Drivers, costs, and revenues of customer satisfaction in a business-to-business context

(1)

Drivers, costs, and revenues of customer satisfaction

in a business-to-business context

- Master thesis -

By

Mathijs Kostermans (0300012)

March 6th 2009

Master Business Economics

Organisation Economics

Faculteit Economie en Bedrijfskunde

Universiteit van Amsterdam

Supervisors:

Dr. ir. S.P. van Triest

(2)

Drivers, costs, and revenues of customer satisfaction

in a business-to-business context

- Master thesis -

By

Mathijs Kostermans

March 6th 2009

Master Business Economics

Universiteit van Amsterdam, the Netherlands

Abstract

Using customer satisfaction survey data of an international steel producer, this study determines drivers of satisfaction in a business-to-business (B2B) context. Drivers of satisfaction in both business-to-customer (B2C) and business-to-business contexts are reviewed, because there are some similar drivers within these two contexts. Differences between these two contexts are the customer, degree of rationality, and term of relationship, causing also differences in satisfaction drivers. Drivers in this specific B2B context are determined by analysis of a satisfaction survey as well as determining the effect of support, reliability scores, complaints and recovery, volume, transportation costs, and production costs. The effect of satisfaction and its drivers on price and margin is the second analysed link in this study. With respect to satisfaction, only satisfaction components product quality and delivery and sales affect overall satisfaction significantly, implicating the rationality of this specific B2B context. With respect to price and margin, satisfaction is not a driver. This study finds a negative correlation of delivery reliability scores with both satisfaction and price, whereas salespeople’s support also affects satisfaction negatively. The price of steel products is also a decreasing function of volume, implicating economies of scale.

(3)

3

1 Introduction

Customer satisfaction has been for decades a topic of academic research and managerial interest, because customer satisfaction is an important driver of profits and future profits (Ittner and Larcker, 1998). In business-to-consumer (B2C) contexts, customer satisfaction is defined by Oliver (1980) as the difference between perceived value and expected value. Reading scientific articles and books about customer satisfaction, customer loyalty, and customer costs, results in many studies applying these concepts on the business-to-customer (B2C) context, like hotels, retail banking, and other consumer products. There is however little research available in business-to-business (B2B) contexts. Although this study analyses customer satisfaction in a B2B context, customer satisfaction in B2C contexts is also reviewed because of the similarities of satisfaction drivers in both contexts. The most important models of customer satisfaction in the B2C context are developed by Oliver (1980), Kano (1984) and Parasuraman et al. (1985, 1988). Only more recently, a satisfaction model for the B2B context has been developed by Homburg and Rudolph (2001). Customer satisfaction in B2B contexts is defined by Homburg and Rudolph as an opinion about a long-term relationship with a supplier providing a desired level of fulfilment related to purchase, including under- or over-fulfilment. Drivers of customer satisfaction in B2C and B2B contexts are the company’s image, expectations of product and service performance, product quality, service quality, and customers’ perceived value

(Anderson et al., 1994). Consequences of customer satisfaction are among others, customer commitment, repurchase intentions, price perceptions and willingness to pay, customer loyalty, word-of-mouth and complaining behaviour, and customer defection (Luo and Homburg, 2007).

The first purpose of this study is to analyse the effect of volume, transportation and production costs, support, delivery reliability scores, and complaints and recovery on satisfaction. The second purpose is to determine the relative importance of the satisfaction components product quality, product packaging and labelling, delivery and sales, technical support, complaint handling, and continuous improvement. The third purpose is to analyse the effect of the above variables and satisfaction

components on the price paid by customers and margin of an international steel supplier. A customer satisfaction survey is used to analyse the effect of volume, transportation and production costs, support of salespeople and technical engineers, percentages of goods produced and delivered on time, and the amount of complaints

(5)

5 and recovery on satisfaction. Subsequently, this satisfaction survey is used to link satisfaction and price.

This study is a contribution to the existing literature about customer

satisfaction and its revenues, because of the variety of data used by this study and its results. Little research is available on customer satisfaction in B2B contexts, making this study in a B2B context a contribution to existing literature. Compared to the INDSAT model of Homburg and Rudolph (2001), expected satisfaction drivers correspond, whereas the importance of these drivers differs. Homburg and Rudolph find all seven expected satisfaction drivers to be significant, whereas this study only finds components product quality and delivery and sales to have a significant effect on satisfaction. Moreover, this study is one of the first to analyse the link between satisfaction and price in a B2B context, resulting in no relation between satisfaction and price. Determining satisfaction drivers and their importance, as well as analysing the link between satisfaction and price is possible because an international steel producer provides a variety of customer specific data including volume, delivery reliability scores, support of salespeople and technical engineers, complaints and recovery, and prices and margins of eight steel products. With this variety of data, this study does not find evidence for a positive link between satisfaction and price.

The construction of this study is constructed as follows: chapter two describes the research question, sub-questions, and methodology of this study. The third chapter contains a literature review on customer satisfaction. This chapter clarifies two

customer satisfaction models in a B2C context, one model in a B2B context, and describes drivers and consequences of customer satisfaction. Chapter four is about the international steel producer Corus, the organisation which provides information and data for the customer satisfaction analysis. The fifth chapter outlines the methodology to analyse drivers, costs, and revenues of satisfaction and describes the data. Chapter six shows the results which lead to an answer to the research question. The seventh chapter concludes results, provides implications to Corus and other companies in B2B contexts, and describes limitations of this study.

(6)

2 Research formulation

This chapter provides the research goal (2.1), the research question, and sub-questions (2.2) leading to an answer on the research question. Paragraph 2.3 describes briefly the methodology of this study.

2.1 Research goal

This study focuses on the relationship between customer satisfaction, costs, and revenues that a company incurs in producing and delivering customer services and products. Purpose of this study is to determine the drivers of customer satisfaction in a business-to-business (B2B) context, analyse the link between expected drivers and satisfaction, and analyse the link between these drivers and satisfaction itself to price and margin. The literature review analyses drivers and consequences of

customer satisfaction in B2C contexts and other B2B contexts. Subsequently, costs of producing and delivering customer services are linked to customer satisfaction

components. In this way the effect of drivers to customer satisfaction is analysed, together with the effect of these drivers and satisfaction itself to the price and margin.

2.2 Research question and sub-questions

Because most studies focus on customer satisfaction in B2C contexts, this study in a B2B context is a contribution to the existing customer satisfaction

literature. Focusing on customer satisfaction is crucial, because satisfaction can result in increased loyalty of existing customers, reduced price elasticity, secretion of existing customers from competition’s effort, lower transaction costs in the future, reduced failure costs, lower costs of acquiring new customers, and an enrichment of the firm’s reputation (Luo and Homburg, 2007). These effects of satisfaction have a positive relationship with profits and future profits. This study tests the drivers of satisfaction found by Homburg and Rudolph (2001) and tests the link between satisfaction and the price and margin. The research question is as follows:

(7)

7 Research Question

What are the drivers, costs, and revenues of customer satisfaction?

