The literature has identified several antecedents to trust: relationship length (Jap 1999), communication frequency (Wuyts and Geyskens 2005), communication valence (positive/negative), and conflict (see Palmatier et al. [2006] for a meta-analysis). Our exploratory studies revealed that horizontal referrals are far more likely to occur when some trust is present. Although negative communications and conflicts are key determinants of trust in many B2B relationships, they do not apply in our context, where such negative elements would preclude a request for a horizontal referral. Therefore, we consider a good relationship a prerequisite for asking for a horizontal referral and focus on relationship length and interaction frequency as antecedents of trust in that relationship (e.g., Doney and Cannon 1997). Indeed: “Both [relationship duration and interaction frequency] provide trading partners with more behavioral information in varied contexts, which allows for better predictions that should increase each party's confidence in its partner's behavior” (Palmatier et al. 2006, p. 140). Confidence in a partner's behavior is central to trust. The literature also shows that dyadic factors such as relationship length and interaction frequency induce parties to act in each other's interests (e.g., Samaha, Beck, and Palmatier 2014), that is, to act as a friend, consistent with our framework.
Antecedents of dependence
The second factor we considered in Study 1 was the referring supplier's dependence on the customer. One of the managers we interviewed stated that a common (and effective) B2B sale strategy was to “land and expand.” 2 The referring supplier's revenues depend on both.
The first component of a referring supplier's dependence is the consequences of losing ongoing revenue (i.e., losing current “land”). Some customers are essential to the financial health of suppliers, whereas others are not. We label the level of this financial dependence as referring suppliers’ economic dependence on the customer, an objective, observable factor. The second component of referring supplier's dependence is the risk of forfeiting future sales (i.e., losing an opportunity to “expand”). Dependence is particularly pertinent for referring suppliers that offer a wide variety of products and solutions and hope to expand sales through cross-selling opportunities. Thus, the referring supplier's product breadth (how broad or narrow the supplier's offerings are) is also an observable, objective antecedent to referring supplier's dependence on the customer (Figure 4).
Theoretical Development: Characteristics of Best Possible Versus Obligatory Horizontal Referrals
Brandt (1988) provides a useful distinction between service attributes, noting that “some … address the minimal requirements and expectations of customers, while others go a step further to add value to the service experience” (p. 35). The literature has offered various terms to characterize these distinctions: must-have versus attractive (Kano et al. 1984), dissatisfier versus satisfier (Cadotte and Turgeon 1988), or minimum requirement versus value-enhancing (Brandt 1988). All make the same distinction. Some characteristics are must-haves: a solution might be deemed unacceptable when it does not reach a minimum threshold but does not increase satisfaction as long as it is present. Other characteristics are attractive but unexpected: the customer is positively surprised by their presence, but their absence has limited consequences.
This distinction has consequences for horizontal referrals: a referring supplier could consider that giving a horizontal referral that meets minimum (i.e., must-have) requirements but does not do well on less essential (i.e., attractive) dimensions fulfills its obligations to the customer. In other words, both obligatory and best possible referrals should meet must-have requirements, but only the latter may exhibit attractive characteristics.
Must-have characteristics
In a B2B context, a referring supplier is expected to give a horizontal referral that achieves satisfactory performance (e.g., Weber, Current, and Benton 1991; Wilson, Weiss, and John 1990). Failing to do so would hurt the referring supplier–customer relationship. Thus, we consider (good) performance to be a referral's must-have characteristic. Furthermore, customers rely on existing suppliers for horizontal referrals because those suppliers have deep knowledge of the customer's specific needs. Therefore, we list the ability to meet customer-specific needs as another must-have characteristic (Figure 4).
Attractive characteristics
In complex B2B markets, customers rely on horizontal referrals to identify firms that offer a suitable solution to meet their specific needs. If a recommended provider offers favorable economic terms, the customer will consider that attribute attractive, if unexpected. The ability to satisfy the customer's future potential needs is another attractive supplier characteristic. Wathne, Biong, and Heide (2001) show that a supplier's product breadth influences the supplier's suitability for the customer, too. Selecting a supplier with high product breadth implies lower search costs in the future for the customer, superior performances due to product bundling, and more efficient buying processes (Kamakura et al. 2003; Tuli, Kohli, and Bharadwaj 2007). Thus, the recommended supplier's ability to meet the customer's future needs is an attractive characteristic that might extensively benefit the customer but will not hurt the customer if absent.
