Designing Critical Success Factors for benchmarking maintenance

(1)

Designing Critical Success

Factors for benchmarking

maintenance

A CONTINGENCY APPROACH

(2)

Designing Critical Success Factors for

benchmarking maintenance

A CONTINGENCY APPROACH

Author : Niels J. Tolboom

Study : Msc Technology Management Student number : 1336762

Email : nj.tolboom@gmail.com

Faculty : Faculty of Economics and Business University Supervisor : Prof. dr. ir. G.J.C. Gaalman

University Co-assessor : Ir. J.Ph.C. Wubben

Company : Heineken Supply Chain Services Company Supervisor : A. Klaver, Maintenance Specialist Company Manager : B. Hut, Manager Packaging Solutions

Date : September 2010

(3)

Management summary

The maintenance section within the Central Engineering department of Heineken worldwide in Zoeterwoude has recently expressed its desire to build a benchmark method for their respective area. This benchmark method should be able to identify the best in class breweries and share their best maintenance practices with other Heineken breweries. In order to improve the Critical Success Factors (CSFs) for benchmarking maintenance that Heineken selected themselves, they have asked to propose a new design of CSFs based on a literature study. The main research question of this research project therefore is: What Critical Success Factors should Heineken use for benchmarking maintenance performance of its breweries worldwide?

Both the maintenance industry, mainly through maintenance consultancy firms, as well as maintenance literature provide many models that subscribe a static set of CSFs that every maintenance organization should apply when comparing maintenance performance of its production plants for benchmarking. The main disadvantage of these ‘static’ models is that they do not integrate the maintenance strategy and other important characteristics of an individual maintenance organization. Although companies clearly pursue different strategies relating to maintenance, both the maintenance industry and literature claim that they should base their maintenance performance measurement on similar CSFs. The approach that is applied to propose a design of CSFs for Heineken is therefore to customize the selection of CSFs to the specific maintenance strategy and other characteristics of Heineken.

The contingency approach claims that the optimal course of action is contingent (dependent) upon the internal and external situation. As the character of the contingency approach is clearly very similar to the selected approach for designing the CSFs as just been described, the contingency theory forms the dominant approach to design a unique set of CSFs for Heineken’s maintenance benchmark method. With the help of system metaphors, the environment, culture, organizational structure, life cycle, technology and strategy contingency factors are identified as being relevant in this situation.

(4)

(5)

Preface

This Master thesis forms the final hurdle in the graduation procedure for the master degree program of Technology Management at the University of Groningen. It is the result of a research project that has been conducted at Heineken Supply Chain Services in Zoeterwoude and that has started in February 2010. During a period of about seven months, I have been involved in a project on developing a benchmark method for comparing maintenance performance of breweries worldwide and I will continue to be a part of this project for the coming months.

Being part of the dynamic environment at the headquarters of an international company like Heineken has been a great experience for me. Besides challenging my academic competences through my research project, I have also been able to learn how to adopt an attitude resulting in tangible benefits for the company by being part of the project team.

Together with Alex Klaver, we started an exciting project from both a theoretical and practical perspective. I sincerely want to thank Alex for giving me the opportunity to become part of the benchmark project team and providing me relevant input for this research project. Moreover, I want to express my sincere gratitude to Berend Hut, for providing me input to my research project in the form of an interview and for offering me the possibility to work for Heineken as both an intern and currently as temporary employee. I would also like to thanks Chris Versteegh for providing valuable input for my main analysis. I am confident that their supervision improved my work on this project in many ways.

Two other persons that deserve attention are University Supervisor Mr Gaalman and University Co-assessor Mr Wubben. As my supervisor, Mr Gaalman has provided the guidance and support a Master student seeks during his research project. The enthusiasm and critical attitude that Mr Gaalman has shown have motivated me to aim for the highest possible result during this project. I would also like to thank Mr Wubben for showing his commitment in such an early stage in this research project. In doing so, he has spent much more time on this project then I could have asked from him. Finally a note of thanks to my dearest friends and family who supported me throughout this project.

Enjoy reading this report.

Zoeterwoude, September 2010

(6)

(7)

List of abbreviations

TPM Total Productive Management SCS Heineken Supply Chain Services department TQM Total Quality Management OPCO Operating company within Heineken JIT Just in Time principle MTBF Mean Time Between Failures

WO Work Order BSC Balanced Scorecard HoQ House of Quality CSF Critical Success Factor

Vocabulary

Best in class Highest current performance level in an industry, used as a standard or benchmark to be equaled or exceeded.

Best practices Methods and techniques that have consistently shown results which are superior than those achieved with other means and which are used as benchmarks.

Consignment stock Consignment stock is stock legally owned by one party, but held by another. Ownership of consignment stock is passed only when the stock is used (issued).

Critical Success Factors Limited number of characteristics that have a direct and serious impact on effectiveness, efficiency and viability of a maintenance organization (e.g. maintenance cost).

Detagging Removing tags that have been placed by operators on machinery that require maintenance.

Draught keg Five liter keg filled with beer to produce a draught beer.

(9)

(10)

1 Introduction

Heineken currently holds varying degrees of ownership in about 125 breweries worldwide. These breweries belong to the 93 operating companies (OPCO’s), usually one per country, within Heineken. Most of these breweries are fully owned by Heineken, but in several occasions Heineken only holds a minority share of less than 50%. The latest large acquisitions of Heineken, the Scottish & Newcastle company in 2008 and the FEMSA company in 2010, show that the size of Heineken’s organization is steadily growing and it has been doing this for the past 145 years.

The Heineken Supply Chain group is responsible for the main process that takes place inside these breweries (see Figure 4-1 Main production process of a Heineken brewery). Heineken Supply Chain Services, one of the main departments within the Heineken Supply Chain group structure, is responsible for providing all kinds of services to the breweries worldwide relating to the operational processes. Among their activities are consultancy work, technical assistance, procurement, providing training programs, research and development for raw materials, processes and products and finally bringing in knowledge and experience during project management of expansions or the building of new breweries.

(11)

2 Research framework

The research framework will commence by analyzing the problem situation. This is done by providing information on what will be designed during this research project, for whom it will be designed, and inside what framework of conditions and restrictions the design should emerge (De Leeuw, 2001).Thereafter, the research methodology and use of data is discussed.

