Creation of customer experience: comparison between Dutch
and Turkish customers
--- ‘‘How do price, assortment, convenience, and supermarket environment influence the customer
experience and subsequently word-of-mouth when comparing Dutch customers with Turkish customers?’’
Zehra Burçin Kurt Master Thesis Marketing
10-01-2016
Abstract
This research has investigated the influences of price, assortment, convenience and supermarket environment on customer satisfaction with regard to a particular supermarket. Besides this, an important point is that two countries - the Netherlands and Turkey have been compared with each other. The main focus is on the culture and disposable income differences between Dutch and Turkish customers. The results have shown that price, assortment, physical aspects and personal interaction have a positive significant influence on customer satisfaction. Moreover, the research proves that Dutch customers value price more than Turkish customers. Furthermore, the results have indicated that the control variable age has an influence on customer satisfaction.
In addition, the influence of customer satisfaction on word-of-mouth has been analyzed. The results indicate that customer satisfaction has a positive significant influence on word-of-mouth. Accordingly, the more a customer is satisfied with a particular supermarket, the more likely is he/she to
recommend this supermarket to others. Besides, the results designate that Turkish customers are more likely to recommend a supermarket when they are satisfied compared to Dutch customers.
Keywords: customer experience, customer satisfaction, price, assortment, convenience,
Preface
After a long period of hard work, finally I have finished my master thesis. Despite the fact that it was difficult, I really have enjoyed writing my research paper. There are several persons who have contributed to my master thesis and to whom I would like to express my gratitude.
First of all, I would like to thank my supervisor prof. dr. P.C. Verhoef for all his valuable feedback, support and assistance during the whole process of Master thesis preparation. I have appreciated being part of his thesis group and I have enjoyed it very much.
Furthermore, I would like to thank my mother and father. Without their enormous support and help, I would not have been able to finish my master thesis!
Besides, many thanks to Serpil for helping and providing me with feedback regarding the analysis of the collected data. Also, a big thank you to Maria for checking my master thesis and also for helping me to find participants for my research.
Last but not least, many thanks to all my family members and friends who have supported me all the time, and also to the people who have participated in my research!
Zehra Burçin Kurt
Table of contents
Abstract……… 1
Preface………. 2
1. Introduction……….. ... 4
2. Theoretical framework………... 8
2.1 Customer experience and customer satisfaction……… ... 8
2.2 Influence of customer satisfaction on WoM………... 8
2.3 Antecedents………..………... 9
2.3.1 Price……….. 9
2.3.2 Assortment………... 10
2.3.3 Convenience………... 11
2.3.4 Supermarket environment………... 11
2.4 Culture and economic situation……….. 12
2.4.1 Culture………. 12
2.4.2 Economic situation: disposable income………. 14
2.4.3 Comparison Dutch and Turkish customers………... 14
2.5 Conceptual model…….………. 16 3. Methodology……….. 18 3.1 Research design………... 18 3.2 Target population………. ……. 18 3.3 Analysis plan……….. 19 4. Results……….. ……. 20 4.1 Sample characteristics……….. 20
4.1.1 Sample characteristics: Dutch participants………... 20
4.1.2 Sample characteristics: Turkish participants………. 21
4.2 Factor analysis: antecedents customer satisfaction………... 23
4.2.1 Factor analysis: customer satisfaction and WoM……….. 25
4.3 Reliability analysis……….... ……. 26
4.3.1 Analysis of the new sum variables……….. 28
4.4 Testing the hypotheses………... 29
4.4.1 Influence of customer satisfaction on WoM………... 29
4.4.2 Influence antecedents on customer satisfaction………... 30
4.4.3 Influence antecedents on customer satisfaction: comparing both groups……….. 32
4.4.4 Moderation analysis: Dutch and Turkish customers…... 33
4.4.5 Moderation analysis including the control variables…... 35
4.4.6 Overview hypotheses………... 37
5. Conclusion……… 38
6. Limitations………. 41
7. Future research………... 43
8. References………... 44
9. Appendix 1: Comparison the Netherlands and Turkey………... 50
10. Appendix 2: Questionnaire for Dutch customers………... 51
11. Appendix 3: Questionnaire for Turkish customers……….. 57
1. Introduction
Nowadays it is extremely important for retailers to create and maintain an excellent customer
experience. Due to the improved technology and increase in the information transparency, customers are more aware of the available different products and services offered by retailers. Customers have the opportunity to exchange their ideas and opinions about the various products and services, which leads to an overall evaluation and a more detailed comparison of the offerings. Consequently,
customers may decide to switch to other more attractive retailers. Therefore, to remain competitive and to retain current and attract new customers, retailers should distinguish themselves from their competitors through continuously adapting their products and/or services.