We list the constructs and their operationalization in Table 3. In solution-based selling, it is difficult to discern performance and economics in absolute terms; such indicators must be evaluated relative to customer experience. Therefore, we describe the horizontal referral's performance and economic terms as better or worse than those of the referring supplier.
Definitions of Constructs and Manipulations (Study 2).
Relative to referring supplier's performance and economic terms.
Hypotheses
We expect a referring supplier to make a recommendation that meets its customer's minimum standards (i.e., must-have characteristics). Failing to do so would hurt its relationship with the customer and jeopardize its current and future revenues, regardless of mitigating factors such as trust or dependence. Therefore, we hypothesize:
Thus, on average, a referring supplier will choose a horizontal referral that delivers the must-have characteristics. The difference between whether the horizontal referral is best possible or obligatory comes from the referring supplier's choice of the horizontal referral's attractive characteristics.
We next hypothesize how the antecedents of trust (relationship length and interaction frequency) and dependence (economic dependence and product breadth) influence a referring supplier's selection of attractive characteristics. We have identified several factors that favor trust, such as relationship length and interaction frequency; trust, in turn, increases the likelihood that the referring supplier will act as a friend and in the customer's best interests (i.e., make a best possible horizontal referral). Since the customer's best interest is to obtain the best recommendation possible, we hypothesize:
Factors that contribute to a referring supplier's dependence (economic dependence, product breadth) will increase the likelihood that the referring supplier will take its own interests into account and not only the customers’ interests (as would a friend). Because it is in the referring supplier's interest to maintain a good relationship with its customer, the supplier is unlikely to make a substandard recommendation. Instead, the supplier must engage in impression management (e.g., Wuyts, Verhoef, and Prins 2009) and make a referral that is at least minimally acceptable (as captured in H4).
However, we expect a supplier that acts as a businessperson to forgo the attractive characteristics: a horizontal referral that offers better economic terms can pose a threat to the supplier's future revenue, as Cannon and Homburg (2001, p. 34) emphasize: “a supplier that enhances customer value by lowering customer costs will increase its ‘share of the customer’ at the expense of suppliers that do not provide such benefit.” Similarly, referrals that offer a broad range of solutions are more likely to reduce the likelihood of the referring supplier's cross-selling opportunities. As Palmatier (2008) notes: “Sellers invest in building relationships with customers because of their expectation that these efforts will increase customers’ contributions to the seller's sales and profits” (p. 77). Wathne, Biong, and Heide (2001) show that customers are more likely to switch to suppliers that offer high product breadth. It is not in the interest of the referring supplier to give a horizontal referral that offers favorable economic terms and extensive product breadth. Since, unlike with must-have characteristics, skimping on attractive characteristics should not affect the supplier’s relationship with the customer, we hypothesize:
Figure 4 summarizes our hypotheses.
Research Design
We sought a research methodology that could directly infer a referring supplier's preferences for horizontal referrals while providing a realistic context for respondents. Conjoint designs are well suited for such assessments (e.g., Wathne, Biong, and Heide 2001). We used the same empirical context as in Study 1: Respondents are invited to imagine themselves as a key account manager at a marketing analytics firm. One of their customers, in the health care industry, asks for a horizontal referral in a related domain (data visualization). (See Web Appendix D for details.)
For each respondent, we manipulate key aspects of the relationship between the referring supplier and its customer (between-subjects manipulation) and, given the specific context described, measure which providers the referring supplier will recommend to the customer (within-subject measurement). We manipulate four referring supplier characteristics: (1) supplier–customer relationship length, (2) supplier–customer interaction frequency, (3) the referring supplier's economic dependence on the customer, and (4) the referring supplier's product breadth. Each dimension can be either high or low (see Table 3) for a total of 24 possible combinations. We run a fractional factorial, 100% efficient design and obtain eight different customer relationship scenarios.
