Master’s Thesis
RELATIONSHIP-SPECIFIC INVESTMENTS AND
SUPPLIER BEHAVIOUR: THE ROLE OF SUPPLY CHAIN
INTELLIGENCE
By:
Lars Mulder
Student number: s2376539
E-mail:
l.mulder.9@student.rug.nl
University of Groningen
Faculty of Economics and Business
MSc. Supply Chain Management
January 29, 2018
First supervisor:
Dr. J. Veldman
University of Groningen, Faculty of Economics and Business
Second assessor:
Dr. ir. N.J. Pulles
Abstract
The effectiveness of relationship-specific investments (RSI) in business-to-business relationships depends on the behaviour of the exchange partner. Therefore, RSI are not always successful. This research investigates how RSI can influence the behaviour of the supplier, focussing on the supplier’s preferential resource allocation and poaching behaviour. Furthermore, this study analyses the role of supplier satisfaction, and examines how supply chain intelligence can play a role in enhancing the effectiveness of buyers’ investments. This paper derives dyadic data from 94 buyer-supplier relationships in the automotive and cycling industry to test the theoretical implications using structural equation modelling. The key results are the negative effects of RSI on the supplier’s preferential resource allocation. Furthermore, this research demonstrates how supply chain intelligence can positively influence the competitive performance. The implications of these findings are that, managers of manufacturing firms should focus on supply chain intelligence to increase the satisfaction of their suppliers and their own competitive performance.
Keywords: Buyer-supplier relationship, relationship-specific investments, supply chain intelligence, preferential resource allocation, poaching.
Table of Contents
1. Introduction ... 4
2. Literature background ... 6
3. Theory and hypotheses ... 8
3.1 Preferred resource allocation ... 8
3.2 Preferred resource allocation and competitive performance ... 9
3.3 Supplier poaching behaviour... 9
3.4 Supplier poaching behaviour and competitive performance ... 10
3.5 Supplier satisfaction ... 11
3.6 Supply chain intelligence ... 13
4. Methodology ... 17
4.1 Sample and data collection ... 17
4.2 Measures ... 19
4.3 Data validity and common method bias ... 21
4.4 Data analysis ... 23
4.5 Control variables ... 24
5. Results ... 25
5.1 The structural model: H1-H4 ... 25
5.2 Mediating effect of supplier satisfaction ... 26
5.3 Multi-group analysis ... 28
5.4 Control variables ... 32
6. Discussion ... 33
6.1 Hypotheses testing ... 33
6.2 Supply chain intelligence ... 34
6.3 Years of supplying ... 35
6.4 Moderating effect ... 36
7. Limitations ... 37
8. Future research directions ... 39
10. Conclusion ... 41
11. References ... 43
12. Appendices ... 50
Appendix A: Survey – supplier ... 50
Appendix B: Survey – buyer ... 52
Appendix C: Multi-group analysis supply chain intelligence – results ... 53
1. Introduction
In highly competitive markets with supplier scarcity, manufacturing firms aim to become the favourite customer of their supplier. Being a favourite or preferred customer comes with advantages that are not as readily available to other customers (Trent and Zacharia, 2012). Accordingly, suppliers provide their preferred customer with the privileged resource allocation (Schiele, Calvi and Gibbert, 2012). To become a preferred customer, firms need to focus on improving supplier satisfaction regarding their mutual relationship (Pulles, Veldman and Schiele, 2016). In order to enhance the satisfaction of the supplier, firms engage in relationship-specific investments (RSI). This, in turn, also leads to an increase in their own performance (Nyaga, Whipple and Lynch, 2010). However, RSI also enhances the likelihood of opportunism, resulting in some harmful consequences for the investing firm (Villena, Revilla and Choi, 2011; Blonska, Storey, Rozemeijer, Wetzels and de Ruyter, 2013). This divergence in outcomes highlights the importance of a better understanding of both the downsides and upsides of RSI. To this end, this research focuses on the roles of supplier satisfaction and supply chain intelligence.
RSI differ from other investments, because these investments only have value in a particular relationship, and lose their value when one of the parties decides to discontinue the collaboration (Luo, Liu, Yang, Maksimov and Hou, 2015). For this reason, investing firms aim to maintain long-term relationships whereby RSI sustain their value. A long-term relationship reduces uncertainty and enhances collaboration, therefore creating benefits for both partners in the relationship (Blonska et al., 2013; Khoja, Adams and Kauffman, 2015). However, due to the increased level of trust in long-term relationships, most companies reduce their monitoring activities, which enhances the risk of suppliers taking advantage of manufacturers’ investments (Villena et al., 2011). Wang, Li, Ross and Craighead (2013) describe this opportunistic behaviour as one of the greatest threats in a buyer-supplier relationship.
Most literature about communication and information sharing in the buyer-supplier relationship focuses on the quantity of information, while few studies emphasise the actual quality or correctness of shared information. This is acknowledged by Blonska et al. (2013), who call for research on how any misalignment in perception and knowledge between buyer and supplier influences the performance of both parties. Meanwhile, Modi and Mabert (2007) request a better understanding of how investing firms can control the behaviour of their suppliers. This research will contribute to the gaps as described by Liu et al. (2009), Blonska et al. (2013) and Modi and Mabert (2007), by investigating the role of accurate knowledge, in this study defined as supply chain intelligence, as a relational governance mechanism.
By possessing accurate knowledge, firms are able to reduce miscommunication (Byron, 2008), prevent conflicts (Hinds and Mortensen, 2005) and increase mutual understanding of the relationship (Ballantyne, 2004), which increases the effectiveness of communication. According to Nyaga et al. (2010), firms with improved information sharing capabilities and accurate knowledge are more able to influence suppliers’ perception of the relationship and predict their behaviour. Therefore, this study investigates how supply chain intelligence supports firms in reducing opportunistic behaviour. Furthermore, the role of supply chain intelligence is investigated where firms with RSI want to increase supplier satisfaction or obtain preferential resource allocation. Finally, this study shows how differences in intelligence can influence the firm’s competitive performance. By doing so, this study examines how supply chain intelligence acts as a relational mechanism to influence the behaviour of an exchange partner and improve the firm performance.
2. Literature background
In a buyer-supplier relationship, firms have the opportunity to invest in the development of their exchange partner (Sillanpää, Shahzad and Sillanpää, 2014). Manufacturing firms make investments with the aim to continue and improve the relationship with their supplier (Crawford, 1990). Strong exchange relationships are characterized by an enhanced level of communication, information sharing and trust (Yang, 2013). The “rich communication” in these interactions helps both parties to create a shared vision and long-standing relationship (Blonska et al., 2013). A long-term and stable relationship between buyer and supplier is crucial, and acts as a source of competitive advantages (Sillanpää et al., 2014; Kohtamäki, Vesalainen, Henneberg, Naudé and Ventresca, 2012).
