1
Master Thesis Supply Chain Management
University of Groningen Faculty of Business and Economics
An analysis of the joint effect of supply complexity and customer order decoupling point on the level of supply chain integration
January 29th, 2018 DENNIS MATHEIS Student number: S3223477 e-Mail: D.Matheis@student.rug.nl Supervisor / University of Groningen
Dr. H. Broekhuis
Co-assessor / University of Groningen Dr. S. Boscari
2
ABSTRACT
3
CONTENT
1. INTRODUCTION ... 4
2. THEORETICAL FRAMEWORK ... 6
2.1 Supply chain integration (SCI)... 6
2.2 Supply complexity ... 7
2.3 Customer order decoupling point (CODP) ... 8
2.4 Joint effect of supply complexity and CODP on supply chain integration ... 9
2.5 Conceptual model ... 11
3. METHODOLOGY ... 12
3.1 Main research methods of this study ... 12
3.2 Study 1: Multiple case research ... 12
3.2.1 Setting and unit of analysis ... 12
3.2.2 Case selection ... 13
3.2.3 Data collection ... 15
3.2.5 Data analysis ... 16
3.3 Study 2: Secondary data ... 18
3.3.1 Data sample ... 18
3.3.2 Data analysis ... 19
4. FINDINGS ... 21
4.1 Multiple case research... 21
4.1.1 Upstream integration efforts ... 21
4.1.2 Supply complexity and supply chain integration ... 23
4.1.3 CODP and supply chain integration ... 28
4.1.4 Joint effect of supply complexity and CODP on supply chain integration ... 29
4.2 Secondary Data ... 30
4.2.1 Downstream integration efforts ... 30
4.2.2 Supply complexity and supply chain integration ... 32
5. DISCUSSION AND CONCLUSION ... 36
REFERENCES ... 41
4
1. INTRODUCTION
Over the last decades, supply chain integration (SCI) has been one of the most important topics in supply chain management (Das, Narasimhan & Talluri, 2006; Germain, Claycomb & Droge, 2008; Van der Vaart & Van Donk, 2008). However, Flynn, Huo, and Zhao (2010) expressed doubts concerning the paradigm that more integration with customers and suppliers is necessarily beneficial. More recent research has shown that the level and the type of SCI depend on the context of the supply chain, such as the customer order decoupling point (CODP) (Van der Vaart & Van Donk, 2006; Van Donk & Van Doorne, 2015) and supply complexity, i.e. supply characteristics, which display the complex process from the buyer’s order to the final delivery of the goods (Gimenez, Van der Vaart & Van Donk, 2012; Van der Vaart & Van Donk, 2008; Van Donk & Van der Vaart, 2004). This paper seeks to extend the earlier findings by investigating the individual impact and the joint effect of the contingency factors of CODP and supply complexity on the level of SCI. Moreover, the study attempts to clarify possible contradictions of the results of previous publications that studied both contingency factors independently.
Recently, Van Donk and Van Doorne (2015) studied the impact of the CODP on the level of SCI. They demonstrated that companies with a made-to-stock (MTS) CODP show a higher level of integration towards the customer (downstream) and a lower level of integration towards the supplier (upstream). The opposite applied to made-to-order (MTO) companies, according to Van Donk and Van Doorne (2015). Van Donk and Van der Vaart (2004) similarly found a positive relation between the level of SCI and supply complexity. Companies with a high supply complexity are more integrated with their downstream supply chain partner than those characterised with a low supply complexity. A complex supply situation is distinguished by different business conditions including small batch production, fluctuating demand, high variety, high expectations about flexibility and quality, and a high level of innovation (Gimenez et al., 2012). Gimenez et al. (2012) investigated the moderating role of supply complexity in the relationship between SCI and supply chain performance. Their research showed that in supply chain links with high supply complexity, high SCI is positively related to performance (Gimenez et al., 2012).
5 low level with their suppliers. However, according to the studies that consider supply complexity, the required level of integration between the two units is expected to be high. Several reasons might explain these contrasting findings. It might be that (all) MTS companies do not have to handle complex supply decisions, or they do not properly deal with them and therefore do not integrate with their suppliers on an appropriate level. It is also possible that a completely different explanation is required for why the results from both studies imply contradicting results in an MTS setting. This research investigates the joint effect of both factors, i.e. supply complexity and CODP, on a buyer-supplier relationship level and clarifies whether the previous results are in contradiction, in harmony, or whether one factor tends to be more dominant.
The novelty of this study is that more than one contingency factor is considered along with their joint effect on the required level of supply chain integration. Moreover, I follow the recommendation of Gimenez et al. (2012) to consider another contingency factor, namely the CODP. Considering more than one contingency factor assists in obtaining more insights regarding which are the most influencing factors on the supply chain and which efforts are commonly used to integrate with other supply chain partners. Furthermore, this study strives to extend the understanding of previous research by investigating the joint effect of two individually explored contingency factors on the level of SCI. In addition, organisations should be provided with recommendations for certain SCI efforts in a particular supply chain constellation. To provide these insights, the research question is as follows:
How do two contingency factors, i.e. CODP and supply complexity, influence the level of supply chain integration?
6
2. THEORETICAL FRAMEWORK
2.1 Supply chain integration (SCI)
Numerous definition and measurements of SCI have been published, but an overarching understanding is still missing (Gimenez et al., 2012; Tsinopoulos & Mena, 2015). For instance, Van der Vaart and Van Donk (2008) analysed 33 survey-based studies with different constructs to measure and define SCI. SCI is usually defined as ‘the degree to which
a manufacturer strategically collaborates with its supply chain partners and collaboratively manages intra- and inter-organization processes […]’ (Flynn et al., 2010: 59).
Inter-organisational collaborative processes in that sense are upstream (towards suppliers) and downstream (towards customer) integration efforts, which improve, according to Frohlich and Westbrook (2001), the performance of the organization. In addition, Flynn et al. (2010) stated that activities within a company (internal integration) are a prerequisite for the integration with other supply chain partner (external integration).
