Do contingencies matter? The effect of process and discrete industry characteristics on Supply Chain Resilience strategies: A multiple case study

(1)

Faculty of Economics and Business

MSc Thesis Supply Chain Management

Do contingencies matter? The effect of process and discrete industry characteristics on

Supply Chain Resilience strategies: A multiple case study

By

Sam van Huet

Student number: S3541916

Email: s.b.h.h.van.huet@student.rug.nl

1

st

Supervisor

Prof. Dr. D.P. van Donk

2

nd

_Supervisor

H. Dittfeld MSc

Co-assessor

Prof. Dr. J.T. van der Vaart

Word Count: 10534

(2)

2

Abstract

Purpose: This study addresses the effect of process and discrete industry characteristics on supply chain

resilience (SCRES) strategies.

Method/Design: An in-depth exploratory multiple case study was carried out across several industries

consisting of 8 companies. 21 semi-structured interviews were executed. Data was analyzed by using a deductive coding approach.

Findings: First, this study shows the importance of categorizing manufacturing companies based on

their industry characteristics. The case findings provide evidence that specific discrete and process industry characteristics could influence SCRES strategies. Variable raw material quality and Make-To-Order (MTO) highly influence collaboration and agility. In addition, trade-offs such as agility versus redundancy in resilience practices are presented. In discrete industries, the level of agility is high versus a low level of redundancy whereas in process industries it is the opposite.

Originality/value: This is one of the first studies to explore the effect of process and discrete

characteristics on SCRES. It provides a contingency perspective to show the relation of specific characteristics on SCRES within several industries.

Keywords: Supply chain resilience, process industries, discrete industries

Acknowledgments

(3)

3

1. Introduction

(5)

5 strategy since the process often cannot be decoupled at an intermediate stage (Van Hoek, 2001). In other operations management research, the difference in process and discrete characteristics already have proven to be significant, such as in the application of lean and in production planning and control (Crama, Pochet, & Wera, 2001; Fransoo & Rutten, 1994; Panwar, Nepal, Jain, & Rathore, 2015). In addition, in terms of sales and operations planning, process industries also have different effects compared to discrete industries (Ivert et al., 2015). According to these examples, it could be noticed that differences in process and discrete industries matter, yet it is unknown how this influences SCRES strategies.

This leads to the following research question: How do process and discrete industry characteristics influence supply chain resilience (SCRES) strategies?

This research will be build up from the strategies identified by Tukamuhabwa, Stevenson, Busby, & Zorzini (2015). The industry-specific characteristics will be distinguished into process and discrete production (Dennis & Meredith, 2000; Müller & Oehm, 2019).

The paper will conduct a multiple case study to explore how different production contexts influence SCRES strategies. The research is qualitative in nature and will be executed by conducting semi-structured interviews at companies across different industries. This study makes three key contributions to SCRES literature. First, it is clear which strategies and practices of SCRES are related to a production context. In addition, the amount and degree of practices will be shown from process and discrete industry perspectives and will shed light on the importance of context-specific characteristics within SCRES. Therefore, this study also contributes to providing a contingency perspective that can be applied in other debates such as the application of Lean practices in different contexts (Chavez, Gimenez, Fynes, Wiengarten, & Yu, 2013). It also contributes to other discussions, such as the role of collaboration and vulnerability in SCRES which also could be different in other contexts (Scholten & Schilder, 2015; Wagner & Neshat, 2012). Moreover, the findings of this study will assist supply chain managers in approaching different strategies according to their own context. Indeed, elements as product life cycles, supply chain characteristics, demand uncertainties, and other industry specifics could influence organizations’ decisions to cope with disruptions. Moreover, today’s fast-changing business environments and the complexity of supply chains requires to apply the right strategy quickly (Bakshi & Kleindorfer, 2009). Therefore, it is important to know the value of specific trade-offs in resilience within certain contexts.

(6)

6

2. Literature

The first section of the literature will focus on supply chain resilience which will be distinguished in four strategies. The next section will elaborate on the differences between process and discrete industry characteristics. The last section concludes this chapter by discussing resilience in each industry.

2.1 Supply chain resilience (SCRES)

Nowadays companies face the challenge to work together more closely than ever due to outsourcing, globalization, and customization (Ponomarov & Holcomb, 2009). This made supply chains more efficient during stable business environments but, in turn, resulted in supply chains being more vulnerable to disruptions (Kamalahmadi & Parast, 2016; Tang & Tomlin, 2008). Supply chain disruptions are unplanned and unanticipated events that disturb the movement of materials, goods and services (Craighead, Blackhurst, Rungtusanatham, & Handfield, 2007).

While there is quite a lot of literature on supply chain resilience, there is no consensus on a formal definition (Dubey et al., 2019). In this study the following definition will be used since it is focused on the SCRES strategies: “the adaptive capability of the supply chain to prepare for unexpected events, respond to disruptions, and recover from them by maintaining continuity of operations at the desired level of connectedness and control over structure and function” (Ponomarov & Holcomb, 2009, p. 131). This definition consists of all the phases of resilience that are related to the SCRES strategies used in this research. Disruptions are inevitable in the volatile business environment of today (Pettit, Croxton, & Fiksel, 2019). Therefore, it is critical that organizations build this adaptive capability to deal with disruptions (Pettit, Croxton, & Fiksel, 2013; Scholten, Scott, & Fynes, 2014).

2.2 SCRES strategies

Tukamuhabwa, Stevenson, Busby, & Zorzini (2015) made a categorization of all the SCRES strategies and summarized four key strategies based on 91 data sources namely: flexibility, redundancy, collaboration and agility. However, Scholten & Schilder (2015) state that flexibility, velocity, visibility and collaboration are the key strategies. Though, according to Tukamuhabwa et al. (2015), velocity and visibility are parts of agility. This research will elaborate on the strategies of Tukamuhabwa et al. (2015) since these are the most cited in the literature and seem to be the most consistent according to their extensive literature review. This is in line with other papers where these strategies were perceived as the main strategies for SCRES (Jüttner & Maklan, 2011; Ponis & Koronis, 2012). Table 1 presents an overview of the SCRES strategies, definitions and corresponding practices.

