REEVALU ATING TH E POR TFOLIO A CASE STUDY AT IS OCAB NV
MASTER THESIS TECHNOLOGY MANAGEMENT 2011
By:
WouterVercauteren
Supervisors:
Ronny Baekelant (CEO Isocab)
Michael Lamprecht (Managing Director Isocab)
Dr. Ir. Hen van de Water (RUG) Dr. R. Rozier (RUG)
University of Groningen Faculty of Economics and Business
MSc. Technology Management
2011
< Page intentionally left blank >
PREFACE
With this thesis I conclude my master Technology Management and put an end to a very important period in any student’s life. As so many students I lived from day to day, not really bothered by future. Although, occasionally I noticed how time seems to fly by and before I knew it I was already halfway through the master and expected to write a thesis!
It seemed like such a daunting challenge; to singlehandedly perform research and advice an organisation on a course of action. However, my doubts were largely unfounded and, in hindsight, I would say that the experience provided me with many lessons that you simply do not learn behind a book. Books and exams do not prepare one for the working life that awaits all students sooner or later; only experiencing it first-hand can. I would not have been able to learn these lessons and complete my thesis if it was not for the support and assistance I received from so many. I would like to use this page to thank them.
First of all, my gratitude goes to Mr. Ronny Baekelant, my company supervisor. Not only did he provide me with the opportunity to conduct my research at Isocab, he was always there whenever I needed his opinions and support, in spite of his busy schedule as Isocab’s CEO. I would also like to thank Mr. Michael Lamprecht, who took over the role as my company supervisor when Mr.Baekelant was unavailable.
This research would not have come off the ground if it was not for the input and feedback from Dr. Hen van de Water, who helped me demarcate the topic, set up a theoretical framework and guide me through the rest of process. His patience throughout this process was greatly appreciated.
I would also like to mention my friends, who test read my thesis and offered many suggestions to improve it. One of them reminded me of what Stephen Hawking’s publisher once said: For every equation, you lose half of your readership. This helped me stick to essence, and not get lost in details.
Finally, I owe a great debt of gratitude for the support, and patience, of my family during my entire study career.
Wouter Vercauteren
Antwerp, 2011
MANAGEMENT SUMMARY
Origin and objective
Over the years, Isocab NV has added many different products to its portfolio, which resulted in an increase of WIP inventory in their CONWIP line. Since too much differentiation is detrimental to the performance of a production line, Isocab wants to reduce the amount of different products offered in order to reduce the capital invested in inventories. The objective was formulated as follows:
To help Isocab NV reduce its product portfolio in order to improve business performance (i.e. reduce WIP)
Due to Isocab’s extensive product portfolio, this research was demarcated to the doors, more specifically the Cold Storage Revolving Door segment.
Preliminary research
First, a literature study was carried out in order to identify different perspectives from which a product portfolio can be evaluated. Each perspective offers its own techniques, each with its own benefits. The different perspectives are the marketing perspective, the production/engineering perspective and the financial perspective. Traditional techniques that correspond with these perspectives are, respectively, the growth-share matrix, the degree of commonality index and the product beta. Which technique is most appropriate depends on the situation at hand.
Choosing an evaluation method
The AHP method was used to determine which of the three perspectives best matches Isocab NV’s situation. The situation at hand was analysed through personal interviews with key individuals, on-site evaluations and using the internal information systems. In order to determine which perspective best suits Isocab’s situation, the AHP method was used. AHP was chosen because:
a. It is a robust algorithm, which is hard to manipulate.
b. It is one of the few methods capable of dealing with many criteria, let alone qualitative criteria.
c. It is very easy to use
In the end, Isocab’s upper management strongly preferred the production/engineering perspective. They strongly valued commonality’s benefits regarding inventory reduction and management compared to alternative approaches.
Evaluating the portfolio
First, a suitable subject for evaluation was selected from the Cold Storage Revolving Door segments. The DFD/DDD model was chosen, as it is in significantly lower demand than its counterparts. Low demand products are wasteful in CONWIP lines, as it involves storing a lot of WIP that is hardly ever used.
The research continued by gathering Bills Of Materials (BOM’s) for the DFD/DDD doors.
In order to compare the doors to the remainder of the segment, the VFD/VDD and
PFD/PDD doors served as a benchmark. As it turns out, the segments enjoy a high
degree of commonality when the DFD/DDD doors are included in the portfolio. This means that many of the components used are shared among all models, which results in:
1. simplified inventory management: fewer parts and tools that have to be managed 2. reduction of safety stock needed, and thus reducing inventory costs
3. less orders have to be placed with suppliers
Removing the DFD/DDD model reduces the commonality index by quite a bit. This implies that only very few components and modules would become redundant.
Removing the model would harm the commonality as a portfolio as whole.
Even though the DFD/DDD model is in low demand, offering it results in only a minor increase in WIP. This is more than offset by the benefits it provides through economies of scale. Removing the model would therefore be ill-advised.
Recommendation
The degree of commonality index has great potential to improve the inventory issue at
Isocab’s CONWIP line. The benefits of commonality are real, and the commonality index
provides a way of measuring those. The index provides an easy to use method of
evaluating a portfolio on the short term. One needs to be aware however that these
short term improvements may not solve the problem in the long run however, and
further, more thorough evaluation will be necessary to assess the strategic implications.
