Positioning the CODP in food processing: a multiple case study

(1)

Positioning the CODP in food

processing: a multiple case study

Master’s thesis 2019-2020

MSc Technology and Operations Management

MSc Supply Chain Management

By: BSc. Jacob Berger

Supervisor: D.P. van Donk

(2)

2

Abstract

Food processing industries face growing logistical demands, a growing variety of products and a high perishable nature of products. While several studies propose methods for positioning the CODP, none of the methods shows to be entirely decisive. Therefore, this research aims to understand how food processing industry-specific issues and characteristics influence the approach to position the CODP. An in-depth case study, involving seven cases have been performed. By using semi-structured interviews and observational data, cases are analyzed to better understand how food processing companies position their CODP. The research revealed that lead-time, delivery lead-time, and especially perishability were important characteristics concerning the positioning of the CODP. Additionally, the research suggests that dynamic environments need more sophisticated tools to monitor, evaluate, and adapt their CODP’s. Finally, the research shows that only a limited amount of characteristics is included to position the CODP. Therefore, a reduced set of characteristics is proposed to position the CODP; lead-time, delivery lead-time, demand volume, and characteristics that contribute to the predictability of demand.

Keywords: Customer-order-decoupling-point (CODP), food processing industry (FPI); make-to-order (MTO); make-to-stock (MTS); Positioning CODP.

Preface

(4)

4

1.Introduction

The food processing industry (FPI) is characterized by high product variety, fast-response times and high perishability (Branco, 2019). These characteristics favor different production strategies. Customized products and products with high perishability can better be handled by a make-to-order strategy, where the customer-order-decoupling point (CODP) is located upstream. Shorter delivery times, on the other hand, necessitate a make-to-stock strategy, where the CODP is located downstream. This dichotomy makes it increasingly challenging to position the CODP, thereby obtaining the right balance between operational costs and service quality while keeping the number of obsoletes to a minimum (Fisher, Ramdas and Ulrich, 1999; van Donk, 2001). Additionally, While the best performance could theoretically be expected by having different production-strategies for each product, this will come at the expense of complexity (Van Kampen and Van Donk, 2014).

There is sizeable literature on factors that influence the position of the CODP (van Donk, 2001; Olhager, 2003). Indeed, it is well-understood whether and how certain factors influence the position of the CODP (van Donk, 2001; Olhager, 2003) and how each characteristics could be weighed (Hemmati, Rabbani and Ebadian, 2009; Hemmati and Rabbani, 2010). There are also studies that aim at positioning the CODP (van Donk, 2001; Huiskonen, Niemi and Pirttilä, 2003; Olhager, 2003; Wu et al., 2008; Yang and Wang, 2014). These involve, strategic approaches, which provide guidelines using knowledge-based systems or conceptual models (van Donk, 2001; Huiskonen, Niemi and Pirttilä, 2003; Olhager, 2003), as well as analytic approaches, which utilize mathematical models (Ji, Qi and Gu, 2007; Wu et

al., 2008a; Hemmati, Rabbani and Ebadian, 2009; Hemmati and Rabbani, 2010). However, studies for

both methods are not entirely decisive when positioning the CODP (van Donk, 2001; Akkerman, Van Der Meer and Van Donk, 2010). Mathematical models are usually too complex, based on too simplistic assumptions or require data too difficult to collect (Soman, Van Donk and Gaalman, 2004; Perona, Saccani and Zanoni, 2009). Where qualitative studies lack in the establishment of decision rules (van Donk, 2001; Akkerman, Van Der Meer and Van Donk, 2010; Van Kampen and Van Donk, 2014).

(5)

5 different CODP’s for each product, it would become too complex to do so. Therefore, as there is no clear trade-off point between both performance and complexity, FPI companies often struggle to position their many different products (van Kampen, Akkerman and van Donk, 2012; Van Kampen and Van Donk, 2014). To understand how the FPI specific issues and characteristics influence the decision to position the CODP, a multiple case study is conducted to answer the following research question:

“

How is the decision to positioning the CODP influenced by product-, market-, and

production-system characteristics of food processing companies“

By conducting a multiple case-study of in total seven cases, data was gathered by means of semi-structured interviews to collect qualitative data about the positioning of the CODP. By answering the research question, this paper contributes to the existing literature regarding the positioning of the CODP (van Donk, 2001; Soman, Van Donk and Gaalman, 2004; Akkerman, Van Der Meer and Van Donk, 2010; Van Kampen and Van Donk, 2014). It seems that there is a lack of guidance as to which techniques should be used to position the CODP and which characteristics should be included under specific circumstances, as most CODP positioning models are too complex, based on too simplistic assumptions or lacking in the establishment of decision rules (van Donk, 2001; Soman, Van Donk and Gaalman, 2004; Van Kampen and Van Donk, 2014). By doing this, the paper contributes by making the tacit CODP-decision-making-criteria more explicit and providing the ground for developing decision rules based on empirical evidence.

(6)

6

2. Theoretical background

The theoretical background will first introduce the main characteristics of the food processing industry. After that, the CODP and its context in the FPI will be covered, followed by an overview of characteristics that influences the position of the CODP. As last, approaches to position the CODP in the food industry will be given.

2.1 Food industry characteristics

The food processing industry (FPI) can be described as companies that involve strategies and systems to change crude ingredients into nourishment or nourishment into different structures for utilization by people or animals either in the home or by the food processing industries (Crama, Pochet and Wera, 2001; Branco, Patricia, Alexandra, 2019). Besides, the FPI is more complex than the manufacturing industry. For example, FPI involves high perishability of products. Based on the work of van Donk (2001), van Kampen and van Donk (2014), Branco (2019) and Soman, van Donk and Gaalman (2004), an overview of the characteristics in the food industry is compiled (Figure 2.1).

FPI characteristics

Product Process Market

- High perishability - High product variety - Variable supply

quality

- Capital intensive, single-purpose capacity - Divergent product structure

- Production rate determined by the capacity

- Long lead- and set-up times

- Varying and increasing demand uncertainty - Frequent deliveries - Short delivery time

Figure 2.1: Food characteristics (van Donk, 2001; Soman, Van Donk and Gaalman, 2004; Van Kampen and Van Donk,

2014; Branco, 2019)

2.2 Customer-order-decoupling point

(7)

7

2.3 Factors affecting the CODP in FPI

Several characteristics that affect the position of the CODP can be found (van Donk, 2001; Olhager, 2003). Van Donk (2001) identified two sets of characteristics that influence the CODP; product- and market characteristics and process- and stock characteristics. Besides, Olhager (2003) identified three sets of characteristics, namely; market, product, and production. They also stressed the fact that these characteristics are interrelated. Based on the work of Olhager (2003) and Van Donk (2001), and insights from Soman, Van Donk and Gaalman (2004), Hemmati and Rabbani (2010) and Romsdal et al., (2013), an overview of characteristics that affect the position of the CODP in the food industry are compiled.

