Business Intelligence and Procurement processes

(1)

Business Intelligence and Procurement

processes

(2)

Study: Business And ICT

Faculty: Economics and Business

University: Rijksuniversiteit Groningen

Supervisor: E.O. de Brock

2nd Reader: D.J.F. Kamman

Company: De Friesland Zorgverzekeraar

Location: Leeuwarden

Supervisor: Gerrit Gerritsma

2nd Supervisor: Gerard Akkerman

(3)

Preface

During the period of five months at De Friesland Zorgverzekeraar and the months hereafter for finishing this thesis, I gained support from certain peo-ple. In this preface I want to thank these specific persons for giving me the strength and optimism to finish the thesis project.

• First of all, I want to thank my parents for their support. By discussion and reason, I gained new perspectives and the motivation to finish this thesis. Thanks!

• My sisters Eline and Suzanne for asking how my thesis project was going and if I was finished already.

• The colleagues at De Friesland Zorgverzekeraar who worked really hard every day.

• My supervisors Gerrit Gerritsma and Gerard Akkerman for their patience and their support. You really gave me self-confidence.

• Wim Dressel for his feedback on procurement. Hope you liked the wine. • My supervisor Bert De Brock for his critical feedback which often put my

feet back on the ground.

• My second supervisor D.J.F Kamann.

• My friends at AMOR for taking my thesis so seriously • All people I forgot to mention...

Thanks!

(4)

Health care insurers in Holland deal with high costs of health care. This is also the case at De Friesland Zorgverzekeraar, a small insurer in the province of Friesland. With a market share of 65% in Friesland, it is market leader in this province. Nationwide, De Friesland Zorgverzekeraar has a market share of 3%. The research was done during a five month period at the procurement de-partement of De Friesland Zorgverzekeraar. The research is targeted towards solutions for decreasing health care costs by effective procurement of health care services and products. With the help of a pilot project, data mining algorithms should be constructed which decreased these costs. However, the pilot project was abandoned because a proper Business Intelligence process is absent at De Friesland Zorgverzekeraar and needs to be present for data mining to work ef-fectively.

The primary data mining objective of De Friesland Zorgverzekeraar was thus changed towards a Business Intelligence research objective. The con-ducted research includes an analysis of the procurement process of De Friesland Zorgverzekeraar and (the information flow of) the data analysis at the procure-ment departprocure-ment. Research was executed with the help of theory from the areas of Business Intelligence and procurement, interviews and a thorough analysis of a relevant health care question.

The goal of the research is to see if De Friesland Zorgverzekeraar can im-plement Business Intelligence to make their procurement process more pro-active. Based on the outcome of the analysis done, weaknesses of De Friesland Zorgverzekeraar for implementation of a Business Intelligence process, are given. These weak spots are:

• Outdated system ’Claim’

• Lack of documentation TPO / statistical employee

• Lack of inter and intradepartmental communication / lack of information requests from management

• Outdated and non-used factbooks

• Lack of feedback towards health care providers

The recommendations to deal with these weaknesses have to do with better departmental information flow with the help of a Business Intelligence team that deals with important health care issues. The communication between the TPO department and the statistical employee can be improved by using the intranet. Also constant automatic monitoring of health care providers should solve the weak points of factbooks and feedback towards health care providers.

(5)

Introduction

In the current, modern world, we become more and more dependent on com-puter technology to execute business processes. Machines are doing an increas-ing amount of work that was once done by humans. Computers can do it faster and cheaper which results in a change of focus of the human worker towards other processes. The result of this increasing usage of computer technology re-sults in the increase of available data. Companies have enormous amounts of data in their database systems. The use of this data however is far less than it can, and maybe should, be. But also in this area, the computer can help. Information technology can be used in the analysis of corporate data. With the help of techniques like Business Intelligence and data mining, correlations in data can be detected and processes can be optimized.

The health insurance company De Friesland Zorgverzekeraar is interested in the possibilities of implementing data mining into their procurement process. This thesis is the result of a five month internship period in which the procurement process of De Friesland Zorgverzekeraar was analyzed and research was done in the area of Business Intelligence (a necessary step for implementing data min-ing). Before data mining can be implemented, processes need to be documented and be executed in the right way. When the processes are at the correct level, Business Intelligence can be implemented into the organization. This thesis deals with analysis of the procurement process and the information acquiring at the procurement department of De Friesland Zorgverzekeraar. Based on the findings of the research, the readiness of the procurement department for Busi-ness Intelligence implementation will be given together with recommendations to solve occuring problems.

Structure

In chapter one an introduction of the company of De Friesland Zorgverzekeraar is given. A short history and organizational structure will clarify the orga-nization. The second chapter will deal with the construction of the research objective together with the research question and according subquestions. The subquestions will be answered throughout the thesis and will guide the reader to possible solutions to the problems posed in chapter two. The third and fourth chapter will deal with the theoretical framework. The third chapter will give the reader insight in Business Intelligence theory. A Business Intelligence example from a real-time case will also be given. Chapter four covers the subject of purchasing / procurement processes. Chapter five will analyse the procurement process at De Friesland Zorgverzekeraar. Chapter six will analyse the infor-mation acquiring process and the most important data source(s) used at the procurement department of De Friesland Zorgverzekeraar. This chapter has the function of exposing the problems at the procurement department that prevent a succesful implementation of Business Intelligence.

The seventh chapter gives recommendations to the found problems. The final chapter, the conclusions, will answer the research question (of chapter two) and give a critical self-reflection of my research and my functioning within the orga-nization of De Friesland Zorgverzekeraar. In my opinion, the skill of adapting to organizational culture and processes is important mainly because it contributes to acquiring the academic level required for writing a thesis.

(10)

ment

In this chapter an overview of the organization of De Friesland Zorgverzekeraar will be given. The overview will give the reader a better understanding of the organizational structure and the market position of De Friesland Zorgverzeker-aar. Also the level of competition within the health insurance market will be analyzed with the help of Porters Five Forces Model. All these aspects should give the reader enough background information for understanding this thesis.