7 2008). In general, the level of integration can be considered as high when more of the named practices are established in the buyer-supplier relationships. A rating or importance of different efforts has not been investigated yet. This paper follows the definition of SC integration of Van der Vaart and Van Donk (2008) because it is the outcome of an extended literature review and has been used in several follow-up studies (Gimenez et al., 2012; Leuschner, Rogers & Charvet, 2013; Vallet‐Bellmunt & Rivera‐Torres, 2013; Van Donk & Van Doorne, 2015). Using the same definition helps to compare the findings of this study with those from previous papers. Furthermore, Vallet‐Bellmunt and Rivera‐Torres (2013) showed that each of the three dimensions represent a different approach to integration: (1) SC practices are often reflected as the interactive dimension and the operational part, (2) SC patterns are related to collaboration dimensions and build the more strategic part, and (3) SC attitudes should be understood as the corporate philosophy and a relational dimension. This division provides a sophisticated way to determine the level of SCI and helps to identify how the interplay of the contingency factors influence the level of SCI in more detail by making it possible to assign specific integration efforts to the influence of different contingency factors.
2.2 Supply complexity
This paper has the aim to investigate the interplay between the CODP and supply complexity and its influence on the level of the SCI. By building the research on the findings of Gimenez et al. (2012), this study considers supply complexity on the level of individual buyer-supplier relationships and follows the definition of supply complexity as ‘the complexity of the
process in which buyer’s orders are converted into the supplier’s manufacturing orders, resulting in the delivery of goods according to the buyer’s expectations (adapted from Welker, 2004)’ (Gimenez et al., 2012: 584). The topic of complexity in supply chains receives,
8 To operationalise the term ‘supply complexity’, Gimenez et al. (2012) used the business conditions defined by Van Donk and Van der Vaart (2004). These conditions were derived from the work of Aitken, Childerhouse, and Towill (2003) and Childerhouse, Aitken, and Towill (2002), who developed the so-called DWV³ approach, which consists of the Duration of the life cycle, time Window for delivery, Volume, Variety, and Variability. According to Childerhouse et al. (2002), these characteristics influence the choices in designing and managing the supply chain. Van Donk and Van der Vaart (2004) followed that line of thought and derived the following business conditions: order winners, volume and variety, lead-time, percentage MTS, and batch size. These conditions consider the uncertainty suppliers have to face regarding their production planning and delivery schedules (Gimenez et al., 2012). More uncertainty requires more flexibility, which is associated with high supply complexity, e.g. small batches, long lead-times, et cetera. Gimenez et al. (2012) felt confident to conclude that the complexity of these business conditions can be understood as supply complexity as they cover their proposed definition.
Since the factors have been used in several other studies (Gimenez et al., 2012; Van der Vaart & Van Donk, 2006; Van Donk & Van der Vaart, 2004), it seems suitable to apply them as well in the present study. Moreover, the usage of these factors helps us to compare existing findings, makes it optional to utilise secondary data that employs the same operationalisation of supply complexity, and expands knowledge based on existing literature. This study strives to investigate Gimenez et al. (2012) findings in relation to the CODP and discover how the joint effect might influence the level of SCI.
2.3 Customer order decoupling point (CODP)
9 of the CODP has an impact on the design of the supply chain, as the characteristics are different. Regarding the characteristics, both upstream- and downstream-orientated supply chains need to be managed differently to increase performance. Moreover, the location helps provide an idea where to implement ‘efficiency- and flexibility-related production
techniques’ (Van Donk & Van Doorne, 2015: 2573). For MTS (upstream-orientated)
companies, a delivery reliability and productivity orientated strategy is often required, which promotes the need for more efficiency with the help of lean and forecasting methods. In contrast, MTO (downstream-orientated) companies concentrate more on delivery speed and flexibility (Hallgren & Olhager, 2006), which implies that more flexible processes are needed with the use of order fulfilment and agile-related mechanics to counter this (Hoekstra et al., 1992; Van Donk & Van Doorne, 2015). Since the requirements in MTO and MTS companies and their recommended integration efforts differ (Van Donk & Van Doorne, 2015), the additional effect of high supply complexity is interesting to observe, as it might contradict the published recommendations.
2.4 Joint effect of supply complexity and CODP on supply chain integration
Today, investigations regarding the influence of both the CODP and supply complexity and their effect on the level of SCI are missing in the literature.10 and Prajogo (2012), MTO companies should integrate suppliers of customised products more intensively to avoid production stops, whereas MTS companies can use standardised products from a limited supply base. Van Donk and Van Doorne (2015) confirmed these findings in their study on the impact of the CODP on the level of SCI, thereby strengthening the literature by stating that MTO companies have a higher integration level towards the supplier (upstream) and MTS less. In addition, the authors explained that MTS companies integrate on high level with their downstream supply chain partner, whereas MTO companies tend to have a lower tendency to do that.
As already mentioned, several studies relating complexity to the supply chain have been published in the last decades (e.g. Vachon & Klassen, 2002; Wilding, 1998). However, research on supply complexity and its effect on the level of SCI is still scarce. Van Donk and Van der Vaart (2004) were the first to prove a relation between supply complexity and the level of integration. The subject of their works was that connections in the supply chain, which reflects a high complexity, are more integrated than links with low complexity (Van Donk & Van der Vaart, 2004). Gimenez et al. (2012) studied this topic and discussed the impact of SCI on performance moderated by the effect of supply complexity. Their study proved that more integration has a positive impact in cases of high complexity, yet their expectations that SCI has no significant influence on the performance in low supply complexity environments were not confirmed (Gimenez et al., 2012).
Ultimately, these two studies are the only studies that link supply complexity with SCI. Despite this research, Gerschberger, Manuj and Feinberger (2017) focused on supplier-induced complexity, i.e. the geographical location of the supplier, the deliver reliability of the supplier, and the dependency on the supplier. With their study, they confirmed the findings of Gimenez et al. (2012) by stating that a high supplier-induced complexity is related to adverse outcomes. In other words, a high supply complexity harms the performance of a company. Gerschberger et al. (2017) also suggested that a high degree of information sharing benefits and increases the supplier reliability as an integration measure.
11 on the supposed contradicting results. This study follows the recommendations of Gimenez et al. (2012) to investigate the influence of another contingency factor in a high complex supply situation. This paper thus investigates the joint effect of both contingency factors (CODP and supply complexity) and seeks to provide an explanation of the contradicting results previous research has delivered.
2.5 Conceptual model
The starting point of this study includes the results from the papers of Van Donk and Van Doorne (2015) and Gimenez et al. (2012) as well as Van Donk and Van der Vaart (2004). This thesis aims to analyse the influence of the contingency factors CODP and supply complexity on the level of SCI.
Based on the described theoretical background, the following conceptual model can be presented, as shown in Figure 1.
FIGURE 1 Conceptual model
The following sub-questions are addressed to obtain a more comprehensive understanding of the effect of the single contingency factors and their joint effect on the level of SCI.