TABLE OF CONTENTS
Preface ... 3
Management summary ... 4
1. Introduction ... 9
1.1 Isocab NV ... 9
1.1.1 Isocab’s History ... 9
1.1.2 As a part of ThyssenKrupp AG ... 10
1.1.3 Isocab’s product portfolio ... 10
1.2 Topic Relevance ... 11
2. Methodology ... 14
2.1 The problem statement ... 14
2.1.1 The research objective ... 14
2.1.2 The research question ... 14
2.2 The theoretical model and concepts... 15
2.2.1 The conceptual model ... 15
2.2.2 Sub-questions ... 16
2.3 Research outline ... 17
2.3.1 Preliminary Research:Motivation and Contents ... 17
2.3.2 Main research ... 19
2.4 Management of used sources of data and information ... 20
3 Prelimenary research ... 21
3.1 Isocab’s current practice ... 21
3.1.1 Product portfolio ... 21
3.1.2 Production process ... 23
3.2 Portfolio Evaluation Methods ... 25
3.2.1 Marketing Perspective ... 25
3.2.2 Production Perspective ... 26
3.2.3 Financial Perspective ... 27
3.3 Selection Technique ... 28
3.3.1 Popular MCDM methods: a short introduction ... 29
3.3.2 Selecting a MCDM method ... 32
4 Analysis: deciding upon an evaluation method ... 34
4.1 Structuring the problem hierarchy ... 34
4.2 Paired comparisons ... 36
4.3 Synthesizing the results... 37
4.3.1 Sensitivity analysis ... 39
5 Design: How to assess a portfolio from a production perspective ... 40
5.1 commonality: a short overview ... 40
5.2 Initial demarcation: Where to focus our efforts ... 47
5.3 Evaluating the portfolio: Applying the Commonality index in practice ... 49
5.4 Conclusions ... 52
6 Final thoughts ... 54
6.1 Reflections ... 54
6.2 Limitations and opportunities for future studies ... 56
References ... 58
Appendices ... 61
1. Isocab’s product portfolio: standard sizes... 61
2 Example of a BOM for the VFD/VDD models ... 63
3. Example of a BOM hierarchy, of a standard cold storage revolving door ... 64
< Page intentionally left blank >
1. INTRODUCTION
This chapter will introduce the context of this research. The first paragraph will briefly introduce the reader to Isocab NV, where the field work was performed. The second paragraph will explain the role this research will play, or in other words, why this research was conducted and is relevant to Isocab. The third paragraph will describe the thesis’s outline and general structure.
1.1 ISOCAB NV
Isocab NV is a successful international manufacturer of industrial sandwich panels, clean rooms, mineral wool panels, LBH houses, etc. for all kinds of markets worldwide, though mainly for the food industry. Isocab NV’s international engineering experience allows them to support projects from start to finish, tailored to the customer’s needs and demands. This high emphasis on customer service and quality products have brought them to the top of the European insulation industry.
Isocab NV’s main production site is located in Bavikhove (Belgium). Other production sites are located in Dunkirk (France) and Perpignan (France), supported by production sites in Vietnam and China.
1.1.1 ISOCAB’S HISTORY
Isocab NV, founded on the 23
rdof November 1973 by Mr. Luc Castelein, started out as a manufacturer of small and medium sized refrigeration rooms, deepfreeze rooms and thermal doors for different branches of the food industry.
In 1984, a second production facility was built at the Bavikhove site.
Isocab NV has been producing industrial panels measuring up to twelve meters since 1986, paving the way for the production of both large cold stores and industrial buildings.
The Dunkirk site (known as Isocab France 1) was opened in 1989. This production line produces isothermal sandwich panels, in whatever size and colour the customer requires. These panels are ready-for-use on delivery.
A second factory was built in Dunkirk (known as Isocab France 2) in 1991. This state-of-the-art production facility produces single roof and façade cladding profiles. This facility is built for speed and flexibility.
Isocab NV was taken over by Thyssen AG in 1991.
In order to stay on top and remain competitive in Southern Europe, North Africa
and other export markets, Isocab NV invested in a production site in Perpignan
(known as Decapanel SAS). Decapanal produces the same isothermal sandwich
panels as Isocab France 1 and a new mineral wool sandwich panel. The factory
was taken into use in 2002. It is currently (Feb. 2011) the largest facility in the world with regards to production capacity.
In 2005, Isocab NV expanded their production capabilities with the Ho Chi Minh site in Vietnam.
Again, in 2006, the production capabilities were expanded with the Rizhao site in China.
In 2007, Isocab NV was positioned as part of the ThyssenKrupp Stahl Bauelemente branch of ThyssenKrupp AG.
1.1.2 AS A PART OF THYSSENKRUPP AG
ThyssenKrupp AG is a conglomerate that operates globally in the fields of steel, capital goods and services. Isocab was taken over in 1991 by Thyssen AG, which later merged in 1997 with Krupp AG to form ThyssenKrupp AG. As a part of ThyssenKrupp Stahl Bauelemente, Isocab’s main activities are the production, sale and delivery of steel industrial sandwich panels, doors and insulated cold storage. Isocab’s placing as a part of ThyssenKrupp AG as a whole is illustrated in Figure 1.
FIGURE 1: THYSSENKRUPP'S MAIN ACTIVITIES: STEEL, CAPITAL GOODS AND SERVICES (SOURCE:
QUALITY HANDBOOK ISOCAB, FEBRUARY MARCH 2011)
1.1.3 ISOCAB’S PRODUCT PORTFOLIO
As explained before, Isocab’sBavikhove site is responsible for the production of several
steel based thermal products, and this paragraph will introduce these. It is important to
note that these are just the different categories. Every category has many subtypes
which are all available in many different sizes. And if the customer still does not find
what they are looking for, Isocab NV will see what they can do to accommodate their needs.
1. Walk in coolers: One of the main products they produce are insulated sandwich panels. These panels are created by injecting a steel skin with isolating foam like polyurethane. These panels can then be connected to each other by using the airtight internal locking system. This modular design allows the cooler to be fully demountable and adjustable according to the customer’s needs.
2. Mini-box: Mini-boxes are small cooling cells, like the walk-in fridges you find in restaurants or hospitals.
3. Doors: Isocab produces a wide range of doors for industrial applications. These are mainly isothermal hinged and sliding doors, but Isocab also produces a series of office and flip flap doors. Their door portfolio consists of the following categories:
- Cool and refrigeration swing doors - Office/factory doors
- Flip-flap doors
- Slightly insulated sliding doors - Heavy sliding doors
- Lift sliding doors
4. Wall and roof panels: The insulated sandwich panels can also be used as wall panels, with their internal locking system allowing the quick construction of both inner and outer walls, and ceilings. It offers a quick alternative to traditional industrial building methods.
5. Decaroc mineral wool panels: These are sandwich panels used both horizontally and vertically that are not filled with a layer of foam, but with mineral wool. This gives the panel exceptional fire resistance, surviving for up to 120 minutes, and acoustic properties.