1.1 History

In 1900, health insurance companies did not exist. People had to pay for their own medicines and medical care themselves. Lack of money for basic health care simply meant you did not receive health care. The poor could ask for money at the local government or the churches. Small funds existed but because of the size of these funds, they were not around for long. Because the deterioration of public health around 1900, laws were enacted in the area of public health care and personal hygiene. Affordable health care was required. In the city of Leeuwarden, these rules and regulations resulted in the establishment of a health care fund for the city of Leeuwarden. It was called ’Eendracht Maakt Macht’. Compared to other funds in Holland, ’Eendracht Maakt Macht’ did not have big problems between private practitioners and clients. A synergetic cooperation of all involved people ensured a healthy fund which could pay their health care providers, ask for a low weekly fee to their clientele and could offer a health care package of good quality. In the beginning of ’Eendracht Maakt Macht’, private practitioners had a large influence in decision making. This influence deterio-rated over time. During the years, the amount of clients increased and also the weekly fee became higher. The size of the health care package also increased. A national health care law was in development but was not implemented because of differences between private practitioners and clients. (the private practition-ers wanted to have a lot of influence in the health funds, the clients wanted to shift the income limit upwards so people with a higher income could enter). The war years and the government of the Germans resulted in het Ziekenfondsbesluit. After the war, small adjustments in this law needed to be made. Government and regulatory organisations were responsible for the regulations of health care funds instead of the local boards. During the 60’s to the 80’s, increasing bu-reaucracy and costs resulted in the increase of the monthly fee for clients and employers. In 1987 minister Dekker and later on (in 1989) minister Simons, proposed a system of competitiveness between health insurance companies. For this proposition to work for the health insurance funds, commercializing of the companies was required. This was not as easy as it looked. With the help of commercials and slogans, the image of the different funds in Holland had to be made clear. The Return on Investment (ROI) for these commercials however was negative. In 1990 the name of De Friesland was introduced. Because of the long tradition (a gentlemen’s agreement) to not compete with each other, the competitiveness between funds in the period 1990-2000 was not fierce. With the appearance of the Health insurance law in 2006, the competition became fierce for a year but stabilized in 2007. Currently, De Friesland Zorgverzekeraar

(11)

1 DE FRIESLAND ZORGVERZEKERAAR AND ITS ENVIRONMENT 11

is the biggest health insurance company of the province of Friesland with more than 500.000 clients. The market share of De Friesland Zorgverzekeraar in the province of Friesland is 65%. Nationwide the company has a market share of around 3% and is in the top five of health insurance companies after big players as Menzis and Achmea. The company has around 500 employees and is still located in the capital of Friesland, Leeuwarden.

1.2 Organizational structure

1.2.1 General

The organizational chart below (in Dutch) gives an overview of the departments and employees present within the organization of De Friesland Zorgverzekeraar.

Figure 1: Organizational Structure of De Friesland Zorgverzekeraar The governance of De Friesland Zorgverzekeraar lies with the executive board. The executive board of De Friesland Zorgverzekeraar consists of two people: dhr. Hans Feenstra (replaced by mrs. Diana Monissen in september 2009) and dhr. Marcel Kuin. Their responsibility is the correct execution of the policy of De Friesland Zorgverzekeraar. Beneath the board of directors is the board of managers. This board is split into four clusters:

• Health Purchasing (Zorginkoop)

• Customer and Service (Klant en Service) • Marketing and Sales (Marketing en Verkoop)

• Finance, ICT and Facilities (Financi¨en, ICT en Faciliteiten)

This thesis was done at the department of Health Purchasing and is therefore focused on this department. The organizational chart of the department of Health Purchasing is shown on page 12 (in Dutch).

(12)

Figure 2: Organizational Chart of the Health Purchasing department Cure

responsible for all the aspects of the curing of clients (hospitals, medicine and doctors). The department Care has its focus on the caring process of clients (clients that cannot be cured like mentally disabled persons or handicapped clients). The Care department also deals with AWBZ-clients (Algemene Wet Bijzondere Ziektenkosten). Because this research is done at the Cure part of the procurement department, the Care department will not be dealt with further. 1.2.2 Department Cure

The Cure department deals with the purchasing of health care from general practitioners, hospitals and pharmacists. Two aspects that play an important role within the department are integrated health care ’ketenzorg’ and Health care vision’zorgvisie’. These terms are explained later on in this chapter. 1.2.3 Mission of the Cure department

The mission of the Health Purchasing Cure department is as follows: • Professionalizing of the health purchasing process.

• Organizing of (integrated) health: ketenzorg. • Improving the health of people.

Professionalization: The professionalizing of the health purchasing process is done by execution of the following steps:

(13)

4. Purchasing 5. Monitoring 6. Evaluation

Integrated Health care: With the help of integrated health care, De Fries-land Zorgverzekeraar is trying to improve the quality of health care. A better coordination of activities and good cooperation are necessary to reach a better fit between the demand for health care and the offering of health care. (Second order fit1_).

Health improvement: The health of people can be improved by making people aware of their own health situation and the prevention of diseases. Since 2004, De Friesland Zorgverzekeraar actively tries to increase the self-consciousness of clients in the area of health.

1.3 De Friesland and its environment

The market of health insurers is not a competitive market. In 2006 the market changed to a competitive market because of the new Health care law but in 2007, the fight was over and the level of switching customers went back to normal. After the introduction of the Healthcare law (Zorgverzekeringswet ), it is obliged for all inhabitants of Holland to be insured for health care. The insurers are competing over clients by offering their basic health package for a price beneath costs price. This paragraph will deal with the most important aspects that are of influence on the competitive position of De Friesland Zorgverzekeraar. These influences are shown with the help of the model that was developed by M. Porter, the Five Forces Model2.

In his Five Forces Model, Porter describes the following factors that influence a company in the area of competition:

1. Bargaining power of customers 2. Bargaining power of suppliers 3. Threat of substitute products 4. Threat of new entrants

5. Competitive rivalry within an industry

These five forces will be dealt with in the following subsections. The forces are applied to the situation of De Friesland Zorgverzekeraar.

1.3.1 Bargaining power of customers

De Friesland Zorgverzekeraar currently has over a half million customers. These customers influence the competitive position of De Friesland Zorgverzekeraar. The customer has bargaining power because it can choose to switch from health

1_{Porter, M. E., ’What is a strategy?’ in Harvard Business Review, November-December} 61-78, 1996.

(14)

insurer whenever he/she wants. The pricing of the health package is the main reason for switching customers. Customers who switch to a competitor, offering a cheaper health package, will decrease the income of De Friesland which, in its turn, can lead to a worsening of the competitive position.