1) How does supply complexity influence the level of SCI? 2) How does the CODP influence the level of SCI?
12
3. METHODOLOGY
3.1 Main research methods of this study
The aim of this paper is to investigate the effect of multiple contingency factors (supply complexity and COPD) and their joint influence on the level of SCI. To gain in-depth knowledge about the topic, a multiple case study approach was chosen and an analysis of the secondary data was executed.
The multi-case study approach was chosen to investigate the unexplored phenomenon of the joint effect of two contingency factors on SCI. Moreover, the approach provides the opportunity to ask in-depth questions about the underlying processes of the interplay of the two contingency factors and how that influences SCI (Eisenhardt & Graebner, 2007; Voss, Tsikriktsis & Frohlich, 2002). The case study approach thus provides the best opportunity to understand the complexity of the different effects on the type and level of SCI.
In addition to the case study approach, an analysis of secondary data was conducted. The secondary data provide information to receive deeper insights regarding how supply complexity affects the level of SCI. Furthermore, the secondary data offered the opportunity to verify the findings from the primary data and guard against observer bias. The setting of the secondary data is slightly different than that of the primary data. As the focal company in the case study is the buying unit, the focal company in the secondary data is the supplying unit. The secondary data consisted of data about the supply complexity from the perspective of the supplying unit. Data on the COPD were not available. Therefore, it is not possible to investigate the joint effect of the CODP (we investigate this from the perspective of the buying unit, not from the perspective of the supplier) and supply complexity (data were only available from the supplying unit). Moreover, the secondary data were already used in the study by Van Donk and Van Doorne (2015) to investigate the effect of the CODP on the level SCI.
3.2 Study 1: Multiple case research
3.2.1 Setting and unit of analysis
13 with three of the companies, and due to anonymity issues, they are named A, B, and C. Other companies did not respond or did not show interest in participating in the research. All companies develop and assemble their bikes on their own and sell it either to the customer directly or to retailers.
The unit of analysis in this research is the relationship between the focal company and its suppliers. Since this study investigates the effect of the CODP and supply complexity on the level of SCI, SCI has to be considered as the dependent variable. Supply complexity and the CODP are independent variables in this paper.
3.2.2 Case selection
To conduct the research, cases had to be selected. The following case selection criteria have been applied:
1) The sample companies should have suppliers with a high supply complexity. 2) The sample companies produce MTS.
3) The sample companies should integrate with their suppliers.
The first selection criterion was the level of supply complexity. Since this research aims to investigate the possible contradictions with high supply complexity in an MTS setting, the desired cases should have a high supply complexity. This helps in obtaining a deeper understanding of how high supply complexity affects the degree of SCI individually and together with the CODP MTS. It was expected to find high supply complexity in the bike industry for several reasons.
We initially aimed to incorporate two cases: one with an MTS approach and one with an MTO approach. This would make it possible to receive deeper insights on the effect of the CODP on the level of SCI. We found that MTO companies do not compete in the same market as the companies for which we were aiming. Made-to-order companies mainly produce customised bikes and do not have the size (in terms of employees) that we wanted for the study.
14 that bigger companies have more developed supply chains, and a certain level of integration can be found.
An overview of the selected cases can be found in Table 1. With respect to the cases, all cases were selected as a literal replication to obtain a comprehensive understanding of the interplay of the observed contingency factors (Eisenhardt & Graebner, 2007). Diversity in the level of supply complexity, different CODPs, and level of SCI was not possible to incorporate in the study. All companies operate in an MTS environment with high supply complexity, and I therefore expected to find similar results.
TABLE 1 Case selection
Supply complexity CODP SCI
Case A High MTS High
Case B High MTS High
Case C High MTS High
In this study, we selected three cases from the bicycle industry. The advantage of selecting from one sector is that all companies offer similar products and work with a similar type of suppliers. Furthermore, it helped to draw conclusions regarding how the dependent variable (SCI) is influenced in that sector by high supply complexity and MTS production.
Further information about the three cases is given in Table 2. Due to anonymity issues, the numbers in Table 2 are only approximate indications.
TABLE 2 Case descriptives
Case A Case B Case C Employees 650-700 50-100 150-200
Revenue (10*^6) 150-200 60-70 60-70
Sales environment B2C B2B B2B
Product families 28 33 17
Position Interviewees Director Production Head of Global Sourcing
15
3.2.3 Data collection
Two different data collection techniques were used: (1) semi-structured interviews and (2) a study of internal data and documents. The main information was derived from the interview with the representatives from the three companies. Additional background information was taken from financial reports, bill of materials, and their websites. These additional data were used to enhance triangulation to increase the reliability and validity of this research (Karlsson, 2016).
16 immediately after the interviews were held by the interviewer. In case of missing information, the representatives were available to answer open questions by mail.
As the interviewees did not want to reveal sensitive numbers, the level of supply complexity was derived by the intervieweraccording to the information the interviewees gave. Specific questions were asked about the different dimensions of supply complexity in relation to the supplied product.
In addition to the execution of the semi-structured interviews, the websites and financial reports were inspected to obtain further background information. The bill of materials could be found on the website and helped to determine the variety of frames, which is ultimately an important factor in determining the supply complexity.
3.2.5 Data analysis
To analyse the data from the interviews and documents in a structured way, this study followed the three suggested steps of Miles and Hubermann (1994): data reduction, data display, and conclusion. The reduction was done by coding-relevant paragraphs, sentences, and quotes (first order). First, first-order categories were added to quotes of the interviewees. This resulted in 48 first-order codes. These first-order codes were subsequently clustered into descriptive second-order codes such as ‘value long-term relationships’, ‘special barcoding’, or ‘face-to-face meetings’. These second-order codes were then assigned to third-order codes. The third-order codes were related to the already pre-determined integration dimensions of Van der Vaart and Van Donk (2008): practices, patterns, and attitudes. Some of the second-order codes were already defined by Van Donk and Van Doorne (2015) and could also be used in this research. Other integration initiatives appeared new in this study.
17 With the help of the total scores, the total integration efforts per case could be identified and compared.
To determine the supply complexity of the supplier, this study used two items. First, specific questions were asked to obtain general background knowledge of the suppliers and how they work the supply chain partner (buying and supply unit) work together. Second, questions about the five suggested dimensions of supply complexity according to Van der Vaart and Van Donk (2006) was used to gain richer data and to quantify supply complexity and compare it to previous research. For the quantification, the coding model of Gimenez et al. (2012) was used, as can be seen in Table 3. Low scores can be associated with high complexity and high scores with a low supply complexity. The scores were derived from the interviews with the company representatives and returned to them to confirm the scoring. The application of that scoring model made it possible to compare the findings from the primary data to the findings of the secondary data.