Strategy Definition Practices

Redundancy “Redundancy involves the

strategic and selective use of spare capacity and inventory that can be invoked during a crisis to cope, e.g. with supply shortages

(7)

7 Table 1: Overview of strategies and corresponding practices (based on Ali et al., 2017)

Redundancy

Adopting redundancy practices can improve the performance of an organization when it is in a changing and complex business environment (Kamalahmadi & Parast, 2017). Implementing safety stock, having multiple suppliers and maintaining capacity to respond before a disruption occurs are examples of redundancy practices (Scholten & Schilder, 2015; Tang & Tomlin, 2008). Another example is when focal companies protect their main suppliers by creating capacity in their operations (Sawik, 2013). Moreover, organizations could also use a back-up supplier in case a primary supplier is disrupted which continues the flow of materials (Sodhi & Lee, 2007). Many researchers focused on the relative importance of redundancy versus flexibility and which of the two should be emphasized more in a certain situation (Christopher & Peck, 2004; Ponomarov & Holcomb, 2009; Zsidisin & Wagner, 2010). According to Kristianto, Gunasekaran, Helo, & Hao (2014), redundancy can also lead to being more flexible and therefore increasing SCRES.

Collaboration

According to Scholten, Scott, & Fynes (2014), collaboration is an important strategy that contributes to resilience. Supply chain collaboration refers to the ability to work together with other players in the supply chain to effectively create mutual benefits such as collaborative planning and joint knowledge creation (Cao et al., 2010; Pettit et al., 2013). Information sharing is a key element of collaboration (Scholten & Schilder, 2015). It refers to engaging in collaborative activities such as forecast-sharing and resources-sharing (Cao et al., 2010). For example, companies could implement information technology

or demand surges”

(Tukamuhabwa et al., 2015, p. 5604).

Collaboration “Ability to work effectively with other entities for mutual benefit” (Pettit, Croxton, & Fiksel, 2013, p. 49).

Information-sharing, resource-sharing, forecast-sharing (Cao, Vonderembse, Zhang, & Ragu-Nathan, 2010; Pettit et al., 2013; Scholten & Schilder, 2015)

Flexibility “The ability to take different

positions to better respond to abnormal situations and rapidly adapt to significant changes in the supply chain” (Kamalahmadi & Parast, 2016, p. 122).

Postponement in production, flexible suppliers, flexible manufacturing processes (Colicchia, Dallari, & Melacini, 2010; Tang & Tomlin, 2008)

Agility “The ability of a supply chain to

rapidly respond to change by

adapting its initial stable

configuration” (Wieland &

Wallenburg, 2012, p. 302).

(8)

8 systems that create connectivity in their supply chain which can help to control their resources during disruptions (Tukamuhabwa et al., 2015).

Flexibility

According to Ponomarov & Holcomb (2009), a supply chain is flexible when it can change quickly and can recover fast after it experienced a disruption. Tukamuhabwa et al. (2015) state that flexibility is the ability to adapt with minimum time and effort in case of a disruption. Examples of flexibility that can improve resilience are a flexible supply base, having flexible transportation systems, flexible production facilities and having flexible product strategies via postponement (Tang & Tomlin, 2008). For example, flexibility in sourcing creates the ability for an organization to quickly change inputs or the mode of receiving inputs (Pettit et al., 2013). In addition, flexible decisions facilitate adjusting to disruptions rather than withstanding it (Wallace & Choi, 2011). It also enables organizations to reallocate labor and transportation resources quickly (Pettit et al., 2013).

Agility

According to Wieland & Wallenburg (2012), agility is the ability to rapidly respond to change by adapting its stable configuration. Wieland & Wallenburg (2013) state that visibility and speed are dimensions of agility. Speed or velocity are related to the pace of the adaptions and determines the recovery time from a disruption (Jüttner & Maklan, 2011). Zara is a good example of an agile supply chain that seeks for short delivery times while facing unpredictable demand (Christopher, 2000). Agile supply chains have a clear view throughout their whole supply chain which enables detecting signals of potential disturbances (Tukamuhabwa et al., 2015). Moreover, agility could contribute to identifying vulnerable suppliers and thereby helping to develop countermeasures to tackle potential disruptions (Jüttner & Maklan, 2011). However, the exact boundaries of agility and flexibility remain unclear in the literature (Purvis, Spall, Naim, & Spiegler, 2016; Tukamuhabwa et al., 2015). In this research, the speed and visibility dimensions will be considered as agility. Practices such as monitoring suppliers, action plans and crisis teams are examples to increase speed and visibility (Knemeyer et al., 2009; Sheffi et al., 2005; Wagner & Neshat, 2012).

2.3 Process and discrete industries

(9)

9 Discrete processing plants are systems that use a combination of machines that work almost independently and are designed for a specific task (Müller & Oehm, 2019). It often produces countable, distinguishable products that are often assembled by the use of bill of materials (Lyons, Vidamour, Jain, & Sutherland, 2013). In these industries, value is mainly added by direct labor and one of the goals is to reduce the work-in-process. Often, discrete manufacturers are focusing on Just-In-Time (JIT) production and therefore material planning is critical (Crama et al., 2001). It is therefore crucial for these manufacturers that all suppliers meet delivery times. Table 2 below presents an overview of the key differences between these industries.

Process industries Discrete industries Relationship with the market

Product type Commodity Custom

Product assortment Narrow Broad

Demand per product High Low

Cost per product Low High

Order winners Price, delivery guarantee Speed of delivery, product

features

Product and product

structure

Frequent shelf-life constraints Limited shelf-life constraints

Manufacturing processes Variable material grade Predictable material grade

Planning & control

Production To stock To order

Long term planning Capacity Product design

Short term planning Utilization capacity Utilization personnel

Starting point planning Availability capacity Availability material

Material flow Divergent + convergent Convergent

Yield variability Sometimes high Mostly low

Pausing production Problematic (losses of the

whole product)

Possible

Revising faults Hardly possible, losses of the

entire product

Possible

Table 2: A comparison of process and discrete industry characteristics (adapted from: Crama et al., 2001; Müller & Oehm, 2019)

(10)

10 for example in semi-process industries (Pool, Wijngaard, & Van Der Zee, 2011). For process industries, perishability, divergent material flow, perishability and the inability to stop production are expected to influence SCRES the most. These factors have also been considered as critical in other phenomena such as the impact on the decoupling point and lean (Van Donk, 2001; Lyons et al., 2013). In case of discrete industries, MTO, convergent material flow, the ability to pause production and material focused planning are expected to be the most critical (Müller & Oehm, 2019; Panwar et al., 2015).

2.4 Supply chain resilience in different industries

The aim of this research is to see how process and discrete industry characteristics influence SCRES strategies. Figure 1 shows the conceptual model which depicts the way of production as independent variables. The framework captures the most important aspects of both process and discrete industries. Expected is that the eight presented factors of the different industries will have the most impact on SCRES strategies. The main question is how these factors influence SCRES strategies and therefore is the dependent variable which includes the four key SCRES strategies. Expected is that SCRES strategies will differ in the process and discrete industries and it could be the case that there are certain boundaries for each SCRES strategy in a certain context. The following section will try to combine the most critical factors of process and discrete industries with the SCRES strategies.