For the horizontal referral characteristics, we again have four factors: the horizontal referral's (1) ability to meet customer-specific needs, (2) performance, (3) economic terms, and (4) product breadth. As each dimension is manipulated to be either high or low (see Table 3), we have 24 possible profiles, and we run a fractional factorial design to generate eight different supplier descriptions. Each respondent is assigned to one of the eight customer relationship scenarios (between-subjects manipulation) and is asked to choose one referral from eight possible suppliers (Figure 5 describes the research design and stimuli order. Web Appendix D provides further details.) As choosing one referral from eight could be onerous for respondents, we split the eight profiles into two blocks of four profiles each 3 and asked respondents to choose among horizontal referral profiles 1–4 and to choose among profiles 5–8. Finally, we asked respondents to choose between their two previous top selections, for a total of three choice tasks per respondent. This stepwise procedure helped minimize response burden and reduced choice-task complexity. The dependent variable is the referring suppliers’ choice of horizontal referral (stimuli in Web Appendix D).
Research Design for Study 2.
Considering the complexity of the conjoint task (highly detailed, scenario-based) and the time availability of professional respondents, we designed each questionnaire to include only three choice tasks (Figure 5). Therefore, we needed a large number of respondents, which we recruited from two sources: Dynata Inc.'s B2B manager panels (113 respondents) and Prolific Academic (250 respondents), for a sample size of 363 respondents.
Model
We estimated the probability of respondent i recommending supplier j by
To estimate βi, we sought a conjoint estimation method that met the following criteria: (1) minimize data collection efforts and the burden imposed on respondents and (2) focus on the relationship between the referring supplier's characteristics and choice of horizontal referral, rather than on individual-preference heterogeneity. Componential segmentation (Green and DeSarbo 1979) meets these two criteria. It explicitly incorporates respondents’ characteristics in the utility function by reexpressing individual respondents’ preference partworths (βi) as linear combinations of descriptor variables. Whereas classic conjoint models compute all vectors βi individually, we reexpress
We calibrated the model on the full sample of 363 respondents. Each respondent viewed eight profiles (horizontal referrals) for a total of 2,904 profiles. Respondents selected their top choices from profiles 1 to 4, from profiles 5 to 8, and then between their two previous top choices. The data set, therefore, includes the results of 363 × 3 = 1,089 choice tasks. We estimated the preference score of each of the 2,904 profiles (Equation 2) to predict the likelihood of the 1,089 observed choices (Equation 1) and found a matrix Ψ that maximized the log-likelihood, using a quasi-Newton search.
Results
We report the parameter estimates in Table 4. As the intercepts (partworths) show, referring suppliers are more likely to recommend competitors that meet customer-specific needs (2.19; p < .01), provide adequate performance (1.47; p < .01), offer better economic terms (.34; p < .01), and have high product breadth (.31; p < .01). On average, it appears that customers are justified in seeking horizontal referrals from their suppliers since the preference partworths' intercepts are positive on all dimensions (i.e., all four desirable characteristics are more likely to be selected than not). If the referring supplier has to trade off between must-have characteristics (meet customer-specific needs and performance) and attractive characteristics (economic terms and product breadth) and offers a less-than-perfect horizontal referral, it favors the must-haves. All two-by-two comparisons among the intercepts (e.g., 2.19 > .34) are statistically significant at p < .01, providing support for H4.
Parameter Estimates (Study 2).
Regarding the influence of trust antecedents (relationship length and interaction frequency) on attractive characteristics (economic terms and product breadth), all four parameters have the expected (positive) sign, and three achieve significance (.38, .17, .09; all p < .01). As predicted in H5, trusted suppliers try to provide the best possible horizontal referral, even on attractive characteristics.
This finding contrasts with the influence of dependence antecedents (referring supplier's economic dependence and product breadth) on attractive characteristics (horizontal referral's economic terms and product breadth). All four parameters have the expected (negative) sign, and three achieve significance (−.33, −.17, −.10; all p < .01). If, given an option, dependent suppliers will only offer obligatory referrals, selecting providers that are more likely to be expensive and offer a limited portfolio of solutions, and limiting future competitive threats. Thus, H6 is supported.