2.1 Problem statement

This research project aims to produce a knowledge product according to the needs of its customer, Heineken. The problem statement will therefore focus on describing what should be designed for whom and for what reasons (de Leeuw, 2001):

What Critical Success Factors should Heineken use for benchmarking maintenance performance of its breweries worldwide?

The Critical Success Factors (CSFs) form a fundamental element of a maintenance benchmark method that is being constructed in parallel to this research project at the SCS department. The comparison of maintenance performances of breweries is based on these CSFs. The Maintenance Specialists of the SCS department have already selected a set of CSFs for this project based on their experience with measuring and benchmarking maintenance performance. In order to have a theoretical validation of these CSFs, the SCS department has asked the author of this research project to review these CSFs with the help of literature.

Figure 2-1 Heineken design of CSFs

(12)

the Installation performance, SHE management and Care of assets CSFs are part of the Output perspective. Chapter 10 will give a more detailed explanation of the content of these CSFs.

The design of the CSFs and therefore the benchmark method itself will have several different customers of which the breweries will be the most important. Benchmarking generally results in the identification of best practices and by sharing them these will enable the underperforming breweries to improve their maintenance performance. The SCS department is not only responsible for the creation and management of the benchmark method, it is a customer as well. As it is their goal to provide consultancy services towards the breweries on production related issues, insights in best practices that can improve maintenance performance can help them improve their services towards the breweries.

2.2 Conditions

The conditions of this research project provide an indication of the limitations that apply to this research project. Conditions in practical research are twofold; there are conditions regarding the result of the research project and conditions regarding the research process itself (De Leeuw, 2001). The conditions regarding the result of the research project, apart from what has been discussed in the problem statement, are difficult to define beforehand because it requires an overview of the proposed design. This section will therefore concentrate on the conditions for the research process itself, whereas the conditions regarding the design of the solution are integrated in part two, the design phase. The ‘Orde voor Organisatie Adviseurs’ (OOA) has produced a code of conduct that includes several issues that have to be taken into consideration when performing research or consultancy work for organizations:

o The period in which the design of the CSFs for the benchmark method should be completed only stretches until September of 2010. As the implementation of the benchmark methodat Heineken will require a substantial amount of time and because the first phase of the project should be finished before the end of 2010, the design of the CSFs must be finished three months in advance of the first benchmark. Consequence of this condition is that the design is thoroughly tested in the field and that the implementation phase of the design will not be included in this research project.

o All aspects of this research project has been carried out by the author himself. There is no budget available to hire experts to assist in some parts of the research project.

(13)

2.3 Research process design

The design of this research project is based on the methodology of design-focused research (van Aken, 2004). The mission of design-focused research is to develop scientific knowledge to support the design of interventions by professionals and to emphasize its knowledge orientation (van Aken, 2004). The design paradigm differs from the traditional explanatory paradigm of research methodology in that it does not only aim at describing, explaining and predicting in order to understand the setting of construction or improvement problems. Instead, it focuses on developing and testing solutions to problems in the field in question.

Figure 2-2 Research process design

(14)

Whereas purely scientific research usually follows the logic of the empirical cycle, practical research most often follows the steps of van Strien’s regulative cycle (van Aken, Berends and van der Bij, 2007). The regulative cycle consists of the diagnosis, design, implementation and evaluation phases. This research project will only focus on the diagnosis and design phase. The change phase will only marginally be involved when the improvements to the Heineken CSFs are clear and the conclusion of the research project is given.

The goal of the diagnosis phase is to produce a problem statement (De Leeuw, 2001). The diagnosis phase (see Figure 2-2 Research process design) commences with the introduction in chapter 1, which will introduce the problem situation. The research framework is given in chapter 2 and will provide the problem statement, conditions, the design of the research process and finally the methods of data collection. An introductory literature research in chapter 3 will introduce the maintenance and benchmarking process. Subsequently, the company at which the research project takes place will be introduced in chapter 4. Chapter five will conclude the diagnosis phase by integrating all important variables in this research project into a research model.

The design phase, whose goal it is to design a solution to the problem (De Leeuw, 2001), commences with a Pluriform analysis in chapter six that will help identify important issues for the design of the CSFs. By describing and introducing the Contingency theory that will form the dominant element in the theoretical framework, chapter seven discusses the theoretical concepts. The theoretical framework, which is explained in chapter eight, will then be applied to the situation of Heineken in chapter nine. The Heineken CSFs will also be introduced in chapter ten. Finally, the comparison between the proposed and Heineken design of CSFs together with the discussion and conclusion in chapter eleven, twelve and thirteen will finalize the design phase.

2.4 Data collection

The question where the data must be gathered for the research project is critical here. When answering this question, one must distinguish two different factors concerning data collection. It is not only important to specify what the source of the data is, the way the data is observed or measured is just as important (De Leeuw, 2001).

There are several different kinds of sources of data when performing practical research. According to De Leeuw (2001), design-focused research is generally based on desk research and real business situations and this particular research project forms no exception. Obviously, documentation in the form of scientific literature and existing Heineken documentation will play an important role. Documentation will not only provide input for designing a theoretical framework for this research project, it also helps in determining the most suitable research process. Thereafter, the actual and real business situation at Heineken will be an important sources of data as well in applying the theoretical framework to the situation at Heineken. This latter source of data will also provide input for formulating the problem statement of this research project.

(15)

(16)

3 Introductory literature research

Before going into details on some of the subjects that have been mentioned in the problem statement in section 2.1, the most important processes that are included in the problem statement from the perspective of literature will be introduced here.

3.1 Maintenance process

According to the Technical Committee CEN/TC 319, part of the European Committee for Standardization (CEN), the maintenance process can be defined as ‘a combination of all the technical, administrative and managerial actions during the life cycle of an item intended to keep it in, or restore it to a state in which it can perform the required function’. According to Djerdjouri (2005), the importance of maintenance is related to the fact that poor maintenance can lead to more frequent equipment failures, bad utilization of equipment, delayed production schedules, higher product variability and hence lower product quality.

Maintenance operations are traditionally divided into three main groups: (1) corrective maintenance, (2) preventive maintenance and (3) improvement maintenance (Komonen, 2002). Corrective maintenance, according to the European standards of CEN/TC 319, is defined as the maintenance activities carried out after a fault has occurred and intended to put an item into a state in which it can perform a required function. Preventive maintenance on the other hand, are those activities carried out at predetermined intervals or according to prescribed criteria and intended to reduce the probability of failure or degradation of the functioning of an item (EN 15341:2007). The activities concerning improvement maintenance focus on improving the availability of an item, without changing the required function.