Crawford (1990) defines investments made to improve a specific relationship with an exchange partner as (RSI). These investments are specific to a particular relationship and are characterized by their difficulty or impossibility to apply to other relationships (Rokkan, Heide and Wathne, 2013). The investments can be physical assets, such as unique machinery to produce new customized products, or human assets, such as task-specific training (Fang et al., 2008). With RSI, firms aim to create more value in the buyer-supplier relationship in order to enhance the performance of the relationship (Henry Xie, Sub and Kwon, 2010; Kothamäki et al., 2012). Firms that are able to fully exploit the buyer-supplier relationship can see an increase in their performance, which may lead to a competitive advantage over its competitors (Blonska et al., 2013).
One important and recently discussed competitive advantage for rival firms is the preferential resource allocation (Pulles et al., 2016a). Manufacturing firms with the preferred customer status are able to receive valuable resources with a higher responsiveness (Pulles et al., 2016a). Recent studies on the topic of the preferential resource allocation have focused on the roles of supplier satisfaction and customer attractiveness. This research aims to contribute to this topic by investigating whether the strategic decisions of manufacturing firms to make RSI create the possibility to attain better resource allocation (Pulles, Veldman, Schiele and Sierksma, 2014).
Hitt, 2004). This opportunistic behaviour may erode previously established advantages: accordingly, RSI do not always result in the preferred outcomes of the buying firm.
Although knowledge about the advantages and disadvantages of RSI is already rather extensive, recent literature has emphasized the lack of knowledge about governance mechanisms available to influence the effectiveness of RSI (Whipple, Wiedmer and Boyer, 2015). Traditionally, manufacturing companies possessed a strong market position in which they were able to use transactional mechanisms and bargaining power to influence their suppliers (Crook and Combs, 2007). However, due to the trend of supplier scarcity and increased dependency on suppliers (Rozemeijer, Quintens, Wetzels and Gelderman, 2012), manufacturing firms are now required to use more relational mechanisms to influence the behaviour of their suppliers (Johnston, Khalil, Jain and Cheng, 2012; Wang et al., 2013). This shift has created new opportunities for future research. For instance, Ahmed and Hendry (2012) state that research is required to investigate how relational mechanisms can encourage suppliers to participate in RSI. Modi and Mabert (2007) highlight that little is known about the use of relational mechanisms as a type of monitoring and safeguarding procedure to control the behaviour of the supplier. This paper explores whether supply chain intelligence can act as a relational mechanism by considering how intelligence relates to the preferred and non-preferred behaviour of the exchange partner.
3. Theory and hypotheses
3.1 Preferred resource allocation
Resource allocation is a selective process performed by the supplier, where competing customers may be treated differently (Mitsuhashi and Greve, 2009). The customer receiving preferential treatment in the allocation of resources by the supplier has the preferred customer status (Hüttinger, Schiele and Veldman, 2012). For this reason, the preferred customer accepts privileges that their competitors do not have, resulting in competitive benefits (Schiele, 2012). For instance, in the situation where the supplier is unable to serve all their customers due to capacity constraints, the preferred customer is served first (Hüttinger et al., 2012). Furthermore, suppliers allocate their resources to the relationships in which they expect the highest value or benefits (Griffith, Harvey and Lusch, 2006; Pulles et al., 2014). Therefore, buying firms that are able to enhance the value of the relationship with their supplier have higher chances of receiving preferential resource allocation (Pulles, Schiele, Veldman and Hüttinger, 2016).
Palmatier, Dant and Grewal (2007) discuss how RSI can improve the interaction between both parties, and as such enhance their relationship performance. For instance, investments in training and communication programmes can improve the interactions between both exchange partners (Palmatier et al., 2007). The interaction helps both parties to align their objectives and reduce miscommunication (Modi and Mabert, 2007). By doing so, investing firms are able to understand their suppliers’ expectations, evaluate their own performance and detect opportunities for improvements in the relationship (Blonska et al., 2013). Eventually, this gives exchanging firms more possibilities to improve the performance and value of their relationship (Whipple et al., 2015).
Kohtamäki et al. (2012) demonstrate a positive relationship between RSI and improved relationship performance due to increased operational performance. For instance, investments in the supplier’s planning system may improve the delivery accuracy, and investments in the supplier’s machinery may improve product quality. When firms are able to create higher performance in the relationship compared to competitor firms, the supplier probably assigns more value to the relationship (Kohtamäki et al., 2012; Whipple et al., 2015). Therefore, by carrying out RSI firms are able to increase the expected value from the relationship, due to increased interaction and operational performances. Subsequently, when firms are able to increase the value of the relationship with their supplier they have a higher chance of receiving preferential resource allocation (Pulles et al., 2016b). Hence:
Hypothesis 1: Relationship-specific investments positively influence the supplier’s
3.2 Preferred resource allocation and competitive performance
As a result of the preferred customer status, buyers are able to obtain the preferred resource allocation and higher responsiveness from their supplier (Hüttinger et al., 2012). “Resources are the tangible and intangible quantities that are available to the organization, which are used to efficiently and/or effectively produce an offering with value to a specific market segment” (Hunt and Davis, 2008:13). Firms with superior resources due to supplier resource allocation are therefore able to offer higher quality and dependability to their customers (Ray, Barney and Muhanna, 2004). Furthermore, the buying firm will possess better capabilities compared to its competitors, resulting in competitive advantages in its market position (Pulles et al., 2014). Finally, these competitive advantages can result in better financial and market performance (Li, Ragu-Nathan, Ragu-Nathan and Rao, 2006). For instance, by offering higher quality to customers, firms demand a higher price for their products, resulting in increased revenues. The organizational performance of a firm refers to how well it achieves both its market-oriented and financial goals (Li et al., 2006), which is affected by the resources it possesses. Hence:
Hypothesis 2: Preferential resource allocation positively influences the performance
of the buyer.
3.3 Supplier poaching behaviour
The shared information and knowledge resulting from social value (whether or not those refer directly to the RSI) may be abused by exchange partners for other purposes (Cai, Goh, de Souza and Li, 2013; Kang and Jindal, 2015). For instance, suppliers could use information about business practices and capabilities of the buyer, which are assets of value, in relationships with competitor firms (Cai et al., 2013). Handley and Benton (2012) define the use by an exchange partner of the information and knowledge gained through interactions for other purposes as poaching. Poaching is self-serving with guile and therefore can be viewed as a form of opportunistic behaviour (Handley and Benton, 2012). Hence:
Hypothesis 3: Relationship-specific investments positively influence the poaching
behaviour of the supplier.