6. Retan clean rooms construction: Isocab also produces complete clean rooms, custom built according to the customer’s demands, to be used for all kinds of controlled environments like for example laboratories in hospitals or chemical industries.
1.2 TOPIC RELEVANCE
Just as many other multinational production companies, Isocab has to manage
increasing global competition, shrinking product life cycles and constantly changing
market needs. This forces many companies to develop and produce more products, in
less time, with lower costs, while maintaining the same quality standards as before.
Isocab, as a market leader, has recognized the need to innovate. Over the years, the product portfolio has been expanded in order to remain competitive and stay on top.
Isocab competes with their experience and high quality products. However, most importantly, they have been able to thrive due to their customer intimacy policy. For every project, a single salesman or project leader is assigned to the job, allowing Isocab to get to know the client’s needs in detail and suggest a tailored solution, or even turn- key projects. Isocab guarantees technical and commercial support on all their products, from start to finish.
However, such a customer intimacy strategy comes at a price. Being able to produce a product for every customer brings with it a complex manufacturing process. More specifically, it results in high amounts of work-in-process stock. For example, at the moment Isocab produces 13 different door models, and each door model has different types. Some door models consist of as many as 12 different types. Every type is available in up to 15 or more different sizes. In order to ensure a timely production and delivery of these doors, stock is carried for many of these variants. Now, some doors are high in demand, and others are not. Still, stock is carried for every product, regardless. Over their entire product portfolio, the location in Bavikhove at times has over 240 000 items in stock.
Isocab is aware of the fact that not all product variants are equally in demand of profitable, but that these are resulting in high amounts of work-in-process inventories.
This requires Isocab to evaluate their product portfolio and identify the ‘sick’ products.
These ‘sick’ products are those that cost resources to produce and hold in inventories, without benefiting in increased revenues. These resources could be better spent on more profitable products.
However, this process is not as straightforward as it may seem at first glance. One
cannot simply perform a revenue contribution analysis and sift through a portfolio
based on financial performance alone. Other product aspects have to be taken into
consideration as well. A product may not be very profitable directly, but it may serve a
greater purpose from a marketing point of view. It may not be very profitable to tailor a
product to a customer’s needs, but if this allows one to form a long term customer
relationship, then the initial loss of profit may very well be offset by the prospect of
future business opportunities. A similar case can be made from a production point of
view; some product variants may not contribute as much to total revenue as others, but
not all product variants are produced as easily as others. Maybe, a lower profit variant
plays an important role in a product family, sharing the total costs over a greater
number of units and reaching more economies of scale. Dropping that product based
upon a lower contribution to total revenue might harm the other members of the family
in the long run. As it seems, the choice whether or not to drop a product from a portfolio
can be a difficult choice indeed. The evaluation process requires one to compare and
evaluate many different product aspects. Fair enough, but how does one go about and
perform such an evaluation in practice? And, more importantly, how can Isocab improve
their portfolio’s performance on the short term, until such a thorough portfolio
evaluation can be conducted?
With my master thesis, I aim to contribute to this evaluation process. This thesis aims to decide upon a quick-and-dirty product evaluation method that can help Isocab through this process. This method should identify those product aspects that are most relevant, and provide Isocab with a means to measure and quantify them. This will allow Isocab to select the best product portfolio evaluation method, on the short term, depending on their current goals and situation, until they have the time available for a full portfolio evaluation on the long term.
Due to the time constraints of this research and the complexity of Isocab’s product portfolio and production process, this research will single out one product group to focus its efforts on: the doors. This product groups has many different varieties and should benefit the most from such an improvement attempt. Also, because it is such a diverse group of products, it should provide a solid testing ground for such a comprehensive theoretical framework. Table 1 contains the door production details from 2007. It shows how every door category contributes to the total door production. It is clear that not every category has the same turnover. It is this research’s goal to identify how each category contributes to the business performance, and from there on suggest a new product portfolio, eliminating truly unprofitable products from the product mix. As an added benefit, Isocab is currently developing a new brochure for their new door offerings. The research’s results are aimed at assisting in that effort.
TABLE 1: PRODUCTION DETAILS DOORS (SOURCE: INTERNAL DOCUMENTATION PRODUKTIONPROZESSE GESPRÄCH 2 2007)
Categories Units produced Percentage
Clean room doors 689 4,72
Company doors 2.473 16,94
Cold storage revolving doors 5.806 39,78
Isolated swing-doors 329 2,25
Mini-box doors 704 4,82
Hatches 983 6,73
Light sliding doors 1.322 9,06
Heavy sliding doors 603 4,13
Polyethylene swing-doors 732 5,02
Others 955 6,54
Total 14.596 100
2. METHODOLOGY
Logically, any research is only as good as the research methodology that backs it up. This chapter will explain the approach taken by this study. De Leeuw (1996) has identified the five elements that make up a methodology: the problem statement, the theoretical concepts and models, used sources of data, observation and measure methods, and the data analysis methods. The following paragraphs will elaborate upon these five elements in detail.
2.1 THE PROBLEM STATEMENT
According to De Leeuw, there are three components to a good problem statement. A problem statement has to state the research objective, the research question and a demarcation.
2.1.1 THE RESEARCH OBJECTIVE
The rationale of the study is simple. Isocab has expanded its product portfolio over years, providing their clients with a wide variety of products to choose from. However, not every product does equally well in the marketplace (anymore): they do not contribute equally to Isocab’s revenue. However, offering so many products results in capital to be tied in WIP (work-in-progress) buffer inventory. As a way of improving business performance, Isocab wants to reduce the amount of products they offer, which will reduce the amount of WIP in the production process while freeing up production capacity for more profitable products.
Therefore, the research’s objective can be formulated as:
To help Isocab NV reduce its product portfolio in order to improve business performance (i.e. reduce WIP)
2.1.2 THE RESEARCH QUESTION
While the research objective details the ultimate goal of this study, it does not tell us how to accomplish this goal. This is specified in the research question:
How can Isocab NV evaluate its product portfolio, taking into account the different facets involved, in order to identify ‘sick’, unprofitable products?