Sub-questions leading to an answer on the research question stated above are described next. The first sub-questions analysed in this study are the following ones:

Sub-questions 1 and 2

1. What are drivers of customer satisfaction according to economic research? 2. What are the consequences of customer satisfaction according to economic

research?

The literature review in this study starts with an overview of drivers of customer satisfaction in B2C and B2B contexts. Customer satisfaction models of Oliver (1980), Kano (1984), and Homburg and Rudolph (2001) are used as theoretical starting points in the literature review. With respect to these models, the literature review describes several studies testing these customer satisfaction models. The second sub-question examines the consequences of customer satisfaction. According to Luo and Homburg (2007), six effects of customer satisfaction can be distinguished: customer

commitment, repurchase intentions, price perceptions and willingness to pay,

customer loyalty, word-of-mouth, and complaining behaviour and customer defection. The literature review describes these consequences and analyses studies investigating these effects of customer satisfaction.

As described above, the first part of the literature review is about customer satisfaction. The empirical part of this study starts with results of a Corus’ satisfaction survey on product and service components. From this survey, customer satisfaction drivers are derived and compared to drivers concluded by other studies. Divisions contributing to customer satisfaction are used to analyse the effect of support on satisfaction. Employees of these divisions are asked about their contribution of time spent to specific customers. The expected link between these two variables is negative, because customers receive more support from a supplier due to problems. When there are no customer problems, support of salespeople and technical engineers is not necessary to both the customer and supplier, except for a little support to

(8)

maintain a good relationship. Other analysed links are between percentages of goods produced and delivered on time and satisfaction, and between complaints and

recovery and satisfaction. Therefore, sub-question three is as follows:

Sub-question 3

3. What is the link of time and money spent at customers, production and delivery on time, complaints and recovery, volume, transportation costs, and production costs

with satisfaction?

The final step is to analyse the link between drivers of satisfaction and satisfaction itself, and profits. The expected link between these two variables is positive, because more satisfied customers are expected to accept higher prices (Homburg, Koschate, and Hoyer, 2005). Therefore, the final sub-question is formulated as follows:

Sub-question 4

4. What is the effect of satisfaction drivers and satisfaction itself on the price and margin?

The results of this analysis are interesting to both scientific researchers, because it contributes to the existing economic literature about customer satisfaction and its revenues, and managers at Corus, because this study generates the most important satisfaction components and analyses effects and revenues of these components.

2.3 Methodology

To answer the research question and sub-questions, a study of Homburg and Rudolph (2001, described in paragraph 3.3) is used as starting point to this study. Their study identifies satisfaction components in B2B contexts. This study links the Corus satisfaction survey components to the satisfaction dimensions of Homburg and Rudolph, in order to find differences and similarities.

(9)

9 The first step of this study is to analyse drivers and consequences of customer satisfaction. Oliver’s disconfirmation paradigm and Kano’s three components model are reviewed in B2C contexts and Homburg’s and Rudolph’s INDSAT model in B2B contexts. The literature review also describes studies testing the results and expected drivers of satisfaction following from these models. Finally, the literature review describes consequences of customer satisfaction.

Corus’ customer satisfaction survey of 2005-2007 is the starting point to the empirical part of this study. In this satisfaction survey, customers rate Corus about their satisfaction with product, delivery, sales, technical services and complaints, continuous improvement, and overall satisfaction. The next step is to identify drivers of satisfaction. This identification is done with factor analysis on the survey answers and with data on divisions’ support to customers, percentage of goods produced and delivered on time, complaints and recovery, production costs, transportation costs, and volume. The correlation of these variables is analysed and means are compared using these variables as two separate groups with low and high values. The final step is to analyse the effect of satisfaction and its drivers to price and margin of steel products. Again, correlations are analysed and means are compared using satisfaction and variables driving satisfaction as grouping variables.

(10)

3 Literature review

The third chapter describes the theoretical framework, which is used as starting point to this study. The topic of this chapter is drivers and consequences of customer satisfaction: three different customer satisfaction models in B2C and B2B contexts are distinguished and tested by other studies (3.1 and 3.3). Although this study uses data from a steel producer in a B2B context, customer satisfaction in B2C contexts is also reviewed, because the B2C customer satisfaction drivers correspond to a large extent to B2B satisfaction drivers. The effect of performance and complaint handling and recovery is analysed in the empirical part of this study, whereas

expectations and disconfirmation are not due to lack of data on expectations. Paragraph 3.2 describes differences between B2C and B2B contexts, whereas paragraph 3.4 describes consequences of customer satisfaction.

3.1 Customer satisfaction models in B2C contexts

The first paragraph of this literature review describes customer satisfaction models in B2C contexts. Two models in the B2C context are analysed and tested by other authors. Drivers of customer satisfaction in B2C contexts resulting from the following review are expectations of product and service performance,

disconfirmation of expectations, performance, and complaint handling and recovery.

3.1.1 Oliver’s expectancy-disconfirmation paradigm

An important customer satisfaction model in B2C contexts is Oliver’s expectancy-disconfirmation model. The following part of the literature review describes Oliver’s model, articles testing this model, and a feature of Oliver’s model which need special interest: the absence of complaint handling and subsequent recovery. The absence of complaint handling is important because Oliver does not include this variable in his model, whereas several studies do find a link between complaint behaviour and satisfaction.

Oliver’s (1980) expectancy-disconfirmation model consists of three

components: expectation, disconfirmation, and satisfaction. Customer satisfaction is defined by Oliver as the difference between perceived and expected product and

(11)

11 service quality. When perceived value is higher than expected, performance is

positively disconfirmed and when perceived value is lower than expected, performance is negatively disconfirmed. When perceived and expected value are equal, performance is confirmed. Customer satisfaction is higher, the more positive the disconfirmation of performance is. Oliver states that expectations of customers are influenced by three factors: former experience with the product, brand knowledge and symbolic components, contact with salespeople and social referents, and individual characteristics. Satisfaction is the additive combination of expectations and resulting disconfirmation. Oliver shows with a questionnaire on a flu vaccination program that the difference between perceived and expected value drives satisfaction. Although Oliver does not test his model in an economic context, others have confirmed Oliver’s model in an economic context. However, Oliver does not include complaint behaviour and subsequent recovery into his disconfirmation paradigm. Several studies do find a significant link between complaint handling and customer satisfaction, providing evidence to take this variable into account when analysing customer satisfaction.