SCRES in process industries

(11)

11 Boekesteijn (2014), safety stock, flexible contracts, portfolio diversification and transportation planning are approaches to enhance resilience in oil and gas supply chains which are typical process industries. SCRES in discrete industries

Discrete industries often are characterized by convergent production systems where all the raw materials, subassemblies and parts will be merged together into one unit (Crama et al., 2001). The diversity of the end products is often high since most of the products are made to order based on specific customer requirements (Müller & Oehm, 2019). Since discrete manufacturers typically apply JIT-production, it is critical that the manufacturer seeks collaborative activities with all the different suppliers to assure on-time delivery (Cao et al., 2010). For manufacturers in this category, it is easier to use the same capacity for a wider range of products. For example, Samsung has a multi-platform system in its manufacturing plants which allows for switching between producing different products at low cost. This could be used in the case of supplier delays as a result of fluctuating market demands (Sodhi & Lee, 2007). As shown in table 2, pausing production is therefore less problematic than in processing industries since these production processes do not involve mixing of ingredients as a part of a continuous process (Müller & Oehm, 2019). This is in line with other researches which examined typical industries such as fashion and textile supply chains where strategic and operational flexibility are crucial for building resilience (Pal, Torstensson, & Mattila, 2014). According to Azevedo, Govindan, Carvalho, & Cruz-machado (2013), a flexible supply base and total supply chain visibility enhance resilience in automotive supply chains which are typical discrete industries. Moreover, Sodhi & Lee (2007) stress that implementing information technology (IT), having low inventories and redundant suppliers are critical in the electronics industry.

(12)

12

3. Methodology

This section will describe the research design, research setting, data collecting and the method of data analysis to conduct valid and reliable research.

3.1 Research design

This research aims to gain insight into how discrete and process context environments affect SCRES strategies. It remains unknown how this relationship works so in line with the research question of this study, an explorative multi-case study will be executed since context-specific SCRES strategies is a relatively new area of research (Eisenhardt & Graebner, 2007). A multiple case-study is chosen since it has advantages over a single case study because of the ability to generalize and compare different results (Gerring & McDermott, 2007). This method of research has been applied earlier in the context of SCRES to explore certain phenomena (Scholten & Schilder, 2015). In addition, multiple cases are selected since it increases external validity. The various SCRES strategies will be examined in their specific context. Multiple-cases are necessary in this study to explore the differences between the cases in relation to the context. The explorative case study is considered as appropriate for this study since it will gain an in-depth insight into a unique phenomenon where the focus is often on how, why and what questions (Yin & Robert, 2009). This way it can compare the SCRES strategies in the different production environments.

The unit of analysis is at firm level within the supply chain. This unit of analysis provides the opportunity to examine individual businesses and will be used to later compare different cases. Considering redundancy as part of SCRES, this research will only focus on the supply side of inventories. More specifically, it will focus on the raw materials dedicated to production and not focus on the finished product inventory.

3.2 Case setting and selection

(13)

13 perishability of the raw materials (Van Donk, 2001). Moreover, material quality could be different and supply and demand markets typically are volatile (Donadoni et al., 2019). Thirdly, two manufacturing plants in the chemical industry are selected since it is a typical process industry and often is a continuous manufacturing process where it is not possible to stop production. Lastly, to cover a wider spectrum of process industries, a packaging manufacturer has been selected that mass produces glass and metal packaging. Moreover, a manufacturer of hygiene and health products has been selected since the fast-moving consumer good (FMCG) industry often involves high volumes that are fast-moving through the supply chain at a high pace with makes it vulnerable to disruptions (Scholten & Schilder, 2015).

Table 3: Case selection criteria

This case selection allowed for theoretical replication which implies that contrary results will be generated for predictable reasons (Pandit, 1996). For example, it could be that companies in process industries could be more constrained in building redundancy compared to discrete manufacturers.

3.3 Data collection

The main source of data in this study are 19 in-depth semi-structured interviews collected in November 2019. By conducting semi-structured interviews, relatively complete capture of the complexity and essentials can be examined, while also keep room for some degree of flexibility (Wilson, 2014; Yin, 1984). It this research, it means that underlying mechanisms between industry characteristics and resilience practices could be related. For each case, multiple individual face-to-face interviews were conducted to be able to triangulate the answers. Interviews were arranged by the researcher and three other fellow researchers that examined the same theme. The researchers made use of their own network to contact the companies. The interviewees were selected based on their experience with disruptions. Moreover, since this research is associated with resilience strategies, managers of Operations, Supply Chain and Production were approached to make sure the interviewees had knowledge on this subject. Table 4 provides an overview of the interviewees.

Discrete Process Food Automotive Chemical Perishability

A: Heavy vehicle manufacturer X  

B: Heavy vehicle manufacturer Y  

C: Health & hygiene production 

D: Chemical manufacturer X  

E: Packaging manufacturer 

F: Food processing   

G: Food & raw material

processer   

H: Chemical manufacturer Y  

(14)

14

Interview protocol development

Based on the literature, an interview protocol (Appendix A) is developed to ensure reliable and valid research (Yin, 1984). The interview protocol was based on the core concepts of the research based on the literature, with open-ended questions and probes to obtain detailed responses. The structure of resilience dimensions is based on earlier work of Tukamuhabwa et al. (2015). Important aspects of process and discrete characteristics are related to findings of Crama et al. (2001); Van Donk (2001); Müller & Oehm (2019).

Interview execution

Each interview started with general questions about the company and the interviewee where after SCRES was introduced. We started with open questions on their experience with previous disruptions to let them explain how they respond or were prepared in that situation. To increase internal validity, two researchers conducted the interviews (Eisenhardt, 1989). Prior to the interview, the interviewees were asked to read and sign the consent form. Moreover, they were informed about the aim of the research and background. They also were asked for permission to record the interview. We often used the critical incident method to zoom in on previous disruptions in order to get detailed information about a specific phenomenon in the company (Flanagan, 1954). In the end, we checked if all the process and discrete characteristics were covered. Otherwise, we asked for these characteristics separately.