We also estimated the effects of the antecedents of trust and dependence on must-have characteristics (meet customer-specific needs and performance). Although we did not propose any specific hypotheses, the results tell an interesting story: Considering the influence of trust antecedents (relationship length and interaction frequency) on must-have characteristics (meet customer-specific needs and performance), the results are underwhelming. Three of the four parameters have a positive sign, and only two achieve significance (.05, p < .05; .11, p < .01). Thus, while referring suppliers systematically offer horizontal referrals that meet must-have criteria (see H4), highly trusted referring suppliers barely diverge from the mean for must-have characteristics.
The upper-right cells of Table 4 report the impact of dependence antecedents (referring supplier's economic dependence on customer and product breadth) on must-have attributes (meet customer-specific needs and performance). The referring supplier's economic dependence has a strong, negative impact even on must-have attributes (−.13, −.43; all p < .01). A best possible horizontal referral poses a threat to a referring supplier, especially if the latter is economically dependent. Consequently, it appears that the referring supplier is willing to risk jeopardizing its relationship with its customer to mitigate such a threat. In contrast, if the referring supplier offers an extensive product breadth, the threat concerns future cross-selling opportunities that may or may not materialize. Thus, the referring supplier's product breadth does not have a clear-cut influence on must-have attributes.
Simulations
We use the preceding results and run a simple simulation wherein only a handful of suppliers are considered by the referring supplier, either because the number of suppliers in the market is limited or because the referring supplier does not know them all (for a different setting, see Web Appendix E). We describe the profile of these three hypothetical referrals in Table 5.
Suppliers in the Referring Supplier's Hypothetical Choice Set (Simulations).
Supplier C is the stereotypical best possible referral. Its likelihood of being recommended will range between 50.4% and 78.1% across various scenarios. Supplier A would be a terrible referral and has a very small chance of being recommended. Finally, the referring supplier has a likelihood between 21.5% and 47.6% to refer Supplier B, a stereotypical obligatory referral. In particular, and in the context of this illustrative choice set, a referring supplier low on trust antecedents (relationship length, frequency interactions) but high on dependence antecedents (product breadth, economic dependence) would be twice as likely to make a suboptimal, obligatory referral than a referring supplier high on trust and low on dependence antecedents.
Study 3: Effect of Horizontal Referrals on Referring Supplier–Customer Relationships
Studies 1 and 2 took the perspective of the referring supplier to ascertain how the referring supplier will act when giving a horizontal referral and to determine the reasons for the choice of referral. In our survey, we found that customers largely expect their referring suppliers to “look out for them [the customer]” when giving a horizontal referral. Given that our empirical studies indicate that referring suppliers do not always do so, a key question is what happens to the customer–supplier relationship when a referring supplier makes a less-than-optimal obligatory referral, and that behavior is detected by the customer. We address this question in Study 3.
Theoretical Development
Research on suppliers' extra-role behavior indicates that depending on the state of the customer–supplier relationship (such as the positive intensity of the relationship), customers have different expectations regarding whether the supplier will go beyond its obligations in the relationship or not (e.g., Mangus et al. 2017). These expectations will likely affect how the customer responds to the knowledge that the supplier acted to fulfill an obligation or went beyond the obligations in the relationship.
Therefore, in Study 3, we study the effect of the supplier–customer relationship (described in terms of mutual trust and the supplier's dependence on its customer, as in Study 1) on customers’ perceptions of the interest alignment in the customer–referring supplier relationship (we refer to this construct as customers’ perceived interest alignment with the supplier). Since Study 3 focuses on the customer's point of view, we also include the customer's dependence on the supplier as a relevant factor. We then study how receiving a best possible referral affects the customer's intent to continue its relationship with its supplier and the moderating role played by perceived interest alignment. We describe the theoretical model in Figure 6.
Conceptual Model for Study 3.