Table 14 of the appendix reports an overview of studies analysing and testing Oliver’s expectancy-disconfirmation paradigm. This table describes seven

investigated drivers of satisfaction, their definition, and their resulting effect on satisfaction. The seven drivers of satisfaction are expectations, performance,

disconfirmation, image, complaint behaviour, justice, and recovery. Results from the analysed studies show the significant effect of expectations, performance,

disconfirmation, and image on satisfaction. Moreover, complaint behaviour, justice, and recovery also affect satisfaction significantly.

3.1.2 Kano’s three-component satisfaction model

Another widely accepted model of customer satisfaction is Kano’s (1984) model. Kano is the first to classify service components into categories affecting satisfaction in different ways. Kano discerns three components of product or service requirements: must-be requirements, one-dimensional requirements, and attractive requirements. Must-be requirements are elementary criteria of a product. Fulfilling these demands does not lead to increased customer satisfaction, because customers take these demands for granted. On the other hand, customers are extremely

dissatisfied when must-be requirements are not met. Therefore, must-be requirements can only result in a state of non-dissatisfaction. The second component of product or service requirements is one-dimensional requirements. One-dimensional requirements

(12)

are generally explicitly requested by the customer. Customer satisfaction is

proportional to fulfilment of these one-dimensional requirements: a higher fulfilment level leads to a proportional higher level of customer satisfaction. The final

component resulting in customer satisfaction is attractive requirements. These requirements have the greatest impact on customer satisfaction. Attractive requirements are neither demanded, nor expected by the customer. Therefore, fulfilment of attractive requirements leads to a more than proportional increase of customer satisfaction. On the other hand, not meeting these attractive requirements does not lead to dissatisfaction, because these requirements are not expected and demanded by the customer. The relationship of these three components to satisfaction is depicted graphically in Figure 1. The link between one-dimensional requirements is linear, while the links between attractive and must-be requirements, and satisfaction are not.

Matzler et al. (1996) state some advantages of classifying customer

requirements by Kano’s method. Product components with the greatest influence on customer satisfaction can be identified, because classification of product and service requirements into Kano’s components of customer satisfaction generates an overview of product and service components to focus on. The second advantage of Kano’s classification is that priorities for product development can be deducted. Kano’s method also generates useful help in critical situations in the product development stage. A fourth advantage of Kano’s model is the possible construction of customer-

(13)

13 specific solutions. The final advantage of Kano’s model is that discovering and

meeting attractive customer demands, creates a wide range for differentiation possibilities.

Table 15 of the appendix depicts an overview of articles on Kano’s

classification of satisfaction components and their effect on satisfaction. This table describes studies dividing satisfaction components into must-be requirements, one-dimensional requirements, and attractive requirements. Results of these studies are reported in the final column.

The next paragraph describes differences between B2C and B2B contexts with respect to satisfaction, in order to explain differences in satisfaction drivers in both contexts.

3.2 Differences between B2C and B2B

The preceding paragraph described two customer satisfaction models and satisfaction drivers in B2C contexts. This study, however, focuses on a B2B context. Therefore, this paragraph describes the differences between these two contexts in order to explain similarities and differences with respect to satisfaction drivers in both contexts.

An important difference between B2C marketing and B2B marketing is the customer. In B2C marketing, a company delivers a product or service to a customer, whereas in B2B marketing, a company delivers to another company. Jackson, Neidell and Lunsford (1995) state four other differences between marketing to industry and marketing to customers. Marketing to industry is aimed at establishing a long-term relationship, whereas marketing to customer is aimed at creating a transaction. This also requires a different strategy to approach customers. Other differences according to Jackson et al. are the more complex organizational buying task, the longer period of time needed, and influence by more forces within and outside the company.

Another important difference between B2C and B2B marketing is the rationality of buyers. In B2C marketing, customers are encouraged to buy products due to advertisements, discounts, and positive word-of-mouth, which implicates their buying decision is not completely rational. The degree of rationality of customers’ decisions in B2C contexts is also reduced by the relative low value of B2C products. In B2B contexts in general and certainly in the case of steel, customers think more rationally before they decide to buy a product. Part of the explanation for the higher

(14)

rationality is the higher value and volume of products in B2B contexts bought by customers.

Customer satisfaction in B2B contexts is defined by Homburg and Rudolph (2001) as an opinion about a long-term relationship with a supplier providing a desired level of fulfilment related to purchase, including under- or over-fulfilment. Customer satisfaction in B2C contexts is defined by Oliver (1980) as the difference between perceived value and expected value. These definitions generate one of the main differences between B2B and B2C contexts regarding satisfaction: B2B contexts are long-term oriented. In general, goods and services sold in B2B contexts are more capital-intensive compared to consumer goods. With these differences in mind, the next paragraph is about satisfaction drivers in B2B contexts.

3.3 Satisfaction in B2B contexts

Paragraph 3.1 described two models of customer satisfaction in B2C contexts and associated satisfaction drivers, whereas paragraph 3.2 described differences between B2C and B2B contexts. The next paragraph describes INDSAT, a model of customer satisfaction in B2B contexts, and articles associated to this model. This paragraph investigates satisfaction drivers in B2B contexts. Resulting satisfaction drivers from the INDSAT model in B2B contexts are products, salespeople, product-related information, order handling, technical services, internal personnel, and complaint handling.

Homburg and Rudolph (2001) provide a model with seven dimensions to measure industrial customers’ satisfaction. In B2B contexts, buyers and sellers are often mutually connected and customized products are necessary to satisfy customer needs and demands. Therefore, the customer is an active partner of a company. As a result, satisfaction plays a crucial role in customer relationships in industrial markets according to Homburg and Rudolph. Another reason for the crucial role of

satisfaction is because costs of searching, marketing, and acquiring new customers broadly exceed the costs of satisfying existing customers.

According to Homburg and Rudolph (2001), customer satisfaction has an evident multi-dimensional nature. Homburg and Rudolph distinguish seven different satisfaction dimensions: products, salespeople, product-related information, order handling, technical services, internal personnel, and complaint handling. The product is the heart of customer relationships and reflects issues as product reliability,

(15)

15 salespeople reflects the interaction between salespeople and customers, salespeople’s knowledge of the product and usage conditions, and friendliness of salespeople. Product-related information reflects information given by technical documentation and information in brochures or prospectuses. Order handling reflects speed of order confirmation and delivery times. The dimension technical services reflects the full range of technical services, such as repair, fitting, maintenance, and the speed of availability of service staff. Dimension internal personnel reflects the interaction with internal staff and consists of the attainability of relevant persons and quality of their answers. Complaint handling reflects product-related complaints and actions taken to deal with complaints. Homburg and Rudolph test a seven-dimension model (with the seven drivers described above) of customer satisfaction against four alternative models with less dimensions influencing satisfaction.