1.3.2 Bargaining power of suppliers

The suppliers of health care (private practitioners, hospitals) have agreements / contracts with the health insurer about the health care they offer to clients. De Friesland Zorgverzekeraar buys health care from these suppliers. The NZa (Nederlandse Zorgautoriteit) prices the offered health care with the help of activity based costing (ABC)3_{. All activities are analyzed, numbered, priced}

and are called DBC’s (Diagnose Behandeling Combinatie). Within these DBC’s there are two segments, the A-segment and the B-segment. The A-segment includes 66% of the activities and the prices in the A-segment are non-negotiable. The price consists of two parts, an institutional part (hospital) and the honorary part (specialist). On an annual basis, the NZa determines the prices for the institutions. This means that the institutional part and the honorary part of a DBC in segment A can increase or decrease per year. Activities that are done on a more regular basis (34% of all the activities) are included in the B-segment and the price of these activities is negotiable. De Friesland Zorgverzekeraar has a large market share in the province of Friesland which results in a good bargaining position with suppliers in this province. In the rest of Holland, the bargaining power is less attractive because of a low market share.

A player in the health market with a good bargaining position whatever the size of the market share of the insurer are the hospitals. Substitution of medical care given by hospitals is difficult and clients therefore are obliged to go to a hospital. (E.g. an X-ray cannot be made in a supermarket). In 2005, health care insurers made an agreement to only provide clients with the cheapest medicines (preference policy). This policy resulted in a price war between pharmaceutical companies and a decrease in prices of medicines. On its turn, the pharmacists saw their profit decrease with enormous amounts because the bonus they re-ceived for buying medicines in large amounts, vaporized. The power of health care insurers in the pharmaceutical area becomes clear when it comes to their preference policy.

1.3.3 Threat of substitute products

Substitute products are not of any influence in the market of health insurance companies. Clients pay a monthly fee to the insurance companies. When a client has a health problem, he/she can go to a hospital or private practitioner. In most of the cases the costs for a consult, medicine or operation are covered by the insurer. These costs are often much higher (e.g. costs for an operation) than the monthly fee. There will be few Dutch people that are not insured and pay for their health care out of private capital or by means of a loan. As mentioned before, the Dutch law also requires inhabitants of Holland to be insured for their health care. Substitute products are therefore not relevant in the case of health care insurers and their products.

(15)

1.3.4 Threat of new entrants

In the Netherlands it is hard to enter the market of health insurance as a new insurer. This is because of the enormous amounts of rules and regulations that are imposed on insurers. A current trend in the insurance business in Holland are mergers. In the last couple of years big mergers took place between insurance companies. In the last couple of years mergers between Univ´e and VGZ, CZ and Delta Loyd and also Achmea and Agis were established. The result of these mergers is the changing of the insurance market into a market with a few big player instead of several smaller ones.

1.3.5 Competitive rivalry

As mentioned earlier, the market of health insurers is a non-competitive mar-ket. The main players which De Friesland Zorgverzekeraar has to deal with are: Menzis, Achmea and Uvit. These insurers have a large market share na-tionwide. In the province of Friesland however, these insurers are smaller than De Friesland Zorgverzekeraar. This results in a bipartite situation. On the one hand, De Friesland Zorgverzekeraar has the role of market leader in the province of Friesland. On the other hand, the role of follower in the rest of Holland is most suitable for De Friesland Zorgverzekeraar. The strategy of De Friesland Zorgverzekeraar is adjusted to this bipartite situation. The strategy of DFZ is to focus on the northern part of Holland (provinces of Noord-Holland, Friesland, Groningen, Drenthe, Overijssel).

All the five forces influence the market share of De Friesland Zorgverzekeraar. The market share of De Friesland Zorgverzekeraar in the Netherlands is visual-ized on the next page4_.

(16)

(17)

2 Research Design

In this chapter, the design of the research will be given based on theory of Verschuren5_{. The research context, scope, research objective, research question,}

according subquestions and the research model will be dealt with. Defining these elements will give the reader a better understanding of the structure of the executed research. The context in which the research was conducted will first be elucidated.

2.1 Research context

After research done by the research company Vektis, it was made clear that the total health care costs of De Friesland Zorgverzekeraar were above the national average. De Friesland Zorgverzekeraar wants to lower these costs to stay com-petitive and therefore needs to acquire insights into these costs.

De Friesland Zorgverzekeraar has an enormous amount of data available that could be used to acquire insights into the high costs. With the help of data mining, De Friesland Zorgverzekeraar hopes to retrieve these insights. There-fore, a project team was constructed and a research objective was defined. The preliminary objective had to do with a data mining pilot project to find correla-tions between health care costs and geographical areas in Friesland. To keep the research project realizable within the given timeframe and because the largest part of health care costs are depending on (successful) procurement, the depart-ment of health care procuredepart-ment was chosen as the scope of the research. During a talk with the project team of De Friesland Zorgverzekeraar, it was indicated that one could not simply do a data mining pilot project. Certain re-quirements need to be present within the organization for data mining to work. Because it was not clear if all the requirements for data mining were met, it was decided to first acquire information about the problems within the procurement department. Based upon this information, a new research objective should be defined.

2.2 Research objective

By conducting interviews with employees and project team members from De Friesland Zorgverzekeraar (see appendix C), the problems within the procure-ment departprocure-ment were analyzed. From the interviews could be extracted that there were problems in communication and information flow throughout the organization. Also the prerequisite for a data mining pilot project, a Business Intelligence process, was absent. Another problem that appeared was the reac-tivity within the procurement process.

After discussion with both supervisors, the decision was made to shift the re-search focus from data mining to Business Intelligence and the company’s readi-ness to introduce Busireadi-ness Intelligence into the procurement process. Because the research should be useful to the procurement department, the technical na-ture of the research was also abandoned and changed into research more aimed at the analysis of the procurement department’s information needs and the

im-5_{P. Verschuren, Hans Doorewaard, ’Het ontwerpen van een onderzoek’, 3e druk, Utrecht} 2000

(18)

pact of Business Intelligence on the procurement process.

The research objective is: ’to analyse the processes of information acquiring and data analysis within the procurement department of De Friesland Zorgverzek-eraar and give recommendations leading to (a more) successful Business In-telligence process implementation in the procurement process of De Friesland Zorgverzekeraar.’

It is important that Business Intelligence has a surplus value for the procure-ment process once it has been introduced. To show the added value of Business Intelligence, the research will also show in which way Business Intelligence can transform the procurement process of De Friesland Zorgverzekeraar into a more pro-active process.

2.3 Research question

To fulfill the research objective, the following research question was developed. ”What changes need to be made in the data analysis and information acquir-ing process of De Friesland Zorgverzekeraar to successfully introduce Business Intelligence into their procurement process and what will the added value of Business Intelligence be for De Friesland Zorgverzekeraar?”