Industry and case name

Type of industry Person Respondent position Length

A. Automotive (Vehicle manufacturer X)

Discrete Interviewee A1 Director Logistic Operations 40:48

Interviewee A2 Manager Innovation & Improvement Operations

49:37

B. Automotive (Vehicle manufacturer Y)

Discrete Interviewee B1 Manager Material supply engineering

50:21

Interviewee B2 Supplier Quality Manager 61:21 Interviewee B3 Manager Logistics Supply 59:51 C. Chemical

(Chemical manufacturer Y)

Process Interviewee C1 Director Technology 61:59

Interviewee C2 Planning & Logistics Scheduler EMEA

52:44

Interviewee C3 Regional Supply Chain Manager (Same interview) Interviewee C4 Supply Chain Director North &

South America

(15)

15 D. Chemical

(Chemical manufacturer X)

Process Interviewee D1 Supply Chain Planner 67:11

Interviewee D2 Supply Chain Planner 45:16 E. Food (Food

processing)

Process Interviewee E1 Manager corporate Demand & Supply

55:48

Interviewee E2 Supply Chain Planner 34:05

Interviewee E3 Manager Logistics 36:48

Interviewee E4 Transport Manager 42:16

F. Packaging (Packaging manufacturer)

Process Interviewee F1 Supply Chain Manager Specialties

47:08

Interviewee F2 Supply Chain Manager N&C 45:16 G. Health & hygiene

(Health & hygiene production)

Process Interviewee G1 Sourcing Director NL 48:30

Interviewee G2 Manager Logistics Planner 46:35 H. Food (Food & raw

material processer)

Process Interviewee H1 Manager Supply Chain Management

55:11

Interviewee H2 Category Buyer Procurement 46:53 Table 4: List of cases and corresponding interviewees

To increase the construct validity of this research, the interpretation of the results was sent back to the interviewee to ask for their interpretation (Yin & Robert, 2009).

3.4 Data analysis

(16)

16

1st order code 2nd order code 3rd order code Dimension Concept

"In our contracts with suppliers, we have flexibility, which means we do not include a standard purchase amount".

Flexible order amounts

Flexible partners

"We freeze our production schedule for 2 weeks, everything is already ordered then so it is not ideal to change the plan on the short term".

Very hard to change

production plan Flexible production process

“We mainly have a single sourcing strategy since we have 500 different suppliers for which we can’t have a second one. We also try to have the best suppliers and

it would be too expensive to have a back-up one”.

Mainly single sourcing

"We have about 1500-2000 suppliers". Many suppliers

“We send automatic orders via EDI and the supplier has to respond within 24 hours if they can handle the

forecast of the next twenty days”.

Sharing forecasts

Information sharing

"For our most important suppliers we have something that is called QBR (Quarterly Business Review), so we come together every quarter to talk about how it goes

with their business".

Resource sharing

Supplier development

"For several scenarios there are emergency plans". Action plans Responsiveness "So, this is our approach: trying to monitor early in the

supply chain to see an issue before it comes up".

Monitoring suppliers

Visibility

"The parts are delivered Just-In-Time, every item will be delivered based on the production sequence".

Just in time

"It is attractive to always have our assets run at maximum capacity". High utilization Flexibility Agility Amount of suppliers Discrete Process Redundancy Collaboration SCRES

Table 5: Excerpt coding tree

After the coding process, all cases have been rated based on the discrete and process characteristics. For each corresponding process or discrete characteristic, one point was assigned to each case. The left side of each manufacturing aspect corresponds to discrete industries whereas the right aspect is related to process industries. Based on this analysis, 3 clusters of similar cases are created since it would decrease clarity to cover all cases separately. Cases A and B consist of only discrete industry characteristics so these are assigned to cluster 1. Case C, D and E consist of both characteristics and are almost even in process and discrete characteristics. Moreover, these companies are not associated with perishability and variable raw material quality and therefore are assigned to cluster 2. Case F, G and H mainly consist of process industry characteristics. Case F and G are associated with perishability and variable raw material quality and therefore are assigned to cluster 3. Moreover, case H is assigned to cluster 3 since this company is connected to suppliers and customers by pipelines which in this research is associated with a process industry characteristic.

(17)

17

Resilience A: Heavy vehicle

manufacturer X

D: Chemical

manufacturer X F: Food processing

• Fl exi bl e order a mounts

• Fl exi bl e pa rtners for converting end product

• Fl exi bl e tra ns port pa rtners

• Pa rts rel a ted to s peci fi c end product, not a bl e to s wi tch or pos tpone • Fi xed production proces s • Fi xed production proces s • Suppl i er hol ds i nventory cl os e to fa ctory • Hi gh a mount of

uni que s upl i ers

• Sa fety s tocks • Stock ba s ed on i mportance of cus tomer • Low i nventory • Si ngl e s ourci ng, dependent on 2 ra w ma teri a l s uppl i ers

• Dua l s ourci ng for cri tica l products • EDI s ys tems , s ha re

forca s ts • devel opi ng s uppl i ers

a nd hi gh res ource

s ha ri ng • Da i l y contact wi th

s uppl i ers wi th EDI s ys tems

• Sha ri ng foreca s ts a nd production pl a n

• Da i l y contact wi th fa rmers • Col l a bora tive s ys tem wi th di s tri bution pa rtner a nd fa rmer • Sha ri ng foreca s ts wi th fa rmers • No res ource s ha ri ng • No s ys tem to s ha re foreca s ts wi th other pa rtners • Moni tori ng s uppl i ers , depa rtment

for qua l i ty • Acting fa s t on

di s ruptions , often before i t i s

probl ema tic • Scena ri o's a nd

a ction pl a ns

• In ca s e of a di s ruption, cus tomer a l l oca tion pri ori ty

• Moni tori ng fa rmers for qua l i ty of ra w ma teri a l • Ba ck-up pl a ns onl y for tra ns port

• No s uppl i er

moni tori ng • No proa ctive a ction

pl a ns

• In ca s e of a di s ruption, cus tomer a l l oca tion pri ori ty Fl exi bi l i ty

Redunda ncy

Col l a bora tion

Agi l i ty

Table 6: Excerpt resilience aspects per case

(18)

18

 Normal information sharing

 Frequent information sharing

 Daily information sharing

A: Heavy vehicle manufacturer X Dicrete: 7 Process: 0 D: Chemical manufacturer X Dicrete: 3 Process: 4 F: Food processing Dicrete: 2 Process: 5

Flexible order amounts   

Postponement in Excess capacity in

production 

Multiple suppliers (for

same product)  Safety stock   Information-sharing    Resource-sharing   Forecast-sharing    Monitoring suppliers   Action plans  Crisis teams  Collaboration Agility Resilience aspects Flexibility Redundancy Low High Medium Low High Low Medium Low Medium Medium Medium High

Table 7: Example score procedure on resilience

For each practice, one, two or three checkmarks are assigned based on the degree or frequency of the practice. An example of information sharing is shown below in table 8.