Factor Explanation Incorporating food industry characteristics

Product

Perishability Perishability causes a higher risk of obsolescence ,

resulting in an upstream placement of the CODP (van Donk, 2001; Hemmati and Rabbani, 2010).

Products are high perishable, therefore upstream placement like MTO is preferable (Soman, Van Donk and Gaalman, 2004; Hemmati and Rabbani, 2010).

Product variety

Low product variety can be provided by an MTS system, however if a wide variety of products have to be produced, MTO is more suitable (Olhager, 2003).

High product variety calls for a more make-to-order strategy and thus locating the CODP more downstream (van Donk, 2001; Olhager, 2003)

Modular design

Modular design enables postponement, which allows products to become customer-specific more downstream (Olhager, 2003).

Food processing companies often use common recipes to produce a wide range of products (van Donk, 2001). This enables the CODP to be placed more upstream (van Donk, 2001; Olhager, 2003).

Holding costs High holding costs leads to an upstream placement

of the CODP, since inventory of products is costly (van Donk, 2001; Hemmati and Rabbani, 2010).

Holdings costs concern two factors; stock value and risk of obsolescence (van Donk, 2001). In FPI, both stock value and risk of obsolescence are high, forcing the CODP more upstream (van Donk, 2001; Hemmati and Rabbani, 2010).

Market

Delivery lead-time

Short-delivery time favors a downstream placement of the CODP (van Donk, 2001; Olhager, 2003). MTO may not feasible due to the production lead-time (Olhager, 2003).

Within the FPI, short delivery time is required by customers (Romsdal et al., 2013; Branco, 2019) and tend the CODP to be placed more downstream and produce make-to-stock(van Donk, 2001; Olhager, 2003).

Demand volatility

Low demand volatility makes forecasting more predictable, making an MTS strategy more feasible (van Donk, 2001; Olhager, 2003). In contrast, for a high demand volatility, forecasting is less

predictable and thus an MTO strategy is preferred (Olhager, 2003; Hemmati and Rabbani, 2010).

FPI is characterized by high demand volatility (Akkerman, Van Der Meer and Van Donk, 2010; Romsdal et al., 2013) forcing the CODP more upstream (van Donk, 2001; Olhager, 2003).

Forecast accuracy

Forecast accuracy is a determinant for producing to stock or to order (Akkerman, Van Der Meer and Van Donk, 2010). When forecast accuracy is low, forecasting is not feasible and therefore an MTS strategy would not be preferred (Olhager, 2003; Akkerman, Van Der Meer and Van Donk, 2010).

Due to the unpredictability of demand in the FPI (Akkerman, Van Der Meer and Van Donk, 2010; Romsdal et al., 2013), the CODP is forced more upstream (Akkerman, Van Der Meer and Van Donk, 2010).

Product range and

customization:

A wide product range and a wide set of

customization required by the customer, would be impossible to provide on an MTS basis (Olhager, 2003). A narrow range and predetermined customer choices could be provided by an MTS strategy (Olhager, 2003).

Market pressure results in customers that demand a wide product range and customized products (van Donk, 2001; Akkerman, Van Der Meer and Van Donk, 2010; Romsdal et al., 2013), forcing the CODP more upstream.

Order size and frequency:

High-frequency deliveries lead to repetitive demand, which makes forecasting more accurate and easier (Olhager, 2003).

Customers demand high frequency deliveries (Romsdal et al., 2013), which has a downstream effect on the CODP (Olhager, 2003).

Manufacturing flexibility

Manufacturing flexibility is a prerequisite for producing to order (Olhager, 2003; Hemmati and Rabbani, 2010).

(8)

8

Production-system

Production lead-times

A high production lead time favors a downstream placement of the CODP (Olhager, 2003). Since long production lead times lead to long delivery time when producing on order (Olhager, 2003; Hemmati and Rabbani, 2010).

The FPI has to deal with high production lead times, which forces the CODP more downstream and produce MTS.

Customization opportunities

Customization opportunities influence the number of potential CODP positions (Olhager, 2003). If customization opportunities can be performed more downstream, products can become customer-specific in a later stage (Olhager, 2003).

Customization opportunities in the packaging stage due to a large number of packaging sizes and labels(van Donk, 2001).

Number of planning points

The number of planning points influences the position of the CODP since it restricts the number of potential CODP positions (Olhager, 2003; Hemmati and Rabbani, 2010).

In FPI there are mainly two planning points, before the processing stage (raw-materials) and before the packaging stage (Soman, Van Donk and Gaalman, 2004).

Controllability Storing a product after the uncontrolled process

safeguards undisturbed delivery (van Donk, 2001). Therefore, uncontrolled processes have a

downstream effect on the CODP (van Donk, 2001).

FPI has an uncontrolled process with variable yield and, due to variability in natural materials, variable quality of products, causing a downstream effect on the CODP(van Donk, 2001).

Traceability Intermediate inventories are less traceable than

inventories of materials and end products (van Donk, 2001).

FPI has to obey to strict regulations which demand HACCP and traceability, restricting the possible location of the decoupling point as materials or finished products (van Donk, 2001).

Figure 2.2: Factors affecting the CODP in FPI

For most companies, decisions regarding the location of the decoupling point relate to the choice between a make-to-order (MTO) policy and a make-to-stock policy (MTS) (Van Kampen and Van Donk, 2014). Figure 2.2 shows that food processing companies face many different, and sometimes contrasting market-, product-, and production-system characteristics. Causing conflicting demands on the position of the CODP. As a result, the decision is surrounded by a discussion between and opposing interpretations by different departments (Van Kampen and Van Donk, 2014).

2.4 Approaches to position the CODP

To position the CODP, literature proposes several qualitative approaches that provide several decision elements (D’Alessandro and Baveja, 2000; Huiskonen, Niemi and Pirttilä, 2003; Olhager, 2003; Donk, Soman and Gaalman, 2005; Wikner and Rudberg, 2005). First, there are qualitative frameworks that aim at positioning the CODP by including the production- to delivery lead-time ratio (P/D ratio) and the demand volatility (Olhager, 2003; Donk, Soman and Gaalman, 2005), as can be seen in the framework of (Olhager, 2003) shown in Figure 2.3.

(9)

9 Another way to position the CODP is by looking at demand volatility and demand volume (D’Alessandro and Baveja, 2000; Huiskonen, Niemi and Pirttilä, 2003). Low demand volatility and high volume would lead to produce-to-stock and position the CODP more downstream (D’Alessandro and Baveja, 2000; Huiskonen, Niemi and Pirttilä, 2003). High demand volatility and high volume, on the other hand, would steer towards a more make-to-order strategy. Additionally, products with a discrete demand and low volume should be discarded due to the uneconomically small production lot sizes for a MTO strategy (Huiskonen, Niemi and Pirttilä, 2003). The above-mentioned characteristics show that the CODP can be positioned by three main characteristics; demand-volume, relative demand volatility and the ratio between lead- and delivery lead-time.