Hypotheses
Consequences of best possible referral
Research suggests that a firm's extra-role behavior prompts gratitude in its partner, leading to positive downstream outcomes, such as the customer’s increased future commitment toward the partner (Mangus et al. 2017). In the context of salesperson–customer interactions, research has also found that salespersons’ extra-role behavior leads to increased customer loyalty (Bock, Folse, and Black 2016). Thus, we expect that a referring supplier who gives a best possible referral (exhibiting extra-role behavior) will increase the customer's intent to continue its relationship with the supplier.
Consequences of perceived interest alignment
If a customer perceives that the referring supplier's interest is aligned with its own, then the customer expects that the referring supplier's behavior can be reliably predicted, generating confidence in the referring supplier (e.g., Scheer 2012). With increased confidence in the supplier, the customer's intent to keep working with the referring supplier should increase as well.
An interesting facet of perceived interest alignment in the context of horizontal referrals, however, is the potential role it plays in moderating how a supplier's best possible referral will impact the customer's intent to continue its relationship with the supplier.
Kim, Hibbard, and Swain (2011) show that customers expect suppliers to act beyond obligations (that is, display extra-role behaviors). The higher the perceived interest alignment between the customer and the supplier, the more likely the customer is to expect the referring supplier to work toward maintaining the relationship and go beyond obligations and make a best possible referral. Hence, we expect interest alignment to negatively moderate the impact of a best possible referral.
While counterintuitive, this prediction becomes more straightforward when framed in the context of interest misalignment. If a supplier makes a best possible referral, it should positively impact the customer's desire to keep the collaboration alive (H7). If the customer perceives that the supplier's interests are misaligned with its own, however, the fact that the supplier decides to act beyond its obligations anyway should come as a positive surprise and hence should increase the customer's intent to continue the relationship. Therefore, we hypothesize:
Research Design
We test hypotheses H7 and H8 with a between-subjects experiment. We manipulate mutual trust (high vs. low), the referring supplier's dependence on the customer (high vs. low), the customer's dependence on the supplier (high vs. low), and the horizontal referral given by the referring supplier (best possible vs. obligatory). We manipulate the first three factors experimentally, which provides the state of the referring supplier–customer relationship to the respondent. Next, we measure the customer's perceptions of the referring supplier's interest alignment. The respondent then sees the referral made by the supplier (best possible vs. obligatory), following which we measure the customer's intent to continue its relationship with the supplier. We use the same empirical context as used in Studies 1 and 2 and adapt it to the customer's perspective.
We manipulate mutual trust (high/low) and the referring supplier's dependence on the customer (high/low) as we did in Study 1. We manipulate the customer's dependence (high/low) on its supplier as follows: [Customer] is (NOT) dependent on its supplier, [referring supplier name]. If this relationship were to end, [customer] would (not) experience significant problems in replacing [referring supplier] as the supplier for its back-end solutions. Best possible referral: [Referring supplier] recommended [horizontal referral firm name], a firm that provides GUI solutions. During the evaluation process you find that out of the possible firms, [referring supplier] recommended a firm that was best for you, [customer], not just an acceptable option. Even though [horizontal referral] also provides back-end analytics solutions and could BE a future competitive threat for [referring supplier] for your share of wallet. Obligatory referral: [Referring supplier] recommended [horizontal referral firm name], a firm that provides GUI solutions. During the evaluation process you find that out of the possible firms, [referring supplier] recommended a firm that was an acceptable option for you, [customer], but not the best. [Horizontal referral] does not provide back-end analytics solutions, and therefore, could NOT be a future competitive threat for [referring supplier] for your share of wallet.
Finally, we measure the customer’s intent to continue the relationship with the supplier with the following two items (Jap 1999) on a seven-point scale (1 = “highly disagree,” and 7 = “highly agree”): (1) “We expect to continue working with them on a long-term basis,” and (2) “Our relationship with them will last long into the future.”
Our sample comprises 818 respondents with at least two years of work experience in B2B industries on Prolific Academic (details in Web Appendix F).
Model
We estimate a system of two equations (see Figure 6). The dependent variable of the first equation is the customer's perceived interest alignment with its supplier. The independent variables are trust, the supplier's dependence on the customer, the customer's dependence on the supplier, and all their respective two- and three-way interactions (see Table 6 and Figure 7).
Results (Study 3).