TABLE 3 Coding model supply complexity (Gimenez et al., 2012: 592)
Indicators Score = 1 (high
supply complexity) Score = 2
Score = 3 (low supply complexity) Percentage MTS ≤25 25< MTS <75 ≥75 Lead-time (days) >15 5< LT <15 ≤15 Volume-variety quotient <100 100< VO/VA <1000 ≥1000 Batch size ≤100 100< BS <1000 ≥1000 Order winners Costs relatively unimportant AND Flexibility important Remaining companies
Cost are important AND Flexibility
relatively unimportant
18 which results in more complexity in the production of the supplier. Order winners were arranged in three groups. High complexity is related to supplying companies where their suppliers value the flexibility. Low complexity is mostly found where low costs are the centre of focus. A moderate complexity can be found in supplying companies that do not set a clear focus on low costs or flexibility. The volume-variety quotient explains how many different products the company has to provide within a certain number of produced items (Van der Vaart & Van Donk, 2006). This research used the numbers of produced bikes as the ‘volume’ and the different stock keeping units (SKU) of frames for the ‘variety’. The complexity increases with a higher quotient because of the higher variety the suppliers have to handle. Afterwards, the effects of supply complexity and the CODP (MTS) were analysed, and the effect on the level of SCI was subsequently derived. The analysis is based on the statements the representatives claimed in the interviews and the documents.
In the beginning, all cases are analysed separately, and afterwards, conclusions are drawn for every case. After the single cases are analysed, an analysis of all cases is executed. In doing so, I have attempted to reveal different patterns that would explain how the level of SCI is influenced by different contingencies and if there is a dominant factor. Moreover, the patterns identified should point to how companies design their integration efforts in relation to the two contingency factors.
3.3 Study 2: Secondary data
3.3.1 Data sample
The secondary data were initially collected for the study of Van Donk and Van Doorne (2015) to investigate the relationship between the CODP and the level of SCI. Data gathering was performed by three researchers in May 2013. The qualitative data were collected with the help of semi-structured interviews with senior managers.
19 determine the supply complexity of the companies. A variety of different CODPs can be found Details about the research sample of the secondary data are located in Table 4.
TABLE 4 Overview cases secondary data (Brolsma, 2012: 9)
Case A B C D E F G H I J K L Employees 90 30 120 400 65 43 90 60 160 60 255 120
Revenue
(10*^6) 27 6 * 60 14 50 12 8 50 * 225 60
CODP MTO ATO MTS
/MTO MTS MTO MTO MTO MTO MTO MTO MTO MTO * Unknown
As the information is provided to perform further research on the relation between supply complexity and the level of SCI, this dataset helps to triangulate the findings from the primary data and obtain a better understanding of the effect of the contingency factor of supply complexity. Moreover, knowing more details about the influence of one contingency factor helps to develop a better understanding of the joint effect of two contingency factors. As no information was provided regarding the CODP of the customer, the joint effect of high supply complexity (supplying unit) and CODP (buying unit) cannot be observed as was done with the primary data.
3.3.2 Data analysis
21
4. FINDINGS
In this chapter, the findings of the multiple case research and the secondary data are presented. The results are shown according to the sub-questions from section 2.5. First, all the findings of the primary data are explained, and the results of the secondary are subsequently shown. A comparison of the results is completed in the discussion section.
4.1 Multiple case research
4.1.1 Upstream integration efforts
At first, the focus was set on the upstream integration efforts of the focal companies. Upon observing Table 5, it becomes clear that all three companies put much effort in integrating with their (frame-) suppliers. Strong similarities can be seen in the integration practices and patterns. In all three cases, the buyers use the following integration practices: special barcoding to improve the physical flow of the goods, frequent deliveries, forecasting from both sides, and joint R&D-activities. Furthermore, a strong focus is set on communication. Cases A and B have scheduled weekly conferences, whereas the representative from case C stated, ‘At the moment we do not do that [weekly conferences], but we try to implement
that’. All companies use intensive communication and face-to-face meetings with their
suppliers. In addition, company B has employees always present at the supplier’s site to conduct quality checks and track the deliveries.
Differences are present in the willingness to cooperate regarding the improvement of processes or the collaboration in general. In cases B and C, the willingness to improve is already present and executed. In contrast, in case A, the willingness to improve is only driven by the focal company. According to the company representative of case A, ‘The suppliers are
more focused on delivering the goods, instead of working together. They do not see that there is potential to work together and to reduce costs in the production’. Even though the
willingness to improve factors together is one sided in case A, the focal company and its suppliers value long-term relationships and strong partnerships.
The interest in long-term relationships is true for all cases, and as an example, case C said,
22 in all, it can be concluded that all of the companies put much effort in integrating with their suppliers. In all three cases, there are similarities in the companies’ strong focus on communication, the exchange of information, and the value of long-term relationships. TABLE 5 Integration initiatives of all three cases
Integration Case A Case B Case C
Practices Physical flow Special barcoding Frequent deliveries Special barcoding Customized packaging Frequent deliveries Special barcoding Customized packaging Frequent deliveries Information exchange continuous forecast weekly update on deliveries access to quality data shared forecast weekly update on deliveries occasional information on stock levels shared forecast weekly update on deliveries Cooperation joint R&D activities joint R&D activities Joint activities in terms of improving the collaboration joint R&D activities Joint activities in terms of improving the collaboration Patterns weekly status conferences continuous communication face-to-face meetings colleagues on site weekly status conferences continuous communication face-to-face meetings Daily contact Face-to-face meetings Attitudes
Value long term relationships
Investment in test equipment
One sided (focal company) willingness to improve collaboration
Value long term relationships occasional investments in specific tools Willingness to improve things
23
4.1.2 Supply complexity and supply chain integration
This section aims to determine the supply complexity of the supplying unit for the purchased frames of all the three cases and to show the effect of supply complexity on the level of SCI. First, the level of supply complexity is described for all cases together, and afterwards, the relation between the supply complexity and SCI is explained per case. The end describes a summary of the findings per case.
From the cases, we derived the supply complexity of the frames based on the coding method of Gimenez et al. (2012). The results per case can be found in Table 6.