In the end, this research will deliver a decision model that will describe a step-by-step
approach for Isocab to rank their products in their product portfolio. This allows them
to identify the least profitable products and consider whether or not dropping the
product could improve business performance on the short term.
2.2 THE THEORETICAL MODEL AND CONCEPTS
In order to guide our attempts at answering the research question, and thus ultimately in achieving the research objective, a conceptual model has been developed. This model, as every model, is an abstract representation of reality in order to provide us with a practical insight. At its core lay the theoretical concepts, field research through personal interviews and common sense. This conceptual model will allow the splitting of the research question into several, manageable, sub-questions.
2.2.1 THE CONCEPTUAL MODEL
The conceptual model guiding this research (see figure 2) consists of four elements: the portfolio’s performance and three perspectives that allow one to evaluate the portfolio’s performance.
Simply put, this models shows that a product’s (and thus a portfolio’s) performance can be evaluated from three perspectives. First, there is the financial perspective, where a product’s performance is evaluated through cost and revenue based methods. Second, there is a marketing perspective, where the product’s performance is evaluated through customer satisfaction and the market role the product plays in the greater portfolio. And finally, the production perspective, which evaluates the product’s performance through the eyes of the production department and the role the products play in their respective product families and such. These three different perspectives have three very different views on what makes a product successful. Which one of these methods is best suited for the occasion, is rather situational. This research will identify Isocab’s situation, which then leads to the suitable perspective. We can then evaluate the product portfolio from this perspective, and come up with a method of improving its performance.
PRODUCT MANAGEMENT
PORTFOLIO EVLUATION - METHOD AND CRITERIA -
PRODUCT PORTFOLIO
BUSINESS PERFORMANCE Strategic Business Unit
performance Leads to
Has impact on
Influence
MARKETING PRODUCTION FINANCIAL
FIGURE 2: CONCEPTUAL MODEL
2.2.2 SUB-QUESTIONS
Based upon the logic of the conceptual model, it allows the formulation of these sub- questions (see figure 3):
1. What are the main drivers behind the current portfolio reduction?
This question is aimed at analyzing Isocab’s motive for a portfolio evaluation/reduction. What is driving them to reduce their portfolio, what does Isocab want to achieve? What does Isocab find important when it comes to portfolio evaluations?
2. Based upon the drivers behind the portfolio reduction, identified in the previous sub-question, which evaluation method identified in the literature review best suits Isocab’s needs?
Depending on what Isocab wants to achieve, different evaluation methods may be better suited for the task at hand than others. By identifying the main drivers behind the portfolio reduction, the research has identified what Isocab wants to achieve, what they find important and what not.
These drivers will serve as a guideline for selecting a portfolio evaluation method. Several methods will identified in the literature review, and this chapter will combine these with the results of the pervious sub-question.
This will serve as the basis for our analysis to determine which method is best suited for the job at Isocab.
3. How should Isocab put this evaluation method into practice?
Here, the research will elaborate upon the chosen evaluation method, and
come up with a concise approach of how this method is put into practice. A
segment of the portfolio will be evaluated as a worked out example. These
findings will be summarized in a final design proposal. This proposal will
elaborate on how Isocab could use the chosen evaluation method on other
segments of their portfolio, and serve as a reference for future projects.
PRODUCT MANAGEMENT
PORTFOLIO EVLUATION - METHOD AND CRITERIA -
PRODUCT PORTFOLIO
BUSINESS PERFORMANCE Strategic Business Unit
performance Leads to
Has impact on
Influence
MARKETING PRODUCTION FINANCIAL
ANALYSIS Sub-question 1
ANALYSIS Sub-question 2
DESIGN Sub-question 3
FIGURE 3: CONCEPTUAL MODEL AND SUB-QUESTIONS
2.3 RESEARCH OUTLINE
This research is divided into three parts. First, there is the literature review, followed by the main research, ending with the conclusions. This structure is illustrated in figure 4:
Chapter 1:
Introduction and origin of research
Chapter 2:
Research methodology and
set-up
Chapter 3:
Preliminary research A) INTRODUCTION AND
TOPIC EXPLORATION
B) MAIN RESEARCH
Chapter 5:
Design (sub-question 3)
C) CONCLUSIONS
Chapter 6:
Conclusions and recommendations Chapter 4:
Analysis (sub-question 1 + 2)
FIGURE 4: RESEARCH OUTLINE
2.3.1 PRELIMINARY RESEARCH:MOTIVATION AND CONTENTS
A preliminary research will be carried out prior to the study’s main research. This
preliminary research will consist of a literature survey that will allow us to get an initial
understanding of the topics involved. This means that several literary sources will be
used to get an idea of the research topic and what factors are in play. The preliminary
research chapter will therefore only summarize major findings relevant to the subject at hand; it will not provide an in-depth literature review. This general theoretical outline of the problem will be used to support and guide the research in the early stages. In the case of Isocab, we would like to know something about:
1. Isocab’s current portfolio and their means of producing it.
This topic is relevant the problem statement, as analyzing Isocab’s current portfolio allows us to narrow it down to a manageable segment. This segment should be small enough to be manageable as a sample, to be used as an example, though still significantly large to represent the portfolio as a whole. Such a demarcation will allow the research to stay within its scope and constraints, while still providing an answer to the research question. Also, Isocab’s production process is relevant, as product mix decisions are often influenced by current and potential production capabilities. Understanding the general outline of the production process therefore could provide some added perspective.
2. Potential product portfolio evaluation methods, from the three perspectives mentioned earlier in the report.
The second sub-question of the research asks which portfolio evaluation tool is best suited to Isocab’s situation, based upon their main evaluation drivers. Earlier in the introduction section, the report mentioned that there are three perspectives from which a portfolio can be set up and evaluated. These are the marketing, production/engineering and financial perspective. Different evaluation tools have been developed for these perspectives. Understanding these perspectives is therefore crucial for answering this sub-question. Before one can select a suitable tool, one must understand its strengths and weaknesses. The preliminary research will elaborate upon these perspectives and an evaluation tool for each – detailing their focus points, i.e. what these perspectives deem is relevant, and how the tools work.