2.3.1 Subquestions

Before the research question can be answered, research has to be conducted in the areas of procurement, business processes and Business Intelligence / data mining. This research will provide a solid foundation upon which answers to the research question will be based. With the help of subquestions, the direction is chosen which will eventually lead to a well-structured and logical research path. As sir Francis Bacon said:

”a prudent question is one-half of wisdom.” The constructed subquestions and their corresponding chapters are:

• What are the prerequisites for Business Intelligence implementation? (Chap-ter 3)

• Which phases can be determined within a procurement process? (Chapter 4)

• How is the procurement process of De Friesland Zorgverzekeraar defined? (Chapter 5)

• How are the information acquiring and data analysis process executed at the procurement department? (Chapter 6)

• Which problems exist in the information acquiring / data analysis process of the procurement department? (Chapter 6)

(19)

2 RESEARCH DESIGN 19

2.4 Research Definitions

In this section, I will mention the research definitions (begripsbepaling ) that are relevant for the research. Clear definitions are important to prevent ambiguity within this thesis.

Definition of Business Intelligence

Business Intelligence is a vague term especially for non-IT people. What defines Business Intelligence? How should it be interpreted? Is it purely an IT term or can it be seen in a broader spectrum? A lot of definitions of Business Intelligence exist, sometimes differing only in a small way but also in large ways. The need therefore exists to give a proper definition of the term Business Intelligence. Although this thesis will be read by people who probably know what Business Intelligence is, the application of theory in chapter six and seven is aimed at employees of De Friesland Zorgverzekeraar and can be seen as an advisory report. To also keep the thesis clear for them, the explanation of definitions is done in chapter three together with an example of a Business Intelligence solution at an insurance company. Definition of Data Mining

Just as with Business Intelligence, the term data mining needs to be elucidated before in-depth research can commence. What is the relation to Business Intelligence? Are there different kinds of data mining?

These questions will be answered in chapter three. Procurement process phases

Theory about procurement is widely available. In this thesis, the model of Van Weele and the NEVI (explained by the chairman of NEVI) will be used and an explanation of each separate step will be given. The models will give the reader an overview of the procurement process. Procurement theory will be dealt with in chapter four. The procurement process of De Friesland Zorgverzekeraar will be analyzed in chapter five.

Pro-activity and reactivity

The ability to respond to events and make the appropriate actions based upon these events is important for companies. When wrong decisions are made, corrections need to be taken to correct the error made. In the chapters five and six, the procurement process and the information acquiring process at the procurement department of De Friesland Zorgverzekeraar will be analyzed. Pro-activity and passivity aspects in these processes will be shown and solutions to make the processes more pro-active will be given.

Definition of continuity

(20)

2.5 Research model

To give a visual overview of the research, a research model is displayed below.

Figure 4: Research Model

(21)

2 RESEARCH DESIGN 21

analysis (dotted arrows).

With the help of both analyses, problems will become known within the orga-nization of De Friesland Zorgverzekeraar in the area of information flow and procurement (chapter seven).

Recommendations (chapter seven) and conclusions (chapter eight) will be given to solve these problems within the organization so Business Intelligence can be implemented and used in the procurement process of De Friesland Zorgverzek-eraar. With this, also the research question of this thesis will be answered.

2.6 Project plan

For the planning of the thesis project, I have partially used the Prince2 methodology.6 _{Prince2 is a project management methodology which I found}

extremely useful for defining the research steps upon which this thesis is build and for planning the defined steps in a timeframe. The project plan, which I made with the help of Prince2, can be found in appendix A. The project plan consists of a project description, requirements analysis, an overview of the external dependencies, a list of assumptions made and a project breakdown schedule (PBS). With the help of the project plan, the boundaries of the project and requirements of all parties involved were made transparent. During the research period, the project plan was adjusted and continuous consent of every party was given to the project plan. This ensured that I could continue with the road taken. Although the project plan was changed from a theoretical to a more practical type and the research question was changed to a clearer question, I still found it necessary to include it in the appendix because the project plan’s global research path is still relevant.

(22)

In this chapter, an explanation of the concept of Business Intelligence will be given together with the concept of data mining. The difference between these concepts will also be dealt with. After the concepts are made clear, real-life situations in the health care / insurance world will be given to show which successes can be reached with Business Intelligence. Finally, the problems (ac-quired from interviews with employees of De Friesland Zorgverzekeraar) within the procurement department of De Friesland Zorgverzekeraar that need to be tackled before a Business Intelligence system can be introduced, are discussed.

3.1 Business Intelligence

The concept of Business Intelligence was first mentioned by H.P. Luhn7 _{in his}

article ’A Business Intelligence System’. In this article, he described a system that kept track of documents requested by specific persons. Based on the stored information, an analysis could be made to better streamline the process of get-ting the suitable documents to the most suitable persons. Luhn already wrote the article in 1958, an era in which computers were not a common good of com-panies and printed documents were used instead of email. The ideas posed in the article, however, are still relevant nowadays. According to Luhn, Business Intelligence has the function of:

’..supplying suitable information to support specific activities carried out by individuals, groups, department, divisions, or even larger units.’

With the fast-paced developments in technology, the concept of Business In-telligence has become more technological throughout the years. Moss and Atre8

define Business Intelligence as:

’..an architecture and a collection of integrated operational as well as decision-support applications and databases that provide the business community easy access to business data.’

An important difference between both descriptions of Business Intelligence is the addition of technological terms as databases and applications by Moss and Atre. Luhn’s definition also lacks to mention the channel(s) in which hu-mans receive information. While Moss and Atre state that users of a Business Intelligence system have easy access to information and therefore an active par-ticipation of users is required to retrieve the information, Luhn does not men-tion either an active nor a passive informamen-tion acquiring approach by users. De Brock9 _{also does not mention the way in which data is obtained. He defines}

Business Intelligence as:

’the process of targeted, pro-active and legal obtainment of data, which is transformed into information and spread to acquire knowledge to support

deci-7_{H.P.Luhn, ’A Business Intelligence System’ in IBM Journal, p. 314-319, October 1958.} 8_{L. Moss and S. Atre, ’Business Intelligence Roadmap: The complete lifecycle for} decision-support applications’, Addison-Wesley, 2003.

9_{E.O. de Brock, ”Theorie en praktijk van Business Intelligence”, to appear.}

(23)

3 BUSINESS INTELLIGENCE AND HEALTH INSURANCE 23

sion making.’