Next to approaches that mainly concentrate on product-analysis, also customer analysis has to be taken into account when positioning the CODP (Huiskonen, Niemi and Pirttilä, 2003). The position of the CODP influences the service level to the customer (van Donk, 2001; Huiskonen, Niemi and Pirttilä, 2003; Perona, Saccani and Zanoni, 2009). To remain customers satisfied, the appropriate service level to each customer must be delivered (Huiskonen, Niemi and Pirttilä, 2003). Therefore, customers should be analyzed and grouped based on attractiveness and judgmental recommendations about the opportunity to maintain, improve, or reduce the service level (Huiskonen, Niemi and Pirttilä, 2003). By linking the product- and customer-analysis, the CODP can be positioned, resulting in the possibility that products may have different CODPS’ for different customers (Huiskonen, Niemi and Pirttilä, 2003).

In contrast to the more qualitative frameworks, also a more quantitative approach can be used to position the CODP. ANP is a multi-criteria decision-making tool that considers both qualitative and quantitative criteria affecting a decision-problem (Hemmati, Rabbani and Ebadian, 2009). By applying ANP, characteristics that influence the position of the OPP are weighed and potential locations of the OPP are evaluated concerning the relative importance of product-, market- and production-system characteristics. Based on the outcome of ANP, the “best” CODP location is selected. Therefore, ANP can be used as a decision-making tool to position the CODP.

2.5 Linking FPI characteristics with positioning methods

(10)

10 2001). This leads to uncontrolled processes with variable yield, causing a downstream effect on the decoupling point, since storing a product after the uncontrolled process safeguards undisturbed delivery (van Donk, 2001). Additionally, manufacturing flexibility is neither addressed in recent frameworks to position the CODP. Due to long lead-times, high sequence-dependent set-up and cleaning times, producing to order is not preferred due to production inefficiencies (Olhager, 2003; Hemmati and Rabbani, 2010). Furthermore, applying ANP to position the CODP is also not fully decisive. First, ANP requires the judgments of a broad level of experts, which is not easily allowed in business (Saaty, 2007). Second, most businesses require a more distinct, easy-to-use explanation of how they should position the CODP (O’Reilley, 2015; Perona et al., 2009).

Concluded can be that different market-, product-, and production-system characteristics of food processing companies influence the approach of positioning the CODP (Figure 2.4). Although, there is no clear way how food processing companies should position their CODP. The qualitative frameworks suggest that the CODP can be positioned mainly based on; “production/delivery lead-time”, “demand volatility”, “product type” and “demand volume”. Yet, their usefulness in the FPI can be questioned since they do not capture specific characteristics related to the FPI. Therefore, some suggestions (perishability, variable quality in supply and manufacturing flexibility) are given which can be a source of inspiration for positioning the CODP. Also, the CODP should not only be positioned by a product-specific policy but also the customer should be taken into account (Huiskonen, Niemi and Pirttilä, 2003). Implying that the same product may have different CODP for different customers. A more quantitative way to position the CODP is by applying ANP (Hemmati, Rabbani and Ebadian, 2009). However, its applicability in the FPI can be doubted (Saaty, 2008; Perona, Saccani and Zanoni, 2009; O’Reilly, Kumar and Adam, 2015). To provide more clarity, this study aims at identifying how different product-, market- and production-system characteristics relate to how companies decide to position their CODPproduct-, which will provide the grounds for decision rules to position the CODP.

(11)

11

3. Methodology

The research question “How is the decision to positioning the CODP influenced by product-,

market-, and production-system characteristics of food processing companies” will be investigated by

adopting a qualitative approach. An explorative approach was chosen since it is the most appropriate to reveal and understand the multiple facets of the positioning of the CODP (Flynn et al., 1990; McCutcheon and Meredith, 1993; Baxter and Jack, 2008). Besides, in-depth analysis of the food process industry-specific characteristics is needed to understand the complex interactions between product-, market- and production-system characteristics and the position of the CODP. Multiple cases of food processing companies are analyzed to find commonalities and patterns within and between the cases of their decision-making process. Analysis of multiple cases provide a stronger base for explorative research and increases validity (Yin, 2003; Baxter, Susan Jack and Jack, 2008). The understanding of how managers decide to position the CODP leads to theory building by primary data (Voss, Tsikriktsis and Frohlich, 2002).

3.1 Case selection

Cases are selected within the domain of companies operating in the food processing industry. The unit of analysis within the case-study is the decision process related to the positioning of the CODP. A variety of product-, market-, and production-system characteristics is necessary to understand how companies position the CODP (McCutcheon and Meredith, 1993). Therefore, cases are selected based on their different nature of perishability (High/Low), sizes (number of employees and profit), and sectors, which shows the use of theoretical replication (Yin, 2003). To further increase the validity of the research, also literal replication logic has been applied by selecting purely cases in the FPI with a hybrid production system (Yin, 2003). Besides, to guarantee a certain management structure, at least a workforce of 50 employees as needed(Van Donk and Van Doorne, 2016).

(12)

12

Case Sector Workforce Perishability

(outputs)

MTO/MTS distribution (Weight)

A Meat substitutes 300 High 60 / 40

B Alcoholic beverages 600 Low 25 / 75

C Confectionery 140 Low 15/ 85

D Bakery-pastry 60 High 10 / 90

E Bakery-pastry 264 Low 65/35

F Flower processing 282 High 40/60

G Bakery-pastry 127 Low 70/30

Figure 3.1: Case sample

3.2 Data collection

Since this research aim is to gain an understanding of how product-, market-, and production-system characteristics relate to how managers decide to position their CODP, data was mainly collected by semi-structured interviews. Semi-structured interviews offer some degree of freedom while still maintaining the structure to acquire the necessary data, resulting in a higher degree of information richness the researcher can gather (Harvey-Jordan and Long, 2001; Voss, Tsikriktsis and Frohlich, 2002). For this reason, an interview protocol was conducted (Appendix 1). The Interview protocol was conducted in cooperation with two other researchers, investigating a similar research topic. To this end, the protocol entails topics related to; production planning, positioning of the CODP, and reclassification. The interviews started with an introduction and the topics to be addressed. After that, more general questions about company aspects were asked, followed by questions concerning production planning. Next, questions about the process of positioning were asked. Including; information about the criteria used, how criteria are weighed, and how the CODP is positioned. As of last, questions regarding the reclassification of products were asked. Besides, interview transcripts from previous studies with a similar topic were used to gather data about the cases.

(13)

13 for each case. Additionally, to enhance the triangulation of the research also documentation and company videos were asked.