3.4 Supplier poaching behaviour and competitive performance
As discussed in section 3.3, opportunistic suppliers may utilize information and knowledge obtained through the relationship with the buyer for other purposes (Handley and Benton, 2012). The information and knowledge that is exchanged outside the intended boundary of the relationship are called spillovers (Fallah and Ibrahim, 2004). Spillovers are particularly important when they contain confidential information, since competitive buyers often share the same suppliers (Cai et al., 2013). Leaking valuable information to competitors may erode the competitive advantages of the buyer (Alcacer and Chung, 2007; Blonska et al., 2013).
When a spillover of valuable information benefits a competitor, the competitiveness of the buying firm could be weakened (Frishammar, Ericsson and Patel, 2015). For instance, when firms are able to receive a high market share due to their knowledge-intensive resources, competitive advantages could be eroded when competitors obtain this knowledge (Ahmad, Bosua and Scheepers, 2014). Consequently, decreased competitiveness has negative consequences for the buyer’s market position and financial results (López-Gamero et al., 2010;
Frishammar et al., 2015). As stated previously, the performance of a firm refers to how well it achieves both its market-oriented and financial goals (Li et al., 2006). Hence:
Hypothesis 4: Poaching behaviour of the supplier will reduce the performance of the
buying firm.
FIGURE 1
The conceptual model of hypotheses 1 to 4
3.5 Supplier satisfaction
The supplier’s decision to continue a collaborative relationship with a buying firm largely depends on their degree of satisfaction (Benton and Maloni, 2005). The satisfaction of a supplier is “the feeling of equity with the relationship no matter what power imbalance exists” (Benton and Maloni, 2005:2). Satisfaction reflects the outcome of the relationship compared to previously established expectations (Schiele et al., 2012). A high degree of supplier satisfaction is achieved when the quality of the collaboration meets or exceeds the supplier’s expectation (Schiele et al., 2012). This quality of collaboration is often measured by factors of a more operational nature, such as delivery, order processes and quality improvements (Hüttinger et al., 2012; Kohtamäki et al., 2012). Operational performances are commonly expressed in financial factors, as well as time aspects, which directly affect the degree of satisfaction (Essig and Amann, 2009).
supplier. Moreover, the buyer’s RSI can enhance the supplier’s appreciation, which may help overcome future concerns about the relationship (Blonska et al., 2013).
Recent literature has focused on the topic of preferential resource allocation in order to reach a better understanding of the mechanisms influencing the choice of the supplier. For example, “suppliers often become highly selective and do not dedicate their resources equally to all of their customers” (Schiele et al., 2012:4). According to Trent and Zacharia (2012:14), “satisfied suppliers are more willing to provide preferential treatment to their preferred customer compared to less satisfied suppliers”. The supplier awards a buying firm with the preferred customer status if the supplier is more satisfied with the customer than with competing customers (Schiele et al., 2012). This statement is supported by the research of Pulles et al. (2016b), who found a strong relationship between supplier satisfaction and the preferential resource allocation. Satisfied suppliers show more commitment to the continuation of the relationship and are, therefore, more likely to respond to the investing firms with certain benefits (Pulles et al., 2016b). Hence:
Hypothesis 5a: Relationship-specific investments positively affect supplier
satisfaction, which in turn positively affect the preferential resource allocation.
When a firm makes RSI, the chance for the exchange partner to act opportunistically arises (Nyaga et al., 2010). According to Das and Rahman (2010), opportunistic behaviour can arise when one of the exchange firms has dissatisfied feelings towards the relationship. This dissatisfaction can occur due to miscommunication, conflicts or goal incompatibility. When this happens, a partner may feel less restriction against acting in a self-serving way and become disloyal (Das and Rahman, 2010). When satisfaction increases, firms perceive the economic exchanges with their customer as more valuable (Ping, 1993). Moreover, satisfied firms respond reasonably to relationship problems and oppose behaviour that is harmful to the relationship, such as opportunism (Ping, 1993; Dorsch, Swanson and Kelly, 1998).
The literature lacks insight into the direct relationship between supplier satisfaction and poaching behaviour. However, since poaching behaviour is a form of opportunism (Handley and Benton, 2012), it is likely that supplier satisfaction also affects poaching behaviour. Hence:
Hypothesis 5b: Relationship-specific investments positively affect supplier
satisfaction, which in turn negatively affects poaching behaviour.
The hypotheses H5a and H5b are visualized in Figure 2.
FIGURE 2
The conceptual model including hypotheses H5a and H5b
3.6 Supply chain intelligence
Jeeva (2008) highlights that only 11% of the manufacturing companies have information sharing capabilities with their direct suppliers. This result is striking as the sharing of information may help both exchange parties to align their objectives and expectations regarding the relationship (Modi and Mabert, 2007). In addition, it enables firms to create a foundation of shared data and knowledge, which helps them to better predict the behaviour of their suppliers and to build effective mechanisms for conflict resolution (Nyaga et al., 2010).
its accuracy, information obtained by a firm should be verified with its exchange partner in the supply chain (Carter, 2000; Sweeney and Webb, 2002). This paper introduces the definition of supply chain intelligence, which is the accurate information and knowledge that buyer and supplier possess about each other’s internal and external environments.
As discussed, RSI represent sunk assets which will lose value if the exchange firm ends the relationship or goes out of business (Rokkan et al., 2013). These risks, together with the difficulty of exactly predicting the future outcome of RSI, create uncertainty in the buyer-supplier relationship (Raman and Shahrur, 2008). Johnston et al. (2012) advocate that uncertainty in a relationship influences the feeling towards the relationship. Accordingly, they found that a reduction in uncertainty leads to an increase in the satisfaction regarding the buyer-supplier relationship. According to Das and Rahman (2010), dissatisfaction in a relationship occurs due to miscommunication, conflicts or goal incompatibility
Modi and Mabert (2007) state that accurate information and knowledge may reduce the uncertainty in a buyer-supplier relationship. Subsequently, this helps firms to create a better common understanding of the relationship, which causes fewer conflicts (Modi and Mabert, 2007). When companies obtain the right information, opportunities arise to collaboratively remove supply chain inefficiencies and improve the relationship (Hsu, Kannan, Tan and Keong Leong, 2008). Therefore, supply chain intelligence enables firms to reduce uncertainty, improve commitment and eventually increase the effectiveness of RSI in the buyer-supplier relationship. According to Essig and Amann (2009), these improvements positively influence the outcome of the buyer-supplier relationship and eventually increase the satisfaction of the supplier (Essig and Amann, 2009). Hence:
Hypothesis 6a: A high degree of supply chain intelligence positively influences the
relation between RSI and supplier satisfaction.
improvements by continuously evaluating the relationship. This in turn, should result in greater motivation to reciprocate investments (Cheung, Meyers and Mentzer, 2011).