Homburg and Rudolph use a pool of 43 items reflecting the seven dimensions described above. Respondents are asked to rate these items from strongly satisfied (5) to strongly dissatisfied (1), to rate overall satisfaction with an organisation, and to rate the likelihood of recommending the company as supplier. The database of this study consists of 5449 surveys sent out to companies in 12 European countries, of which 1679 surveys are returned. From these 1679 surveys, a 29-item instrument (INDSAT) is developed to determine customer satisfaction in B2B contexts. Homburg and Rudolph define overall satisfaction as a construct reflecting the global satisfaction rating of a customer with respect to the customer-supplier relationship, without considering specific satisfaction aspects. Using causal modelling, Homburg and Rudolph show that satisfaction with order handling and satisfaction with salespeople are the most important factors in determining overall satisfaction. The least important factors are satisfaction with product-related information and communication with internal staff. Table 16 of the appendix reports studies analysing drivers of

satisfaction in B2B contexts. The second column of this table reports the definition of satisfaction components, whereas the third column reports results of the studies.

Above paragraphs describe three models of satisfaction in both B2C and B2B contexts. The following paragraph 3.4 describes the importance of focusing on customer satisfaction, by reviewing results of studies analysing consequences of satisfaction.

(16)

3.4 Consequences of customer satisfaction

The preceding two paragraphs describe satisfaction drivers and satisfaction models in B2C and B2B contexts. Focusing on customer satisfaction in both B2C and B2B contexts is important because of several reasons described below.

According to Anderson et al. (1994), customer satisfaction is likely to increase loyalty of existing customers, reduce price elasticity, result in secretion of existing customers from competition’s effort, lower costs of transactions in the future, reduce failure costs, lower costs of acquiring new customers, and result in an enrichment of the firm’s reputation. Increased customer loyalty results in a longer and more

probable purchase pattern from the same supplier. Reduced price elasticity arises, because satisfied customers have an increased willingness to pay for received benefits and they are more likely to be permissive to price increases. Satisfaction results in customer retention and therefore a firm does not have to spend much money on acquiring new customers, resulting in lower future transaction costs. Another reason for lower future transaction costs is that satisfied customers are likely to purchase more often, greater volumes, and other offered products and services. Satisfaction should result in lower costs of acquiring new customers because satisfied customers are more likely to spread positive word-of-mouth. Moreover, satisfaction could make advertising more efficient and allows a company to offer attractive guarantees. The overall reputation of a company is also positively affected by customer satisfaction, which helps in introducing new products, and building and protecting relationships with key customers, distributors, and subcontractors. Because these consequences of satisfaction are likely to lead to profits and future profits, focusing on customer satisfaction is crucial.

Luo and Homburg (2007) provide an overview of outcomes of empirical studies about customer satisfaction. Results of these studies lead to eight customer-related outcomes of satisfaction: three behavioural intentions, three customer behaviours, one financial performance, and one nonfinancial performance.

Behavioural intentions are customer commitment, repurchase intentions, and price perceptions and willingness to pay. Customer behaviours are customer loyalty and repurchase behaviour, word-of-mouth and complaining behaviour, and customer defection. Luo and Homburg test the outcome of customer satisfaction with respect to efficiency-related outcomes and employee-related outcomes, which is irrelevant to this study. However, from their literature review, Luo and Homburg conclude that satisfaction increases customer loyalty and influences future repurchase intentions and

(17)

17 behaviour. Moreover, very satisfied customers are willing to pay higher prices and their sensitivity to prices is lower.

This study analyses the link between expected drivers and satisfaction. The second link this study analyses is the link of satisfaction and its drivers with the price and margin of five steel products. Finding literature about this topic is difficult and to the best knowledge of this study, no study analyses this link. There are however, some studies analysing the link between satisfaction and willingness to pay. Homburg, Koschate, and Hoyer (2005) is one of the studies analysing this link. Homburg, Koschate, and Hoyer execute two experimental studies: a lab experiment and a real experiment with usage experience over time. In their lab experiment, they manipulate experiences with an Italian restaurant in relation to food, ambience, and service, rated by + and –. This definition results in eight different satisfaction scenarios, with corresponding satisfaction scores. Willingness to pay is measured by asking how much participants would pay to visit the restaurant. The result of this lab experiment is that willingness to pay increases when customer satisfaction increases, so satisfied customers are willing to pay more. The real world experiment is about a CD-ROM tutorial providing academic assistance in a class about pricing. Experience is manipulated by adapting the tutorial into an easy and difficult to read version. Feedback of instructor’s assistants is also adapted into positive and negative

performance feedback. The result of this experiment is the same: higher satisfaction corresponds to more willingness to pay. The study of Homburg, Koschate, and Hoyer approaches this study’s purpose the most.

Table 17 of the appendix reports an overview of satisfaction consequences, and shows the importance of focusing on customer satisfaction, both in B2C and B2B contexts. The second column of this table reports the definition of variables used to analyse the consequences of customer satisfaction, whereas the third column reports results.

(18)

4 Corus

The preceding chapter described the theoretical starting point of this study. The purpose of this study is to determine drivers of satisfaction in a B2B context, to analyse the link between these drivers and satisfaction, and to analyse the link of these drivers and satisfaction itself with price and margin. The empirical data of this study comes from an international producer of steel. This steel producer’s strategy for 2010 is to become “Best supplier to best customers”. Because the focus at this company is on satisfied customers, Corus is an excellent company to analyse drivers, costs, and revenues of customer satisfaction. This chapter describes the organisation of Corus (4.1) and Corus’ production process and customers (4.2). A paragraph about the history of Corus is part of the appendix as well as some subsections of paragraph 4.1 and 4.2.

4.1 Organisation

Corus’ business division CSPIJ (Corus Strip Products IJmuiden) produces strip steel for applications in packaging, construction, and automotive. Important divisions of CSPIJ with respect to this study are logistics and transport (L&T), commercial sales (COM), commercial services (COS), and customer technical services (CTS). Commercial services and customer technical services are sub-divisions of commercial sales, whereas logistics and transport is a sub-division of CSPIJ Services.

CSPIJ distinguishes nine industries to which steel is supplied: automotive, steel service centres (SSC), mill to mill, end users, USA, narrow strip, rest of the world (ROW), construction, distribution and building systems (CDBS), and other. The automotive industry consists of vehicles moving by itself, like automobiles, trucks, airplanes, etc. Steel service centres operate as warehouses in order to deliver steel to end markets. The mill to mill industry consists of strip mills and plate mills in which steel is processed further. The branche end users includes precision tubes (automotive and furniture are main segments), tubes, radiators, profiles, racking, and drums. Corus characterises customers in USA as a separate market. USA customers are supplied with yellow goods, automotive goods, and construction goods. To supply USA customers, steel service centres are used to deliver steel to end users. Narrow strip steel is top quality ultra clean steel with a superb surface, of which batteries are made. CDBS consists of yellow goods, steel for truck and trailers, automotive, and

(19)

19 white goods. ROW (rest of the world) includes deliveries which are not in Europe, UK, or USA (rest of the world). The industry other consists of deliveries which cannot be placed into one of the preceding categories.