Case Nr. Interviewees Function Duration Data

gathered A 1 Continuous improvement manager 101 minutes Interview

B 1 Planner 92 minutes Interview

C 1 Supply chain manager 71 minutes Interview

D 2 Production planner

51 minutes

Interview + tour

Head business office

E 2 Production manager

70 minutes

Interview Logistics manager

F 1 Plant manager 75 minutes Interview

G 3 VP-CEO 70 minutes Interview

Quality manager Sales manager

Figure 3.2: Interview details

3.3 Data analysis

First, the semi-structured interviews were transcribed and summarized. Summaries of the interviews were sent to the people involved to validate the interviews and if needed correct for misunderstandings. Next, the semi-structured interviews were coded with the help of Atlas_TI. Codes contain a mix of both inductive and deductive coding. First, deductive coding was used to explore the aspects mentioned in the literature. Therefore, FPI characteristics were coded in a deductive manner, in which the characteristics identified by (van Donk, 2001; Olhager, 2003) are used as a starting point. Furthermore, to identify approaches to position the CODP, an inductive approach was used to maintain the richness of data. The codes for each case were grouped in different categories: 1st_{order, 2}nd_order,

and 3rd_{order codes to consolidate meaning and explanation and facilitate a cross-case analysis (Flynn}

et al., 1990; Karlsson, 2016). The codes used are written to gather more insights into general aspects of

(14)

3th order 2nd order 1st_order _Quotations Product-related criteria perishability Modular design Product variety

(E): “Now regarding the due date of the product we don’t have an issue because the product is sterile and it has a due

date of 9 to 12 months for the Greek market or even 15 months for the exports….”

(A): “There is always a generic component in which you can say that the large mix is modular and changeable. So an

item can be produced on multiple lines”

(F): “Yes there are at least 120 products. We have 3 categories one being the retail products that will end up in

supermarkets, either with our name or as a private label with the supermarket's name. “ Market-related criteria Demand variability

Delivery-time Order frequency

(A): “The question of volatility is very important in this, because it indicates how much a customer wants to purchase.” (D): “The delivery lead-time for MTS is 48 hour and for MTO it will be 2 weeks”

(F): “if we know that the client C biscuits obtains every week 250 tons, even if they haven’t informed us today for next

week we know that the requested quantity will be more or less similar if something extraordinary hasn’t happened.”

(B): “In principle, it is a process towards a new product where those agreements are made where, based on volume

determination, it is said that we will produce these products by means of this production strategy.”

Production-system related criteria

Controllability

Lead-time

Flexibility

(G): “Yes of course this is vital for us, because we don’t always have stable quality of raw materials and a percentage

of raw material is lost, due to the fact that they have dirt and foreign seeds inside.”

(A): “So if I get an order from order today and deliver tomorrow, within that time window I will be able to pack the

product. If not, I would have to stock the product.”

(B): “But in general, we don't have many options to use other production lines to meet market demand, the lines are

very specific.” Process of positioning Approaches to position CODP Method to position CODP Stakeholders

(A): “Yes, we have 10 factors that determine whether an article is suitable for MTO or MTS. We have set a range in

our system for all these factors."

(C): “That is a bit of cooperation between the planning, sales assistants, supply chain and myself. Then we look at an

article and discuss how it goes and then the decision is made.”

Figure 3.3: Overview codes

(15)

By insights gained by the company-videos, and the codes from Figure 3.4 and Appendix 4 and 5, the presence of product-, market-, and production-system characteristics have been analyzed and scored together with other researchers. Presented in Figure 3.4. Scores range from (1) to (5) high, based on the comparison of the cases. For instance, when you look at lead-time, Case B mentions that “The

process from brewing till fermenting takes regularly about 2 till 3 weeks, and sometimes even 4 weeks.”,

“The minimum lead-time is four weeks”, resulting in a score of (5) high. On the other hand, Case F mentions that “Production lead time is very important and is always taken into account, because if you

say today that you want flour A, you have to calculate going backwards at least 20 hours before to start production for the production of that flour.” resulting in a score of (4) medium. Regarding demand

predictability, a score of 5 was allocated to Case B “We ask our clients yearly how much beer they are

going to sell. Based on that forecast we will create a schedule due to budgetary reasons. “, “We get from multiple markets a demand forecast, usually at Thursday or Friday. “, “We have weekly scheduling cycles which determine the production planning for the next 14 till 78 weeks.”. Also, Case A was

allocated a score of (1) low “That is hard in the determination how much capacity you will need in a

year. We think we have enough capacity but from our experience, we know that the market can act very unpredictable”, “Meat replacement products is a relatively new market, which makes it hard to predict.”. Since market demands are hard to predict and forecast focuses on the short-medium term, a

score of (4) was allocated to Case E where multiple data indicates medium-high demand predictability. “However generally demand is relatively stable for most products.”, “Essentially it’s the sales data

everything is related to the forecast and past sales data and of course experience so we know that specific customers are going to order for certain some quantities, because we also have a contract with them. So the sales forecast is transformed into a detailed production schedule. Correct me if I am wrong”. To ensure transparency about the codes, additional information on how the scores were established are given in Appendix 3.

(16)

16

4. Findings

In this chapter, the findings will be presented. First, different FPI characteristics that are incorporated in the positioning of the CODP will be identified. After that, approaches to position the CODP as MTO or MTS will be investigated.

4.1 Identifying characteristics

Looking at characteristics incorporated in approaches to position the CODP, the following results were found. First, for perishability, it was found that Company A, D, and F, all with a medium till high perishability, included the characteristic. For Company A, the perishability of end-products is seen as a major determinant due to their high nature of perishability in comparison to intermediate products. (A) “Perishability is taken into account, however, if you know your product has a long shelf

life, it will be of less importance”, “For finished goods, it is a different case, since they are highly perishable products, perishability as factor is a major determinant for determining to keep the product on stock or not.“. Also, Company D uses perishability as a boundary condition to position the CODP.

(D) “Related to perishability, each product has a certain best-before date, therefore perishability can

be seen as a boundary condition for determining the CODP.”. Contrary, companies with lower

perishability (B, C, E, and G), did not incorporate perishability into their approach to position the CODP since the relatively long shelf-life of their products did not force them to do so. Company C mentions “The perishability of our products is fine, our products mainly consist of sugar, glucose, and gelatin,

resulting in high shelf life of about 1 until 2 years. Thus, we run a little risk on the perishability of our products”. Also (G) mentions that “Additionally based on tests we have performed on our products and their durability, our product even after 2 or 2.5 years from a microbiology sense there is no issue with its consumption from a healthcare perspective. So we don’t really have perishability as a factor affecting our product.”.