To obtain supply chain intelligence, firms continuously assess their own information alongside the information of their partners (Sweeney and Webb, 2002). By doing so, firms can provide directions for improvement and reduce miscommunication and confusion in their relationship (Modi and Mabert, 2007; Hsu et al., 2008). According to Trent and Zacheria (2012), without continuously evaluating and possessing the right information, firms are unable to assess how well they are performing in the eyes of the supplier. Following the line of reasoning of Krause et al. (2000) and Cheung et al. (2011), supply chain intelligence can help firms to increase the supplier’s intrinsic motivation, which can eventually result in higher chances of suppliers embracing RSI. In addition, accurate information helps firms to detect the best improvement opportunities in the relationship (Blonska et al., 2013). Finally, firms can more readily meet the needs of a partner, for example by, “better understanding technological demands related to techniques, methods and product design applications of the partner firm”(Cheung et al., 2011:1067). To conclude, supply chain intelligence can help firms to increase the supplier motivation, and to find possibilities to increase the value in the relationship when making RSI. Firms that are able to enhance the value of their buyer-supplier relationship are more likely to receive the preferential resource allocation (Pulles et al., 2016b). Hence:
Hypothesis 6b: A high degree of supply chain intelligence positively influences the
relation between RSI and the preferential resource allocation.
To prevent RSI from losing value, firms aim to create a long-term relationship with the supplier (Luo et al., 2015). However, long-term relationships may enhance mutual trust and thus automatically reduce the effort made to monitor the other party (Ritter and Ellegaard, 2007; Cai et al., 2013). By doing so, firms save monitoring costs (Cai et al., 2013), but are less likely to detect cheating or self-seeking behaviour from their exchange partner (Villena et al., 2011). As discussed in section 3.3, opportunistic behaviour arises when the company is more motivated to achieve gains at the expense of the exchange partner rather than creating mutual benefits (Das and Rahman, 2010). As discussed, supply chain intelligence reduces miscommunication and friction and enhances the motivation of the supplier. Supplier motivation is an important indicator of a supplier acting opportunistically or exhibiting poaching behaviour (Blonska et al., 2013). Hence:
Hypothesis 6c: A high degree of supply chain intelligence negatively influences the
relation between RSI and suppliers’ poaching behaviour.
FIGURE 3
4. Methodology
4.1 Sample and data collection
The unit of analysis in this research is the buyer-supplier relationship. This study focuses on the impact of supply chain intelligence on the effectiveness of RSI. To identify this cause-and-effect relationship, confirmatory research is used (Karlsson, 2016). Survey-based research is preferable in this situation, because survey research is able to test whether the hypothesized relationships hold in different contexts. In addition, compared to other methods, survey-based research is less costly, gives a higher generalizability, and is preferred for sensitive topics due to anonymity (Nardi, 2015). The sample for this research was drawn from companies in the automotive and cycling industry, and the data was collected with the online survey tool Qualtrics.
Between September and October 2017, around 20 manufacturing companies in the northern part of the Netherlands were approached by telephone with the question of whether they were willing to participate in this research. Two large manufacturing firms agreed to participate and share a list of suppliers. Both companies selected those suppliers that were the most important to them in terms of spending and turnover. In the second week of November, the online questionnaire was distributed to the selected suppliers via e-mail, three days after the buying firms informed their suppliers about the research. The suppliers were informed that all data supplied would be held confidentially, and that only the research team would see the direct answers from each supplier. Consequently, the case company cannot trace back the individual answers and only receives the aggregate results. Participating suppliers were also promised a summary report of the research, to increase their motivation to answer the questionnaire.
Unfortunately, due to internal circumstances, one company did not succeed in completing all surveys. Therefore, 94 completed surveys are used for this study, which yields an effective response rate of 33.6% (Table 1). To determine whether the current sample size is large enough, sample size recommendations for PLS-SEM were used from Hair, Hult, Ringle and Sarstedt (2014:21). For a statistical power of 80%, a significance level of 5% and a minimum R2 of .25, this study (with a maximum of four arrows pointing at one construct) would need a minimum sample size of 65. Therefore, it can be concluded that this study has enough respondents to have statistical power. On average, the respondents were personally involved in the relationship with the buying firm for 10.3 years. Moreover, 70.2% of the respondents have worked together with the buying company for at least 5 years. Table 2 shows the demographic profiles of the respondents.
TABLE 1 Response rates
Company A Company B Total
Participants contacted 109 171 280
Supplier survey accessed 90 96 186
Removed responses 37 33 70
Useable surveys 53 63 116
Response rate 48.6% 36.8% 41.4%
Completed buyer surveys 53 41 94
Effective response rate 48.6% 24% 33.6%
TABLE 2
Demographic profiles of respondents
Frequency Frequency Sector Bicycle manufacturing 44% Automotive manufacturing 56% Country Belgium 4.3% Slovenia 1.1% China 7.4% Sweden 1.1%
Czech Republic 2.1% Taiwan 17.0%
Finland 3.2% Thailand 1.1%
France 3.2% The Netherlands 39.4%
Germany 6.4% Turkey 2.1%
Italy 5.3% United Kingdom 3.2%
Japan 1.1% United States 1.1%
Portugal 1.1%
Number of employees Annual turnover
0-50 27.7% 0-50 Million 58.5% 51-100 11.7% 51-100 Million 9.6% 101-500 31.9% 101-500 Million 7.4% >500 18.1% >500 Million 6.4% Unknown 10.6% Unknown 18.1% Years of collaboration 0-5 13.8% 26-50 13.8% 6-10 26.6% >50 2.1% 11-25 37.2% Unknown 6.4%
4.2 Measures
This research contains six main variables: RSI, supplier satisfaction, poaching, preferential resource allocation, buyer performance and supply chain intelligence. The scale for all question ranges from 1, “no strongly disagree”, to 5, “yes, strongly agree”, except for poaching, which Likert-scale ranges from 1, “very likely”, to 5, “very unlikely”.
satisfaction of the supplier regarding their current relationship with the buyer. The dependent variable – competitive performance – is measured with items of supplier competitive performance. These measurement items are based on the study of Pulles et al. (2016a) and measure how well the company is able to create competitive advantages due to its relationship with the supplier, based on achieving its market and financial orientated goals (Liu et al., 2006).
The last construct, supply chain intelligence, measures the accuracy of information that the buyer and supplier possess about each other. This study measures the degree of supply chain intelligence by focussing on the information that both exchange partners possess about their share in turnover, number of competitors, level of trust and product specification. “Share in turnover” measures to what percentage both buyer and supplier believe the customer accounts for in the turnover of the supplier. Two items of number of competitors measure how many customers the firms have that are identical to the company considered in this research. Similarly, two items of trust measure how both buyer and supplier view the level of trust that the supplier possesses about the buyer. The measurement scale for trust uses percentages and was developed by Pulles et al. (2014). Finally, the intelligence variable of product specification measures how much influence both parties believe the buyer has on the product design. Kotabe, Martin and Domoto (2003) developed the measurement scale for this single item.