The organisation of commercial sales (COM), commercial services (COS), and customer technical services (CTS) enables this study to analyse customer satisfaction on customer-level. These three divisions are organised in branches and within these branches into customers. Commercial sales is divided into six branches: mill to mill and rest of the world, USA and narrow strip, end-users, steel service centres, automotive, and distribution and building systems. Within these branches, each commercial sales employee is responsible to at least one customer, so each customer has one salesperson as a contact. This organisation form enables the analysis of customer satisfaction and associated costs and revenues at the customer level, because employees can estimate their time spent at each customer.

Commercial services is also divided into branches, whereas customer technical services is not. Commercial services consists of three branches: automotive, steel service centres, and end-users, mill to mill, and rest of the world. Commercial

services is also organised in the way that each employee is responsible to at least one customer, enabling the analysis of satisfaction on the customer level. Customer technical services is not divided into branches, but employees are responsible to one or more customers too.

4.2 Production process and customers

Corus’ production process consists of several large production units with heavy machinery. Substantial investments are needed to build these production units. With these production units, producing strip steel takes about six weeks from iron ore to the end product steel. These features of Corus’ production process contribute to the characterisation of the process as capital-intensive. Within this process, planning and transport is crucial. Because of the process of continuous production, Corus tries to prevent stagnation as long as producing is profitable. This study analyses the effect on the prices and margins of eight products: slabs of steel, hot rolled dry steel, hot rolled pickled and oiled (p/o) steel, DSP dry steel, DSP p/o steel, full hard steel, cold rolled steel, and galvanised steel. This paragraph describes the production process of these products.

The Oxy steel factory turns steel into one of the best qualities steel in Europe. The steel is adjusted with added substances such as aluminium, manganese, or

(20)

phosphorus, to meet the specific demands of the customer. The Oxy steel factory supplies slabs of steel to the Hot Strip Mill. The slabs are rolled out in here, resulting in hot rolled steel. The Direct Sheet Plant (DSP) is the second client of the Oxy steel factory. Using a continuous casting machine, the DSP produces sheets of steel, which are rolled up after production. The Cold Strip Mill rolls out steel into very thin sheet steel. After every production step, steel can be supplied to customers or supplied to the last link in the chain: Coated Products (CPR). When customers want their steel galvanised, the Cold Strip Mill delivers steel to one of the hot-dipped galvanising lines of CPR. Rolls are welded together and cleaned. Another part of CPR is the paint line, which provides the steel at the customer’s request with a layer of paint. This production process results in eight main steel products analysed in this study: slabs of steel, hot rolled dry steel, hot rolled pickled and oiled (p/o) steel, DSP dry steel, DSP p/o steel, full hard steel, cold rolled steel, and galvanised steel.

(21)

21

5 Data and methodology

This chapter describes the data and methodology used to answer the research question and sub-questions on determining drivers, costs, and revenues of customer satisfaction. Paragraph 5.1 describes the data and analyses differences and similarities with the studies reviewed in the third chapter. Paragraph 5.2 describes the used methodology. The methodology is based on the study of Homburg and Rudolph (2001).

5.1 Data

The data used in this study comes from Corus, a subsidiary of the international steel producer Tata Steel. Corus’ customer satisfaction survey from 2005 to 2007 inclusive is used to identify satisfaction components, to analyse the link between customer costs and customer satisfaction, and to analyse the link of customer satisfaction and its drivers with revenues. This satisfaction survey consists of six components: product, delivery, technical support and complaints, sales, continuous improvement, and surplus value. Each component includes questions on how important a sub-component is to the customer on a 1-4 scale from not important to very important, the rating of how well Corus is doing on an item on a 1-5 scale from not satisfied (with a sad face) to very satisfied (with a happy face), and how Corus performs on each survey component in comparison with other steel suppliers (better, worse, or equal). The final question is whether a customer would recommend Corus to others. This design corresponds to the design of Homburg and Rudolph: their respondents also rate items on a 1-5 scale, from strongly satisfied to strongly dissatisfied, with no verbal labels for scores 2-4. Homburg and Rudolph include a rating of overall satisfaction in their study. This study uses the average of the items of the component surplus value as measure for overall satisfaction. Homburg and

Rudolph also ask respondents about the likelihood of recommending the company the others.

Importance scores allow this study to make a weighted score of satisfaction on each component. An importance score of 1 corresponds to a weighted score of 0.7, an importance score of 2 corresponds to a weighted score of 0.8, an importance score of 3 corresponds to a weighted score of 0.9, and an importance score of 4 corresponds to

(22)

a weighted score of 1. These weights are arbitrarily chosen, whereas weight factors of 4 to 1 instead of 1 to 0.7, are also analysed to check for differences.

The survey concludes with a question about the recommendation of Corus when asked (yes or no). The satisfaction component product includes nine items,

delivery includes four items, technical support and complaints includes six items, sales includes six items, continuous improvement includes five items, and surplus value includes six items. Table 1 gives a further explanation of these six components

and their items.

Satisfaction component Questions (importance scores on a 1-4 scale, performance scores on a 1-5 scale)

1 Products (a) quality of hot rolled (pickled) products (b) quality of cold rolled products

(c) quality of hot galvanized products (d) quality of YMAGINE products (e) product range offered

(f) protection of packaging to products (g) safety of opening and unpacking products (h) disposability or recyclability of packaging (i) clarity and completeness of labelling

2 Deliveries (a) punctual delivery

(b) delivery of agreed quantity (c) lead times

(d) notification in case of delivery delays

3 Technical support and complaints

(a) quality of technical services (b) value of technical services (c) speed of response

(d) timeliness of complaint settlement (e) removal of rejected material (f) reaching a satisfactory solution

4 Sales (a) clarity of order related documents

(b) flexibility relating to orders and order changes (c) speed of order booking and confirmation

(d) user friendliness of customer information network (YMonline) (e) attainability of CSPIJ contacts

(f) service provided by CSPIJ contacts

5 Continuous improvement

(a) new product launches

(b) efforts taken to improve your production process and yield (c) joint development of new and existing markets

(d) reduction of transactional costs (e) realisation of logistical improvements

6 Surplus value (a) commitment shown by CSPIJ (b) value of CSPIJ as reliable supplier (c) rating of CSPIJ regarding to flexibility

(d) figuration of CSPIJ in long-term purchasing plans (e) value of the support of local offices

(f) recommendation of CSPIJ if asked by someone (yes or no)

(23)

23 The component surplus value is a measure of overall satisfaction. Because

recommendation is answered by yes or no, this question is deleted from the overall satisfaction construct. Rating the value of local offices’ support is also deleted, because of few answers. Therefore, overall satisfaction in this study is the average of survey ratings on commitment, reliability, flexibility, and figuration in long-term purchasing plans.