Table 8: Example classification information sharing

(19)

19

4. Results

(20)

20

Cluster Case MTO MTS Convergent Divergent Labour

intensive Machine intensive Not perishable Perishable No variable raw material quality Variable raw material quality Pausing production possible Pausing production not possible Suppliers/ customers not connected by pipeline Suppliers/ customers connected by pipeline

A: Heavy vehicle manufacturer X

Dicrete: 7 Process: 0       

B: Heavy vehicle manufacturer y

Dicrete: 7 Process: 0       

C: Health & hygiene production

Dicrete: 3 Process: 4        D: Chemical manufacturer X Dicrete: 3 Process: 4        E: Packaging manufacturer Dicrete: 4 Process: 3        F: Food processing Dicrete: 2 Process: 5       

G: Food & raw material

processer

Dicrete: 2 Process: 5       

H: Chemical manufacturer Y

Dicrete: 2 Process: 5       

Manufacturing aspects Dicrete / Process

3 2 1

4.1 Assessment cases on process and discrete industry characteristics

(21)

21 Below in table 10, the classification from pure discrete to pure process cases is presented.

Table 10: Cases per cluster Cluster 1: Pure discrete industry

The first cluster consists of cases A&B which are two heavy vehicle manufacturers. The companies in this cluster only have discrete manufacturing characteristics. First of all, the automotive industry consists of some unique characteristics. Firstly, the companies apply JIT-production strategy since there is no room for storage of the parts at the production line. Interviewee A2: “Nothing is pushed through our production process, no vehicle is standardized and every truck is unique. Every part is being delivered based on the production sequence”. This because the parts generally are large and not easy to move. The production system consists of a convergent product flow where lots of different parts are assembled into one end product. Pausing production is possible since most of the value is added through tabor. However, it is very expensive to pause the production line since hundreds of people are unable to assemble the vehicles.

Cluster 2: Discrete and process industries

The companies in the second cluster are characterized by discrete and process elements. All the companies are process industries but are not associated with perishability and variable raw material quality and therefore are categorized in cluster 2. All the products are MTS and the material flow is divergent. In addition, for companies C and D it is not possible to pause production without waste of material since these production systems are associated with substances and chemicals that have to go through the whole process.

Cluster 3: Pure process industry

The third cluster consists of pure process industry cases. These production systems have a divergent material flow and apply an MTS strategy. Both cases F and G are associated with perishable and variable raw material quality and therefore are located to the pure process industry spectrum. Case H is a chemical manufacturer that is connected to a network of other companies by pipelines. This unique characteristic is labeled as process industry characteristic and therefore this case is assigned to cluster 3. Moreover, for this company, it is also not possible to pause production since chemicals are involved in the production process.

Discrete Discrete/Process Process

Cluster 1 Cluster 2 Cluster 3

(22)

22

Resilience A: Heavy vehicle

manufacturer X B: Heavy vehicle manufacturer y C: Health & hygiene production D: Chemical manufacturer X E: Packaging

manufacturer F: Food processing G: Food & raw

material processer H: Chemical manufacturer Y

Fl exi bi l i ty

• Fl exi bl e order a mounts • Pa rts rel a ted to s peci fi c end product, not

a bl e to s wi tch or pos tpone

• Fl exi bl e order a mounts • Pa rts rel a ted to s peci fi c

end product, not a bl e to s wi tch or pos tpone • Agreed wi th s uppl i ers on 5-10% s wi tch i n vol ume • Fi xed production proces s • Fl exi bl e pa rtners for converting end product • Fi xed production proces s

• Fl exi bl e pa rtners for s tora ge a nd tra ns port

• Fi xed production proces s

• Fl exi bl e tra ns port

pa rtners • Fi xed production

proces s

• Fl exi bl e tra ns port pa rtners • Fi xed production

proces s

• In ca s e of a di s ruption,

ma chi ne s uppl i ers ca n del i ver fa s ter • Fi xed production

proces s • Rel a tivel y fi xed

vol umes compa ni es connected

Redunda ncy

• Suppl i er hol ds

i nventory cl os e to fa ctory • Hi gh a mount of uni que

s upl i ers • Low i nventory

• Suppl i er hol ds

i nventory cl os e to fa ctory • Hi gh a mount of uni que

s upl i ers • Low i nventory

• Sa fety s tocks • Dua l s ourci ng

for cri tica l products • Sa fety s tocks • Si ngl e s ourci ng, dependent on 2 ra w ma teri a l s uppl i ers • Sa fety s tocks • Ba ck-up s uppl i ers

• Stock ba s ed on i mportance of

cus tomer • Dua l s ourci ng for

cri tica l products

• Dua l s ourci ng for cri tica l products • Hi gh s a fety s tocks

• Bui l di ng i nventory • Hi gh s a fety s tocks

• Ma i nl y s i ngl e s ourci ng, few ma i n s uppl i ers

Col l a bora tion

• EDI s ys tems , s ha re

forca s ts • devel opi ng s uppl i ers

a nd hi gh res ource

s ha ri ng • Da i l y contact wi th

s uppl i ers wi th EDI s ys tems

• Sha ri ng foreca s ts , da i l y

contact wi th EDI s ys tems • Orga ni zed s uppl i er

performi ng progra ms • Hi gh res ource s ha ri ng

• Provi de tra i ni ng on a wa renes s a nd qua l i ty for

cri tica l s uppl i ers • Not s ha ri ng foreca s ts wi th s uppl i ers • Sha ri ng foreca s ts a nd production pl a n • No res ource s ha ri ng • Frequent communi ca tion wi th di s tri bution pa rtners

on vol umes • Sha ri ng foreca s ts

wi th cri tica l s uppl i ers • Revi ew meetings for

i mprovement of s uppl i ers

• Da i l y contact wi th

fa rmers • Col l a bora tive

s ys tem wi th di s tri bution pa rtner

a nd fa rmer • Sha ri ng foreca s ts wi th fa rmers • No s ys tem to s ha re foreca s ts wi th other • Cl os e col l a bora tion, joi nt

devel opment of bl ends a nd da i l y contact wi th fa rmers • Sha ri ng foreca s ts wi th fa rmers • Tra ns pa rent communi ca tion wi thi n