Holding costs were included by Companies A, E, and G into their approach. Where products with a high stock value steer towards a more MTO strategy. (G) “If a product contains complementary

expensive raw materials, we make it last minute (MTO).”, (E) “Now regarding holding costs these are of course also taken into account for more expensive products”. On the other hand, companies B, C,

and F did not include holding costs into their approach. Where Company C did not include it since it is already calculated in the cost-price of the product, (C) “Holding costs are already taken included in our

cost price” and (D) “Our commercial margin is much higher. As a result, we do not include holding costs in our decision to keep products in stock or to make to order.”. Remarkable is that Company F

did not include holding cots despite their high perishability. (F) “for us, it isn’t that relevant because

we have high inventory turns.”. Where high inventory turnovers seem to mitigate the risk of

(17)

17 Considering market-characteristics, the following characteristics are taken into account. First, demand variability was found for al companies except case B. In which, demand variability indicates to what extent it is reasonable to produce forecast or order-driven. Low demand variability implies that the product can be produced to stock. High variability increases the difficulty of forecasting whereas such items typically need to be produced to order. (A) “To determine the “risk” of keeping a product on

stock, demand volatility is an important determinant since it indicates how much a customer will purchase. If demand volatility is high, it can be the case that the customer suddenly would not purchase anymore.”. Besides, for cases A, C, D, and F were found that demand variability is often used in

combination with order frequency. (D) “That is right, we mostly look at order frequency and demand

variability for positioning the CODP as MTO or MTS”. Noticeable is that in contrast to the other cases,

Company F and G, increased their stock levels to absorb demand variability (G) “A product that has

stable demand can also be produced more to order and have less stock, whereas a product that has variability in demand should have more stock to account for this uncertainty. Next, all companies

included demand volume. Where low volumes steer towards a more MTO strategy and high volumes toward MTS. (B) “new customers, operating mostly in new markets, are mainly MTO due to their low

volumes. However, when the volumes become higher, and/or pressure for a shorter-lead time, we will discuss MTS”. Or, (E) “On some of our standard products however we need to have stock because they are produced in large quantities and have high demand.“. Additionally, forecast accuracy is included

by companies C, D, E, F and G. Whereas forecast accuracy indicates to what extend producing to stock is reasonable. (C) “For example, if sales say we are going to sell 2 million bags and someone else tells

the planner that he has to make that and then the customer will only take 100 000 bags. You will have huge financial consequences. So the most important thing for us is the forecast accuracy”. (F) “The most important criteria that relate to the MTO/MTS decision is the predictability of demand as it defines whether a product can be produced with smaller stock or needs higher stock or without any stock if it is a specialized product.”. As of last, all companies included delivery lead-time into their decision to

position the CODP. In which a balance must be found between delivery lead-time and inventory costs (F) “In general choosing MTO/MTS and where to separate it is complex as it brings conflicting interests

namely production and sales and the ability to produce and deliver on time and at the same time hold minimal inventory”. Additionally, products with a required short delivery lead-time are usually provided

on a MTS basis. (B) “If the customer demands a lead-time of two weeks, I will say that will be fine, but

in that case, I will produce MTF for you and the consequences related to that decision are for the customer”, (D) “If we produce the product to stock, we will be able to deliver the product within two days.”. On the other hand, products with a longer delivery lead-time can be provided on a MTO basis

(B) “MTO has a higher lead-time because we are more conservative when producing raw materials for

a MTO product.”. (D) “For MTO products customers have to order two weeks in advance.”. Yet, in

(18)

18

I have to deliver something to the customer, but I have two days' lead time of the product. In that case, I will not put the product in stock. Because then I will only produce the product a few days before the delivery date.”.

Concerning production-system characteristics, the following characteristics are included. First, lead-time was included for most companies except Company B. Where lead-time initiates the time of production. (C) “After that, I will see what is the lead time per product. Subsequently, the program will

calculate it back until the time when I should have the product produced to stock or order”. Second,

controllability was included for Company E, F, and G. Where, uncontrolled processes lead to a more downstream position of the CODP to be able to safeguard the processes. (G) “Yes of course this is vital

for us because we don’t always have stable quality of raw materials and a percentage of raw material is lost, due to the fact that they have dirt and foreign seeds inside”. Besides, also intermediate storages

can be used to increase the controllability of the processes, as seen for Company (F) and (G). As of last, only Company B and D included customization opportunities in their approach to position the CODP. Where the position of the CODP is restricted by the possibilities to customize the product (B) “as the decoupling point follows from the product, we have to add that flavoring, that fruit mug, after

filtration”.

Next to existing characteristics, also new characteristics have been found. Company A, B, and E included financial liability in the decision of positioning the CODP. In which financial liability can be seen as a boundary to produce MTS. (A) “It could be the case that a customer would not order

anymore if you did not make financial agreements about that, the risk will be for you as a producer”.

Next, for companies B, C, E, and G contractual agreements are incorporated in the decision to position the CODP. To include customer-wishes (B) “The customer has a very large influence on the decision to

produce MTO or MTF. Regularly, we agree with the customer whether a product will be positioned as MTF or MTF”, or more restrictive by limiting the possible CODP positions. (E) ”The decision of MTO/MTS is also guided through a contractual agreement up to a point. So you cannot make MTO into MTS”. As of last, the balance of MTO/MTS products was found to be an important characteristic that

is included in the approach to position the CODP. Where the mix of MTO/MTS products should ensure a stable level of the required capacity to increase efficiency. (A) “The mix of MTO/MTS products is

(19)

19

Cases A

B

C

D

E

F

G

Characteristics

Perishability Order frequency Demand volume Demand variability Forecast accuracy Lead-time Delivery lead-time Financial liability Customer contracts Holding costs Controllability Mix MTO/MTS Product customization

Table 4.1: Characteristics used to position the CODP

Table 4.1 provides an overview of characteristics that are used by FPI companies to position their CODP. Every “ ” denotes that the characteristic is included in the decision to position the CODP as MTO or as MTS. Additionally, since not every characteristics is observed to be equally important to the company, the bold icons indicate that the characteristic is a major determinant, and the normal icon indicates a regular determinant. To distinguish between both, Appendix 4 and 5 have been analyzed. For instance, lead-time was found to be the number one criterion for Company (F) The number one

criterion we look at when deciding on MTO/MTS is the processing time and this is logical as the processing time might be a restricting factor towards MTO or MTS.”. However, for Company A,

demand volatility is seen as a major determinant. “As soon as those risks increase, you prefer to not

keep the product in stock. Herein demand volatility is a very important characteristic.”.

By analyzing all seven cases, it can be found that companies with high perishability use similar characteristics to position the CODP. As can be seen in Table 4.1, Company A, D, and G all include perishability, order-frequency, demand volume, demand variability, lead-time, and delivery lead-time. Only for forecast accuracy there was found that Company A did not include it. However, contrary to Company D and F, Compay A has low demand predictability, which strongly decreases forecast accuracy. Yet, while all three companies include the same characteristics, they do not consider their characteristics equally important. For instance, order frequency and demand variability are seen as major determinants to Company D, “That is right, we mostly look at order frequency and demand variability

for positioning the CODP as MTO or MTS”. In contrast, Company A sees lead-time as a major

(20)

20

and this is logical as the processing time might be a restricting factor towards MTO or MTS”. Strikingly,

companies with medium until low perishability did not show similar characteristics in their approach to position the CODP.

Besides perishability, also lead-time and delivery lead-time seem to be an important characteristic that influences the approach to position the CODP. Where a high lead- and required lead-time forced companies to include lead-time and/or delivery lead-time in their approach to position the CODP. Despite its high lead-time, Company B did not include the characteristic in its approach. A possible explanation for this could be that mainly customers determine the position of the CODP as MTO or MTS, and therefore takes the corresponding lead-time for granted.