Customers in the Corus satisfaction survey belong to six branches: automotive, steel service centres (SSC), mill to mill, end users, USA, and narrow strip. In 2005 and 2006, the branche tubes/profilers also consists, but after 2006 the branche end users absorbed this branche. Table 2 depicts the number of customers and subsidiaries of Corus within each branche in 2007. In 2005 and 2006, these numbers do not differ much.

Branche Customers (subsidiaries)

Automotive 15 (1)

End users 7 (1)

Mill to mill 3

Narrow strip 5 (1)

Steel service centres 18 (6)

USA 11

Table 2: branches and customers of Corus’ satisfaction survey.

The customers of the satisfaction survey are the basis of this study. The purpose is to link support to these customers and their ROTIF and DOTIF scores to the satisfaction scores from the survey, and subsequently to customer revenues. Every employee of commercial sales, commercial services, and customer technical services is responsible to at least one customer. Therefore, when the amount of full time employee (FTE) per customer per division is known, it is possible to link costs and satisfaction for commercial sales, commercial services, and technical support. Other customer specific available data available is the production costs in 2007, as well as the 2007 transportation costs. Furthermore, all complaints in 2007 for the customer satisfaction survey customers are available, as well as the recovery paid to

compensate these complaints. Customers’ volume is available for the period 2005-2007. This data is used to analyse the effect of these variables on satisfaction.

The satisfaction component deliveries corresponds to logistics and transport (L&T). L&T takes care of intern transport on the Corus site and the delivery of steel from the gate of Corus to the customer’s demanded location of delivery. Costs made

(24)

by L&T therefore reflect the cost of customer satisfaction regarding to deliveries. Because L&T is organised in another way as commercial sales, commercial services, and technical customer services, support of L&T per customer is not available. Therefore, this study uses an alternative variable to analyse the effect of deliveries on satisfaction: the 2006 and 2007 ROTIF and DOTIF scores, which reflect the

percentage of tons ready on time in full (ROTIF) and dispatched on time in full (DOTIF).

The satisfaction survey component technical support and complaints corresponds to the division Customer Technical Services (CTS) and Commercial Services (COS). CTS provides technical support and service to customers, whereas COS handles customers’ complaints. The satisfaction component sales corresponds to the division Commercial sales (COM), whereas services corresponds to Commercial Services (COS).

Continuous improvement is done by all three divisions COM, COS, and CTS. COS tries to lower customers’ transaction costs and realise logistic improvements, whereas CTS tries to improve customers’ production process and develops new and existing markets with customers.

To summarize, the data used in this study consists of customer satisfaction survey scores on importance and performance, time spent to customers by employees of the divisions commercial sales, commercial services, and technical services, ROTIF and DOTIF scores, the number of complaints and following recovery, and the price and margin per ton steel. All this data is available on customer level, which makes this study a contribution to the extant customer satisfaction literature. Table 3 depicts an overview of the operationalisation of used variables in this study.

Subsequently, Table 18 of the appendix depicts a summary of the used data in this study and its descriptive statistics. Each variable is described by the number of observations, average, median, minimum, maximum, sum (when relevant), and standard deviation.

(25)

25

Variable Operationalisation

Average customer satisfaction

Average of performance scores on the questions of satisfaction components 1-5

Weighted average customer satisfaction

Weighted average of performance and importance scores on the questions of satisfaction components 1-5 Comparison with other

suppliers

Average of comparison questions at the end of satisfaction components 1-5

Overall satisfaction Average of performance scores on the questions of satisfaction component 6

Recommend Answer on the question to recommend CSPIJ to others

Price Selling price of a ton steel for eight products: slabs, hot rolled pickled and oiled (p/o), hot rolled dry, DSP p/o, DSP dry, full hard, cold rolled, and galvanised steel

Margin Profit margin of sold ton of steel for above stated eight products

ROTIF (ready on time in full)

Percentage of orders ready (produced) on time in full during a year

DOTIF (dispatched on time full)

Percentage of orders dispatched on time in full during a year

Volume Volume of steel delivered to customer during a year Transportation costs Transportation costs (on Corus’ site and from Corus to customer) of delivered tons of steel during a year Production costs Production costs of delivered tons of steel during a

year

Support COM Time spent by COM employee to a specific customer, based on the assumption that (s)he spends 100 percent of her Corus-time to customers

Support COS Time spent by COS employee to a specific customer, based on the assumption that (s)he spends 100 percent of her Corus-time to customers

Support CTS Time spent by CTS employee to a specific customer, based on the assumption that (s)he spends 100 percent of her Corus-time to customers

Complaints Number of complaints of a customer during a year Recovery Recovery paid to a customer during a year due to

complaints

Table 3: operationalisation of used variables.

With respect to the literature review in the third chapter, the available data does not enable testing Oliver’s disconfirmation paradigm, because data on expectations and disconfirmation are not available. However, in B2B contexts, expectations are expected to be an important driver of satisfaction too, because long-term relationships and salespeople raise expectations about quality, delivery, and support. The correlation of expectations and disconfirmation with satisfaction could be a topic for future research. In B2B contexts, performance or perceived value is also expected to be a driver of customer satisfaction. In contrast with expectations and disconfirmation, the variable performance is analysed in this study. Performance of

(26)

the product, logistics and transport, salespeople, and technical engineers are expected determinants of customer satisfaction in this study, and are tested on their effects to satisfaction.

Oliver’s model ignores complaint handling and recovery. Several authors find a significant relationship between these variables and customer satisfaction, for example Bearden and Teel (1983) and Bitner et al. (1990). These studies reveal the importance of complaint handling and associated compensation in order to restrict customer dissatisfaction. In case of product or service failure, several recovery options are possible. Smith et al. (1999) suggest compensation, response speed, apology, and recovery initiation to bound the negative effect of failure on satisfaction. These solutions result in a positive effect on satisfaction after product or service failure. In B2B contexts, service failure recovery seems even more important, because of the long-term relationship between supplier and customer. Repeated product or service failure, which is not compensated by the supplier, could lead to termination of the relationship. Termination of this relationship results in substantial opportunity costs to the supplier, because future profits are foregone. Therefore, product or service failure recovery is expected to be an important driver of satisfaction in both B2C and B2B contexts. This study analyses the effect of complaint handling to customer satisfaction in a B2B context, using the number of complaints and the amount of recovery paid to customers to analyse the effect on satisfaction.