cl us ter of compa ni es • Not s ha ri ng foreca s ts /res ources

Agi l i ty

• Moni tori ng s uppl i ers ,

depa rtment for qua l i ty • Acting fa s t on

di s ruptions , often before

i t i s probl ema tic • Scena ri o's a nd a ction

pl a ns

• Moni tori ng s uppl i ers , ca pa ci ty ri s k

ma na gement tea m • Cri s i s tea m • i ncentivi ze s uppl i ers to

crea te a n a ction pl a n • Suppl i er moni tori ng ma i nl y on qua l i ty • Joi nt contingency pl a ns wi th s uppl i ers • Cri s i s tea ms i n ca s e of a di s ruption • In ca s e of a di s ruption, cus tomer a l l oca tion pri ori ty • No s uppl i er moni tori ng • No proa ctive a ction pl a ns • KPI da s hboa rd to moni tor cri tica l

s uppl i ers • Contingency pl a n for

certai n probl em ca tegory

• Moni tori ng fa rmers

for qua l i ty of ra w

ma teri a l • Ba ck-up pl a ns onl y

for tra ns port • In ca s e of a

di s ruption, cus tomer a l l oca tion pri ori ty

• Moni tori ng del i ver rel i a bi l i ty a nd qua l i ty of ra w ma teri a l • Si mpl e moni tori ng i n Excel , not a utoma tica l l y i n s ys tems • No a ction pl a ns for di s ruption

• For s ome s cena ri o's there a re a ction pl a ns ,

not for di s ruptions

• No s uppl i er moni tori ng • In ca s e of a di s ruption,

cus tomer a l l oca tion pri ori ty

4.2 Resilience practices per case

Below in table 11, the different resilience practices are shown per cluster.

(23)

23 The specific resilience aspects per case which will be rated to a specific score below in table 12. On the next page, in figure 2, these results are visually

illustrated.

Table 12: Resilience practices per case

A: Heavy vehicle manufacturer X Dicrete: 7 Process: 0 B: Heavy vehicle manufacturer y Dicrete: 7 Process: 0 C: Health & hygiene production Dicrete: 3 Process: 4 D: Chemical manufacturer X Dicrete: 3 Process: 4 E: Packaging manufacturer Dicrete: 4 Process: 3 F: Food processing Dicrete: 2 Process: 5

G: Food & raw material processer Dicrete: 2 Process: 5 H: Chemical manufacturer Y Dicrete: 2 Process: 5

Flexible order amounts        

Postponement in production Excess capacity in

production  

Multiple suppliers (for

same product)     Safety stock       Information-sharing         Resource-sharing       Forecast-sharing       Monitoring suppliers       Action plans      Crisis teams    Collaboration Agility

Cluster 1 Cluster 2 Cluster 3

Flexibility

Redundancy

(24)

24 Flexibility

Redundancy

Collaboration Agility

Resilience per industry cluster

Pure discrete industries Discrete and process industries Pure process industries

4.3 Cross-cluster analysis

In this section, the clusters will be compared based on the resilience aspects. Below in figure 2, the resilience dimensions per cluster are shown. Interesting trade-offs can be observed across all clusters. The analysis showed that discrete manufacturing cases adopt higher agility and collaboration practices whereas process industry companies are more focused on redundancy strategies at the expense of agility practices. At the end of each cluster section, the process and discrete characteristics will be related to the resilience dimensions.

Figure 2: Visual representation resilience dimensions per cluster Pure discrete industry: Collaboration and Agility

(25)

25 proactive in avoiding a potential disruption. Both companies are big players in the industry and have a lot of knowledge regarding the optimization of processes and therefore are adopting high resource-sharing activities. Interviewee A1: “We have a six-sigma organization and we also stimulate our suppliers to apply that knowledge, they can attend a course at our organization to improve their processes”. Key informants have identified resource-sharing as an important aspect of collaboration since it leads to mutual benefits. Both cases have organized supplier development programs to increase delivery reliability. In addition, both cases have systems to share forecasts. Interviewee B3: “We send forecasts for the next twelve months. For the first eighteen days of that, are trucks that already are sold to our customers”.

In terms of agility also significant results are found. As mentioned by several interviewees, visibility is really important in the automotive industry, since it involves hundreds of suppliers that have to deliver JIT. Both cases adopt supplier monitoring practices in order to increase visibility at the supply side. According to the interviewees, it is important to have real-time insight into the status of orders and measure the performance of suppliers. Both companies use EDI and customized systems to monitor their suppliers. Case B adopts a supplier rating program that is used by purchasing and logistics. Interviewee B2: “Right now we have a tool on supplier level and parts level with which we can see if volumes increase or not. We also have a supplier rating, which we already make before the ordering. We rate them on performance like delivery, communication, flexibility and EDI reports”. This way they are able to continuously asses the performance of their suppliers which increases the visibility in their supply chain and enables the organization to proactively prevent possible disruptions. Both cases also highly adopt action plans and task force teams in case of a disruption. Interviewee B2: “We use an official letter with signature for this, which includes the reason for the escalation and expectations of setting up an action plan”. In addition, the convergent product flow which, in this case, is related to a

high amount of unique suppliers also has a high effect on collaboration and agility. Therefore, both

cases highly collaborate in terms of information, resource and forecast sharing to be able to communicate with hundreds of suppliers. Interviewee B3: “We form a task force team and then we look what actions are necessary at that supplier”. This way they are able to communicate fast which increases the responsiveness.

In terms of flexibility, both companies are limited in flexibility practices associated with their production process. No vehicle could be made on stock or that parts of the production could be postponed since each part of the production line has a specific function. Therefore, it is hard to change production schedules in the short term.

(26)

26 for our critical items we only have one and a half or two days of stock”. This shows that the level of safety stock is low which decreases redundancy.

Discrete and process industries: Mainly Redundancy and average on other resilience dimensions

In this cluster, mostly redundancy practices related to inventory and the number of suppliers are observed.

Safety stock is often used by all the cases in this cluster to prevent or limit the impact of a disruption and therefore increases resilience. Case D implemented high safety stocks in the first part of the process. Interviewee D2: “In the first part of our total process we can handle disruptions well by holding stock since we don’t have risks”. Here the risks are related to the degree of variety of products which is low in this stage. Case C has quite high safety stock levels due to the lead-time of the raw materials. Case E also implemented safety stocks. Interestingly, they also take past disruptions into account. Interviewee F1: “Based on that, you also try to create safety stocks, because you know that you are so dependent and you got disappointed in the past”. For most of the cases, divergent material flow is related to the number of suppliers. A low amount of unique suppliers decreases the risk of having too much inventory of a specific material since it is used for most of the products.