Different maintenance methodologies have been developed in order to manage the complexity and criticality of technical systems and their functions. An example of an established maintenance methodology is Total Productive Maintenance or Total Productive Management (TPM) (Nakajima, 1988). This methodology emphasizes continuous improvement and the close cooperation between different stakeholders such as production, maintenance, and system design (Soderholm, Holmgren and Klefsjo, 2007). Other popular maintenance concepts are Reliability Centered Maintenance (RCM) and World class maintenance.

(17)

The relevance of performance measurement systems for the maintenance function is obvious. Nowadays maintenance budgets are large, maintenance related cost are at least as high as the direct costs included in these budgets and maintenance labor cost increase faster than the total operating cost index. Furthermore, the impact of maintenance on production and operation aspects such as capacity, quality, costs, environmental and employee safety is considerable, according to Pintelon and van Puyvelde (1997).There are several reasons why maintenance performance measurement is more difficult than for other processes. First, as maintenance and production are closely related because the former provides a service to the latter, the merits and shortcomings of the service are not immediately apparent. Another issue is the time lag effect, which makes it difficult to specify the amount and intensity of the service needed for ensuring proper plant performance. Processing data to obtain useful management information also is a time consuming business (Pintelon and van Puyvelde, 1997).

3.2 Benchmarking process

In this particular research project, benchmarking focuses on enabling breweries to learn from each other. Therefore, Watson’s (1993) definition of the benchmarking process is most suitable here: ‘a process of organizational adaptation where the focus is not simply on copying others but on learning how to improve organizational performance by sharing ideas’ (p.1).

Many different types of benchmarking have emerged in literature. All of these new types are a variation of the four essential types of benchmarking, defined by Zairi (1994):

o Internal benchmarking: intra organizational exemplars of replicated activities provide a trigger for improving anomalous performance

o Competitive benchmarking: an organization’s business practices are re-evaluated in the light of knowledge that their primary competitors have been observed to demonstrate superiority in some important elements of performance

o Functional benchmarking: an organization’s business practices are re-evaluated in the light of knowledge that non-competitor organizations demonstrate superiority in some common elements of business practice

o Generic benchmarking: an organization’s business practices are purposefully compared with organizations having demonstrably superior performance from broadly similar practices or dispositions

The reason why companies like Heineken use benchmarking in their production process is described by Asrofah, Zailani and Fernando (2010): ‘a better understanding of strengths and weaknesses of processes, improved cycle time, improved supplier’s management, reduced production costs and overcoming paradigm blindness’ (p.121). Soni and Kodali (2010) add some more detailed advantages of benchmarking in a production context:

o Provides a framework for making not just incremental, but also quantum leaps in improvement

o Provides a structured approach to meeting goals and objectives which link into the overall mission or vision of the organization

(18)

o Can help to motivate the team

o Provides opportunities to develop staff by broadening their knowledge, expanding horizons and offering them opportunities for wider experience

o Breaks the introspective approach to improvement

o Creates pride at the place of work, because it is seen to provide the best service, product or practice

The benchmarking process can be described with help of Figure 14-2 Figure 14-2in Appendix B by Binder, Clegg and Egel-Hess (2006). This model is mainly focused on the implementation of a benchmark and consists of three phases: planning, realization and implementation. During the planning phase, the goals, project team are determined first. Following, the process, benchmark partners and type of benchmarking will have to be agreed. The planning phase is finalized by selecting the Critical Success Factors of the process that will be benchmarked and the development of clusters of object types like factories (breweries) or production lines. During the realization phase, the actual benchmark is carried out. First, the performance metrics on which the comparison will take place must be determined, followed by the collection of the data. After all the data has been gathered, the data analysis can take place and best practices can be defined. The final implementation phase concentrates on implementing the best practices in the organization.

(19)

4 Company introduction

Riis et al. (1997) claim that it is impossible to discuss a maintenance system meaningfully without first having clearly defined what the object system is. This chapter aims to do just that.

4.1 Main production process

The main process which takes place inside the Heineken breweries is visualized at a high level of aggregation in Figure 4-1 Main production process of a Heineken brewery. Incoming goods like barley, malts and hops are brought in to produce a brew. These raw materials are then filtered out of these brews in the filtration phase. After filtration and fermentation with the help of yeast, the brew has transformed into beer and is ready to go to the packaging process. Here the beer is filled into various forms of cans, kegs and bottles and made ready to be transported to the customer. Utilities has a supporting role to these processes by providing water, energy, CO2, electricity and steam.

Figure 4-1 Main production process of a Heineken brewery

4.2 Maintenance organization

(20)

installation management. Subsequently, the departments indicated by number three, the Central Technical Support department and Maintenance Group, carry out all tasks relating to preventive maintenance. The shift technicians and operators, indicated by numbers one and two, are not part of the official maintenance crew but do carry out simple maintenance tasks concerning corrective and autonomous maintenance.

The corporate organization of maintenance is relatively small. Heineken Supply Chain Services, which is part of the Heineken International head office, has the role of a central engineering department for all breweries worldwide. Within this department, a maintenance specialist team is present that supports the breweries worldwide on maintenance related issues. This team is also closely involved in this particular research project.

4.3 Total Productive Management within Heineken

After a worldwide benchmark study with competitors, which showed substantial improvement potential for Heineken in terms of productivity and cost leadership, the Heineken Board of Directors decided to make Total Productive Management (TPM) a company strategic program in 2003.

The new created Heineken corporate TPM office created a roadmap to implement TPM in all breweries worldwide. To monitor the implementation of the TPM program, the TPM corporate office audits all breweries on their TPM progress twice a year. The structure of the TPM organization is visualized in Appendix D . Separate TPM steering committees have been organized at the corporate, OPCO and brewery level. These committees monitor the progress made on implementing TPM. They consist of a team where each pillar (visualized in Appendix E ) is represented by one person. These nine Heineken TPM pillars are applied to coordinate the implementation of TPM and structure the activities related to TPM.