With respect to Kano’s model, Kano’s satisfaction requirements differ for individual companies and industries, not only in B2C contexts, but also in B2B contexts. Theoretically, must-be requirements in this study are the satisfaction components product and delivery, the elementary criteria of the product itself. These components are expected by the customer and can only lead to a small increase in satisfaction. One-dimensional requirements are the satisfaction components

commercial sales, commercial services, technical support, and complaint handling, because these services are explicitly requested by customers and needed to sell and produce steel. This human interaction between supplier and customer can lead to a proportional increase or decrease in satisfaction. Part of the activities of technical engineers can be classified as attractive requirements. Corus’ technical engineers work closely together with their customers. They try to improve customers’

production processes, to develop existing and new markets together with customer, to lower customers’ costs, and to realise logistic improvements. These activities can be done by customers on their own, but they receive help from technical engineers of Corus. Therefore, a part of the technical engineers’ activities, called continuous improvement, can be classified as attractive requirements. This classification into

(27)

27 Kano’s requirement is a forecast of what the division would be in the case of Corus. However, this classification is not statistically analysed.

With respect to satisfaction in B2B contexts, this study follows the

methodology of Homburg and Rudolph, because of the similarity of satisfaction data and purposes. Using factor analysis, underlying components are distinguished. The design of INDSAT, with seven drivers of satisfaction, is compared to Corus’ satisfaction survey, with five drivers of satisfaction: product, delivery, technical support and complaints, sales, and continuous improvement. Comparing these two designs leads to the following differences: (1) the survey components delivery and sales are reflected by order handling in the INDSAT model, (2) the survey component technical support and complaints are divided in the INDSAT model into two

dimensions: technical services and complaint handling, (3) the INDSAT dimension product has more items in the Corus satisfaction survey, whereas the INDSAT dimension salespeople has less items in the survey, and (4) the Corus satisfaction survey includes another dimension: continuous improvement by Corus and jointly with customers.

Besides the analysis of satisfaction drivers, this study analyses the effect of support of salespeople and technical engineers on satisfaction, the effect of the percentage of goods produced and delivered on time on satisfaction, and the effect of volume, production costs, transportation costs, the number of complaints, and the amount of recovery on satisfaction. Subsequently, this study analyses the link of these variables with price and margin of steel products, as well as the link of satisfaction itself with price and margin. Results of these analyses are compared to Homburg, Koschate, and Hoyer (2005), who analyse the effect of satisfaction to willingness to pay.

Corus’ 2010 strategy of becoming best supplier to best customers, makes Corus an excellent company to analyse drivers, costs, and revenues of customer satisfaction, because their focus is on customer satisfaction. Corus’ strategy is based on six focus areas: top service to selected customers, growth in high value-adding products creating value in the supply chain, strengthening market positions in the supply chain, rapid market oriented innovations, optimal utilisation of current plant, and continuous improvement. Focusing on these six areas should lead to becoming: “Best supplier to best customers” in 2010. With “best supplier” Corus means best on every component, the best quality, best service, best performances, best neighbour (environmental), and best employees. CSPIJ wants to build an outrageous and durable relationship with her most prosperous customers. Corus wants to strengthen her position in western Europe, the United States of America, and the automotive sector.

(28)

Corus also chooses growth in high-quality products and selective growth in eastern Europe. Further strategic headlines are: involvement of all employees, value creation with products and services, strengthening and expansion of the market position, optimal utilization of production capacity, an optimal supply chain performance to customers, and a return on investment (ROI) of more than fifteen percent. In 2010, according to Corus’ strategy, the purpose of CSPIJ is to score 8.0 on average in their annually customer satisfaction survey.

5.2 Methodology

The methodology of this study in order to answer the research question and sub-questions, uses the study of Homburg and Rudolph (2001) as example, because their data and goals correspond to the ones in this study.

The first step is to distinguish satisfaction components. These components are derived from Corus’ customer satisfaction survey of 2005-2007, and are related to satisfaction components derived from the literature review in B2C and B2B contexts. Processing the satisfaction survey data is the next step. From this survey, five

variables are deducted: average customer satisfaction, weighted average customer satisfaction, comparison with other suppliers, overall satisfaction, and

recommendation. Average customer satisfaction is the average of the performance scores (on a 1-5 scale) on the questions of the satisfaction components product, delivery, technical support and complaints, sales, and continuous improvement. The weighted average customer satisfaction is the average customer satisfaction corrected by importance scores. The variable comparison with other suppliers is the average of the six comparison questions at the end of the satisfaction components. To quantify these scores, a score of “CSPIJ better than competitors” corresponds to a value of 2, a score “both equal” corresponds to a value of 1, and a score “CSPIJ worse than

competitors” corresponds to a value of 0. The variable overall satisfaction is the average of performance scores on the questions of satisfaction component surplus value, whereas the variable recommend consists of the single answer on the question whether a customer would recommend Corus to others when asked. A

recommendation corresponds to a value 1, whereas not recommending corresponds to a value of 0. The (weighted) average satisfaction scores are the basis of the

subsequent factor analysis. Factor analysis reveals underlying components which correlate with the customer satisfaction survey items.

(29)

29 The next step in order to determine the cost of customer satisfaction, is the identification of divisions creating customer satisfaction. The support given by these divisions is linked to the satisfaction scores. The divisions logistics and transport, supply chain planning, commercial sales, commercial services, customer technical services, and production units are the ones contributing to customer satisfaction. Support of commercial sales, commercial services, and customer technical services, as well as production costs are used to analyse the effect on satisfaction. This study does not analyse total customer satisfaction costs, because costs of every step in the production process are not available. Only above described costs with respect to support, transportation, and production are available.

The following step is to identify costs made by these divisions per customer in order to satisfy the customer. With respect to commercial sales, commercial services, and customer technical services, employees are asked to estimate their time spent to specific customers, assumed that they spend 100% of their worktime to customers. Production and transportation costs per customer are known. Moreover, customer specific ROTIF (ready on time in full) and DOTIF (dispatched on time in full) scores are available, as well as the number of complaints and following recovery to these complaints.

With this data, the link between the above described variables and customer satisfaction is analysed, by calculating correlations between these variables and satisfaction scores and comparing means using these variables as grouping variables. Linear regression is not used to analyse the effects on satisfaction, because using divisions’ support and ROTIF/DOTIF scores as independent variables results in multicollinearity. However, linear regression is used to determine relative importance of satisfaction components, using overall satisfaction as dependent variable as

Homburg and Rudolph (2001) do. With customer-specific prices and margins of eight different steel products, the link of satisfaction and its drivers with price and margin is analysed. Price and margin of the following eight products are available: slabs of steel, hot rolled pickled and oiled steel (p/o), hot rolled dry steel, DSP p/o steel, DSP dry steel, full hard steel, cold rolled steel, and galvanised steel. By calculating

correlations and linear regression, the effect of customer satisfaction on the price and margin per ton steel is analysed. The following chapter describes results of the applied factor analyses, coefficient alphas, correlations, comparison of means, and linear regression.