The results from the supplier and buyer on the above six items are compared, and the absolute difference gives the supply chain intelligence. For instance, the supplier may trust the buying firm by 70%, while the buying firm believes that the supplier possesses 100% trust. In this case, the absolute difference, or supply chain intelligence, is 30%. The items of supply chain intelligence are measured separately from each other, while the constructs measure different aspects of intelligence and have their own measurement scale. Tables 3 and 4 provide an overview of the items used in this study.
TABLE 3
Measurement Items – Supplier Questionnaire
Constructs Measurement item (1 = very likely; 5 =
very unlikely)
Factor loadings Poaching (Handley and Benton,
2012)
Cronbach’s alpha =.87; Composite reliability =.92; Average variance extracted =.79
How likely is your firm to use information obtained from this customer...
…to gain favor with other clients. .93 …to help win new business with other
customers.
.92
…to develop new services that you can offer in the marketplace
.82
TABLE 3 (continued)
Constructs Measurement items (1 = strongly disagree;
5 = strongly agree)
Factor loadings Preferential resource allocation
(Physical)(Pulles et al., 2014) Cronbach’s alpha = .85; Composite reliability = .91; Average variance extracted = .77
Compared to other customers… …we grant this customer better utilization of our production facilities.
.89
…we would choose to give this customer priority in the allocation of our products in the case of extreme events.
.91
…we allocate our scarce materials to this customer in case of capacity bottlenecks.
.84
Supplier satisfaction (Cannon, 1998; Pulles et al., 2016b) Cronbach’s alpha =.89; Composite reliability =.92; Average variance extracted =.75
We are very pleased with what this customer does for us.
.88
Our firm is NOT completely happy with what this customer does for us.
.83
Our firm is satisfied with the value we obtain from the relationship with this customer.
.88
Our firm is very satisfied with the relationship with this customer.
.88
4.3 Data validity and common method bias
To assess the measurement instruments in terms of reliability and validity, several tests are conducted before analysing the data. Composite reliability (CR) is used to evaluate the correlation of the survey items. According to Hair et al. (2014:8), “the CR provides a more appropriate measure of internal consistency reliability compared to the traditional assessment of the Cronbach’s alpha”. All CR values ranged between .89 and .92, therefore exceeding the recommended threshold of .70 (Hulland, 1999). To test for convergent validity, this study considers the indicator reliability and the average variance extracted (AVE). The outer loadings
TABLE 4
Measurement Items – Buyer Questionnaire
Constructs Measurement items (1 = strongly disagree;
5 = strongly agree)
Factor loadings Supplier Competitive
Performance (Pulles et al., 2016a)
Cronbach’s alpha = .82; Composite reliability = .89; Average variance extracted = .73
The relationship with this supplier… …has provided my firm with strategic advantages over competitors.
.92 …enabled my firm to reduce cost to a highly
competitive level.
.83
…enabled my firm to defend against competitive threats.
.82
Relational Investment (Liu, 2012)
Cronbach’s alpha = .89; Composite reliability = .92; Average variance extracted = .75
If we switch to another partner, we would lose a lot of the investment we’ve made in this
relationship.
.91
We have made a substantial investment in personnel development dedicated to this partner.
.71
We have invested a great deal in building up the relationship with this partner.
.91
Our firm is very satisfied with the relationship with this customer.
of the five constructs range between .71 and .93, comfortably exceeding the recommended threshold of .708 (Hair et al., 2014). The AVE values range from .73 to .79, so the constructs explain 73% of the variance of the indicators at the minimum. Both tests for convergent validity are above the threshold, so no items are eliminated from the scale.
Discriminant validity measures the extent to which a construct is truly distinct from the other constructs (Hair et al., 2014). The Fornell-Larcker criterion, whereby the square root of each construct’s AVE must be greater than its highest correlation with any other construct, is used to assess discriminant validity. Table 5 demonstrates that all square roots of the AVE on the diagonal are higher than the other correlation; therefore all constructs fulfil the requirement of discriminant validity. Finally, the common method bias is determined to check whether the respondents gave overly positive answers due to sociability (Chang, van Witteloostuijn and Eden, 2010). This is important to consider as this could potentially threaten the validity of the results. The model can be considered free of CMB if all variance inflation factors (VIF) resulting from a full collinearity test are equal to or lower than 3.3. As can be seen in Table 6, two VIF values of poaching behaviour and supplier satisfaction are slightly above the restriction of 3.3. The values are all below 5.0, therefore indicating a minor CMB. However, this limitation should be taken into account (Kock, 2015).
TABLE 5
Mean, Standard deviation, Squared AVE and Correlation table.
Mean Std. dev. 1. 2. 3. 4. 5. 1. Competitive performance 2.72 .96 0.925 2. Poaching Behaviour 3.11 1.24 -0.088 0.945 3. Preferential Resource Allocation 3.86 .83 0.036 0.075 0.937 4. RSI 2.53 1.10 0.726 -0.079 -0.157 0.931 5. Supplier Satisfaction 3.86 .77 0.135 0.084 0.390 0.117 0.930
Notes: Mean = sample mean. Std. dev. = standard deviation. Numbers on the diagonal show the
square root of the AVE. The numbers below the diagonal that are not bold show the correlation
TABLE 6
Collinearity statistics (VIF)
1. 2. 3. 4. 5.
1. Competitive performance 3.762* 3.308* 1.039 3.748*
2. Poaching behaviour 1.019 1.020 1.022 1.017
3. Preferential resource allocation 1.331 1.515 1.244 1.186
4. RSI 1.100 3.997 3.274 3.842
5. Supplier satisfaction 1.321 1.323 1.039 1.278
4.4 Data analysis
The hypotheses of this research are tested with structural equation modelling (SEM). SEM is a general term describing numerous statistical models, which is used to evaluate the validity of theories with empirical data (Lei and Wu, 2007; Karlsson, 2016). This paper tests whether the theoretical model as outlined in section 3 is consistent with the collected empirical survey data. This is in line with the goal of SEM, whereby “causal relationships among variables are tested on prior theoretical assumptions against empirical data” (Hoe, 2008:81). With SEM, research questions are answered by systematically modelling the relationships between independent and dependent variables (Hair, Ringle and Sarstedt, 2011). Furthermore, SEM is able to study relationship that are not directly observable (latent variables), but which are measured by numerous items (Lei and Wu, 2007). By doing so, measurement errors of the latent variables can be analysed, thus improving the model’s reliability (Gefen, Rigdon and Straub, 2011).