(21)

5 Research model

The research model will visualize the problem situation and its most important variables. The research model is build according to the universal ‘steady state’ modeling language of In ‘t Veld (2002). He provides a language that can be applied to describe any business process, independent of its field. By doing so, the problem situation and its variables are integrated into an overview together with the dominant process in this research project, the maintenance process. The fact that generic functions like input, output and transformation will be identified will also make the model easier to understand for non-maintenance experts.

In order to gain a clear sight on the different perspectives of the problem situation, several system metaphors have been included in the research model as well. The metaphors enable one to have a pluriform look at the problem situation. The metaphors and their specific influence on the selection of contingency factors and CSFs will be discussed in more detail in chapter 6.

5.1 Explanation of research models

The research model is split up into two different figures. Figure 5-1 First research model provides a high-level black box approach of the main production process within breweries and a supporting process, in this situation the maintenance process. The main production process is further elaborated in section 4.1. As the specifications of the main production process do not play an important role in this research project, this process is projected as a black box. The maintenance process does play an important role though and is therefore projected in more detail in Figure 5-2 Second research model.

Figure 5-1 First research model

(22)

levels. The lowest level only concerns the maintenance process and all the activities that belong to this process. The brewery level is the intermediate level and includes all the activities that are carried out within a brewery. The highest aggregation level in the research model is the corporate level, which are the activities that take place at the head office of Heineken, where all the breweries worldwide are managed and controlled. The different functions of the research model that have been included will be described starting with the lowest aggregation level.

Figure 5-2 Second research model

Input

According to Pintelon and Wassenhove (1990), the input of a maintenance process is generally in the form of manpower, work orders and cost. As Hoekstra (2008) describes, the input for a work order can be provided by a work request, breakdown or by a preventive maintenance plan.

Planning

(23)

Maintenance execution

The maintenance execution step is similar to the transformation function in the generic steady state model. The transformation function converts the input into the desired process output (In ‘t Veld, 2002). For example, a maintenance mechanic fixes a packaging machine, based on the work orders that have been planned for that particular day.

Evaluation

According to In ‘t Veld (2002), the second filter function should test the output of the transformation function on quality. If the output does not conform to quality standards, it could lead to rejection. It could then be send to the environment as waste or it can be repaired by bringing it back to the transformation function again. In Figure 5-2 Second research model, the second filter function is carried out by an evaluation of the quality of the maintenance execution. The quality standards are deducted from the initiating function again. When the quality of a certain action is insufficient, the evaluating function can determine to carry out the activity once more and thereby repair the quality of the activity or remove the activity out of the system. In a Heineken brewery, the evaluation will usually be carried out by a chief mechanic, based on a written report of the mechanic itself on the repair activity that he has been doing. The chief mechanic can decide to approve or disapprove the report, based on the quality standards from the maintenance policy. In the latter case, the responsible mechanic will have to carry out additional tasks to complete the repair.

Output

The output of a maintenance process comes in the form of a certain plant condition and plant performance according to Pintelon and Wassenhove (1990).

Measurement function

The measurement function should check whether the norms are still correct and whether they should be improved (In ‘t Veld, 2002). In this particular situation, the measurement function measures the performance of the maintenance process based on a fixed set of CSFs. These CSFs are determined by the evaluating function, represented by the performance analysis in Figure 5-2 Second research model. In a Heineken brewery, this step would be similar to the maintenance data system (e.g. SAP) which stores all the data concerning maintenance and translates them to quantifiable performance indicators. For example, it can measure and quantify to what extent spare parts management has been outsourced.

Maintenance policy

(24)

activities. The evaluating function, represented in Figure 5-2 Second research model by the performance analysis, will periodically confirm or disconfirm that the brewery has actually outsourced 50% percent of its civil maintenance activities. The evaluating function will also provide information on whether 50% percent is still the value Heineken should aim for.

Performance analysis

Translating vague demands of the environment into practical norms can only be useful when these norms are frequently measured and evaluated (In ‘t Veld, 2002). In ‘t Veld (2002) claims that changes in the demands of the environment and changes within the system itself can be the reason for adapting the norm. This is the responsibility of the evaluating function within the generic steady state model, which is represented by the performance analysis in Figure 5-2 Second research model. When the norms are no longer correct or are obsolete, the evaluating function will pass this on to the initiating function.

(25)

Corporate Strategy

The strategy of Heineken is created at the highest level within the company and therefore lies in the corporate aggregation level in Figure 5-2 Second research model. In the generic steady state model of In ‘t Veld (2002), the corporate strategy has the initiating function of a higher echelon. The corporate strategy will not be sufficiently detailed and customized that it can be applied to create maintenance norms directly. The initiating function therefore has to translate these high-level goals into a maintenance policy.

Corporate maintenance benchmark

As already explained above during the evaluating function part, the Heineken Supply Chain Service department, which can be perceived as the central engineering at head office, will be responsible for carrying out the corporate maintenance benchmark based on the performance analysis of all the breweries worldwide.

Contingency factors

The environment function in this situation is represented by the contingency factors just discussed in the performance analysis section. In ‘t Veld (2002) claims that the environment provides signals to which the system should adapt by changing the norms.These contingency factors play an important role in this research project and will be discussed in more detail in later stages. The motivation for the selection of these specific contingency factors will be discussed in chapter 6.

System metaphors: Brain, Machine, Culture and Organism

The system metaphors that have been included in figure 5.2 are the brain, machine, culture and organism metaphor. With help of these metaphors, the most relevant contingency factors have been selected. An analysis of what metaphors have led to the selection of which contingency factors is provided in chapter 6.

5.2 Sub questions

The research model has provided insight into the problem situation and most important variables. Therefore, the sub questions that this research project will have to provide an answer to can now be formulated:

1. What contingency factors can be identified and how do they influence the design of the CSFs?

2. What theoretical framework can be used to design the CSFs for benchmarking maintenance performance?

3. What is the Heineken design of the CSFs?

4. What are the differences between the Heineken and proposed design of the CSFs? 5. How can the proposed design of CSFs improve the Heineken design of CSFs?

(26)

(27)

(28)

6 Pluriform analysis

This pluriform analysis will help to gain a clear sight on the different perspectives of the problem situation as already mentioned in the previous chapter. More concrete, in this research project the metaphors are applied to look at the problem situation through the perspectives of these metaphors and thereby identify relevant contingency factors that can influence the design of the CSFs. In this case, Morgan’s (2006) ‘images of organization’ will be applied because of their applicability and popularity among system thinkers. Because metaphors represent a specific view of the situation, not all metaphors are appropriate in every situation. Only metaphors that directly resulted in the selection of certain contingency factors are discussed. Besides their relevance to the problem situation, the extent to which the contingency factors are discussed in literature is also a criterion for selecting these contingency factors.