3. A method of analysis to determine which of the previously identified evaluation methods best serves Isocab’s needs.
The second sub-question also requires us to select a tool suitable for
Isocab. A technique is needed to identify which tool is the best line with
Isocab’s objective. It is important for us to understand which technique
can assist in deciding on an evaluation tool. The technique must be able to
rank the evaluation drivers according to their relative importance. These
preferences can then be translated into which evaluation technique is best
aligned with Isocab’s objective. We will explore such a technique in the
preliminary research.
2.3.2 MAIN RESEARCH
This study will follow the Diagnosis, Design and Implementation approach as specified by De Leeuw (1996).
Diagnosis
The Diagnosis section can be seen as an extension to the preliminary research, and will continue to describe and analyze the current situation at Isocab NV. Where the preliminary research provides the study an initial introduction to Isocab in order to get a better understanding of the research objective, the diagnosis will analyse specific aspects of Isocab in order to answer the research question that follows the research objective. Chapter 4 will deal with the first two sub-questions. First, it is important to identify what Isocab aims to achieve by reducing their product portfolio; what drives the portfolio evaluation effort. The second sub-question will use a selection technique identified in the preliminary research to determine which one the portfolio evaluation techniques are best aligned with the drivers identified earlier. Once these two questions are answered, the diagnosis has ended and the report can move to the design section.
Design
In the Design section the report will describe the proposal best suited to Isocab NV in the reasoning of the author. Chapter 5 will build upon the results from the diagnosis. In this chapter, the study will elaborate on how Isocab should go about the portfolio evaluation process, with the technique previously selected in the diagnosis. This framework, taking shape as a decision model, will provide Isocab with a step-by-step approach. This will allow them to evaluate their product portfolio without neglecting any aspect relevant to the situation.
Implementation
The next step would be managing the implementation of the design into practice. This would imply offering the newfound product portfolio, and dropping any unattractive products. Implementing this would also imply dealing with the side effect from dealing with a new product portfolio. For example, if some products are not longer offered, then the production line would have to be rebalanced, in order to deal with the extra capacity.
Inventory management protocols, such as re-order points and safety stock, will have to
be adjusted, as some components will not be needed anymore while others will be more
important than before. Perhaps some staff will have become redundant, while other
positions might require and extra pair of hands. All these things and more have to be
taken into account when implementing changes to the product portfolio. Managing these
changes lies outside the scope of this research. This research will limit itself to offering
advice on how to perform a portfolio evaluation, outlining the steps that one has to go
through. This process will be illustrated by a worked out example, a segment of the
doors. This answers the research question from Isocab, explaining how to go about
performing an evaluation. Based upon this decision model, they will be able to evaluate
their full portfolio on their own, and fully implement their findings.
2.4 MANAGEMENT OF USED SOURCES OF DATA AND INFORMATION As with any research, this thesis stands or falls with the data that lies at its foundation.
This research project also relied on several information sources that might not seem reliable for an in-depth analysis. However, it is my understanding that these sources are appropriate nonetheless, as it is often tricky at best to find formal, ready-to-use data during “field work” like internships. Certain demarcations and assumptions have to made, based upon the information available. The following passages are meant to account for this, by explaining why these sources can be reliable enough to work with.
Literary source
Literature will be gathered through the use of the search engines of the University of Groningen: RUG Combine and Purple Search, and books relevant to the subject matter at hand.
Personal interviews with the people involved
The door production managers have been closely involved in all stages of this research.
Although they have expended a lot of time and effort in guiding me and giving me access to all the information I needed for my research, their opinions and insights are always
‘coloured’ to a certain extent. Nonetheless, the information they provided can be considered reliable, as they are knowledgeable individuals in their respective fields and the information they provided can be validated by others. In order to ensure this accountability, the information source has always been mentioned. Question asked during interviews were semi-structured, and closely related to the sub-questions.
Motivation of data sources
A lot of the information needed to make quantitative statements came from internal
information systems. Isocab has an extensive information system that keeps track of
almost everything. This made my research considerable easier on one level, while
making it harder on the other. The focus shifted from data gathering to data sifting; what
information is most relevant for this research? Of course, the information contained by
the system is only as accurate as the users. However, in Isocab’s case, the information
found in the systems was easily verified by the users when ambiguous and thus can be
considered reliable. This research only relied on information provided by and verified
by the system’s users.
3 PRELIMENARY RESEARCH
The first chapter introduced the reader to Isocab and the relevance of this research topic. This chapter summarizes the major findings from our preliminary research. As said earlier, this research consists of a literature survey that allows us to get an initial understanding of the topics involved. The chapter will only explore relevant topics; it will not provide an in-depth literature review. This general theoretical outline of the problem will be used to support and guide the research in the early stages. In the methodology, the following questions were identified:
1. Isocab’s current portfolio and their means of producing it. (see Chapter 3.1) 2. Potential product portfolio evaluation methods, from the three perspectives
mentioned earlier in the report. (see Chapter 3.2)
3. A method of analysis to determine which of the previously identified evaluation methods best serves Isocab’s needs. (see Chapter 3.3)
3.1 ISOCAB’S CURRENT PRACTICE
Chapter 3.1 will elaborate upon Isocab’s product portfolio, and its means of producing it.
This expands on the initial impression from the introduction earlier, and is helpful in a further demarcation of this research.
3.1.1 PRODUCT PORTFOLIO
As explained earlier, this report focuses only on the door-segment of Isocab’s portfolio.
This chapter will take a closer look at these products. Figure 5 illustrated Isocab’s
current product portfolio. As one can see, there are six categories of doors: Cool and
refrigeration swing doors, Office/factory doors, Flip-flap doors, slightly insulated sliding
doors, Heavy sliding doors and Lift sliding doors. These six categories each consist of
several models, and these models are all available in different types, depending on the
materials used. These types are also available in various sizes.