Parameter Estimates (Study 3).
*p < .1. **p < .05. ***p < .01 (one-tailed tests).
Notes: N = 818. See Figure 7 for illustration.
The dependent variable of the second equation is the customer's intent to continue with the referring supplier. The predictors are the referral given (best possible = 1, obligatory = 0), customers’ perceived interest alignment (previously described), and their two-way interaction. We also account for the direct influence of trust, the supplier's dependence on the customer, the customer's dependence on the supplier, and all their respective two- and three-way interactions (see Table 6 and Figure 7). As a robustness test, we also accounted for the interactions of the three relational variables with the referral given, and our results hold.
We estimate the two equations simultaneously. We use bootstrapping to estimate the indirect and total effects of trust, the supplier's dependence, and the customer's dependence on customer's intent to continue with the referring supplier, mediated by the customer's perceived interest alignment (for estimation approach, see Hayes [2015]).
Results
We report the direct effects (top two panels) and the indirect and total effects (bottom panel) in Table 6. We find that receiving a best possible referral (vs. an obligatory one) increases the customer's intent to continue its relationship with the referring supplier (1.73; p < .01), confirming H7, and so does customers’ perceived interest alignment (.34; p < .01), as expected.
As hypothesized, interest alignment moderates the impact of a best possible referral on the customer's intent to continue its relationship in the expected, negative direction (−.23; p = .01), supporting H8. In Figure 7, Panel B, we plot the total effect of the referral received (best possible in dark, obligatory in dashed gray) on intent to continue the relationship with the supplier as a function of interest alignment. When interest alignment is low, making a best possible referral has a positive effect on the customer's intent to continue the relationship. When interest alignment is high, the customer expects to receive a best possible referral, hence its positive effect is attenuated and only impacts the customer's intent to continue the relationship minimally.
Several insights emerge regarding the aspects of the customer–supplier relationship that drive the customer's perceived interest alignment with its supplier. First, as expected, the effect of mutual trust is positive and significant (.72; p < .01). Second, the supplier's dependence on its customer is positive but barely significant (.23; p < .10), whereas the customer's dependence on its supplier is negative and strongly significant (−.50; p < .01). The interaction of these two terms is positive and strongly significant as well (.93; p = .01). We plot the total effects Figure 7, Panel A. If the customer's dependence on its supplier is low (dashed gray line), the supplier's dependence only minimally impacts the customer's perceived interest alignment. If the customer's dependence on its supplier is high (dark line), however, the total effect depends on whether that dependence is reciprocal or not. If both the customer and the supplier depend on one another, the customer perceives a higher interest alignment. If the customer is highly dependent on its supplier but not the other way around, the customer may rightfully fear that the supplier could take advantage of it, hence perceiving a lower interest alignment.
Discussion
Theoretical Implications
Research on referrals
Our work contributes to research on referrals in B2B markets (e.g., Hada, Grewal, and Lilien 2014; Kumar, Petersen, and Leone 2013) and the important role of referrals in reducing customers’ uncertainty regarding their future suppliers. In the small stream of research on B2B referrals and the much larger stream on B2C referrals, the focus has mostly been on customer-to-customer referrals. Although we have shown that horizontal referrals in B2B markets are commonplace, we have also demonstrated that the quality of those referrals—that is, whether referring suppliers will give an obligatory or best possible referral—is uncertain. Relying on role theory (e.g., Heide and Wathne 2006), we show that the uncertainty in the horizontal referral's quality can be reduced by better understanding the role a referring supplier adopts when making a horizontal referral (i.e., friend vs. businessperson).
Research in economics
The economics literature on horizontal referrals largely considers only the economic trade-offs facing the referring supplier, and concludes that the referring supplier will dominantly act in its own interest (e.g., Arbatskaya and Konishi 2012; Park 2005). By relying on the socioeconomic paradigm, we show that when there is high trust in the relationship, a referring supplier is likely to act in the customer's interest and give a best possible horizontal referral. However, we also find that the referring supplier's dependence is more diagnostic than trust between the two firms in predicting referring supplier behavior. Therefore, our research informs the economics literature about the contingencies in predicting suppliers’ behavior. And we show that the contingencies arise from the referring supplier's own characteristics (e.g., product breadth) and the interfirm relationship (e.g., relationship length), which are readily observable.