Machine

According to Jackson (2000): ‘organizations viewed from the machine metaphor are constructs that should enable maximum efficiency’ (p.27). Control is exercised through rules, targets and a strict hierarchy of authority and information is processed according to the arrangement of tasks and by exception reporting up the hierarchy (Jackson, 2000). Riis et al. (1997) acknowledge the important role that the rules and hierarchy play within an organization. They therefore argue that there is a need to link manufacturing to the perspective of organizational structure. The machine metaphor also focuses on the arrangement of certain tasks. The arrangement of tasks is very much dependent on the concept of manufacturing (technology) that is applied, whether that is TQM or JIT for example. Riis et al. (1997) therefore state that the advancement of maintenance technologies requires companies to re-examine the role of their maintenance systems. Looking at the problem situation from the perspective of the machine metaphor also accentuates the role of targets or objectives. Riis et al. (1997) for example claims that: ‘the maintenance system is very much dependent on the specified corporate objectives’. The machine metaphor has helped in identifying three important factors that will be taken into account during the design of the CSFs, the influence of organizational structure, technology and strategy on the maintenance organization.

Organism

(29)

acknowledges the important influence of this angle on manufacturing performance of organizations moving through different life cycles. The organism metaphor has helped in identifying the important role of the environment and organizational life cycles in designing CSFs.

Brain

The brain metaphor emphasizes active learning and focuses on decision making and information processing (Morgan, 2006). Organizations must be able to perform both single and double loop learning according to Jackson (2000). Decision-making relies heavily on the objectives and strategy that top management has created. The role of strategy in a maintenance environment has already been discussed in the machine metaphor section by quoting Riis et al. (1997). The brain metaphor has acknowledged the important role of strategy in designing CSFs for maintenance performance, that was also identified by the machine metaphor.

Culture

(30)

7 Theoretical concepts

This chapter will provide theoretical concepts relating to the issues and concerns that have been identified in the Pluriform analysis. These theoretical concepts are combined in the subsequent chapter to form a theoretical framework that will be applied to the situation at Heineken. First, a motivation is provided of why these specific theoretical concepts have been selected. Following, the theoretical concepts are introduced separately.

7.1 Motivation

The theoretical framework that is presented in the next chapter is mainly based on the Contingency theory. Before this theoretical concept will be described, this approach will be compared with other well-known maintenance benchmark models.

Mainnovation1 is a Dutch consultancy firm that is receiving increased attention in the industry through the establishment of their Value Driven Maintenance (VDM) methodology, which is described in their eponymous book. The basic principle of the VDM methodology is to transform the maintenance function within a company from a cost centre to an economically value adding centre. Benchmarking maintenance performance forms an integral part of the general consultancy trajectory of Mainnovation. When a company hires Mainnovation to improve their maintenance performance, the first activity that is undertaken is measuring their maintenance performance and comparing these figures against the maintenance performances of other companies in their database. Based on this performance analysis, the subsequent consultancy trajectory is determined, in which underperforming parts of the maintenance function will be improved.

The maintenance benchmark studies that Mainnovation executes are based on their own performance measurement system, called the VDM Control Panel (see Appendix F ). Haarman and Delahay (2004), founders of Mainnovation, claim that ‘the strength of the VDM Competence Model stems from the completeness of the CSFs and their mutual relations’ (p.37). The VDM Control Panel uses a combination of different performance indicators, following a trend that has been initiated by the BSC of Kaplan and Norton (1992). The following ten CSFs are included in the VDM Control Panel:

o Cost control managing maintenance cost

o Asset utilization the technical availability of the equipment

o SHE management complying with SHE laws and regulations

o Reliability engineering application of preventive maintenance

o Planning & preparation work order management

o Maintenance execution the productivity of the maintenance mechanics

o MRO Supply Chain management spare parts management

o Service Supply Chain management outsourcing of maintenance activities

o Skill & Tool management training and education of personnel

(31)

Although the VDM control panel is gaining acceptance in the industry, several flaws of the model can be identified. The main disadvantage of the VDM control panel is that the CSFs are not adapted to the maintenance strategy and other characteristics of the company at which it is deployed. Every company, whether that is a highly capital intensive oil refining plant or a low volume brewery producing specialty beers with all their different characteristics, will be assessed based on these same ten CSFs. Haarman and Delahay (2004) discuss this issue by stating that not all CSFs in their VDM Competence Model are equally important to an individual maintenance organization. According to Haarman and Delahay (2004); ‘a maintenance organization should concentrate primarily on those CSFs that are critical to creating value’ (p.38). They do provide some guidance how to select those critical CSFs in their VDM Competence model, but this selection is only based on the consideration of one variable. Furthermore, by stating that a company should focus on critical CSFs only, they also acknowledge that only a small part of their model is actually relevant. The model for benchmarking maintenance performance at Heineken will aim to compensate this flaw by producing a unique design of CSFs for an individual maintenance organization and align those CSFs with the characteristics of this company.

Literature has only until very recently picked up the contingency approach to selecting maintenance performance measures. In 1990, Pintelon still suggested a static set of CSFs for measuring maintenance performance in her article on the Maintenance Management Tool (MMT). In an article that will be publicized later this year though (Muchiri, Pintelon, Gelders and Martin, 2010), Pintelon actually criticizes these static models because they lack a methodological approach of selecting the most important CSFs. Pintelon and her fellow authors therefore propose a maintenance measurement model that integrates maintenance strategy in selecting CSFs for maintenance performance measurement. Muchiri et al. (2010)’ most current model is a first small step towards a contingent approach on creating performance measurement frameworks for maintenance. The difference with the method presented in this paper is that Muchiri et al. (2010) only integrate the strategy contingency factor, whereas this paper also integrates the culture, technology, life-cycle, structure and environment contingency factors. This paper therefore presents a more holistic description of factors that influence the selection of CSFs for performance measurement frameworks used in maintenance organizations.

(32)

operational objectives and measures is reported in a survey conducted in the Belgian manufacturing industry (Gelders et al., 1994).