Cooling and refrigeration
swing doors Office/factory doors
Flip-flap doors Slightly insulated
sliding doors
Heavy sliding doors Lift sliding doors
One leaf swing door (D)
Double leaf swing door (DD)
One leaf service door (B)
Double leaf service door (DD)
Insulated flip-flap door (F)
Polyethylene flip-flap door (T) Polyethylene flip-flap
door (N)
Sliding door (E)
Slightly insulated sliding door (S)
Heavy sliding door (A)
Heavy sliding door (R)
Heavy sliding door (X-I)
Lift sliding door (H) Two types:
Positive temps (60mm) Negative temps (100mm)
Two types:
Positive temps (80mm) Negative temps (80mm)
Two types:
One leaf door Double leaf door
Two types:
One leaf door Double leaf door
Two types:
One leaf door Double leaf door
Two types:
Positive temps (60mm) Negative temps (100mm)
Two types:
One leaf door Double leaf door
Four types:
Positive temps (60mm or 100mm) Negative temps (100mm or 140mm)
Four types:
Positive temps (60mm or 100mm) Negative temps (100mm or 140mm)
Four types:
Positive temps (60mm or 100mm) Negative temps (100mm or 140mm)
Four types:
Positive temps (60mm or 100mm) Negative temps (100mm or 140mm)
FIGURE 5: ISOCAB'S PRODUCT PORTFOLIO
Note that this overview does not yet include the different materials and sizes that these doors are available in. Every door is available in various materials in 4 to 15+ different sizes. These sizes are not included in the overview as they lay outside of this research’s scope. This research will analyze the current portfolio in order to diagnose any unprofitable products or product groups. Offering these doors in different sizes does not impact production in any noticeable way, as the steel plates required are produced internally. All the other steps are basically the same. Producing an unprofitable product type however, has a greater impact on the production process, taking production time and scarce resources away from product types that are more profitable. Therefore, this research will aggregate the portfolio on the product type level. Appendix 2 contains an overview of the available standard sizes.
3.1.1.1 SCOPE OF THE RESEARCH
At this point in the preliminary research it is important to keep the research’s
constraints in mind. Evaluating every door category falls outside of the scope of the
research. As stated in chapter 2, methodology; a full implementation of the findings and
an evaluation of the complete door portfolio is not part of this research. The intent is to
offer Isocab a step-by-step approach, rooted in academic research, which will allow
them to evaluate their own portfolio. In order to achieve this goal, the research has picked one door category that will serve as an example for the other products in the portfolio.
According to table 1 on page 13,and in conjunction with interviews conducted at Isocab, further demarcation of the research is allowed. In absolute production numbers, the cold storage doors make up the bulk of the door production, followed by the company doors. Therefore, selecting the cold storage doors as an example for the rest would have two advantages. First, any improvement to the cold storage doors would immediately have an impact on the door production as a whole. The sheer volume results in this category having a greater impact relative to the others. Second, because this category is produced and sold in such high numbers compared to the rest, there is more production and sales data available for this research to work with, resulting in more accurate findings.
3.1.2 PRODUCTION PROCESS
As with any other manufacturer, the process starts with a customer that has a need for your product. The salesman identifies these needs and, in concertation with the sales department, offers the customer a quotation. Should the customer place an order, the sales department then passes the order to the production department. Once the order has been produced, it gets delivered to the customer and assembled on-site. So much for the general steps that guide the production process, let us now zoom in on the production department itself.
As an international manufacturer of industrial steel plating, Isocab performs most of the door production steps in-house. Let’s start at the beginning: the production of all steel plates starts with steel coils of various sizes. These coils then get unrolled and flattened to get the curve out. These flat plates get cut and folded into the right sizes. These plates then get transferred to the sheet-metal workshop, where they are worked according to the end-product they will be part of, like doors or sandwich panels. So far, the general steps have been the same for every metal sheet. Remember, this research focuses specifically on the production process of the doors; other products lay beyond our scope.
The door production process consists of 5 consecutive steps. First the metal sheets get composed, meaning that all the layers that make up the door panel are put on top of each other, insulated, etc. After that these layers get compressed to make up an actual door panel. These door panels are then prepared for assembly, like adding holes for door handles and locks. The production process ends with the doors being assembled and packaged. Interesting in how the production process is actually controlled. Isocab has two ways of using the production line, depending on the doors they are producing;
standard doors, or doors that exist in the brochure and have standard dimensions and characteristics, and special doors, which are built according to the customer’s wishes.
The special doors’ production process is pretty straightforward, at least in terms of the
production steps. As they are built to customer specifications, the production starts from
the very first production step, and gets pushed through the process after that. The production time varies greatly depending on the capacity and product complexity.
The standard doors’ production process is managed a bit differently. These doors are built to relatively standard dimensions and are in higher demand; they get ordered frequently. That is why Isocab maintains a constant WIP inventory of every standard door before the assembly step. Once the customer places an order, this sends a signal to the assembly step to start producing the ordered product. After assembly, the doors are immediately packaged and ready for delivery. So the assembly step gets fed from a constant WIP inventory. This inventory should always contain at around 10 units of every standard door. This might be 8 units, or could be 13; this depends on the current capacity and demand. Now, once the inventory reaches a certain threshold, this sends a signal all the way to the beginning of the production process. This allows the production of the depleted product, and should refill the constant WIP inventory before it runs out.
Managing their production process like this allows Isocab two things. It allows them to quickly deliver their products, while at the same time maintaining production efficiency.
Isocab’s production process is based upon the CONWIP principle. CONWIP, which stands for constant work in process, can best be explained as follows. A CONWIP system covers several of the production process steps. For all these production steps, there is a limited amount of WIP allowed, a WIP cap. When the preset amount of WIP is met, no new production jobs can be released into the CONWIP system; only after one of the jobs is finished and leaves the system, can a new job will be allowed to enter. This will result in a WIP level that remains very stable over time, and will never exceed a preset maximum:
hence the name CONWIP (Hopp et al., 2008). Figure 6 illustrates Isocab’s production process in the form of a value stream map. A full-page copy can be found in appendix 2.
Delivery
Finger-plate
composing
Door layer
Compressing
Compressed door layers
Preparing
Ready for assembly
Assembling Packaging
Finished doors
FIFO FIFO FIFO
Standard door stock +/- 10 units of every type
Customer order (standard door) Stock replenishment signal
Other production departments
Customer
Demands
Sales Production Expedition
On-site assembly
Salesman/
Foreman
order Quotation
Implementation
Customer service
Research scope Research scope
FIGURE 6: VALUE STREAM MAP
3.2 PORTFOLIO EVALUATION METHODS
Chapter 3.2 will focus on several product portfolio evaluation techniques found in relevant literature. A technique from every perspective will be chosen and introduced.