The term pro-activity is mentioned in the definition of De Brock. With this term (to elucidate the definition further), he does not mean the way in which data is acquired but when it is acquired. According to De Brock, the data should be obtained with a specific purpose in mind (targeted) in a pro-active manner. Although it is not mentioned in his definition, reactive acquiring of data is also possible although a pro-active approach should be pursued whenever possible.

Although there are more definitions of Business Intelligence available, one definition has to be chosen. As can be seen from all the different definitions mentioned earlier, Business Intelligence is seen by authors in different ways. Basically, there are three types of Business Intelligence definitions that can be found: Business Intelligence as a technology, as a process or as information. Although differences exist in the definitions of Business Intelligence mentioned earlier, they also have a commonality. This is that the data acquired from the Business Intelligence system has the function of supporting the decision making process (a so-called Decision Support System (DSS)). The definition that will be used throughout the thesis to describe Business Intelligence will be the definition of De Brock. The choice for not using the definition of Moss and Atre is that the definition has a too technical approach while the definition of De Brock is more process-based. De Brock also included the transformation process from data to information, which in my opinion is an important process.

3.2 Data mining

The term data mining is often mentioned together with concepts as Knowl-edge Discovery in Databases (KDD) and KnowlKnowl-edge Management. According to Fayyad et al.10 _{KDD and data mining are related. He defines KDD as:}

’a nontrivial process of identifying valid, novel, potentially useful, and an ultimately understandable pattern in data.’

As a part of KDD, Fayyad et al. mention data mining and define it as: ’..consisting of applying data analysis and discovery algorithms that, under acceptable computational efficiency limitations, produce a particular enumera-tion of patterns over the data.’

When looking at data mining in a broad way, it includes strict data mining (discovery-based data mining) and statistics (hypothesis-driven data mining). In the given definition of Fayyad et al., it appears as if the authors use the data mining process on data without having a hypothesis beforehand. The definition is aimed at strict data mining in which a (large) data set is mined and patterns are evaluated on usefulness afterwards.

Another type of data mining is statistics (hypothesis-driven data mining). With this type of data mining a hypothesis (target) is defined which one tries to val-idate by doing data analysis. This type of data mining is thus more targeted

(24)

than discovery-based data mining.

I developed a broader and clear definition that encompasses both hypothesis-based and discovery-hypothesis-based data mining.

’The process in which a data set is analyzed and data patterns are outputted upon which insights can be obtained and/or defined hypothesis can be validated.’

The size of the data set (depending on the type of data mining) is not mentioned to keep the definition as clear as possible.

3.3 Business Intelligence versus Data Mining

After talking about both Business Intelligence and data mining, a clear distinc-tion should be made between both concepts. What is the difference between Business Intelligence and data mining? Both have to do with data but what is the difference exactly?

The most important difference between both concepts is their relation towards each other. When making a hierarchy (i.e. for implementation purposes), the concepts of Business Intelligence and data mining should not be placed on the same level within this hierarchy. The discovery of patterns in a data set (data mining) is a step in the KDD process. KDD (data analysis) on its turn is a part of the Business Intelligence cycle which contains among other things: need anal-ysis, data collection, data structuring/elimination, data analysis and intelligence utilization11_{. When introducing Business Intelligence into an organization like}

De Friesland Zorgverzekeraar, data mining can be a method for data analysis within this cycle. However, the previous steps in the Business Intelligence cycle should have been properly executed first before data mining can commence. For a Business Intelligence cycle to work, requirements have to be met. These requirements are: a clear business case, trustworthy data, a well-structured in-formation flow, uniformity in data collection / analysis and the proper use of retrieved information.

3.4 Business Intelligence / Data Mining and Insurance

The amount of data available at insurance companies is large. However, the usage of this data and acquiring of insights based on this data is often not at an optimal level. Implementing a Business Intelligence cycle and applying data analysis techniques like data mining, could help in optimizing the data utiliza-tion of insurance companies. In this paragraph a practical example of Business Intelligence or data mining in the area of insurance is given.

Smith et al.12 _{describe how data mining techniques are used at the Australian}

Associated Motor Insurers Limited insurance company. The researchers want to find out what the optimal price for an insurance policy is. A price which is too high will result in leaving clients, a price which is too low will result in an increase in losses. Smith et al. used the percentage of clients that had

11_{Vitt, E., Luckevich, M., Misner, S. , ’Business Intelligence} _{Making Better Decisions} Faster’, Microsoft Press, Redmond, 2002.

(25)

3 BUSINESS INTELLIGENCE AND HEALTH INSURANCE 25

to renew their policy but instead terminated their policy. This percentage was 7.1%. They analyzed demographics, the policy holder history and policy details. Also the difference between the old premium and the new premium price was acquired. With this information, data mining techniques were used to classify client groups according to the risk on policy termination. From now on, the in-surance company could see if a user had a high risk of terminating their policy or renewing it.

The writers also used data mining to analyse these claim costs. They did this by using clustering techniques which grouped clients according to cost ratio (total cost of claims / total insurance sum) and frequency (how often does a client claim).

Based on the results of the applied data mining techniques, an optimal pricing of policies is possible. By looking at the predicted claim costs of clients and their behaviour towards adjustment of their insurance policy premium, an optimal price for insurance premiums can be acquired without losing profitable clients. Another example of data mining in the area of insurance is the use of the Intelligent New Business Electronic Application and Automated Underwriting Solution (iNB) at the Western-Southern Life Insurance Company13_{. Using their}

laptop, insurance agents can fill in insurance application papers at the clients home, data is sent to the iNB system which checks the data on errors and analyses the data on patterns (is the client profitable or not?). Based on the found pattern(s), the iNB system gives the results to the agent which approves or declines the application. By using the iNB system, the insurance company increased the integrity of the application data and increased the speed of the insurance policy application process. The above examples both show that the information delivered by the Business Intelligence system is used to base a deci-sion on. Human thinking keeps playing an important role, although, in the last example, automatic decision making could be applied.

3.5 Business Intelligence at DFZ

Companies that introduce Business Intelligence into their organization with-out proper preparation, will end up with a badly functioning system and more dangerously, decisions made upon bad information. Before Business Intelli-gence could be implemented in the procurement department of De Friesland Zorgverzekeraar, problems within this department had to be analyzed. After conducting interviews with employees from De Friesland Zorgverzekeraar, some problems appeared.