SEM is divided into the approaches of covariance base analysis (CB-SEM) and the variance-based approach or partial least squares (PLS-SEM) (Gefen, Straub and Boudreau, 2000). To answer the question of whether to use CB-SEM or PLS-SEM, researchers should take into account the characteristics and objectives that distinguish the two methods (Hair et al., 2011). PLS-SEM is favourable when the goal of the study is to predict key target constructs instead of theory comparison or confirmation. Furthermore, PLS-SEM is favourable when the sample size is relatively small and the structural model is complex (i.e. contains many constructs and indicators) (Hair et al., 2014). The goal of this research is to explain the dependent latent variable of competitive performance and to predict the influence of supply chain intelligence on different relationships. The structural model is made up of many constructs and indicators, and the sample size of 94 is considered to be small. Therefore, this study uses the PLS-SEM approach to estimate relationships in the structural equation model. The software programme SmartPLS 3.0 is used to analyse the measurement and structural model (Ringle, Wende and Becker, 2015). This research follows the multi-stage process of PLS-SEM as described by Hair et al. (2014) and Henseler, Ringle and Sarstedt (2012). First, the inner and outer models are set up based on the theory described in section 3. Second, the outer model is evaluated on its reliability and validity as discussed in section 4.4. Once the constructs of the outer model are trusted, it is possible to accurately measure the variables and relationship in the inner circle. This evaluation of the inner model is performed in the next section.
intelligence. The moderating effect is tested with the multi-group analysis (MGA) approach of Henseler, Ringle and Sinkovics (2009). The continuous variable of supply chain intelligence is divided in two categories: “high supply chain intelligence” and “low supply chain intelligence”, with the median split approach used to dichotomize the variable (Hair et al., 2014).
4.5 Control variables
To control for the possible effect of other factors on the relationships of this study, two control variables are taken into account. The first control variable is the length of the buyer-supplier relationship. When relationships continue over time, collaborating partners have more knowledge about how to meet the quality of their products and services to the expectations of their exchange partner (Venetis and Ghauri, 2004). Therefore, those firms have improved capabilities to make RSI in order to improve supplier satisfaction and preferential resource allocation (Schiele et al., 2012). In addition, companies that have worked successful together in the past are less likely to act opportunistic and harm the relationship (Fynes and Voss, 2012).
5. Results
5.1 The structural model: H1-H4
As depicted in Figure 4, the four direct effects are not significant at the 5% confidence level. Therefore, this study is not able to support hypotheses H1-H4. Although limited and insignificant, some of the results are of interest. First, there is a negative effect (-.18) between RSI and the preferential resource allocation (physical). In addition, the direct effect between RSI and poaching behaviour is also found to be negative (-.08). These results are interesting, because they suggest that the investments made in a relationship with the supplier decrease the chance of preferential resource allocation. Besides, this also suggests a decrease in the chance of poaching behaviour when investments are made in the buyer-supplier relationship. The results are summarized in Table 7.
TABLE 7
Significance Testing Results of the Structural Model Path Beta
(β) t-value Significance level
p-value 95% confidence intervals RSI Preferential
resource allocation
-.18 1.23 NS .22 [-0.38, 0.18]
RSI Poaching behaviour -.09 .58 NS .56 [-0.32, 0.22] Preferential resource allocation Competitive performance .04 .24 NS .81 [-0.22, 0.29] Poaching behaviour Competitive performance -.09 .61 NS .54 [-0.33, 0.22]
Note: Beta (β) = path coefficient, NS = not significant
FIGURE 4
Results of the structural model
Notes: ns = not significant. The structural paths show the standardized beta coefficients and t-values
between brackets.
5.2 Mediating effect of supplier satisfaction
To test hypotheses H5a and H5b, the variable of supplier satisfaction is added to the model of Figure 4. To test for mediating effects, this study follows the approach of Preacher and Hayes (2008). This approach is favourable over the most commonly used Sobel test, as it can be applied to rather smaller sample sizes with more confidence. In addition, the approach exhibits higher levels of statistical power compared to the Sobel test and is “perfectly suited for the PLS-SEM method” (Hair et al., 2014:223).
(p<.10). Therefore, there is no significant indirect effect and no mediation effect of supplier satisfaction on the relation between RSI and poaching behaviour.
In conclusion, there is no evidence to suggest any mediating effect of supplier satisfaction in this model. For this reason, hypotheses H5a and H5b are rejected. However, interesting findings arise when considering Figure 5. First, there is a significant and negative effect of RSI on preferential resource allocation (β=-.21, p<.05). Second, the relation between supplier satisfaction and preferential resource allocation is significant and positive (β=.41, p<.01). In the new situation, the coefficient of determination (R2) of the endogenous construct preferential resource allocation is .19. Meanwhile, the effect sizes (f2) for the relation between with RSI and preferential resource allocation could be considered small to medium (f2=.05). The relation between the constructs of supplier satisfaction and preferential resource allocation has a medium to strong effect (f2=.21).
TABLE 8
Significance Testing Results of the Structural Model Path – Mediator effect Beta (β) t-value Significance level p-value 95% confidence intervals Relational investment Preferential resource allocation -.21 1.97 ** .05 [-0.40, 0.03] Relational investment Supplier Satisfaction .12 0.97 NS .33 [-0.11, 0.35] Supplier Satisfaction Preferential resource allocation .41 5.1 *** .00 [0.26, 0.58] Relational investment Poaching behaviour -.09 0.67 NS .54 [-0.32, 0.19] Supplier satisfaction Poaching behaviour .10 0.73 NS .53 [-0.18, 0.36]
FIGURE 5
Results for the mediator effect of supplier satisfaction
Notes: ** p<.05, *** p<.01, ns = not significant. The structural paths show the standardized beta
coefficients and t-values between brackets.
5.3 Multi-group analysis
Hypotheses H6a-H6c state that high supply chain intelligence positively influences the relationships between RSI and the constructs of preferential resource allocation, supplier satisfaction and negatively the relation with poaching behaviour. In the questionnaires, supplier and buyer were asked six questions related to the topic of supply chain intelligence. Two questions were asked about the number of competitors, two related to the level of trust, one related to share in turnover and one about product specification (see Appendices A and B).
TABLE 9
Median split – Supply Chain intelligence items
High intelligence Low intelligence
Item Median Average n Average n
Number of competitors (1) 8 2.5 31 92 32 Number of competitors (2) 50 11.7 32 302.1 31 Product specification 26 10 45 50.7 45 Share in turnover 5 2 47 13.8 41 Trust (1) 21 11.2 44 38.4 45 Trust (2) 20 10.2 39 40.6 45
Note: n = sample size
First, the PLS algorithm was run for the two groups of high and low supply chain intelligence, to examine whether the groups differ in terms of path coefficients (∆ß). Second, the MGA was run with a bootstrapping of 2,500 subsamples. The measurement invariance was consulted, to ensure “that dissimilar group-specific model estimations do not result from distinctive content and the meanings of the latent variables across groups” (Henseler, Ringle and Sarstedt, 2016:409). If there is enough confidence to believe that an effect exists due to trait differences rather than measurement differences, the different models can be compared by looking at the significant levels. Third, the significance level between the groups of high and low supply chain intelligence was determined.