The contingency theory will be applied here to enable the proposed design of the CSFs to be aligned with the characteristics of an individual maintenance organization. Based on the contingency factors discussed in chapter six, the most suitable design of CSFs will be produced for the maintenance organization of Heineken in this case. The influence of contingency factors on performance measurement within companies has been widely discussed by researchers in the area of accounting. Miles and Snow (1978) were among the first to claim that performance measurement should rely on a company’s strategy and organizational structure. This idea was later popularized by Kaplan and Norton (1992) in their Balanced Scorecard model. The influence of the contingency factors that are selected in this research project will be discussed in the following subparagraph, where the contingency theory will be introduced more elaborately.

Although the contingency theory has been applied in the area of accounting, the literature study that has been conducted during this research project has not identified an already existing example of such an approach in the area of maintenance performance measurement.

7.2 Contingency theory

The contingency theory emerged in the 1960s, mainly as a result of the work of Woodward (1965), who investigated the connection between organizational structure and the success of industrial organizations. By the 1970s, it had become the established paradigm in organization theory.Jackson (2000): ‘contingency theory views organizations as consisting of a series of interdependent subsystems, each of which has a function within the context of the organization as a whole’ (p.110). The approach additionally rests upon the open systems view that regards the organization as dependent upon the wider environment in order to survive (Jackson, 2000).

There are four main hypotheses upon which the contingency theory rests (Jackson, 2000):

o There is no one best way to structure the activities of an organization in all circumstances

o Certain contextual factors determine the nature of the structure because of the constraints they impose. If organization structure is not adapted to context, opportunities are lost, cost rise and the maintenance of the organization is threatened

o Depending on the context, some form of organization structure is likely to be more effective than others

(33)

Heineken’s breweries are facing. The system metaphors in the Pluriform analysis have been applied to satisfy both requirements.The following subparagraph will provide insight into which factors are applied in this research project and what their effect is on the use of performance measures.

7.2.1Environmental uncertainty

The environment is an important contingency factor that lies at the heart of the contingency approach. Perhaps the most widely researched aspect of the environment is uncertainty. Chenhall (2003) defines environmental uncertainty as situations in which probabilities cannot be attached and even the elements of the environment may not be predictable. Govindarajan (1984) claims that these elements of the environment are customers, suppliers, competitors and regulatory groups such as government agencies. Companies can either face a highly uncertain environment or a stable environment and everything in between.

Several authors agree about the effect that an uncertain environment has on performance measurement (Chenhall, 2003; Gosselin, 2005; Hoque, 2004; Govindarajan, 1984). Firms that face a higher level of environmental uncertainty focus more frequently on non-financial measures whereas firms that face a lower level of environmental uncertainty tend to focus on financial and internal indicators. Govindarajan (1984) adds that the greater the environmental uncertainty, the greater would be the need of superiors to rely on subjective rather than objective approaches in evaluating performance.

7.2.2Organizational culture

Henri (2006) states, assuming performance assessment is an exercise primarily founded on values, that the attributes of performance measurement should reflect aspects of organizational values. Control and flexibility represent two competing organizational values, which are considered attributes of organizational culture. The control/flexibility dilemma, which refers to preferences about structure, stability and change, is part of Quinn and Rohrbaugh (1983)’ Competing Values Model. Control values refer to predictability, stability, formality, rigidity and conformity while flexibility values refer to spontaneity, change, openness, adaptability and responsiveness.

Henri (2006) has found proof that flexibility value firms are associated with a broad set of financial and non-financial performance measures. These findings suggest that flexibility value firms tend to further integrate performance measurement in their organizational processes and use more performance indicators than control value firms do. Control value firms emphasize mainly financial measures. The emphasis on financial information may be consistent with the importance placed on conformity and stability within control value firms (Henri, 2006).

7.2.3Organizational life cycle

(34)

most heavily upon informal controls that exemplify an interpersonal control strategy. Associated with the simple structure with short span of control, only a minimal amount of information is used for decision-making. What formal control systems they do used tend to be based on simple and narrowly-defined performance measures.

The growth phase is expected to occur once the firm had established its distinctive competences and enjoyed some initial product-market success. The emphasis is upon achieving rapid sales growth and amassing resources in an attempt to realize advantages accruing to larger scale. Typically, a functionally based structure is established, some authority is delegated to middle managers, and procedures are formalized. Decision-making tends to focus on a broad array of factors, demanding a larger amount of information than during birth stages. Performance measurement for growth firms are expected to be more formal than in other stages. To facilitate growth firms adapting to the more diverse and competitive situation, managers rely heavily on environmental scanning and quality controls.

The maturity stage is expected to follow the growth stage as sales levels stabilize, the level of innovation falls, and a more bureaucratic organization structure is established. The goal becomes smooth and efficient functioning. Consistent with their emphasis upon efficiency and profitability, mature firms seek to stabilize their operations by strategies that replace innovation with consolidation. Maturity firms therefore concentrate on internal, financial and future related performance measures.

The revival phase is typically a phase of diversification and expansion of product-market scope. It shows firms adopting divisionalized structures for the first time in order to cope with more complex and heterogeneous markets. Associated with speed, adaptability and intuition as required in revival firms, decision-making styles tend to be flexible, using minimal data for loose and fluid pattern of control.

The final decline phase reveals encroaching stagnation as markets dry up and firms begin to decline with them. Profitability drops because of the external challenges and because of the lack of innovation. There is a tendency towards internal orientation to preserve resources and locus of control is at the top level of management that may rely on a minimal amount of information for decision-making.

7.2.4Technology

Technology refers to how the organization transforms inputs into outputs and includes hardware, materials, people, software and knowledge (Chenhall, 2003). One of the aspects of technology to describe the influence on performance measurement is the manufacturing concept applied. Whether a company uses Just in Time (JIT), Total Quality Management (TQM) or Flexible Manufacturing (FMS) determines what kind of performance measures they should use.

(35)

measures focused on time, quality, operating efficiency and flexibility. JIT is mainly associated with organic performance measurement.

7.2.5Organizational structure

Organizational structure is about the formal specification of different roles for organizational members or tasks for groups, to ensure that the activities of the organization are carried out (Chenhall, 2003). Chia (1995) has looked into the effects of decentralization on performance measures. Decentralization is one type of organizational structure, which refers to where decisions are taken within the organization, i.e., the level of autonomy that is delegated to managers for their decision-making. The higher the degree of decentralization, the lower the hierarchical level in the organization where the decisions are taken.