These techniques will be used in the analysis, where the study will determine which fits best with Isocab’s goals.
3.2.1 MARKETING PERSPECTIVE
Product portfolio planning problems have generally been considered from a dominantly marketing perspective. In this perspective, the focus lies on customer concerns, such as analyzing how different sets of product attributes and options interact and compete amongst each other in the targeted market segments. As a result, the primary focus of marketing research lies on the measurement of customer preferences with regards to alternative sets of product attributes. When it comes to product portfolios, the goal is to find a balance of products that together achieve the maximum gains on the long run with minimum use of scarce resources (Day, 1977).Therefore, a product should only be dropped if its long term contribution to profit is below what the company could attain from producing an alternative with the same resources.
Day (1977) argued that common to all portfolio portrayal techniques is the understanding that the value of a competitive market share depends on the structure of the competition on the one hand and the stage of the product life cycle on the other. This is as true today as it was then. The earliest and most commonly used tool is the share/growth matrix that was developed by the Boston Consulting Group (see figure 7).
This 2x2 matrix classifies an organisation’s products on two dimensions. The vertical axis represents a products current or forecasted market growth. This serves as a proxy measurement for the product’s life cycle. Early on in the product’s life cycle, the growth rate will be fast, only stabilizing later in the ‘mature’ phase. The horizontal axis represents a measure of market share dominance. This serves as a proxy for a product’s profitability (Day, 1977). Together, these two dimensions provide an excellent platform for a product portfolio analysis and serves as a basis for strategy evaluations.
FIGURE 7: THE BCG GROWTH-SHARE MATRIX [SOURCE: DAY (1977)]
After analysing the current portfolio, analysts can plot the products on a scatter plot.
This allows an organisation to decide which products to fund (and how much) and which products should be discontinued. It is a portfolio balancing tool, combining profitability with a product’s life cycle. These two dimensions can be translated into a product’s role, which helps a company set product management strategies.
For Isocab, where the product portfolio has grown over the years, it could be a very valuable exercise to analyze their product’s roles. Some of these products have been offered for quite some time, and compete in a very mature and saturated market, while their market share has declined. It would be interesting to check to what extent their portfolio is balanced. Such an exercise would provide valuable information on their portfolio composition with regards to market performance. It could also focus attention and funds away from the dogs to where they are most needed, extending the life of cash cows and nurturing stars.
3.2.2 PRODUCTION PERSPECTIVE
This chapter will deal with product portfolio planning from a production or engineering perspective. This perspective is perhaps less obvious than the other two, but nonetheless has a great impact on production management overall.
Production management is a part of operations management, and is responsible for maintaining a manufacturing process that is both efficient and effective. A manufacturing process is efficient when it uses the minimum amount of resources needed, and it is effective when it meets the costumer’s requirements (Aquilano et al., 2006). This logically implies that, in order to sustain a long term competitive advantage by combining two strategies, it is necessary to set up an optimal product mix (Jiao and Zhang, 2005).
The act of balancing a product’s distinctiveness (customer intimacy) versus low costs through commonality (efficiency) is done through the use of product platforms. Muffatto (1999) has defined a product platform as follows:
“A relatively large set of product components that are physically connected as a stable sub-assembly and are common to different final models”.
To put this into laymen’s terms, it means that a product platform can be any technology, design, module etc. that can be shared among different products within the same product families. With a product platform, companies are able to manage the basic trade-off between distinctiveness and commonality (Ulrich et al. 2007).
From such a production or engineering perspective, product portfolio decisions have to be made based upon costs and flexibility in a multi-product manufacturing environment;
how much does it cost to produce a certain set of products due to the inherent
complexity, and how do their interactions influence the production process in general
(Jiao and Zhang, 2005)? In practice, this comes down the analyzing the economies of
scale that can be achieved from commonality among products, compared to the marginal
profit lost from the lack of distinctiveness.
In the case of Isocab, where the product portfolio has grown over the years and where it is no longer as feasible to offer this many variations, it becomes necessary to identify which variation contributes the least to the door department’s profitability. In theory, the best practice would be to map the entire cost structure of all products, and from there determine which is the least efficient to produce, i.e. benefits the least from economies of scale and comes with the most production complexity. In practice, it is often very time consuming and downright impossible to map the costs for so many products with so many parts. A shortcut is needed.
In his work, Collier (1982) introduced the degree of commonality index to calculate the degree of commonality between product structures. The benefits of commonality can be summarized as:
1. simpler inventory management: fewer parts and tools that have to be managed 2. less orders have to be placed with suppliers (higher quantities per order perhaps,
but total quantity order is lower)
3. reduction of safety stock needed, and thus reducing inventory costs
As commonality leads to the most cost efficient and easiest to produce product portfolio, identifying which product has the least in common with its product platform can be used as a shortcut product mix evaluation tool. If it is certain that products have to be dropped from the product mix then it makes sense, from a production perspective, to focus on the products that contribute the least to economies of scale and are the most complex to produce, such as indicated by a low degree of commonality.
3.2.3 FINANCIAL PERSPECTIVE
From a financial point of view, it has been argued that product mix decisions bear some striking resemblance to financial portfolio decisions. When performing a portfolio analysis, an investor wants to know which portfolio will provide him with the most optimal combination between risk and return (Wind, 1972). Investments are always evaluated together as a whole, as a portfolio. While a portfolio’s return is of course directly related to the individual investments’ return, its riskiness is not. Risk is pooled across the entire portfolio. The same applies to product mixes: spreading out across different products and markets minimizes risk. Logically, investors seek out high returns while avoiding risk as much as possible.