The problems I discovered by interviewing employees from De Friesland Zorgverzek-eraar were shortly mentioned in chapter two and will be further analyzed in chapter six. De Friesland Zorgverzekeraar wants to introduce a Business Intel-ligence process to acquire more insights into health care costs. However, before a Business Intelligence process can be introduced, present pitfalls have to be taken care of first. After reading documents about cost control at De Fries-land Zorgverzekeraar, I noticed two pitfalls that were present. Although other pitfalls exist throughout the organization, I decided to mention these specific

(26)

pitfalls because they were dealt with during the research.

• Too rapid introduction of data mining

• Only stick to the functionality of available software tooling An explanation of these problems is given now.

3.5.1 Too rapid introduction of data mining

De Friesland Zorgverzekeraar posed an initial research proposal in which the concept of data mining was present. The concept was also mentioned in project plans of projects which were already running during the writing of this thesis. While the projects were in the initiation phase, it is hard to talk about employ-ees who are data mining when data mining and Business Intelligence are not yet introduced. Advice was given to the members of the project team to take the introduction of data mining as a step-by-step process which should be exe-cuted in an orderly and timely fashion (structured and not too fast). Business Intelligence is a prerequisite of data mining and therefore is more suitable to be used as a concept instead of data mining (the team agreed on this subject). 3.5.2 Only stick to the functionality of available software tooling At De Friesland Zorgverzekeraar, software tooling is available for data analysis. SPSS, Microsoft’s SQL Server and also Excel and Access are licensed software. SQL Server is not used yet but employees are informed about the software and its possibilities. Although education of employees should be seen as a positive aspect, a presentation of one software package has a negative influence on the objectivity of these employees. The choice of a software package should be dependent on the suitability to fulfil the requirements of the organisation. When employees are involved in the process of choosing a software package (which they should be), objectivity is important. A software package that is doing the job best should be chosen instead of the situation in which jobs are chosen based on the functionality of the software.

(27)

4 Procurement

In this chapter, the concept of procurement will be handled. First, different definitions of procurement will be discussed. Secondly, existing procurement process models will be shown and will be compared with each other. This chapter is meant to give a solid overview of the procurement concept. The next chapter is more in-dept and will deal with the procurement process of De Friesland Zorgverzekeraar.

4.1 Procurement defined

In procurement literature, authors discuss terms like procurement, purchasing, contracting out and supply chain management. The use of these terms is done intermittently, resulting in the blurring of the conceptual meaning of the dis-tinct terms. In this paragraph, different terms will be clarified, concluding with a definition of procurement that will be used throughout this thesis. According to Heinritz et al.14 purchasing is:

”.. the management of materials in flow, from the establishment of sources and ’pipelines’, through inventory stores, to the ultimate delivery at production stations as needed.”

As can be seen in the definition of the purchasing term mentioned above, Heinritz et.al. see purchasing as a managing function. The focus of the function is aimed at the operational area and is more suitable for the purchasing of products then that of services because of the mentioned ’inventory stores’ in the definition (services cannot be stored). A definition more aimed at services is given by R.H.A. Van Stekelenborg15_{. Besides the term purchasing, which}

refers to products, he mentions the term ’contracting out’. According to Van Stekelenborg the term ’contracting out’ is used for the acquiring of processes or activities. Although both terms differ from each other, Van Stekelenborg uses the term purchasing throughout his dissertation to include both products and services. He chooses to do so because, according to him, both concepts ”..try to satisfy the needs of an organization by making use of external resources in a cost-effective way”.

14_{S.F. Heinritz et al.,”Purchasing: Principles and Applications”, p. 7, 7th Edition, Prentice} Hall, New Jersey, 1986.

15_{R.H.A. Van Stekelenborg, ”Information Technology for Purchasing”, p.} _{105, KPMG} Management Consulting, Eindhoven, 1997.

(28)

The term procurement is mentioned by S.F. Heinritz et al. as a purchasing activity with corresponding sub-activities. Other purchasing activities defined by Heinritz et al.16 _{are research, materials management and basic information}

maintenance. It is thus seen as a part of purchasing. Van Weele17 _{does not}

agree with Heinritz. He sees purchasing as a part of procurement. Van Weele defines procurement as:

”All activities, aimed at the control and guiding of the stream of incoming goods from the moment they leave the supplier until they are used in production.” When looking at this definition, it appears as if Van Weele sees the procure-ment process, just as Heinritz et al., as a process focused at the operational area. Activities that need to be executed beforehand like market analysis and requirements specification are neglected in his definition. Also activities after the products / services are delivered, like evaluation, are absent. Van Weele however, does includes these activities in his model (shown later on in this chapter). The definition of procurement and the procurement model of Van Weele therefore do not fully match. A broader definition should be constructed that also includes phases like market analysis and specifying of to-be delivered goods / services. It should also contain services because, just as the definition of Heinritz et al., the definition of Van Weele is only aimed at the processing of ”incoming goods”. We constructed a new definition which both includes services and products and also matches with the model of Van Weele. The definition of procurement that will be used throughout the rest of the thesis is given below. ”All activities, aimed at the control and guiding of the stream of incoming goods and/or acquired services reaching from the analysis of business needs to the evaluation of the provided good/service.”

The change from the term supply into provide is made because a provider can provide both goods and services while the term supplier is more used when one talks about products.

4.2 Procurement models

Throughout the years, many procurement models have been developed by var-ious authors and institutions. In this paragraph, two procurement models will be shown and described shortly. A comparison between the models is also made. Finally, the procurement model developed by the NEVI will be further elucidated.

4.2.1 Van Weele’s Model

A.J. Van Weele developed his procurement model in 1999. In his book ”Inkoop in strategisch perspectief: analyse, planning en praktijk”18, he describes a pro-curement model consisting of six phases. These six phases are: specify, select,

16_{S.F. Heinritz et al., p. 12-13.}

17_{A.J. Van Weele, ”Inkoop in strategisch perspectief: analyse, planning en praktijk”, 3e} druk, p. 11-13, Kluwer, 1999.

(29)

4 PROCUREMENT 29

contract, order, monitor and evaluation/follow-up. The whole process runs from provider to customer. The model is visualized below:

Figure 5: Van Weele’s Procurement Model

4.2.2 The NEVI Model

The Nederlandse Vereniging Voor Inkoopmanagement (NEVI) has developed the so-called NEVI-model. The model is displayed below.