The first step was more exploratory and visualized the differences in terms of path coefficients between the high and low supply chain intelligence groups. All supply chain intelligence variables showed somewhat interesting results. Therefore, it was decided to conduct an MGA for every intelligence variable. After running the MGA, a critical issue that must be addressed is the measurement invariance (Latan and Noonan, 2017). The invariance test is performed, by checking whether the outer loadings of the items are significant. For the variables of share in turnover and number of competitors, two items were found to be significant (p<.05). Therefore, these variables were removed from this research. The remaining four intelligence variables have items with significance above .05.
which expects a positive effect for the high supply chain intelligence group. The other intelligence variable of trust, the belief that the other party would make sacrifices for the other, did not show any significant relationship.
The second intelligence variable, which relates to the knowledge on the number of competitors in the market that buy similar products, also gave some remarkable results. As depicted in Figure 6.2, the high intelligence group has a significant (p<.10) negative relation between RSI and poaching behaviour. The negative relation for the low intelligence group is less strong. Therefore, these results support H6c: “a high level of supply chain intelligence negatively influences the relationship between RSI and poaching behaviour”. Next, while the full sample provides no evidence to support H3 and H4, this MGA shows that this is not the case for the high intelligence group. There is a significant positive relationship between the preferential resource allocation and competitive performance (p<.10) and a significant negative relationship between poaching behaviour and competitive performance (p<.10). Another interesting finding is the significant positive relationship between supplier satisfaction and the preferential resource allocation for the group with low intelligence. This result was not found in the sample with high supply chain intelligence.
The third intelligence variable, product specification, examines how buyer and supplier assess the degree to which they believe the buyer can influence product design specifications. Figure 6.3 shows a significant positive relation between supplier satisfaction and poaching behaviour (p<.10) for the high intelligence group. This result opposes H5b, which expects a negative relationship between both constructs. Similarly to the intelligence item of competitors in the market, a significant negative relationship exists between poaching behaviour and competitive performance for the high intelligence group (p<.05).
FIGURE 6.1
Multi-group analysis – Trust 1
High supply chain intelligence Low supply chain intelligence
Notes: ** p<.05, *** p<.01, ns = not significant
FIGURE 6.2
Multi-group analysis – Number of competitors (2)
High supply chain intelligence Low supply chain intelligence
Notes: * p<.10, ns = not significant
FIGURE 6.3
Multi-group analysis – Product specification
High supply chain intelligence Low supply chain intelligence
5.4 Control variables
6. Discussion
Building on the existing theory about the buyer-supplier relationship, this paper discusses the relevance of supply chain intelligence when making relationship-specific investments (RSI). This study defines supply chain intelligence as the accurate of information and knowledge firms possess about their exchange partner’s internal and external environment. First, it was examined whether RSI could result in preferential resource allocation and poaching behaviour of the supplier. In doing so, this research aims to see if the buyer’s investments could generate the desired outcomes for the buyer, namely improved competitive performance. Next, the mediating effect of supplier satisfaction was added to the research, to see which role satisfaction plays for the effects of RSI. Finally, the hypothesized relationships were tested for six different variables of supply chain intelligence. Accordingly, this research explores if and how these differences in intelligence contribute to the effectiveness of the investments. An overview of the hypotheses and the results is given in table 10.
TABLE 10
Overview of the hypotheses testing
Hypotheses Results
H1: RSI (+) the preferential resource allocation Not supported
H2: The preferential resource allocation (+) competitive performance Not supported
H3: RSI (+) poaching behaviour Not supported
H4: Poaching behaviour (–) competitive performance Not supported
H5a: RSI (+) supplier satisfaction, which (+) the preferential resource allocation
Partly supported H5b: RSI (+) supplier satisfaction, which (–) poaching behaviour Not supported H6a: Supply chain intelligence (+) moderates the relationship of H1 Partly supported* H6b: Supply chain intelligence (+) moderates the relationship of H2 Supported* H6c: Supply chain intelligence (–) moderates the relationship of H3 Partly supported*
Note: *based on the intelligence variables of trust, number of competitors and product specification.
6.1 Hypotheses testing
between RSI and preferential resource allocation, this study suggests that this is certainly not the case for every investment, and therefore other factors need to be taken into account.
6.2 Supply chain intelligence
This study did so, and posited that the possession of accurate knowledge would be an important factor to influence the effectiveness of RSI. When examining the relation between RSI and preferential resource allocation, the intelligence variable of trust showed a significant difference between those companies possessing high or low supply chain intelligence. In contrast to what was hypothesized, a significant negative relationship was proven for the group with high intelligence, while a positive (non-significant) relationship was demonstrated for those with low supply chain intelligence. This does not imply that a high level of trust is negative for the preferential resource allocation; rather it suggests that when buyer and supplier are aware of the level of trust both parties possess about each other, RSI have lower chances of giving the favourable customer status.
Tsai and Ghoshal (1998) investigated how the three dimensions of social capital (structural, cognitive and relational) influence resource exchange and combination. At a dyadic level, i.e. similar to the methodology of this research, they found trustworthiness to be significantly and positively associated with resource exchange. However, they found no evidence to support a direct effect between cognitive social capital and resource exchange. Although cognitive social capital, defined as resources that provide a shared representation and interpretation among parties (Nahapiet and Ghosal, 1998), may reduce uncertainty and conflict while enhancing commitment (Modi and Mabert, 2007), Villena et al. (2011) support the finding of this study by showing “the dark side” of cognitive social capital. When cognitive social capital rises to high levels, “risks of groupthink and isomorphism becomes stronger, creativity in the relationship reduces and costly investments to improve the relationship may end up being harmful for the overall performance” (Villena et al., 2011:564). In other words, a shared representation and interpretation might only be beneficially to a certain level, suggesting an inverted u-shaped relationship.
exists. Meanwhile, the group with high intelligence supported the third and fourth hypotheses of this research. The results showed a significant positive relation between preferential resource allocation and competitive performance and a significant negative relation between poaching behaviour and competitive performance. In other words, the possession of accurate knowledge contributes to a company’s competitive performance. These results support previous studies, such as Paulraj, Lado and Chen (2008) that investigate inter-organizational communication as an antecedent of performance outcome in buyer-supplier relationships. However, where most of these studies focused on the quantity of information, this research shows that qualitative or accurate information is an important attribute to enhance performance outcomes in the buyer-supplier dyad.
The intelligence variable of product specification also showed a significant negative relation between poaching behaviour and competitive performance. Subsequently, the results of the MGA revealed significant differences between the groups of high and low intelligence when focussing on the relationship between RSI and supplier satisfaction. This result suggests that when both parties are on the same page about the degree of influence that the buyer has on the product specification, the supplier will be more readily satisfied with RSI. In the context of new product development, a similar result was found in the research of Fang, Palmatier and Evens (2008). They show that specific investments in the relationship result in increased satisfaction, product value and performance, when the contribution of the buyer to the product development is categorised as “fair”.