In his work, Wind (1974) noted how similar the goals and portfolio evaluation decisions
of an investor and an organisation actually are. Both an investor and an organisation are
looking for the optimal portfolio balance (i.e. what proportion of the total investment (or
product mix) should be allocated to each securities (or product)) in order to achieve the
best risk return ratio. Even though there still are some differences between securities
and products (for example, alternative products to offer are rarely readily available), but
these do not negate the usefulness of financial portfolio analysis techniques for product
mix decisions. According to an empirical study conducted by Cardozo et al. (1983),
portfolio theory has found its way in commercial product mix decisions, albeit in a
modified form.
Rabino et al. (1985) have moulded the financial portfolio theory into a practical form that allows it to be used more easily for product mix evaluation decisions. They argued that it is also necessary to look at the other impact that products have on others in the line, something that traditional risk/return calculations not fully capture. Rabino et al.
have identified three contexts in which product evaluations are required, and which technique is most suited (see table 2).
TABLE 2: PRODUCT EVALUATIONS [SOURCE: RABINO ET AL. 1985]
Decision Context Method of Analysis
- Screening new products (identifying risk and returns) - Identifying and diagnosing
“problematic” products (e.g.
products with unstable patterns of sales growth, market share performance, etc.)
- Monitoring existing products (Comparing actual performance to budgeted or planned performance)
- Product Beta or NPV
- Product contribution and residual income
- To a lesser extent, product Beta
- Product contribution and residual income
- Product Beta
This study aims to come up with a method of identifying sick products at Isocab.
According to table 2, this is best done through the calculation of production contribution and residual income. However, such a financial evaluation requires time and expertise, which is not always available. Isocab, which is currently redefining its product mix, requires a short, quick but reliable evaluation method. Rabino et al. suggest that the product beta concept can be applied with relative ease, providing a quick measurement of the riskiness of offering a certain product. It also takes into consideration the uncertainty of future cash flows of both the product and the product line (Rabino et al.
1985). The main benefit of using the product is its ability to judge a portfolio’s risk and return as a whole. This allows a very straightforward approach to portfolio management.
3.3 SELECTION TECHNIQUE
So far we have identified three different perspectives on evaluating a product portfolio,
each with its own techniques and strong points. The following step in this research was
to decide which one is most suited for Isocab’s situation. In other words, which method
or point of view best captures the wide variety of aspects that come into play. However,
bounded rationality and our own limitations make it impossible for a decision maker to
truly keep all factors in mind. To overcome these, this research turned towards multi-
criteria decision methods (MCDM methods). These are techniques that allow a decision
maker structure a decision problem in such a way that it becomes easier to solve. By
breaking up a problem into a series of smaller sub-problems, the initial problem
becomes less daunting and easier to handle. Over the years, many methods have been developed and, based upon literary findings, have been applied successfully in many fields of study. In this chapter, we will have a closer look at several of these methods, elaborate upon their strengths and weaknesses, and motivate which one is best suited for this decision problem.
3.3.1 POPULAR MCDM METHODS: A SHORT INTRODUCTION
These are several popular MCDM methods out there, such as AHP (and its ANP derivative), ELECTRE, MacBeth, Topsis, etc. (Montevechi et al., 2001). All these MCDM methods have been developed with the same goal in mind: when multiple criteria have to be considered, which alternative is better? Essentially, these methods use the same general steps: structuring the problem, weight valuation, weight aggregation and a sensitivity analysis (Labib et al, 2009). They all use a decision matrix, which represents how each alternative performs according to each criterion. The methods differentiate in how they approach the decision matrix. Some methods (eg. ELECTRE) just provide a ranking of the alternatives, while others (eg. AHP) measure the alternatives performance per criterion, and then provide a ranking based upon these performances (Montevechi et al. 2001).
In this chapter we will compare several MCDM methods to one another, in order to find the most appropriate method for Isocab’s situation. This evaluation is performed for ELECTRE, AHP and its variant ANP. Other methods do exist, but this short comparison was limited to the most widely known and reported methods. These methods were chosen based upon their popularity in trade journals and industry reports. Important to note is that extensive comparisons have been conducted in the past. These comparisons have given rise to a general criticism to MCDM methods, namely that different approaches to the same problem may yield different results. However, it has been proven that the results from AHP/ANP and ELECTRE do not differ significantly from one another. This is important, as it allows us to compare these methods to each other without influencing the research’s final results (Montevechi et al. 2001, Dublish et al.
1998). Our comparison simply aims at finding an appropriate method for Isocab to apply; we do not dispute that each of these methods could get the job done.
ELECTRE
The ELECTRE method is the brain child of Bernard Roy, which he developed as an alternative decision making solution method. The ELECTRE method has grown into several versions, though they are all based upon the same reasoning and concept. What sets ELECTRE apart from other methods (such as the AHP and ANP, discussed further in this chapter) are the concepts of indifference thresholds and outranking (Buchanan et al., Dublish et al, 1998).
Traditionally, MCDM methods take a very numerical approach to decision problems. For
example, if option A has a 5% higher preference score than option B, than logically
option A is better and stands on top of the ranking. However, is 5% difference really so
significant? As a decision maker, will you be able to convince yourself that this small
number makes option A superior? The ELECTRE method tries to solve this decision dilemma by introducing indifference threshold. These thresholds are set by the decision maker, and act as a pivot point, where a decision maker is either indifferent to the solution or strictly prefers it. In practice, it is common to place two thresholds (one for indifference and one for preference) with an indecision buffer zone between them. By using these thresholds, the ELECTRE method incorporates the decision maker’s real life uncertainty, and tries to work around it.
By using such thresholds, ELECTRE then develops an outranking relation between the alternatives. It starts with the hypothesis “alternative A is at least as good as alternative B”, and does this for all criteria. ELECTRE then proceeds by testing for whether or not this hypothesis holds true, for every alternative and criterion. This is done through the calculation of the concordance index, providing a numerical interpretation of how often this hypothesis was true. For example, a concordance index of 0.80 for alternative A and B implies that option A was at least as good as option B, for 80% of the criteria. These concordance indexes allow the ranking of alternatives according to preference.
ELECTRE’s strengths and weaknesses
Like every multi-criteria decision analysis method, ELECTRE has its strengths and weaknesses. The following table lists the most important ones (Buchanan et al, Kangas et al. 2001):
TABLE 3: ELECTRE STRENGTHS VS WEAKNESSES