Research on interfirm relationships
We contribute to research on supplier–customer relationships where interests are misaligned, which is the case for horizontal referrals (for other examples of interest misalignments, see, e.g., Scheer [2012]). Although extant research in prosocial theory has considered how suppliers go above and beyond, that work generally neglects specific situations and responses. We specifically look at a situation in which interests between the supplier and customer could be misaligned and how these misaligned interests affect whether the supplier goes above and beyond the relationship or not.
We also contribute to the research on the dark and bright sides of close/trusting relationships. Researchers argue that partners might act opportunistically in close relationships because opportunistic behavior can go undetected in these relationships (e.g., Anderson and Jap 2005). We consider a substantive situation (horizontal referrals) in which the quality of the given referral is not evident, allowing us to consider how suppliers will behave in a situation that is possibly devoid of the chance of detection. The consideration of the chance of detection is the key difference between the “dark side of relationships” perspective and the role theory perspective. In research studying the dark side of relationships (e.g., Noordhoff et al. 2011), suppliers behave opportunistically or not depending on the likelihood of detection; according to role theory, suppliers behave opportunistically or not because of how they identify with the role, and not because of the chance of detection. By integrating social exchange theory with role theory, we provide a theoretical framework to study situations of misaligned interests in relationships.
We find that the referring suppliers’ dependence on the customer increases the chance that the referring supplier will act as a businessperson, increasing the likelihood of an obligatory referral. This finding conflicts with research that has found positive effects of dependence on a supplier's commitment to the customer. For example, Scheer, Miao, and Garrett (2010) show that higher economic dependence increases a supplier's long-term orientation toward the relationship. To maintain the benefits obtained from the relationship, dependent parties are expected to adhere to rules and procedures and refrain from opportunistic behaviors (Brown et al. 2016). Our research shows that these findings might not hold in situations of misaligned interests in which the supplier's opportunistic behavior could go undetected.
Managerial Implications
Relying on suppliers’ market knowledge in complex B2B markets can dramatically reduce customers’ search costs and uncertainty. In our survey, customers report that the horizontal referrals they received were, on average, of high quality (M = 5.40, SD = 1.40).
Our first two studies confirm this finding. In Study 1, 61% of the respondents playing the role of referring suppliers stated that they would make a best possible referral; however, 39% reported that they would make only an obligatory referral. Because customers do not know the target market well, and even obligatory referrals generally meet must-have requirements, obligatory referrals are difficult to spot. Reassuringly for customers, however, we show that their occurrence is predictable. As we show in Study 2, the chance of receiving an obligatory referral is largely influenced by the (observable) characteristics of the referring supplier: trusted suppliers should be sought after for referrals, but more importantly, dependent suppliers should be avoided.
The latter conclusion may seem surprising. Interviewed customers reported that, on average, the referring suppliers they relied on were economically dependent on them (M = 4.72, SD = 1.71). We demonstrate that this choice of referring supplier is misguided. In Study 2, not only do we show that economically dependent referring suppliers are the most likely to make an obligatory referral, but they also have a higher chance of making a referral that does not even meet must-have requirements. Furthermore, the purpose of a horizontal referral is to rely on the referring supplier's market knowledge to reduce search costs and uncertainty. Although we did not ask specifically in our exploratory survey, customers would likely heavily rely on referring suppliers with large product breadth as these suppliers’ knowledge would span the industry more than niche suppliers’ knowledge. We find that this strategy could backfire. A referring supplier with a large product breadth is more likely to offer an obligatory referral to mitigate the risks of losing cross-selling opportunities in the future.