TABLE 6 Supply complexity scoring
Indicators Case A Case B Case C
Scoring Value Scoring Value Scoring Value
Percentage MTS 1 <25% 1 <25% 1 <25% Lead time (days) 1 90-120 1 90 1 100-110 Volume-variety quotient 1 (88.000/886) =99,32 2 (80.000/466) =171,67 2 (30.000/128) =234,38 Batch size 1 <100 1 <100 1 <100 Order
winners 2 Mixed 2 Mixed 1
Focus on flexibility
Total score 6 7 6
Nearly all of the frames were MTO by the suppliers, but the supplier used the buyer’s forecasts to order material in advance, as Case A revealed: ‘They produce [the frames] on
order, but they [the suppliers] order the material for the frames based on our forecasts’.
24 The lead time of the production of frames can be found in Table 6. In all three cases, the lead-time can be considered as very high and leads to high supply complexity.
The volume-variety quotient is in all cases is close to 100 (see Table 6). In case A, the variety of different frames is very high, and that leads to high supply complexity. The number of needed frames in case B is similar to the number in case A, but the variety is significantly lower, which results in a moderate complexity. The total number of produced bikes and the variety are lower in case C. The quotient can be categorised as a moderate complexity. The batch sizes were derived by the order quantities of the focal companies. For example, the representative of case A stated that the order quantities ‘…are completely mixed. From
one to eighty’. Case B added that ‘…it is possible that we order just 50 frames’. Statements
like these were made in all interviews and lead to the conclusion that the suppliers have small orders and subsequently, small batches of production. Small production batches require a high flexibility because of the frequent changes in production. That results in a high complexity.
When it comes to the order winners, case A and case B preferred lower costs instead of flexibility. As an example, case A said, ‘The additional costs for an expensive frame are hard
to save in the remaining supply chain. Therefore, it is preferable to have lower costs for a product and build up a slightly bigger stock’. Case C was conversely more in favour of
flexibility and delivery reliability: ‘Yes the costs are also important, but you have different
priorities. We want to have at first a very high delivery reliability and flexibility’. Therefore,
the costs seem to be more important in cases A and B. Whereas case C values a high flexibility ‘…due to the low pre-order quote’. Even though the costs were important for case B, the representative stated a restriction: ‘I would like to pay a bit more for a frame with
high quality’. Case A was not fully in favour of the costs and stated, ‘In the end, it is a mix out of both factors’, but the costs were slightly more important.
The supply complexity for the bike frame in case A and C is rated as a 6 and a 7 in case B. A value of 5 is the highest possible supply complexity. The supply processes for these companies can therefore be seen as very complex.
25 several objective measures of Gerschberger et al. (2017) to determine the supplier-induced complexity. First, the representatives explained how long it takes to let an already developed frame produce by a new supplier. The representative from case B provided an indication by saying, ‘You really have to consider one to two years’. The interviewee from case A stated, ‘I
think it would take a year to do that [switch the supplier]. The whole R&D work has to be done again’. In comparison, case A said, ‘We expect at least a time of six months to ensure the right quality’. It became clear from the interviews that all of the companies put a lot of
emphasis on getting the right quality from the suppliers. To ensure this, in case of a new supplier, the production process has to be redeveloped. The switch to a new frame supplier is thus lengthy and complex.
Second, the representatives provided insight on the reliability of the suppliers. Table 7 shows an estimated number of ‘perfect’ deliveries from the suppliers regarding quality, quantity, and time.
TABLE 7 Percentage of perfect deliveries
Case A Case B Case C
Percentage 50% 70% No identification
Especially in case A, the number of perfect deliveries is low, this increases complexity. Case B demonstrated that 90% of the deliveries do not have quality issues. The result of 70% is thus mainly triggered by quantity and time deviations. So, insufficient supplier reliability causes problems in companies and increases the complexity. Case C could not provide insights on the specific percentage but stated that ‘…reliability is very important’ to them.
Third, the geographical location of the suppliers, i.e. often abroad, was considered as a factor that increases complexity. To determine a level of complexity related to the
geographical location, we used the LPI data from 2016
26 cooperating with Chinese and Taiwanese countries seems to be of low complexity. However, working with companies from Vietnam and Cambodia can be more challenging and can be considered to have moderate complexity (LPI score between 3.4 and 2.5). Even though the LPI score may be high, e.g. for China, the representative of case B stated that ‘…everyone is a
bit different and you have cultural differences. That is also a challenge to find a good way to understand each other in the right way’. Because of this, the cultural differences must also
be considered when working with Asian suppliers. TABLE 8 LPI data about the countries of the frame-suppliers
Case A Case B Case C
Suppliers’ location LPI (Rank/ Score) Shipping (Days) Suppliers’ location LPI (Rank/ Score) Shipping (Days) Suppliers’ location LPI (Rank/ Score) Shipping (Days)
China 9/4.07 37 China 9/4.07 37 Asia */ * 35-40
Taiwan 25/3.70 39 Vietnam 64/2.98 35 Cambodia 73/2.80 35
The shipping time from the countries was determined with the help of Searates (https://www.searates.com/de/). A long shipping time increases complexity because it adds more time to the already existing lead time of the supplier. The shipping time is only the time it takes from the biggest harbour of the country to Hamburg (i.e. the main harbour for the three case companies).
In conclusion, the supply complexity of the frame suppliers from the conducted cases are very high. The scoring model of Gimenez et al. (2012) provides an objectified impression of the supply complexity, and the suggested categories of Gerschberger et al. (2017) adds further information regarding the supplier-induced complexity.
In the interviews, the representatives were asked about the relationship between supply complexity and SCI. In the following section, the outcomes are shown case-by-case.
The representative from case A stated, ‘We are following the approach of industry 4.0 and
trying to integrate as much as possible. We want to have real-time data…’. A high level of
27
delivery?’. This statement can be understood as a desire for a high delivery reliability, which
is underlined by continuous communication and tracing orders. Furthermore, the representative of A stated that ‘forecasting’ is their most important practice to counter long lead times: ‘The forecasts from our side are the basis to sensitise the supplier how much we
need at which point in time. And they take the forecasts very seriously’. With the help of the
forecasts, case A gives their supplier a base to prepare for their demand and order production material. The order of material in advance helps to reduce the lead times and also the supply complexity. Therefore, ‘…it should be the aim to reduce the lead-time’. From case B, it became clear that ‘The biggest challenge is, having the right goods, at the
right time and the right price’. Since the company values a long-term relationship and sticks
to its suppliers, the right price is a matter of negotiation. The right time and the right goods can be referred to as supply complexity. To handle that challenge, ‘the forecast is one of the
most important things to do’. The importance of the forecast can be underlined with the
following statement: ‘Without forecasts, it is not really possible anymore to order things or
work together with suppliers’. With the help of the forecasts, the suppliers prepare their
production facilities and order material to reduce the lead times. Besides the forecasting, the intensive communication is another important integration initiative as stated by the representative: “The daily work we do in our purchasing department is the tracing and
maintenance of delivery dates’.