Bu purely focusing on the characteristics used to position the CODP (Table 4.1), the following can be concluded. First, lead-time, delivery lead-time, demand volume, and other characteristics that contribute to the predictability of demand are important as they are in most cases used to position the CODP. Another interesting finding was found concerning delivery-lead time and lead-time, and especially the relation between both. As can be seen in almost all cases, companies incorporate both lead-time and delivery lead-time. About this, Case A mentions that “I have a 30-day order lead-time

for example, so in thirty days I have to deliver something to the customer, but I have two days lead time of the product. In that case, I will not put the product in stock. Because then I will only produce the product a few days before the delivery date.”. Implying the ratio between time and delivery

lead-time determines if products should be produced to stock or order. Additionally, also Case C mentions “I start with the sales forecast from a certain period, then I will look at the lead-time per product and

processing. After that, the program will calculate it back to point till when I have to produce the product to stock, and when I have to produce the product on order.”. Both citations suggest that if the lead-time

is lower than the delivery lead-time, products can be produced to order. Which is also suggested in prior literature (Olhager, 2003; Donk, Soman and Gaalman, 2005). Yet, to capitalize on economies of scale, some products may be produced to stock, even if the ratio between lead- and deliver-time is smaller than one. (A) “If perishability allows, you want to prevent those shocks in your production in such a way that

you are not continuously driven by the customer's demand, but you can also efficiently organize your process. So if the product its perishability permits, we try to make it make-to-stock”. Thus, by placing

the CODP more downstream to capitalize on economies of scale, perishability should allow for it. Additionally, also company C argues to place the CODP more downstream to capitalize on economies of scale, (C) “adjust product classification from MTO to MTS, will enable you to run larger production

(21)

21

4.2 Approaches to position the CODP

To position the CODP, several approaches were identified in which each approach showed a different level of sophistication. First, Company D was found to mainly rely on experience to position the CODP. In which Company D used a few characteristics and a rule-of-thumb to position products as MTO or MTS. Products being sold regularly are positioned as MTS, and products not being regularly sold are positioned as MTO. (D) “if the product is being sold regularly, it will be produced to stock, if

the product is not being sold regularly, but more sold promotional or occasionally, the product will be produced to order.”. Second, in comparison with Company B, Company C, E, F, and G also used

experience to position the CODP. However, next to experience, they also used forecasting tools to observe demand. (C) “It is a decision that you have to estimate”, “The forecast tool will qualify your

demand and its pattern.. and it guides in this”. (G) “It is of course a combination of experience ….So it is a combination of experience and also programs that allow you to monitor what you sold last year to a customer”. Additionally, both Case C and F differentiate between positioning new or existing

products. (F) “After we develop the product we will start with some small initial stock…. from there we

will observe how the product does.” And (C) “You will think like, first we will produce to stock for two months…after that you will check if demand matches the expectations, and based on that you can adjust your production-strategy”. Where new products are initially positioned as MTS and the CODP will be

adjusted if needed. Existing products are positioned based on experience and forecasting tools. (F) “older products however we use previous experience and of course sales prediction.”. Next, in contrast to the other companies, Company A used a multi-criteria decision tool to position the CODP. (A) “We

have 10 factors that determine to position the CODP as MTO or MTS. For al these factors we have introduced a certain range.”. This enables Company A to continuously monitor the dynamic

environment and adjust the production strategy to keep the right product mix. (A) “This enables us to

keep our product-mix up-to-date. By continuously reviewing products by the tool. Where aspects like, do we have the right production strategy for the product related to the scores of the tool monitored. This could enforce you to change your strategy to MTO or MTS. “. As of last, for Company C, the CODP is

primarily determined by its customer. (C) “The customer has a great influence in the decision to

position a product as MTO or as MTS. In principle we agree together with the customer if a product is MTF or MTO, since most implications are for the customer.”.

(22)

22

is not being evaluated regularly”, or (C) “You do not want to evaluate the production-strategy of the assortment every week. That should not be necessary, because then you keep changing your production-strategies and not everybody has time for that”. Additionally, also Company F does not continuously

monitor and evaluate its CODP. (F) “As I said earlier every month we perform a meeting and we plan

ahead.”. The above suggests that companies with a lower need to evaluate and monitor their CODP’s

have less sophisticated tools to position the CODP in comparison with companies that have a lower need to monitor and evaluate their CODP’s. Additionally, companies with a higher need to monitor or evaluate the CODP’s also have a lower demand predictability in comparison to companies with a lower need to monitor the CODP’s. This finding supports both studies of van Kampen, Akkerman and van Donk, (2012) and Yang and Wang, (2014). In which is argued that an unpredictable demand creates a more dynamic environment which increases the need to monitor, evaluate, and adapt the CODP.

Case Positioning by

A Multi-criteria tool

B Contractual agreements (Customer) C Experience supported by forecasting tools D Experience

E Experience

F Experience supported by forecasting tools G Experience supported by forecasting tools H Experience supported by forecasting tools

(23)

23

5. Discussion

The previous section presented the findings with regard to characteristics and approaches used to position the CODP. In this section, we aim to extend the literature on the positioning of the CODP. First, approaches to position the CODP will be discussed, followed by the characteristics included in the decision to position the CODP. As of last, the influence of different FPI characteristics on the approach to position the CODP will be discussed.

The first observation that can be made from our study is that the adaptation of sophisticated tools to position the CODP in food processing companies is still limited. While literature suggests various sophisticated ways to position the CODP (Wu et al., 2008b; Hemmati, Rabbani and Ebadian, 2009; Rafiei and Rabbani, 2014). Our results show that most companies rather rely on experience and forecasting tools instead. Overall, these findings are in line with previous research (Kerkkänen, 2007). Arguing these decision problems like the positioning of the CODP call for more simple approaches.

5.1 Characteristics for positioning the CODP

It was made clear that different product-, market- and process-related characteristics affect the position of the CODP (van Donk, 2001; Olhager, 2003; Rafiei and Rabbani, 2014). Yet, a first finding is that due to the many proposed characteristics, only a limited set of product,- market,- and process-characteristics in included in the decision to position a product as MTO or MTS (Velev, Andreev and Panayotova, 2011; van Kampen, Akkerman and van Donk, 2012; Rafiei and Rabbani, 2014). Since too many characteristics increase the complexity of the process of evaluating the efficiency of alternatives. Making it increasingly challenging to take all characteristics simultaneously into account (Velev, Andreev and Panayotova, 2011; van Kampen, Akkerman and van Donk, 2012; Rafiei and Rabbani, 2014). A second finding is that companies with a high-perishable nature of products include a similar set of characteristics in their decision to position the CODP. These results oppose the findings of (Kerkkänen, 2007), mentioning that the applicability of a set of general characteristics to determine the CODP for a particular industry is restricted. Additionally, As in agreement with (D’Alessandro and Baveja, 2000; Huiskonen, Niemi and Pirttilä, 2003; Olhager, 2003; Donk, Soman and Gaalman, 2005) the characteristics lead-time, delivery lead-time, demand volume and characteristics that contribute to the predictability of demand are included in most positioning approaches.

(24)

24 is expressed through the RDV. Comparing our results to those of (Olhager, 2003), it must be pointed out that food processing companies should not only consider the P/D ratio and the RDV, but also include perishability in this trade-off. Where perishability should allow for a MTS strategy.