An interesting result in terms of the number of suppliers is found for case D. This company is dependent on two raw material suppliers for their entire production process since the raw materials are very specific. This increases their vulnerability in case of a disruption which is also mentioned by the supply chain planner. Interviewee D1: “Last year, a fire took place at our main supplier, so they couldn’t deliver which caused our inventory level to drop really fast”. Overall, these practices increase redundancy and therefore resilience.

For case C and D, it is not possible to pause production since the machines have to run to process the materials. For example, in the case of an inventory outage, all the materials in progress will become worthless since the chemical processes will be disrupted. Although case E is able to pause the production process, no effects on redundancy practices are observed. All companies adopt high redundancy practices in terms of safety stock to prevent idle production systems in case of a disruption.

Both agility and collaboration practices are medium in this cluster. Interviewee D1: “We monthly share our 13-week planning”. Interviewee F1: “You actually want to have a dashboard with your critical suppliers, with a composed set of KPI’s”.

Pure process industry: Redundancy and Collaboration

(27)

27 the degree of redundancy. For these companies, the risk is lower since all their products are made out of a few raw materials. Interviewee E2: “Most of the time we only need 80% of what we order, that way we create a buffer which we can also use for other products”. Overall, redundancy practices significantly increase resilience and are crucial to prevent and mitigate disruptions in this cluster. Interestingly, collaboration is observed as an important resilience dimension in this cluster too. As briefly discussed before, cases F and G are associated with natural raw materials which could differ in

quality which increases the level of collaboration. More specifically, the level of information sharing

with critical suppliers is high. Case F implemented a web application to share information to farmers regarding transport times and possible problems to easily communicate changes in case of a disruption. In terms of agility, all cases in this cluster are not adopting a high degree of practices to increase responsiveness. In case of a disruption, all cases apply a reactive strategy where they allocate products to the most important customers. Therefore agility is classified as low in this cluster.

Two cases have to take into account the perishability of their raw materials. Case F and G have to consider a trade-off between keeping high inventory levels and the risk of obsolete products. However, both cases only have a couple of raw material variations which decreases the risk of obsolescence. No significant effects on safety stocks are found since both cases still hold relatively high stocks.

Both case F and G work with natural raw materials that have implications on the quality of the product. Interviewee H1: “We have daily contact with the farmers. That is with our AGO people. These AGO people drive to our farmers and get in touch daily. Not only about volumes and expectations”. Noticed is that information-sharing is critical in terms of the quality of the product to prevent disruptions. Case G has a specific department that is in contact with the farmers concerning the quality of the product. They use their knowledge to support the farmers. Interviewee H2: “We try to help the farmers to increase the output”. This characteristic increases the need to share information and therefore increase the degree of collaboration.

For case H, both suppliers, customers and other companies are part of this complex pipeline network of

companies so collaboration in terms of information sharing is important. Information sharing for this

(28)

28

4.4 Within-cluster analysis

In this section, remarkable possible differences in resilience within a cluster will be highlighted and will be related to differences in industry characteristics. Table 13 below shows the corresponding score per resilience dimension on case level.

Table 13: Within-cluster-analysis Flexibility

All cases are limited in flexibility practices associated with the production process since it is not able to postpone production. Within cluster 3, differences in flexibility are noticed for case H compared to case F and G. Case H is bounded to relatively fixed volumes since the companies are connected by pipelines which constrain the supply of raw materials. Other than previously described factors, homogeneous results are found in terms of flexible order amounts across all clusters. Interviewee B1: “We should strive to tell them our daily volume and within upper and lower limits they should be able to move along with us. Like 20% more or less than the agreed volumes”. This practice increases flexibility and therefore increases resilience.

Redundancy

Cluster 1 is characterized by low redundancy. Adopting redundancy practices in terms of inventory is hard for both cases A and B. However, both cases have strong requirements for their suppliers in terms of inventory and delivery times. Instead of having inventory at the factory, both cases require their suppliers to store parts close to their factories. Interviewee A2: “We try to have our inventory as low as possible. Our suppliers have to hold a certain inventory level close to our plant”. This way, the inventory buffer is located upstream in the supply chain. Both managers from case A and B mention that inventory at the plant is only hiding problems and is not preventing disruptions or resolving problems. In cluster 2, homogeneous results are observed in terms of inventory.

Discrete Process

Cluster 1 Cluster 2 Cluster 3

A B C D E F G H

Flexibility Medium Medium Medium Medium Medium Medium Medium Low

Redundancy Low Low High Medium High High High Medium

Collaboration High High Medium Low Medium High High Medium

(29)

29 In cluster 2, differences are noticed in the number of suppliers. Case D only has two raw material suppliers since it involves a very specific product and therefore is classified as medium compared to case C and E. In cluster 3, case H has lower degree of redundancy compared to case F and G. This because they only have a couple of main suppliers whereas case F and G have a lot of suppliers that deliver the same raw material. Safety stock is often used by all process industry cases to prevent or limit the impact of a disruption and therefore increases resilience.

Collaboration

Within cluster 1, similar results are found in terms of collaboration. Both cases A and B adopt a high degree and amount of information, resource and forecast sharing which increases the level of collaboration and therefore improves supply chain resilience. Within cluster 2, case E adopts collaboration practices in terms of information, resource and forecast sharing. On the contrary, case C does not actively share forecasts and case D is not involved in resource sharing with suppliers. Case D is scored as low on this dimension. It could be the case that this is related to the low amount of unique

suppliers. Interestingly, within cluster 3, a higher degree of collaboration for both cases F and G are

observed compared to case H. This because both cases have variable raw materials and are highly involved in information and resource sharing with their farmers to prevent any disruptions related to product quality. Case H adopts transparent communication within the cluster of companies in the long term to inform each other on maintenance activities since they are connected by pipelines. However, case H adopts low collaboration practices with suppliers in the short term. In this case, it could also be the case that the low amount of unique suppliers is related to the low degree of collaboration.