Case C puts much emphasis in tracing the delivery dates to ensure delivery reliability. For case C, the main challenge is the following: ‘It is about the lead time and also about the
delivery reliability. And also that the supplier gets the time to prepare the materials and prepare themselves for our demand’. To let suppliers prepare themselves, ‘a really good forecast’ is provided and labelled as the most important part. Furthermore, ‘If no one from our side is tracking that [delivery lists], the delivery reliability gets worse’. The control of the
lists is done on a daily base, according to the interviewee.
All companies have joint R&D activities and a high degree of communication, including face-to-face meetings and conferences to make those activities as efficient as possible.
28 companies communicate often with their suppliers to trace the delivery dates and help to keep the delivery reliability high. With the continuous communication, the companies increase the safety for their own production and avoid missing components.
4.1.3 CODP and supply chain integration
In the interview, we asked the companies how they plan their production. Based on their responses, we determined their CODP in Table 9.
TABLE 9 CODP of the cases
Case A Case B Case C
MTS 100% 90% 100%
ATO 0% 10 % 0%
Case A uses a sales forecast and plans its production according to that. Since the company has a direct sales approach, it adapts its forecast as soon as it has real sales data. Case B sells bikes to retailers. The first pre-orders from the retailers arrive in September and October. The production planning has to be started earlier because of the long lead times. The company therefore plans production until a certain percentage before the retailers send the first orders. As soon as the company receives the first orders, it plans the whole production year and adjusts it according to the demand for the first orders.
Case C is similar to case B in that the company sells bikes to retailers and receives the first orders in September and October. Based on the orders, the production plan will be adjusted. All of the cases follow mainly an MTS approach. According to the representatives, a clear connection between the CODP and the level of integration is missing. Since the lead times are that long, the companies are forced to plan their production before real sales data are available: ‘The lead time is still the dominant factor. The lead time decides if you can make
the step from MTS to MTO’ (case A). Therefore, an MTO approach is not applicable without
long customer waiting times or a high stock of parts, which would be economically inefficient. In addition, as Case B has indicated, ‘For the customer, it is hard to understand
why he should wait five months for the bike’. For these reasons, no integration efforts are
29
4.1.4 Joint effect of supply complexity and CODP on supply chain integration
This section provides insights into the interplay of the two observed contingency factors. From the cases, a joint effect of the two factors was not observed as described in chapter 2. In the interviews, it became clear that the long lead time for the components as a dimension of supply complexity is the most challenging factor of the two observed contingency factors. All representatives stated that this has the largest effect on the level of SCI. These long lead times not only affect the level of SCI but also the CODP.
Moreover, the CODP is not just triggered by the lead time. In the interviews, the representatives claimed that in this retail industry, the customers do not want to wait for a long time for the bikes, as indicated in chapter 4.1.3. According to a representative of case B, it would be possible to produce MTO for ‘… smaller manufacturers with a deluxe bike brand
and a small number of bikes’. An MTO approach is thus not applicable for bike companies
with a larger number of bikes. The companies are forced to work in an MTS setting to shorten the time for the consumer to gain market share, and according to the representative of case B, ‘… the customer is able to buy it everywhere and in solid quality’. An MTO approach would only be possible if the companies store a lot of frames and take the financial risk. Furthermore, as case A indicated, the geographical location of the supplier is an additional important factor: ‘It means that the transportation time from Asia makes the
difference. If you would switch to the upcoming production facilities in Europe, that scenario [producing MTO] would become more realistic’. Considering all of the aforementioned
aspects, the relationship between the contingency factors can be seen in Figure 2.
30
FIGURE 2 Interplay of Industry and lead time
4.2 Secondary Data
4.2.1 Downstream integration efforts
This section strives to obtain insights on the downstream integration efforts of the focal companies. In contrast to the companies of the primary data, the focal companies of the secondary data are considered as suppliers, and we are therefore focussing on the downstream integration efforts to compare the primary data with the secondary data by looking at the same buyer-supplier relationships. The downstream integration efforts of all 13 companies can be found in Table 10.
Cases C, D, and E have the highest level of integration. Integration efforts can be found in all three sub-dimensions (Practices, patterns, and attitudes).
Cases A, B, F, G, and I have a moderate level of integration with the downstream supply chain partner. A wide variety of integration practices through all the dimensions can be found.
31
TABLE 10 Downstream integration efforts (Van Donk & Van Doorne, 2015: 2577-78)
Case Practices Patterns Attitudes Level of SCI
A Shared product identification
Shared forecast
Face-to-face meetings Reliable forecasts of the customer
2+1+1=4
B
Dedicated transport Daily contact Discuss pricing
Discuss yearly sales volume
likely they will form joint-venture
1+1+3=5
C
Shared packaging
Shared product identification
order entry through EDI
receive forecasts Face-to-face meetings Continuous communication Exchange of knowledge 4+2+1=7 D JIT frequent deliveries
tracking and tracing of reverse flow
online ordering system
customer complaint system
shared forecast
Face-to-face meetings IODM on R&D, price, promotion
6+1+3=10
E
Customised packaging
VMI
Shared product identification
Share information about: stock levels, productions plans, and sales forecasts
Face-to-face meetings IODM on R&D 6+1+1=8
F
Customised packaging
Shared product identification
Shared production plans
‘with some customers there is very frequent contact’ 3+1+0=4 G ‘Inter-organizational Kanban system’ customised packaging
shared product identification
IODM on production plans 3+0+1=4 H Customised packaging Shared forecasts 2+0+0=0 I • Joint investments • VMI
• ‘Access to ICT systems of some main buyers’ • Shared forecast
Frequent contact 4+1+0=5
J No integration 0+0+0=0
K Shared product identification
‘ICT integration’
Daily contact 2+1+0=3
32
4.2.2 Supply complexity and supply chain integration
This section aims to show the supply complexity per case and afterward describe its relation to the level of SCI. Table 11 shows the supply complexity factors per case.