Next to the product-, market-, and production-system characteristics identified by van Donk (2001), Olhager (2003), Hemmati, Rabbani and Ebadian (2009), also other characteristics emerged. First, financial liability emerged as a characteristic that influences the position of the CODP. Where financial liability transfers the risk of obsolescence from the producer to the customer, making producers more willing to produce to stock. Another important emerging characteristic was the use of contractual agreements. Where, contractual agreements increase the predictability of demand by sharing point-of-sale (POS) data, shifting the CODP more downstream (van Donk, 2001), or have a more restrictive function, in which the customer determines the position of the CODP. As of last, to increase efficiency required capacity should be kept stable by having the right mix of MTO/MTS products.

5.2 Influence FPI characteristics on approaches to position CODP

While we expected that many food-specific characteristics would influence the positioning of the CODP. We found that in particular lead-time, delivery lead-time, and especially perishability are important influential characteristics. Overall these findings are in accordance with van Donk (2001), Olhager (2003) and Soman, van Donk and Gaalman (2007), which argue that the perishable nature of products, lead-time, and delivery lead-time are important characteristics that may constrain the possible CODP positions. Remarkable is that for the other product,- market-, or production-system characteristics, no clear patterns were found.

(25)

25

6. Conclusion

The research looked at how product-, market- and production-system characteristics relate to how food processing companies position their CODP. Through analyzing seven cases operating in the food-processing-industry, the research investigates which characteristics are used to position the CODP, which approaches are adopted to position the CODP, and how product-, market- and production-system characteristics relate to this. The contributions of this study are threefold. First, the research adds to the literature by relating product-, market-, and production-system characteristics to approaches to position the CODP. Second, this led to new findings related to characteristics that influence the positioning of the CODP. Third, the study provides empirical evidence that supports previous findings of positioning methods (D’Alessandro and Baveja, 2000; Huiskonen, Niemi and Pirttilä, 2003; Olhager, 2003).

By qualitatively analyzing the approaches to position the CODP, this study adds to the knowledge base of decision-making processes and providing the ground for developing decision rules based on empirical evidence. The adoption of sophisticated tools to position the CODP is limited. Instead, FPI companies rather rely on experience, in which a limited set of characteristics is included to reduce the complexity of the process. Additionally, empirical evidence was found that perishability, lead-time and delivery lead-time are the main characteristics that influence the approach to position the CODP. In which especially perishability seems a major determinant. Besides, demand volume, lead-time, delivery lead-lead-time, and characteristics that contribute to the predictability of demand commonly used to position the CODP.

(26)

26

7. Limitations and further research

The empirical results reported herein should be considered in light of some limitations. First, due to the corona-virus several companies decided to drop-out of the research. This limited the number of cases participating in the study. To increase the reliability of the study, several cases from previous studies have been used. Second, regarding the cases, it cannot be guaranteed that a different case-selection would have generated the same results. Despite the efforts that have been spent to select cases from different sectors, varying in a different product-, market-, and production-system characteristics, different cases may lead to different outcomes. As third, due to the corona-virus, physical interviews and company-tours were mostly unavailable. Therefore, data had to be gathered by video-calling, which limited the possibilities to ask additional questions and gather data. Additionally, while interviewees have been selected on their experience concerning the positioning of the CODP, differences in experience have led to different levels of comprehensive discussions. To increase the richness of data, company-videos and additional questions were asked during follow-up questions by mail.

While we found that different product,- market,- and production-system-characteristics are included in approaches to position the CODP. Gathered data inhibited the ability to deeply investigate how food processing companies deal with trade-offs between these characteristics. Therefore, it would be interesting to investigate how food processing companies deal with these trade-offs and to develop appropriate decision rules related to that. Furthermore, since the research had an explorative nature, factors, and relations that merged from this study need further investigation to understand their explanatory value. Meaning that the factors and relations that emerged from this study require further investigation. Directions for future research could, for example, be to investigate how perishability influences the positioning of the CODP

(27)

27

References

Akkerman, R., Van Der Meer, D. and Van Donk, D. P. (2010) ‘Make to stock and mix to order: Choosing intermediate products in the food-processing industry’, International Journal of Production

Research, 48(12), pp. 3475–3492. doi: 10.1080/00207540902810569.

Baxter, P., Susan Jack and Jack, S. (2008) ‘Qualitative Case Study Methodology: Study Design and Implementation for Novice Researchers’, The Qualitative Report Volume, 13(4), pp. 544–559. doi: 10.2174/1874434600802010058.

Branco, Patricia, Alexandra, B. (2019) Food industry : Processes and technologies. 1st edn. Delve Publishing. Available at: ttps://ebookcentral.proquest.com.

Crama, Y., Pochet, Y. and Wera, Y. (2001) ‘A discussion of production planning approaches in the process industry’, Science.

D’Alessandro, A. J. and Baveja, A. (2000) ‘Divide and conquer: Rohm and Haas’ response to a changing specialty chemicals market’, Interfaces, 30(6), pp. 1–16. doi: 10.1287/inte.30.6.1.11627. van Donk, D. P. (2001) ‘MTS or MTO : The DP in the food processing industry’, International

Journal of Production Economics, 69(3), pp. 297–306. doi: 10.1016/S0925-5273(00)00035-9.

Van Donk, D. P. and Van Doorne, R. (2016) ‘The impact of the customer order decoupling point on type and level of supply chain integration’, International Journal of Production Research. doi: 10.1080/00207543.2015.1101176.

Donk, D. P. Van, Soman, C. and Gaalman, G. (2005) ‘a Decision Aid for To-Order and Make-To-Stock a Decision Aid for Make-To-Order and Make-Make-To-Stock’, EurOM International Conference

on Operations and Global Competitiveness, (May 2014).

Fisher, M., Ramdas, K. and Ulrich, K. (1999) ‘Component sharing in the management of product variety: A study of automotive braking systems’, Management Science. doi: 10.1287/mnsc.45.3.297. Flynn, B. B. et al. (1990) ‘Empirical research methods in operations management’, Journal of

Operations Management. doi: 10.1016/0272-6963(90)90098-X.

Harvey-Jordan, S. and Long, S. (2001) ‘The process and the pitfalls of semi-structured interviews’,

Community Practitioner.

Hemmati, S. and Rabbani, M. (2010) ‘Make-to-order/make-to-stock partitioning decision using the analytic network process’, International Journal of Advanced Manufacturing Technology, 48(5–8), pp. 801–813. doi: 10.1007/s00170-009-2312-4.

Hemmati, S., Rabbani, M. and Ebadian, M. (2009) ‘Positioning of order penetrating point in hybrid MTS/MTO environments’, EMS 2009 - UKSim 3rd European Modelling Symposium on Computer

Modelling and Simulation, pp. 263–268. doi: 10.1109/EMS.2009.82.

Hoekstra, S. and Romme, J. (1992) ‘Integral Logistic Structures: Developing Customer-oriented Goods Flow’, Industrial Press.