Chia (1995) has found that the higher the level of decentralization of a structure, the broader the scope of the performance measures. This means that both internal and external, economic and non-economic and forecast related performance measures will be included. Gosselin (2005), who found that centralized firms tend to use more financial measures while decentralized firms tend to use more non-financial measures, confirms this.

7.2.6Strategy

Strategy is somewhat different from other contingency variables. In a sense it is not an element of context, rather it is the means whereby managers can influence the nature of the external environment, the technologies of the organization, the structural arrangements, the organizational culture and performance measurement. Notwithstanding the strategic direction selected by the organization, contingency-based research predicts that certain types of performance measures will be more suited to particular strategies (Chenhall, 2003). Several generic taxonomies have been developed including entrepreneurial-conservative, prospectors-analyzers-defenders, build-hold-harvest and product differentiation-cost leadership.

(36)

8 Theoretical framework

By combining the theoretical concepts that have been discussed in the previous chapter, a framework is created that integrates these concepts. Figure 8-1 Theoretical framework visualizes the theoretical framework, which consists of a combination of the contingency theory and House of Quality.

Figure 8-1 Theoretical framework

Figure 8-1 Theoretical framework visualizes how the contingency theory will finally lead to a design of CSFs for Heineken. The current maintenance strategy of Heineken will be the starting point of the theoretical framework, just like Tsang et al. (1999) suggest in their approach. Although Tsang et al. (1999) apply the Balanced Scorecard model from Kaplan and Norton (1992), which is a more mechanistic approach than the contingency approach applied in this paper, their approach does provide an effective procedure to translate a strategy to CSFs. Tsang et al. (1999) describe this step as following:

(37)

Because the Heineken maintenance strategy will be divided into CSFs similar to those mentioned by Tsang et al. (1999), it is important that the maintenance strategy is well formulated and includes several maintenance priority areas. These CSFs will then be divided among the three perspectives from the input-output model presented in section 2.1.

Subsequently, the six contingency factors that have been selected in chapter 6 will be characterized for Heineken’s specific situation. For example, it must determined to what extent the environment in which the Heineken breweries operate can be described as being uncertain. Literature on the effect of environmental uncertainty on designing a performance measurement framework claims that companies operating in highly uncertain environments prefer non-financial indicators. When all other contingency factors are characterized for Heineken as well, the contingency factors will be translated to preferences for performance indicators with the help of literature, as for the example of environmental uncertainty. This will lead to a list of preferences for CSFs, as explained in section 7.2, which are most suitable to Heineken. These preferences, together with the CSFs, will then be integrated into a House of Quality (HoQ) matrix. Kutucuoglu et al. (2001) also used the HoQ concept to link certain preferences of performance measures to CSFs in order to select the most relevant CSFs for a specific company. Considering the amount of key design features for a performance measurement system and the complexity of the maintenance system, Kutucuoglu et al. (2001) found the matrix structure of the HoQ suitable for developing a performance measurement framework that embraces features identified in literature. It is clear the HoQ concept is used here for a different purpose than its original purpose, which is translating customer preferences into product specifications. Its practical appeal in this setting was therefore the main reason to select the HoQ concept in this case.

(38)

9 Applying the theoretical framework to Heineken

The theoretical framework that has been introduced in the previous chapter will be applied to the situation at Heineken here. First, the Heineken maintenance strategy will be explained, followed by the contingency factors and their influence on performance measures in this specific situation. Finally, the filled in HoQ will be described and the proposed set of CSFs that will be the result of this exercise.

9.1 Heineken maintenance strategy

The Heineken Maintenance strategy is based on an internal Heineken document from 2001. Heineken is currently developing a new updated maintenance strategy, but as the updated maintenance strategy has not been completed yet, it is decided to apply the theoretical framework to the current maintenance strategy. Heineken’s maintenance vision is: ‘to support cost leadership and maintain quality through effective maintenance’. The goal Heineken has formulated for maintenance is to ensure that physical assets continue to do what the users want them to do.

Figure 9-1 Heineken maintenance strategy house

(39)

9.2 Heineken contingency factors

Based on the description and dimensions of the contingency factors that are selected in this research project, they will now be applied to the situation of Heineken’s breweries in general. An interview with Heineken’ Manager of Packaging Solutions Berend Hut provides the input for this analysis. Berend Hut was selected for this interview because of his experience in working in multiple breweries within Heineken, both in the Netherland and abroad. The contingency factors were filled in for Heineken from the perspective of an average Heineken brewery.

9.2.1Environmental uncertainty

Environmental uncertainty has been defined as situations in which probabilities cannot be attached and even the elements of the environment (customers, suppliers, competitors and regulatory groups such as government agencies) may not be predictable. Companies can either face a highly uncertain environment or a stable environment and everything in between.

The degree of environmental uncertainty the breweries of Heineken face can be considered as relatively low. The demand from the market has quite a stable pattern. Seasonal influences on demand return each year and are therefore reasonably well predictable. Obviously, the demand for beer differs per region. Europe is considered a mature market whereas Africa is characterized as a developing market. As this does not negatively influence the predictability of the demand for beer, the environment seen from the customer perspective is considered stable.

The technology and products delivered by suppliers have changed to a certain degree, although no fundamental changes have emerged. Where the brewing process of beer remained largely unchanged, the packaging process faced some changes due to innovating packaging types. Apart from the dominant glass bottle and aluminum can types of packaging, smaller draught kegs and even aluminum and PET bottles have been introduced. These innovations are mainly driven by marketing and environmental demands.

Although a large consolidation has taken place in the beer market, not many new competitors have entered the beer market. The newly created companies are still made up of the same companies that have been active in the beer market for a long time already.

Concerning the changes in governmental regulations, some changes have emerged. Surely, the safety and environmental rules have become stricter in general. Non-alcohol campaigns are also influencing the beer market, but these factors have not fundamentally changed the production process at Heineken’s breweries.

9.2.2Organizational culture

(40)

formality, rigidity and conformity while flexibility values refer to spontaneity, change, openness, adaptability and responsiveness.

This means that its institutional characteristics are more towards formalization, structure, bureaucracy, production and getting the job done instead of the breweries being a personal, entrepreneurial and dynamic place. The leader of the Heineken breweries are characterized by being coordinators, organizers, technicians and hard-drivers instead of mentors, innovators and risk takers. Finally, the emphasis within the breweries lies on stability, efficiency and achievement and not on human resources, morale and new challenges.