TABLE 11 Supply complexity factors (Brolsma, 2012: 13)
Case Percentage MTS Lead time (Days) Number of products Sales
Volume Batch Size Order Winners
A 0 7 Customized 850 1 Quality, innovation, flexibility B 0 21 Unknown 5700000 125 Flexibility, Reliability
C 60 56 1500 2200 ton 5 tons (per
day) Flexibility, Lead time, Costs D 100 42 1500 Unknown 1000-infinite Flexibility, lead time E 15 28 6000 13000 Unknown Flexibility F 1 1 2000 33000 10 Speed of delivery and flexibility G 30 21 15000 100000 50 Product variety, Quality H 1 7 40000 100000 50 Reliability, Flexibility, Good relations I 1 1 23000 100000 3 Delivery reliability, flexibility J 5 10 Indefinite 100000 15 Flexibility K 0 12 1000 150000 22 tons Independent player, flexibility L 0 98 10000 100000 1 Quality, delivery time, flexibility
To analyse the data, in Table 12, the scoring of every case according to the coding model of Gimenez et al. (2012) can be found.
33 Differences can be found in the lead times. Cases A and J have significantly lower lead times than E and L.
TABLE 12 Scored supply complexity factors
Case Percentage
MTS Lead time
Volume/Variet y
Order
Winners Batch Size Total score
A 1 2 1 1 1 6 B 1 1 3* 1 2 8 C 2 1 2* 2 2* 9 D 3 1 2* 1 3 10 E 1 1 1 1 2* 6 F 1 3 1 1 1 7 G 2 1 1 2 1 7 H 1 2 1 1 1 6 I 1 3 1 1 1 7 J 1 2 1 1 1 6 K 1 2 2 1 2* 8 L 1 1 1 1 1 5
* Scores derived from the context of the cases
Cases A, E, H, J, and L are considered to have the highest complexity. All produce mainly on an MTO basis. Furthermore, their customers value their flexibility. The flexibility and the related supply complexity are also expressed by the small batches and the low volume-variety quotient, which stands for frequent product changes in the production process. Differences can be found in the lead times. Cases A and J have significantly lower lead times than E and L.
Cases B, F, G, I, and K have a moderate complexity. Across all companies, a wide variety of different complexity characteristics promotes the level of supply complexity. The complexity in cases B and I is induced more by the lead time than it is in the remaining cases. Furthermore, case G is the only company in this category where the customers do not mainly value the flexibility.
Cases C and D have the lowest complexity. Both cases have a relatively high volume-variety quotient and a high batch size in common. Even though the overall supply complexity is low, both cases have long lead-times.
34
FIGURE 3 Relation between supply complexity and SCI (secondary data)
A clear pattern cannot be derived from this figure. For instance, cases C and D, with a low level of supply complexity, have a high level of integration. In contrast, case J does not integrate at all with the buyer but has a moderate supply complexity.
After not finding a clear connection between complexity and SCI, the single supply complexity dimensions were set in relation to the level of SCI. By observing all five supply complexity dimensions, it seems that the level of SCI is influenced by the lead times, as is demonstrated in Figure 4.
FIGURE 4 Relation between lead time and SCI
0 2 4 6 8 10 12 0 2 4 6 8 10 12 A B C D E F G H I J K L Level o f su p p ly c o m p le xi ty Level o f SCI Cases
Level of SCI Supply complexity
0 10 20 30 40 50 60 70 80 0 2 4 6 8 10 12 A B C D E F G H I J K L Lead t im e ( D ay s) Level o f SCI Cases
35 It can be seen that cases B, C, D, and E with long lead-times also have a high level of integration. All have regular daily contact, either on a daily basis, face-to-face meetings, or both. Furthermore, besides case B, the others receive and share forecasts with their customers. The results for case B can be explained by the fact that rough information is still provided by the buyers, which replaces the need for forecast in that case.
Case L has, with 98 days, the highest lead time of all companies but a low level of integration. This company takes 84 of the 98 days to receive the material. In addition, the products are built especially for one customer. The customer is therefore aware of the lead time and by knowing the time to get the material, the lead-time for the buyer is also predictable.
The remaining cases of A, F, G, H, J, and K have lower lead times and also a lower level of SCI. It is thus demonstrated that there is not only a positive relation between long lead times and a high level of integration but also between low lead times and a low level of integration. The activities around VMI (Investments, access to ICT) of case I explain their somewhat higher level of integration.
36
5. DISCUSSION AND CONCLUSION
Discussion
The possible contradicting results in literature about the single effect of the CODP and supply complexity on the level of SCI, and if their might be a joint effect were the starting points for this research. The findings of this study do not confirm all the results of two main previous studies that triggered this research of Gimenez et al. (2012) and Van Donk and Van Doorne (2015). The analysis of the primary and secondary data gave partly mixed and contradicting results. In the following section, the results of the three investigated relationships, (1) supply complexity and SCI, (2) CODP and SCI, and (3) the joint effect of both supply complexity and CODP on SCI are discussed.
First, the study investigated the relation between supply complexity and the level of SCI. The case research conducted confirmed the findings of Gimenez et al. (2012) and Van Donk and Van der Vaart (2004) that buying companies, which have to face a high level of supply complexity, integrate on a high level with their upstream supply chain partners. In addition to these studies, I found in my study that all the companies stated that lead time –one dimension of supply complexity- is the most challenging contingency factor in planning their production. Further, I found that they use sharing forecasting models as main way to integrate with their suppliers and to reduce uncertainty for their supplier. The forecasts are used to help the suppliers order material and plan their production capacity. In addition, the suppliers provide forecasts about their planned production to keep the buyer informed. Furthermore, intensive communication using different media (face-to-face, conferences, and daily contact) helps to reduce uncertainty and provides a good opportunity to keep track of the deliveries and guarantee delivery reliability.
37 helps both sides to plan their business (production, sales, et cetera) and be constantly informed. All in all, a clear relation between high supply complexity and a high integration could not be proved, but long lead times are a factor that promotes the integration with supply chain partners. Lead time was already mentioned by Manuj and Sahin (2011) as a main driver for supply complexity. Furthermore, Droge, Jayaram, and Shawnee (2004) stated that closer collaboration with the supplier could reduce the lead time significantly. This finding is in line with findings of my study that the exchange of forecasts reduces lead times, as it provides the opportunity for the suppliers to order material and prepare the production facilities in advance. According to Hill, Doran and Stratton (2012) exchanging forecasting is a viable method to reduce uncertainty (complexity) if inventory buffers are not feasible. Inventory buffers would be too costly in the cases from my case research and forecasting is therefore the appropriate method to reduce lead-times and uncertainty. Moreover, to make the demand visible for the upstream supply partners, Arshinder, Kanda, and Deshmukh (2008) recommend, besides IT integration and SC contracts, the use of intensive communication as it also executed in the observed cases.