Huiskonen, J., Niemi, P. and Pirttilä, T. (2003) ‘An approach to link customer characteristics to inventory decision making’, International Journal of Production Economics, 81–82, pp. 255–264. doi: 10.1016/S0925-5273(02)00367-5.

Ivert, L. K. and Jonsson, P. (2014) ‘When should advanced planning and scheduling systems be used in sales and operations planning?’, International Journal of Operations and Production Management. doi: 10.1108/IJOPM-03-2011-0088.

Ji, J. H., Qi, L. L. and Gu, Q. L. (2007) ‘Study on the CODP position of the process industry implemented mass customization’, Xitong Gongcheng Lilun yu Shijian/System Engineering Theory

(28)

28 van Kampen, T. J., Akkerman, R. and van Donk, D. P. (2012) ‘SKU classification: A literature review and conceptual framework’, International Journal of Operations and Production Management. doi: 10.1108/01443571211250112.

Van Kampen, T. J. and Van Donk, D. P. (2014) ‘When is it time to revise your SKU classification: Setting and resetting the decoupling point in a dairy company’, Production Planning and Control, 25(16), pp. 1338–1350. doi: 10.1080/09537287.2013.839063.

Karlsson, C. (2016) Research Methods for Operations Management, Research Methods for

Operations Management. doi: 10.4324/9781315671420.

Kerkkänen, A. (2007) ‘Determining semi-finished products to be stocked when changing the MTS-MTO policy: Case of a steel mill’, International Journal of Production Economics. doi:

10.1016/j.ijpe.2006.12.006.

McCutcheon, D. M. and Meredith, J. R. (1993) ‘Conducting case study research in operations management’, Journal of Operations Management. doi: 10.1016/0272-6963(93)90002-7.

O’Reilly, S., Kumar, A. and Adam, F. (2015) ‘The role of hierarchical production planning in food manufacturing SMEs’, International Journal of Operations and Production Management. doi: 10.1108/IJOPM-04-2014-0157.

Olhager, J. (2003) ‘Strategic positioning of the order penetration point’, International Journal of

Production Economics, 85(3), pp. 319–329. doi: 10.1016/S0925-5273(03)00119-1.

Perona, M., Saccani, N. and Zanoni, S. (2009) ‘Combining make-to-order and make-to-stock inventory policies: An empirical application to a manufacturing SME’, Production Planning and

Control, 20(7), pp. 559–575. doi: 10.1080/09537280903034271.

Rafiei, H. and Rabbani, M. (2014) ‘Hybrid MTS/MTO order partitioning framework based upon fuzzy analytic network process’, Applied Soft Computing Journal, 19, pp. 312–321.

Romsdal, A. et al. (2013) ‘Tactical and operational issues in a hybrid MTO-MTS production environment: The case of food production’, in IFIP Advances in Information and Communication

Technology. doi: 10.1007/978-3-642-40361-3_78.

Rudberg, M. and Wikner, J. (2004) ‘Mass customization in terms of the customer order decoupling point’, Production Planning and Control, 15(4), pp. 445–458. doi: 10.1080/0953728042000238764. Saaty, T. L. (2008) ‘The Analytic Hierarchy and Analytic Network Measurement Processes:

Applications to Decisions under Risk’, European Journal of Pure and Applied Mathematics. doi: 10.1007/0-387-23081-5_9.

Sharman, G. (1984) ‘The rediscovery of logistics.’, McKinsey Quarterly.

Soman, C. A., Van Donk, D. P. and Gaalman, G. (2004) ‘Combined make-to-order and make-to-stock in a food production system SOM-theme A: Primary processes within firms’, Int. J. Production

Economics, 90, pp. 223–235. Available at:

https://ac.els-cdn.com/S0925527302003766/1-s2.0-

S0925527302003766-main.pdf?_tid=6feda083-4556-4d68-adeb-55f5900770b6&acdnat=1550064497_92e671d303d4c83d8b06938caa2a5030.

Soman, C. A., van Donk, D. P. and Gaalman, G. J. C. (2007) ‘Capacitated planning and scheduling for combined make-to-order and make-to-stock production in the food industry: An illustrative case study’, International Journal of Production Economics, 108(1–2), pp. 191–199. doi:

10.1016/j.ijpe.2006.12.042.

Velev, M., Andreev, O. and Panayotova, T. (2011) ‘Using AHM approach to position CODP’, in

BMSD 2011 - Proceedings of the 1st International Symposium on Business Modeling and Software Design.

(29)

29

International Journal of Operations and Production Management. doi: 10.1108/01443570210414329.

Wang, Y. and Chen, Y. (2016) ‘Multi-CODP adjustment model and algorithm driven by customer requirements in dynamic environments’, Cluster Computing. doi: 10.1007/s10586-016-0661-y. Wikner, J. and Rudberg, M. (2005) ‘Introducing a customer order decoupling zone in logistics decision-making’, International Journal of Logistics Research and Applications. doi:

10.1080/13675560500282595.

Wu, M. et al. (2008a) ‘Study on the customer order decoupling point position base on profit’,

Proceedings of 2008 IEEE International Conference on Service Operations and Logistics, and Informatics, IEEE/SOLI 2008, 1, pp. 44–47. doi: 10.1109/SOLI.2008.4686359.

Wu, M. et al. (2008b) ‘Study on the customer order decoupling point position base on profit’, in

Proceedings of 2008 IEEE International Conference on Service Operations and Logistics, and Informatics, IEEE/SOLI 2008. doi: 10.1109/SOLI.2008.4686359.

Yang, K. and Wang, E. (2014) ‘Optimal control of decoupling point with deteriorating items’, Journal

of Industrial Engineering and Management. doi: 10.3926/jiem.1205.

Yin, R. K. (2003) ‘Case study research: Desing and Methods’, Applied Social Research Methods

(30)

30

Appendices

Appendix 1: Interview protocol

Opening

1. Introduction of interviewers and interviewee 2. Confidentiality assurance

3. Permission to audiotape General

1. Interviewee job-title

2. Could you describe the company?

a. Characteristics such as; size, market share and locations b. Number of employees (plant and organization)

c. Turnover plant in terms of money and volume d. Location in entire supply chain

3. Could you give a rough description of the manufacturing process from raw material to final products?

4. What makes your plant more successful than your competitors? a. What is the driving force in your production planning? 5. At what point do your products become customer specific? 6. Do you produce private label products?

7. Do you sub-divide products in MTO, MTS or Mix-to-order (ATO)? a. Approximately what percentage of products is produced to order?

Product planning

1. Do you subdivide products in product families or groups?

2. Which steps do you take to create your production planning per product group? a. What is the timeline

b. Who are involved?

c. How do you deal with capacity constraints? d. What are your process constraints?

e. How do you deal with these constraints? f. What is limiting the output of the factory?

3. What criteria or guidelines do you use to determine which product to allocate capacity to first? 4. Are there certain production lines that are specific to one product or are they flexible?

Positioning the CODP in food processing: a multiple case study