9.2.3Organization Life Cycle

Organizations move through the following life cycle phases: birth, growth, maturity, revival and finally through the decline phase. The maturity phase best fits the characteristics of Heineken at this moment. Heineken recently finished their second large acquisition, being the FEMSA company in Mexico, and Heineken’s CFO has recently expressed that Heineken will now focus on the integration of these new companies instead of seeking new acquisition objectives. This is consistent with the description given of the maturity phase, which states that after a period of growth, the emphasis will lie on efficiency and profitability.

9.2.4Technology

Heineken’s technology will be defined here by selecting the manufacturing concept that is applied. Whether a company uses Just in Time (JIT), Total Quality Management (TQM) or Flexible Manufacturing (FMS) determines what kind of performance measures they should use. TQM closely resembles the TPM concept that Heineken applies, because they both stress that employees must be empowered to initiate corrective action. Obviously, TQM has a much wider scope than TPM, which is mainly focused on equipment. As the TPM organization and background has already been provided in earlier chapters, the selection of TQM as the technology that fits Heineken best requires no further clarification.

9.2.5Organizational structure

Organizational structure is assessed by characterizing the level of decentralization at Heineken. Decentralization refers to where decisions are taken within the organization, i.e., the level of autonomy that is delegated to managers for their decision-making. The level at which decisions are taken within an average brewery is relatively high. The organizational structure of Heineken can therefore be characterized as centralized. The operators and mechanics that work in a brewery do not have much decision-making autonomy in general. Most of the power is situated at the management levels within brewery because operators in most countries have low educations.

9.2.6Strategy

(41)

ratio between the commercial and production side within a large company are quite common, as the production side generally is perceived as a cost centre.

9.3 House of quality

By linking the contingency factors that have been applied to Heineken from the previous paragraph to the effects of contingency factors on performance measurement described in section 7.2, a HoQ can be constructed which integrates these effects with the CSFs from the Heineken maintenance strategy. The CSFs that show the highest correlations with the described preferences for performance measures will be included in the proposed design. Only those preferences for performance measures that were formulated clearly have been selected here. In some cases, the authors of the articles to which is referred is section 7.2 used such vague terminology that it required too much interpretation to understand what the requirement was about. These specific preferences have been excluded from the analysis that will take place in this section.

An interview with Maintenance Specialist Alex Klaver and Engineering Policy Manager Chris Versteegh provided the input for filling in the HoQ that is shown in Table 1 House of Quality. The depth of the descriptions of the CSFs of the Heineken maintenance strategy in the internal documentation was not sufficient to be able to carry out this analysis without these interviews. Alex Klaver, in his role as Maintenance Specialist, has extensive experience with applying the principles of the Heineken Maintenance Strategy in multiple Heineken breweries. Chris Versteegh on the other hand is more closely related to the theoretical side of the Heineken Maintenance strategy as Engineering Policy manager.

Per CSF, a short description will be provided here on the correlations that have been appointed. As the financial requirement occurs five times, it will only be discussed once under stable environment because the other four will be appointed with similar correlation scores. A similar approach is used for the internal and external preferences.

Before Table 1 House of Quality will be discussed, the definitions of the preferences in the HoQ are provided:

o Financial: an indicator expressed in a cost value

o Objective evaluation of performance: an indicator which can be evaluated based on an objective and transparent measurement technique

o Internal : an indicator focused on everything that takes place within the maintenance department

o Future: an indicator that is used to predict future performance

o External: an indicator focused on everything that takes place outside the maintenance department

o Non-financial: an indicator that is not expressed in a cost value

o Operating goals: an indicator which is used to formulate operating goals

(42)

CSFs from Heineken maintenance strategy

Input Process Output

M a in te n a n ce c o st S p a re p a rt s m a n a g e m e n t O u ts o u rc in g p o li cy a n d p ro ce ss W o rk m a n a g e m e n t M o d if ic a ti o n c h a n g e m a n a g e m e n t M a n a g in g m a in te n a n ce p la n s T e ch n ic a l p u rc h a si n g M a n a g e m e n t o f te ch n ic a l d o cu m e n ta ti o n a n d i n fo rm a ti o n O p e ra ti o n a l in v o lv e m e n t o f p ro d u ct io n i n m a in te n a n ce Le a rn in g f ro m f a il u re s a n d c o n ti n u o u s im p ro v e m e n t T e ch n ic a l a n d t e a m t ra in in g E a rl y e q u ip m e n t m a n a g e m e n t C a re o f a ss e ts In st a ll a ti o n p e rf o rm a n ce Stable environment Financial + + + 0 + 0 + 0 0 0 0 0 - 0

Objective evaluation of performance + + 0 + + + + + 0 0 0 + 0 +

Internal + 0 0 + + + 0 + 0 + 0 0 + +

Control culture

Financial + + + 0 + 0 + 0 0 0 0 0 - 0

Maturity Organizational life cycle

Internal + 0 0 + + + 0 + 0 + 0 0 + + Future + + + + + + + - - - - 0 + + Financial + + + 0 + 0 + 0 0 0 0 0 - 0 TQM Technology External - 0 0 - - - 0 - 0 - 0 0 - - Non-financial - 0 0 + 0 + - + + + + + + + Centralized structure Financial + + + 0 + 0 + 0 0 0 0 0 - 0 Conservative strategy Financial + + + 0 + 0 + 0 0 0 0 0 - 0 C S F P re fe re n ce s b as e d o n c o n ti n g e n cy f ac to rs Operating goals + 0 0 0 0 0 0 0 0 0 0 0 0 + Total scores: 8 8 6 4 8 4 6 2 0 1 0 2 -2 5

Table 1 House of Quality

Management of Maintenance Cost (eight points)

Designing Critical Success Factors for benchmarking maintenance

Designing Critical Success

Factors for benchmarking

maintenance

A CONTINGENCY APPROACH

Designing Critical Success Factors for

benchmarking maintenance

Management summary

Preface

Table of contents

List of abbreviations

Vocabulary

1

Introduction

2

Research framework

3

Introductory literature research

4

Company introduction

5

Research model

6

Pluriform analysis

7

Theoretical concepts

8

Theoretical framework

9

Applying the theoretical framework to Heineken