Figure 6: The NEVI Procurement Model

As can be seen, the NEVI-model consists of six phases: sourcing, specifi-cation, selecting, contracting, providing, added value by organization. The six phases are framed by a framework consisting of four influential factors. Two of these factors are actors namely the procuring organization and the providers of the (to-be) procured goods/services. The other two factors are the strategy of the procuring organization and its information infrastructure.

4.3 Comparing the models

(30)

the procurement phases will be given in the next paragraph). Another differ-ence between the models is the absdiffer-ence of influential factors on the procurement phases in the model of Van Weele. These factors are visualized in the NEVI model by the yellow framework.

Other differences can be found when one looks at the procurement phases. The NEVI model describes phases where business needs are analyzed and market analysis is done (sourcing). This phase is absent in the model of Van Weele. His model starts with the specifying of product/service criteria. The phases specifying, selecting and contracting are mentioned in both models. Another difference can be found in the ordering phase mentioned by Van Weele. The NEVI model does not mention this phase and goes directly to the providing phase. The monitoring phase of Van Weele is also not mentioned in the NEVI model. A difference also occurs in the last step of the NEVI model and Van Weele’s model. Van Weele describes the final step as an evaluation step which is a broad concept. This evaluation phase is also present in the NEVI model but is more aimed at the evaluation of the own organization and its influence on the procured products/services than of the evaluation of both the providers and the own organization.

Another difference is the absence of the external customer in the model of Van Weele. In his book, he mentions that the procurement process runs from sup-plier to customer but these actors are not shown in his model. In the NEVI model, the external customer is present.

4.4 The procurement phases

To get a better understanding of the NEVI model and the procurement pro-cess, an interview with the chairman of NEVI, J.W. Dressel, was conducted. He reflected on the procurement process of health care services and gave an explanation of the NEVI model. The model will be clarified now.

The (yellow) framework of the NEVI model consists of four factors: pro-curement organization, suppliers, strategy and information infrastructure. Just as in the model of Van Weele, the procurement process is a process done by the procuring organisation and suppliers, the actors. The choice for the acquired products or services is influenced by the demands of the external customer. As the model visualizes, the six phases all evolve around this external customer. The customer is therefore an important aspect of the procurement process and should not be ignored.

The other two parts of the frame, strategy and information infrastructure, are the influences that come from both inside and outside the company. The com-pany’s overall strategy should influence the procurement strategy. The overall business strategy is composed after the external market has been analysed (by i.e. using Porter’s Five Forces Model). So, in an indirect way, the external sources present in this market, have an influence on the procurement process.

(31)

4 PROCUREMENT 31

the more effective and efficient the procurement process can be executed because information is used in an orderly fashion.

4.4.1 Sourcing

In this phase, the defining of needs is done by the procuring organization. An external scan (who are the providers, what are they providing?) and internal scrutiny (what do we want to procure?) are executed. The total spend analysis is also done in this phase. Sourcing should be seen as the phase in which a broad, market-wide, analysis is done. Vollman et al.19 _{define sourcing as:}

”the market research of potential delivery sources, the insurance of the conti-nuity of the supply, the search for alternative sources and the maintaining of according knowledge”.

4.4.2 Specifying

The organization analyzes the products / services that fulfill the needs analyzed in the sourcing phase. Also the providers of to-be procured products and/or ser-vices are analyzed (market analysis). The analysis done in the specifying phase zooms in on possible providers (thus more narrow than the sourcing phase). According to Kamann20_{, providers cannot be equally treated. Every provider}

is different and the organization should be aware of this dissimilarity. Kamann proposes the use of the Kraljic Matrix in this phase. Dressel also obliges em-ployees to use this Matrix. With the help of the Kraljic Matrix, the products / services can be classified. Kamann says:

”the Kraljic matrix is an instrument to continuously evaluate the position of suppliers and the most suitable strategy”.

4.4.3 Selecting

In this phase, the selection of suppliers is done. The most suitable suppli-ers are selected based on the results from the request for information (what are the specifications of possible to-be procured product or service? (quality, quantity)) and request for quotation (what is the price for the to-be procured product/service?).

4.4.4 Contracting

After the supplier has been chosen, the contract is devised. In this contract, a service level agreement (SLA) is included. According to Dressel, risk man-agement is also important in this phase. By making a settlement about the responsibilities of both the supplier and the own organization, costs of repair-ing a defect in a product or adjustment of a certain service, are dealt with. A balance between value, risk and costs needs to be found. Also responsibilities

19_Th. _{E. Vollman, W.L. Berry, D.C. Whybark, ”Manufacturing Planning and Control} Systems”, McGraw-Hill, Illinois, 1984

(32)

have to be made transparent in the contract to overcome friction between both players and ensure cooperative success.

4.4.5 Providing

In the providing phase, the products or services are delivered to the organiza-tion or to the customer. During this phase, monitoring is done to check if the SLA is met. When monitoring, the organization should not only be checking the agreements of the supplier but should also be looking inward, into the or-ganization itself. Is the oror-ganization executing the contract with suppliers as promised? Important is to not only check the organizations performance or the supplier performance but also the interaction between both actors.

4.4.6 Organizational value added

In the final step, an evaluation is executed of the delivered products / services, the suppliers and the own organization. In this evaluation, the added value of the organization to the procured products / services is analyzed. Due to the cyclical process in the NEVI model, the results of the evaluation are used in sourcing phase.

(33)

5 Procurement at De Friesland Zorgverzekeraar

In this chapter, first the role of De Friesland Zorgverzekeraar in the health care market will be explained. After this is done, the procurement process of De Friesland Zorgverzekeraar will be analyzed with the help of information acquired from various interviews with employees of the procurement department of De Friesland Zorgverzekeraar.

5.1 The role of De Friesland Zorgverzekeraar

De Friesland Zorgverzekeraar has an intermediary function in the health care market. The organization makes deals with all kinds of providers of health care services and health care products. By making agreements with these providers, De Friesland Zorgverzekeraar ensures the availability of health services and products to its clients. Van Montfort21 _{visualizes the role of health insurers}

in his three market model displayed below.

Figure 7: The three market model

The squares in the model of Van Montfort are split into two parts. This is done to show the different roles and relationships between health care providers, clients / patients and insurers. E.g. the relationship between a health care organization and a person is that of a health care provider and patient. When looking at the relationship between that same person needing health care and that of a health care insurer, the person is not a patient but a client of the health care insurance company. The roles are thus dependent on the relationship between the three entities.