Thus, while our research confirms that most horizontal referrals tend to be of high quality, it also provides guidance about the characteristics of the referring suppliers that customers should rely on to increase their chances of obtaining a best possible referral. Study 3 sheds light on the customer's point of view and demonstrates that, if detected, an obligatory referral can hurt the supplier–customer relationship. This effect is negatively moderated by the customer's perceived interest alignment. But the data from Study 3 also reveal an interesting facet of customers’ expectations and how the supplier's dependence alters them. When asked which suppliers are most likely to recommend the best supplier possible, customers unequivocally identify trusted (.67; p < .01) and dependent (1.69; p < .01) suppliers. We demonstrate that this belief is misguided, as dependent suppliers are far less likely to make best possible referrals than nondependent suppliers. However, when the same customers are asked when suppliers are likely to recommend a competitor, the respondents’ anticipated impact of the supplier's dependence reverses (−.92; p < .01), whereas the role of trust remains positive and significant (.31; p < .01).
Thus customers do not seem to naturally think of how potentially damaging recommending a competitor might be for the referring supplier. Instead, customers focus on their relationship with their supplier, their perceived alignment of interests, and the expectation that their supplier will try to please them by going out of their way to recommend a best possible referral. But when we nudge customers into thinking about the competitive environment in which their suppliers evolve and the potential risk posed by referring a competitor (even if ideal for the customer), customers revise their expectations. Interestingly, when referring suppliers are highly dependent on their customers, customers understand their suppliers’ necessity to protect themselves from future competitive threats and tend toward maintaining the relationship.
On the one hand, customers usually ask for referrals for products or services that current suppliers do not offer. Consequently, customers perceive as irrelevant the competitive threat posed by a best possible referral, and they may incorrectly choose the supplier they ask for a horizontal referral. On the other hand, suppliers that focus on a “land and expand” strategy better gauge the current and future dangers of referring a potential competitor. This difference in perspective leads to misaligned recommendations. The good news is that this difference in perspective, when pointed out explicitly to customers, becomes obvious to them as well, suggesting that customers only need to train their intuition to select the right suppliers for horizontal referrals and avoid making costly mistakes.
Limitations and Further Research
As with all research, ours has limitations. There are numerous possible antecedents to referring suppliers’ role adoption and horizontal referral choice that we did not consider. We relied on academic research on interfirm relationships (e.g., Palmatier et al. 2006; Ulaga and Eggert 2006) and our interviews to build our conceptual framework and considered a mix of perceptual and observable characteristics. It would be fruitful for researchers to consider characteristics beyond those we studied here. Some factors that could change the referring supplier's recommendation are (1) the extent to which there is customer lock-in for the suppliers’ current offering, (2) the extent of possible competitive overlap between the referral and the referring supplier, and (3) other nuanced parameters that vary from one purchasing situation to another.
We based our research conclusions on both survey and laboratory-based data collection procedures in a cross-sectional format, an approach that always has questionable external validity. Tracking real horizontal referrals over time would be a useful future research avenue.
Conclusion
We began by noting a common phenomenon: More than 70% of B2B customers in our sample report ask their suppliers for referrals in new purchase situations, often placing referring suppliers in a quandary. Referring suppliers want to maintain their positive relationship with their customers by giving horizontal referrals that are best for their customers but, at the same time, want to protect their revenue by giving a horizontal referral that minimizes the threat to their business. Our findings, in contrast with what the economics literature predicts, indicate that referring suppliers do try to act in the client's interest, often to the detriment of their own, at least in the short term. However, many B2B customers are unknowingly forcing some conflicted suppliers (the ones dependent on them now or ones that wish to cross-sell to them in the future) to choose between satisfying their own and their customer's best interests. Our research provides a framework to study this problem (and other situations of misaligned interests) and provides guidelines for customers to follow concerning which suppliers they should rely on in such situations.
Supplemental Material
sj-pdf-1-mrj-10.1177_00222437231175415 - Supplemental material for Horizontal Referrals in B2B Markets
Supplemental material, sj-pdf-1-mrj-10.1177_00222437231175415 for Horizontal Referrals in B2B Markets by Mahima Hada, Arnaud De Bruyn and Gary L. Lilien in Journal of Marketing Research
Footnotes
Acknowledgments
The authors thank Rajdeep Grewal and Rick McFarland for their valuable feedback on this article.
Coeditor
Vikas Mittal
Associate Editor
Satish Jayachandran
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research received financial support from the Institute for the Study of Business Markets (ISBM).
Notes
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