6.3 Years of supplying
The additional analysis for the control variable of years of supplying showed fairly similar results as the intelligence variable of product specification. There is a strong positive relation (ß=.35) between RSI and supplier satisfaction for those suppliers working less than 11 years with the buying firm, and a strong negative relation (ß= -.22) for those suppliers who have supplied the buying firm for more than 11 years. This result suggests that RSI are more effective in relatively young buyer-supplier relationships, when the aim is to improve the satisfaction of their supplier. This result is somewhat surprising, as firms make RSI with the aim of continuing, improving and building long-term relationships with their suppliers (Crawford, 1990). Certain consequences of RSI, namely increased dependency and higher switching costs, could explain why RSI decrease the satisfaction of suppliers in a long-term relationship (Lam, Shankar, Erramilli and Murthy, 2004; Villena et al., 2011).
and establish trust, dependency and assurance, RSI give confidentially to both parties in the dyad (Brennan and Turnbull, 1999). However, when a relationship reaches the phase of maturity or decline, relational norms and commitment are more effective than RSI and additional dependencies (Jap and Ganesan, 2000). Therefore, RSI should be made in the developing phase of the relationship, when the aim is to increase supplier satisfaction and establish potential long-term relationships.
6.4 Moderating effect
7. Limitations
As with all research, there are limitations that need to be addressed. First, the results of this study uncover only a snapshot of the potential variation in outcomes and intelligence in the different buyer-supplier relationships. With this study, it was not possible to examine whether the answers of both supplier and buyer change or hold over time, or whether they are reciprocal longitudinally. The process of supply chain intelligence is continuous and changes over time (Rouach and Santi, 2001). For instance, Jap and Anderson (2007) recognised that the perception of trust fluctuates over time. Therefore, it must be taken into account that the results of this study might be different when conducting the same research another time.
Another limitation of this research is the operationalization of buyer performance. This study focused on competitive performance and investigated whether buying firms were able to obtain competitive advantages from the relationship with their supplier (Pulles et al., 2016a). Although competitive advantages are closely related to performance improvements (Newbert, 2008; Zhou, Brown and Dev, 2009), usually they do not immediately enhance the organization’s overall performance (Li et al., 2006). Therefore, the competitive advantages of the buying firm may be influenced by RSI, preferential resource allocation and poaching behaviour, but these factors will not directly improve or reduce the overall performance of the firm. For example, spillover effects due to poaching behaviour reduce the competitive advantage of the firm; however, the resulting economic performance and other consequences will only be noticeable after some time. Therefore, this research measures competitive advantages, which may give performance fluctuations in the future, but does not measures the current operational performance of the firm. This is a notable limitation of the study.
8. Future research directions
This study points to a number of areas where future research would be valuable for academicians and managers. Some significant results were found for the moderating effects of the intelligence variable of supply chain intelligence. For instance, it was revealed that high intelligence about the product specification is important when making investments with the aim to improve the satisfaction of the supplier. Although the six items of intelligence all measured the degree to which buyer and supplier possess the same knowledge about each other, they did not provide the same outcomes, let alone significant results. The literature of competitive or supply chain intelligence is relatively limited, thus the aim of this study was to explore different variables of intelligence and identify similarities and differences. However, an interesting field of research emerges, where researchers can investigate the underlying reasons and mechanisms that explain the differences in intelligence variables. Future research could elaborate on the intelligence variables presented in this study or focus on other intelligence variables such as commitment, sustainability or power. By doing so, researchers could elaborate even further on the gaps in the literature as discussed by Blonska et al. (2013) and Whipple et al. (2015), and subsequently establish which supply chain intelligence variables are the most relevant for influencing the RSI effectiveness and overall performance of the firms.
9. Managerial implications
This study also provides information to assist managers in their decision-making. The literature gives several tools for managers to increase information sharing capabilities in the buyer-supplier relationship. However, as previously mentioned, most studies neglect the quality or accuracy of information. This study focused on how the supplier’s behaviour influences the effectiveness of RSI, and investigated the role of supply chain intelligence. The analysis carried out here showed positive outcomes for the group with high supply chain intelligence regarding the competitive performance. Furthermore, the likelihood of poaching behaviour decreased and supplier satisfaction increased, indicating that supply chain intelligence influences the supplier’s behaviour. Therefore, this study states that firms which possess accurate information and knowledge about their exchange partner’s internal and external environment are able to make more effective RSI. This finding could help managers to improve the impact of RSI in the relationship with their supplier. Managers can improve their supply chain intelligence in a number of ways, such as by providing immediate feedback (Maddox, Ashby and Bohil, 2003), sharing relevant and critical information (Paulraj et al., 2008), investing in information processing capabilities (Grover, Teng and Fiedler, 2002), effective conflict management (Kale, Singh and Perlmutter, 2000) or simply increasing regular (face-to-face) contact by plant visits (Bell, Oppenheimer and Bastien, 2002; Zhou et al., 2014). By doing so, managers could increase the accuracy of the information and knowledge they possess, in turn helping to receive beneficial outcomes from RSI.
10. Conclusion
This research contributes to the literature through several aspects. First, this research considered and investigated both the supplier and buyer perspective in the buyer-supplier relationship. Where many similar types of research only focus on one side of the relationship (Blonska et al., 2013, Wang et al., 2013), this dyadic research acknowledged the perspectives of both parties in the relationship. Second, Blonska et al. (2013) noticed a gap in the literature about how a misalignment in perception and knowledge between buyer and supplier influences the performances of both parties. In addition, Whipple et al. (2015) identify a lack of knowledge about mechanisms to influence the effectiveness of RSI. This research found significant differences in competitive performances between relationships where dyadic partners have a high and low degree of supply chain intelligence. For example, it was highlighted that in relationships where both parties have considerable knowledge about the competitiveness of their market, RSI have a higher chance of resulting in preferential resource allocation, and eventually in creating higher competitive performance. Opposite results were found for the group with a high misalignment in knowledge about the market and competitors.
In addition, Modi and Mabert (2007) demand further understanding of how firms with RSI can control the behaviour of their supplier. This research showed that supplier satisfaction has an important moderating effect on the relationship between RSI and poaching behaviour of the supplier. Of course, there are many ways to improve the satisfaction of the supplier, but this research indicates that RSI can contribute to the supplier’s satisfaction if investments are made in the developing phase of the relationship. In addition, relationships with high intelligence for the three supply chain intelligence variables of product specification, number of competitors and trust all showed improved outcomes for the prediction of poaching behaviour. Accordingly, this research states that firms with accurate information are able to predict, or control, the behaviour of their direct suppliers when making RSI.