38 (Gunasekaran, Lai & Edwincheng, 2008). The three cases display close partnerships with their suppliers, which include trust and long-term relationships. In long-term relationships, intensive and rich communication methods, as face-to-face meetings, are commonly used (Ambrose, Marshall, Fynes & Lynch, 2008). Furthermore, our case study supports the findings of Forslund and Jonsson (2007) that for MTO companies (supplying unit), receiving forecasts is important when planning the production. These are all explanations for the high level of integration in my three cases, besides the findings about the effect of the lead time on the level of SCI. It seems that the factors of strategical importance, joint R&D process, long lead times, and long-term relationships have a more dominant and positive influence on the level of SCI than the CODP.
Third, the expected contradictions derived from the results of Gimenez et al. (2012) and Van Donk and Van Doorne (2015) were supported by this study. From the case study, it became clear that the CODP does not affect the level of SCI. The representatives of the companies stated that the supply complexity and especially the long lead times affect the integration efforts most. Oppositely, I found that the long lead times do not only affect the integration efforts but also influence the location of the CODP. According to Olhager (2010), in MTO environments, it is assumed that material must be available at the CODP to meet delivery promises. In the cases from my study, this is not possible because the lead times are long and stock buffers would be too expensive. Furthermore, If the delivery time for the customer must be shorter than the time a company needs to receive the material and produce the product, the companies are forced to produce MTS (Olhager, 2003). The kind of CODP is therefore determined by the lead time and delivery time to the customer. The focal companies operate in the retail industry, and their customers expect the products to be available as soon as possible. That makes it impossible for the focal companies to produce the majority of their bikes on an MTO/ATO basis. In the end, lead time (supply complexity) has an effect on the level of SCI, but a joint effect could not be proven.
Limitations
39 different industries. Furthermore, cases A and B of my own case research were based on single interviews and documents. The fact that they were completed with senior managerial interviewees makes these data more valuable, because the managers have long-term experience. In addition, the determination of the supply complexity of the supplying unit was done from the perspective of the focal companies. Even though the numbers are based on the experience of the representatives, the numbers are reliable, as the collaboration has been close for a long time, and due to the joint R&D process, the focal companies have insights in the production processes of their suppliers.
The secondary data set also provided some limitations. Since the data were not gathered to investigate the interplay of two different contingency factors, no specific questions were asked about that topic. Therefore, in-depth knowledge concerning that topic was not provided.
In general, the supply complexity scoring method that we applied tends to be simple and neglects factors such as uncertainty and geographical location of the supplier. These factors were not analysed in the secondary data and could reveal a different level of supply complexity per case. That could lead to a different outcome regarding the relation between supply complexity and SCI.
Implications for future research
Considering the aforementioned limitations, a suggestion for future studies would be to research a compressive supply complexity scoring method that incorporates objective and subjective measures. This would provide a more valid base to determine and compare the level of supply complexity. An additional avenue for further research would be to connect the outcomes of this study to performance. In particular, it would be interesting to show how forecasting and intensive communication could reduce the lead time and increase delivery reliability. Including the performance would have exceeded the scope of this paper but would certainly bring additional insights on how effective some integration efforts are.
Managerial implications
40 managers are aware of individual contingency factors they have to face and what literature suggests counteracting them. For instance, this study reveals that forecasts and intensive communication help to counter long lead times of the suppliers. Integration efforts such as these should be considered by the managers and evaluated to determine if they are appropriate for their purposes.
Conclusion
The results of this study mainly contradict those of previous publications. A clear influence of each of the individual contingency factors, CODP and supply complexity, on SCI could not be proven. In addition, a joint effect on the level of SCI does not exist. Still, an interesting finding of the study is that long lead times as a dimension of supply complexity have a dominant and positive influence on the level of integration. The dimension “lead time” showed significant influence on the level of SCI in both data sets.
41
REFERENCES
Aitken, J., Childerhouse, P., & Towill, D. (2003). The impact of product life cycle on supply chain strategy. International Journal of Production Economics, 85(2), 127–140.
Alfalla-Luque, R., Medina-Lopez, C., & Dey, P. K. (2013). Supply chain integration framework using literature review. Production Planning & Control, 24(8-9), 800–817.
Ambrose, E., Marshall, D., Fynes, B., & Lynch, D. (2008). Communication media selection in buyer‐supplier relationships. International Journal of Operations & Production
Management, 28(4), 360–379.
Arshinder, Kanda, A., & Deshmukh, S. G. (2008). Supply chain coordination: Perspectives, empirical studies and research directions. International Journal of Production Economics,
115(2), 316–335.
Blome, C., Schoenherr, T., & Eckstein, D. (2014). The impact of knowledge transfer and complexity on supply chain flexibility: A knowledge-based view. International Journal of
Production Economics, 147, 307–316.
Bozarth, C., Warsing, D., Flynn, B., & Flynn, E. J. (2009). The impact of supply chain
complexity on manufacturing plant performance. Journal of Operations Management,
27(1), 78–93.
Brolsma, J. (2012). Supply chain integration and reverse logistics performance: The
moderating effect of supply complexity. MSc Thesis, University of Groningen, Groningen.
Childerhouse, P., Aitken, J., & Towill, D. R. (2002). Analysis and design of focused demand chains. Journal of Operations Management, 20(6), 675–689.
Das, A., Narasimhan, R., & Talluri, S. (2006). Supplier integration—Finding an optimal configuration. Journal of Operations Management, 24(5), 563–582.
Droge, C., Jayaram, J., & Vickery, S. K. (2004). The effects of internal versus external
integration practices on time-based performance and overall firm performance. Journal of
Operations Management, 22(6), 557–573.
Eisenhardt, K., & Graebner, M. (2007). Theory building from cases: Opportunities and challenges. Academy of Management Journal, 50(1), 25–32.
Fabbe‐Costes, N., & Jahre, M. (2008). Supply chain integration and performance: a review of the evidence. The International Journal of Logistics Management, 19(2), 130–154.
Flynn, B. B., Huo, B., & Zhao, X. (2010). The impact of supply chain integration on performance: A contingency and configuration approach. Journal of Operations
Management, 28(1), 58–71.
Forslund, H., & Jonsson, P. (2007). The impact of forecast information quality on supply chain performance. International Journal of Operations & Production Management, 27(1), 90– 107.