Olmsted Teisberg et al.22 describe the situation of the actors in the three markets. On the one hand, the health care providers often have incentives to provide as many services as possible because they receive premiums for every service provided. The health insurer on the other hand wants to spend less on patient care than was received in premiums. The patient wants to find the best quality care regardless of costs. The actors thus all have different incentives. All parties have to find the golden mean for cooperation to work. In the case of De Friesland Zorgverzekeraar, the focus is aimed at the insured client. If the client is treated the right way and stays insured at De Friesland Zorgverzekeraar, the

21_{R.M. Lapre, A.P.W.P. van Montfort, ”Bedrijfseconomie van de gezondheidszorg”, Reed} Business Information, 1997.

22_{E. Olmsted Teisberg et al., ”Making competition in Health Care Work” in Harvard} Busi-ness Review, Juli - Augustus 1994.

(34)

company has done a good job. However, the total costs of treatments provided to clients should be lower than the total amount of premiums paid by these clients. De Friesland Zorgverzekeraar is losing money when the total amount of premiums paid is lower than the total costs of provided treatments which is a non-desirable situation. The costs for procured health services / products should therefore be as low as possible and be perceived by the insured client as high quality products / services. A well-defined and executed procurement process of the health care products / services is required to reach this goal.

5.2 Procurement at De Friesland Zorgverzekeraar

After conducting interviews with various employees of De Friesland Zorgverzek-eraar, the procurement process was made clear. At De Friesland Zorgverzekeraar the model of Van Weele (elucidated in chapter four) is used for the procurement process. The model is shown below once again:

Figure 8: Van Weele’s Procurement Model

The period in which the procurement process of De Friesland Zorgverzeker-aar is executed is a year and must therefore be seen as a long running process. The specifying phase starts in January and the contracting phase ends in De-cember. Because the research period was shorter than the duration of the whole procurement process of De Friesland Zorgverzekeraar, it was impossible to fol-low the whole process in real-time. Fortunately, the conducted interviews gave me a thorough insight into the organisation’s procurement process. A detailed description of the procurement process of De Friesland Zorgverzekeraar will be given in the next section. The description covers the overall procurement process of De Friesland Zorgverzekeraar and is therefore not specified on the procurement teams existent at De Friesland Zorgverzekeraar (i.e. pharmacists or hospital care). It is however specified towards Cure and not Care (explained in chapter one).

5.2.1 Company wide activities

Although the procurement process is done at the procurement department of De Friesland Zorgverzekeraar, overall company activities are influencing this process. First these global activities will be explained. This is done because knowledge of these dependencies will result in a better understanding of the procurement activities.

Longterm policy planning

(35)

5 PROCUREMENT AT DE FRIESLAND ZORGVERZEKERAAR 35

Annual Business Plan

In the annual business plan of De Friesland Zorgverzekeraar, the strategic goals per department are defined together with the Key Performance Indicators (KPI) and Critical Success Factors (CSF). For 2008, the CSFs for the procurement de-partment are to make the quality of health care more transparent for clients, enable segmented and differentiated approaching of health care providers with the help of acquired information through clients and benchmarking of health care costs.

Budget

The annual budget is made between August to October. The amount of avail-able money is an important aspect when selecting and contracting health care providers.

Product adjustment

From February to September adjustments are made to the health care packages from which clients can choose (for the following year). Changes in health care laws and the available amount of money, influence the specification of these health care packages.

The phases of the procurement process (according to Van Weele) of De Friesland Zorgverzekeraar will be dealt with now.

5.2.2 Specify External scan

In the specifying phase, De Friesland Zorgverzekeraar orientates itself on the market of health care. The specifying phase starts in January with the analysis of supply and demand in the health care market. De Friesland Zorgverzekeraar needs to know who the players in the market are and what these players have to offer (external scan). The portfolios of different suppliers are asked and sent to De Friesland Zorgverzekeraar. With the help of these portfolios, De Friesland Zorgverzekeraar plots the different products and services of the suppliers. The plotting of the provided information of the suppliers results in an overview of products and services available on the market.

Use of factbooks

The procurement department uses factbooks which contain information about health care providers (how many treatments did they execute, how many pre-scriptions did they write). With the facts in the factbooks, the employees of the procurement department can acquire information about the performance of health care providers.

Evaluations processing

(36)

was the overall treatment?). Not only the results of the evaluations done by the clients are processed but also the results of the evaluations done by De Friesland Zorgverzekeraar are included.

Specification formulation

The two phases above result in an overview of the products and services avail-able on the market. However, a closer look should be given to these products and services. This is done in the specification (re)formulation phase executed in March to June. Because the specifications of products and services change over time, an adjustment of specifications by De Friesland Zorgverzekeraar is performed. De Friesland Zorgverzekeraar analyzes the products and services on an individual basis (How is a certain service executed? What does the prod-uct do?). After the analysis of services and prodprod-ucts is done, De Friesland Zorgverzekeraar decides per health category (medical equipment, dental care, physical therapy etc.) which services and products suit these categories best. This results in defined packages with services and packages per health category. Defining of strategy and policies

The external scan, evaluations processing and specification (re)formulation phases result in an overview per health category of the available suppliers and their products, services and their quality. With the help of this information and the overall strategy and policy of De Friesland Zorgverzekeraar, the strategy and policy per health category are defined. This is done with the help of Porter’s Five Forces Model (strategy) and the Kraljic Matrix (category management). The defining of strategy and policies has influence on the selection of providers of health care products and services. Based on the defined strategy/policy, De Friesland Zorgverzekeraar will or will not select and order products/services from these providers. To keep the relationship between providers, client organi-zations and De Friesland Zorgverzekeraar as good as possible, the providers and client organizations are involved in the defining of the policies of De Friesland Zorgverzekeraar. In August, the public version of De Friesland Zorgverzek-eraar’s health care policies are made and are published on the website of De Friesland Zorgverzekeraar.

Select

Formulating procurement criteria

Throughout the months May and June, the procurement criteria of De Fries-land Zorgverzekeraar are formulated. This phase is closely related to the defin-ing of strategy and policy, done in the previous mentioned phase. The chosen strategy and policies influence the procurement criteria. E.g. If De Friesland Zorgverzekeraar’s strategy is to increase the number of clients and wants to do this by keeping the monthly fee beneath that of its competitors, this decision influences the procurement criteria. De Friesland Zorgverzekeraar can choose to not reimburse expensive medicines, that previously were reimbursed, to de-crease costs.

Selecting providers

Business Intelligence and Procurement processes