Agility

Within cluster 1, homogeneous results in terms of agility are observed. Both cases adopt supplier monitoring practices in order to increase visibility at the supply side. According to the interviewees, it is important to have real-time insight into the status of orders and measure the performance of suppliers. Both companies use EDI and customized systems to monitor their suppliers. Case B adopts a supplier rating program that is used by purchasing and logistics. Remarkably, in cluster 2, different levels of agility for all cases are noticed. Case C has a high level of agility, since it is adopting monitoring practices, created contingency plans and can implement crisis teams in case of a disruption. Case D is classified as low since it does not monitor suppliers and did not create action plans. Interestingly, the

amount of unique suppliers could be related to the level of agility. The low amount of suppliers for

(30)

(31)

31

5. Discussion

The goal of this study was to examine the influence of process and discrete industry characteristics on supply chain resilience. According to Crama et al. (2001) & Ivert et al. (2015), process and discrete characteristics are important determinants in production planning and control and it is already argued that resilience practices could be context-specific across different industries (Christopher & Peck, 2004; Donadoni et al., 2019). The analysis shows that the effects of discrete and process industry characteristics indeed have an effect on supply chain resilience strategies. This discussion is divided into two parts. First, trade-offs between resilience dimensions such as redundancy versus agility and the effect of specific industry characteristics are discussed. Then the process and discrete industry classification are highlighted where certain characteristics seem to play another role than expected.

5.1 Trade-off between resilience dimensions

Redundancy versus Agility

(32)

32 showed high inventory levels in both cases since their raw material is a base product necessary for all products. It could be argued that the degree of perishability is relatively low in these cases. However, variable raw material quality is a critical aspect of food processing companies which is also observed as important in this research (Van Donk, 2001; Ivert et al., 2015). Variable raw material quality influences monitoring practices since focal companies seek for visibility regarding their raw material. Contrary to the level of visibility, the level of responsiveness is low for all pure process companies. In case of a disruption, all companies in cluster 3 look at customer priority to allocate the products. It could be argued that the low level of responsiveness is related to the high level of technology and the fixed production process which is typical for process industries (Crama et al., 2001).

These trade-offs challenge earlier work of for example Kochan & Nowicki (2018) that examined resilience for different industries but did not look at the effect of those differences on resilience.

Collaboration

Overall, collaboration is an important resilience dimension to increase resilience which is in line with findings of Scholten & Schilder (2015). Interestingly, pure discrete companies are adopting high collaboration practices to reduce the risks of disruptions. High levels of information sharing are observed by implemented EDI systems to increase visibility. Similarly, Carvalho, Cruz-Machado, & Tavares (2012) stress that visibility is a critical factor in automotive supply chains to disrupt and mitigate disruptions. In addition, in line with our findings, they stress the importance of supply chain information systems as resilience practice. It could be argued that the MTO strategy and a high number of unique suppliers is related to the high degree of collaboration practices.

In the case of pure process industry companies, also high levels of collaboration are found. In this research, the high number of suppliers delivering the same raw material which also is pointed out by Van Kampen & Van Donk (2014), was of critical importance for process industries. With respect to resilience, the amount of suppliers has a high effect on collaboration. Both cases try to increase the quality and availability of raw materials by a high degree of information and resource sharing with their raw material suppliers. This finding is in line with Ivert et al. (2015) that argue the purchasing process as critical in production planning for food producers. In contrast, for both chemical manufacturers, relatively lower collaboration practices are observed. Interestingly, it could be argued that a low amount of unique suppliers negatively influence the degree of collaboration. This could be the case since it is easier to control a smaller supply base.

5.2 Level of process/discrete characteristics

(33)

33 characteristics and adopt different levels of resilience practices. For example, within process industries variable raw material quality could have impacts on collaboration. It has been highlighted that SCRES could be context-specific. Indeed, this research contributes to a more contingent perspective of resilience. This is in line with Sousa & Voss (2008) that mention contextual factors significantly influence performance outcomes of OM practices. However, current literature is mainly focused on antecedents, capabilities and definitions of SCRES by for example, Kamalahmadi & Mellat-Parast (2016); Kochan & Nowicki (2018), this research shows it is important to know in which context certain resilience practices are valuable. Below in figure 3, a new conceptual is proposed to illustrate the effects of specific industry characteristics on resilience.

(34)

34

6. Conclusion

This exploratory research has led to interesting insights into how process and discrete industry characteristics influence supply chain resilience strategies. First of all, process and discrete characteristics influence SCRES. It showed that variable raw material quality influences both information and resource sharing and therefore showed the importance of collaboration in such process industry contexts. In addition, variable raw material quality also influences monitoring practices and therefore the importance of visibility as part of agility. Both perishability and the inability to stop production did not have significant effects on safety stock levels.

In discrete manufacturing contexts, the findings showed that MTO highly influences the degree of information and forecast sharing since end products are highly customized and involve many unique suppliers that have to deliver Just-In-Time. Therefore, these suppliers need to know exactly what to expect in terms of requirements. In addition, from a focal company perspective, companies in these contexts have to adopt monitor practices to increase visibility to detect possible disruptions.

The amount of unique suppliers also influence certain SCRES strategies which are applicable for both process and discrete industry contexts. It positively influences all collaboration practices and also increases the need for monitoring practices since the complexity of the supply base increases when the amount of unique suppliers is high.

These findings shed light on how to increase resilience in certain process and discrete manufacturing contexts. It refines the understanding of underlying mechanisms of specific process and discrete aspects on specific resilience strategies.

Managerial implications

(35)

35

Limitations and suggestions for further research

Despite the findings and implications highlighted above, this study also presents limitations. First, this research was limited in the number of cases regarding the time constraint. A larger sample could improve the generalizability of the findings, although an effort was made to entail 8 cases that consist of all the defined characteristics. In addition, other industries such as the electronic and textile industry consist of other characteristics and could bring other insights into the effect of those characteristics on resilience (Pal, Torstensson, & Mattila, 2014; Sodhi & Lee, 2007). These industries could be highlighted in future resilience research.

In this research, discrete manufacturing companies adopt an MTO policy. However, this is not always the case for every company. In addition, this research focused on seven manufacturing aspects but it could be the case that external factors also could influence resilience. Future research, for example, could examine discrete manufacturing companies that adopt an MTS policy and possibly could integrate external factors. Considering the underlying research of differences in process and discrete industries (Crama, Pochet, & Wera, 2001; Müller & Oehm, 2019; Panwar, Nepal, Jain, & Rathore, 2015), this could be the first start to further investigate other characteristics on SCRES.

The scope of this research was at the supply side of the companies and was focused on the buyer-supplier relationship. It is discovered that in some cases, customers could influence the resilience level of the focal firm. Future research could shed light on the demand side in relation to resilience since demand variations are common in process industries (Dittfeld, Scholten, & Van Donk, 2018; Pool et al., 2011). Another limitation could be the unit of analysis which is at firm level for one location. It is expected that considering multiple plants of a company could increase flexibility since production could be split across different plants.

Do contingencies matter? The effect of process and discrete industry characteristics on Supply Chain Resilience strategies: A multiple case study

Faculty of Economics and Business

MSc Thesis Supply Chain Management