(30)

6 Results

This chapter describes the results of the analysis of Corus’ customer

satisfaction survey and the effects of variables on satisfaction. Paragraph 6.1 describes results of factor analysis on the satisfaction scores, analysis of the effect of

satisfaction drivers, and determination of the relative importance of satisfaction drivers. Paragraph 6.2 describes results of analysing the link of satisfaction and its drivers with price and margin of eight steel products, resulting in the revenues of satisfaction. Paragraph 6.3 describes effects of price and margin as a whole, whereas paragraph 6.4 describes results of satisfaction scores on recommending Corus as supplier.

6.1 Satisfaction drivers

This paragraph uses factor analysis to identify underlying components of the satisfaction survey with initial components product, delivery, technical support and complaints, sales, and continuous improvement. The sixth survey component surplus value is transformed to a measure of overall satisfaction. Factor analysis is applied on the survey items, to check the reliability of Corus’ survey components. The design of Corus’ satisfaction survey is compared with the INDSAT model of Homburg and Rudolph (2001). Sub-paragraph 6.1.1 identifies satisfaction components using factor analyses, whereas sub-paragraph 6.1.2 and 6.1.3 analyse the effect of variables as support and reliability scores on satisfaction. Sub-paragraph 6.1.4 analyses

satisfaction drivers using linear regression of the satisfaction components on overall satisfaction. Support results to have a negative effect on customers satisfaction, whereas delivery reliability scores result to have a negative effect on overall

satisfaction. Satisfaction components product quality, delivery, and sales result to be the most important drivers of satisfaction, implicating a high degree of rationality in the steel market.

6.1.1 Satisfaction components

This sub-paragraph describes the factor analysis of the customer satisfaction survey scores. First, Cronbach’s alpha is calculated for the six survey components

(31)

31 product, delivery, technical support and complaints, sales, continuous improvement, and overall satisfaction. Subsequently, factor analysis is applied to check if different components reflect different underlying factors, and to see if all items really belong to the component they suppose to reflect. These analyses result in a mixed model: delivery and sales are merged, whereas the components product and technical support and complaints are divided. These results correspond to the INDSAT model, which merges delivery and sales into the dimension order handling.

Cronbach’s alpha measures inter-item correlation, with the number of items taken into account. Cronbach’s alpha is a score between 0 and 1, with higher scores indicating higher levels of reliability. A common used rule of thumb is that the score should be at least above 0.5 to reflect reliability, whereas a score above 0.7 is

preferable. Table 4 reports Cronbach’s alpha for the six satisfaction survey components, using normal and weighted scores. With respect to weighted scores, weights of 1 to 0.7 are used. Using weights of 4 to 1 does not result in any different results. Because of the similarity of normal and weighted scores, normal scores are used further. Component Number of items Cronbach’s alpha (normal scores) Cronbach’s alpha (weighted scores) Product 9 0.823 0.823 Delivery 4 0.573 0.581 Technical support and complaints 6 0.764 0.762 Sales 6 0.777 0.781 Continuous improvement 5 0.722 0.677 Overall satisfaction 4 0.755 0.748

Table 4: Cronbach’s alphas of initial five satisfaction components and overall satisfaction.

The Cronbach’s alpha of the components indicates reliability: only delivery is below 0.7 but still above 0.5, resulting in the conclusion that all components have inter-item correlation. However, the scores of Homburg and Rudolph (2001) are much higher (all above 0.84). Because all Cronbach’s alphas are above 0.5, factor analysis is applied to the survey scores of the components product, delivery, technical support and complaints, sales, and continuous improvement. Overall satisfaction consists of four items: commitment shown by Corus, reliability of Corus, flexibility of Corus, and figuration in future purchasing plans. Because these items are an overall measure of several satisfaction items, factor analysis is not applied to the component overall satisfaction. The overall satisfaction items are further used as separate items.

(32)

The rotated component matrix (only showing loadings above 0.6) of this calculation is reported in Table 19 of the appendix. Factor analysis shows that three underlying components explain 65 percent of the variance in these items (Table 20 of the appendix). The other four components only reflect one item, so these are not taken into account. The first component reflects the three items of complaints and the five items of continuous improvement. The second component reflects three out of four delivery items, the item product range, and four important sales items (flexibility, speed of order booking, attainability, and service). The third component reflects two out of three technical support items and four out of nine product items: three out of four product quality items and the most important item of packaging and labeling, namely the amount of protection of packaging. When these results are compared to the INDSAT model, the following conclusions arise: (1) the survey components delivery and sales belong to the same underlying components, corresponding to the INDSAT model, where delivery times and order processing and confirmation belong to the dimension order handling, (2) the survey component technical service and complaints can be divided into two separate components reflecting technical service and complaints, as in the INDSAT model, (3) product items belong to the same underlying component as technical service, which is not the case in the INDSAT model, (4) the survey component complaints belongs to the same underlying component as the survey component continuous improvement, but because the INDSAT model has no continuous improvement items, these results cannot be compared.

Following the methodology of Homburg and Rudolph, items with low

loadings on all components or high loadings on more than one component are deleted, resulting in deleting the seventh item of product, reflecting safety of unpacking

products, and the first and fourth item of sales, reflecting clarity of order related documents and user friendliness of YMonline (customer information system). Results of factor analysis without these three items are reported in Tables 21 and 22 of the appendix. The result of this factor analysis is more elegant: four underlying

components reflect the items and explain about 69 percent of the items variance. The first component reflects again product range, delivery, and sales. The second

component reflects product quality and technical service. The third component reflects continuous improvement, whereas the fourth component reflects complaints. Product packaging and labeling, labeled by product_6 to product_9, does not

correspond to an underlying component, but to more than one component. Therefore, the initial component product is divided into product quality and product packaging and labeling. Comparing to the INDSAT model, product and technical service could

Drivers, costs, and revenues of customer satisfaction in a business-to-business context

Drivers, costs, and revenues of customer satisfaction

in a business-to-business context

- Master thesis -

By

Mathijs Kostermans (0300012)

March 6th 2009

Master Business Economics

Organisation Economics

Faculteit Economie en Bedrijfskunde

Universiteit van Amsterdam

Supervisors:

Dr. ir. S.P. van Triest

Drivers, costs, and revenues of customer satisfaction

in a business-to-business context

- Master thesis -

By

Mathijs Kostermans

March 6th 2009

Master Business Economics

Universiteit van Amsterdam, the Netherlands

Abstract

Table of contents

1 Introduction

2 Research formulation

3 Literature review

4 Corus

5 Data and methodology

6 Results