Based on a study of Verhoef et al. (2009), the customer experience is a holistic concept that is
defined as the reaction of a customer to the retailer in terms of ‘‘cognitive, affective, emotional, social and physical responses’’. Customer experience is a broad concept which can be measured by
several outcomes such as customer loyalty, customer satisfaction, brand equity, and Net Promoter Score (Chahal and Dutta, 2014; De Haan et al., 2015). Due to the broadness of this concept, the focus of this research will be on the outcome customer satisfaction.
Customer experience has several antecedents. According to Verhoef et al. (2009) some of these antecedents are within the control of the retailer, for example the provided price of a product or service. However, they also refer to factors which are uncontrollable for the retailer such as the influence of the social environment of customers (Verhoef et al., 2009). In this research the focus will be on the antecedents price, assortment, convenience and store/supermarket environment. These antecedents have been chosen as these factors are crucial for the customers’ choice of a specific supermarket, and their satisfaction of relevant needs and requirements.
A (perceived) high or low price influences the satisfaction of the customer with the relevant product or service (Gocek and Beceren, 2012). A high price refers to a good quality of the
Convenience is also an antecedent of customer satisfaction. It has a crucial influence on
customer satisfaction and the study of Hunneman et al. (2015) proves a positive relationship between convenience and satisfaction.
The last antecedent in this research is supermarket environment. According to Baker et al. (2002) store or supermarket environment can be based on three important cues, namely ambient, design and social factors. In the current research the focus lies on the social and the design factor, whereby the social factor corresponds to the service employees in the supermarket, and the
interaction with them. Zomerdijk and Voss (2010) argue that the available service personnel is the key to satisfy the customer. This is based on the fact that personal interaction with the employees has a direct significant influence on customer satisfaction (Beneke et al., 2012). The design factor of the supermarket environment has also an explicit influence on the customers. Taking into account the design factor, Huang (2009) and Beneke et al. (2012) refer to the physical aspects in the supermarket environment which are indicated as the appearance and the layout of the supermarket. Beneke et al. (2012) have found a significant influence of the physical aspects in the supermarket on customer satisfaction.
Furthermore, customer satisfaction has a crucial influence on word of mouth (WoM) (Confente and Russo, 2015). WoM is one of the options to share experiences with other customers. Negative WoM of the corresponding retailer and products may also negatively influence customer satisfaction of other customers.
This research focuses on the comparison between Dutch and Turkish customers regarding their experience and satisfaction with a particular supermarket. The expectation is that this research will show considerable differences in the customer satisfaction between these two groups due to the cross-cultural and economic situation differences between the Netherlands and Turkey. According to the data of the World Economic Situation and Prospects (WESP) (2014), the Netherlands belongs to the developed countries and Turkey to the developing countries. Accordingly, The World Bank (2015) indicates that the Netherlands belongs to the high-income economies, and Turkey to the middle-income economies. The disposable middle-income of the customers plays an important role in the
purchasing process. For instance, customers will be more satisfied if they have the opportunity and the money to buy unique products that better meet their needs and requirements.
in more detail, namely individualism versus collectivism, masculinity versus femininity, power
distance, uncertainty avoidance, and long-term versus short-term orientation (De Mooij and Hofstede, 2010). Additionally, the culture of customers might influence their shopping behavior and
subsequently the customer satisfaction. For example, in a collectivistic culture, customers will be more likely to share their opinions about products or services with each other. This is in line with the social influence theory, which states that the actions or behavior of other customers lead to a
conformity by other customers (Shrihari and Srinivasan, 2012).
Several countries have been researched by means of the earlier mentioned cultural dimensions. A further examination of these dimensions have provided remarkable cultural differences between the Netherlands and Turkey (Hofstede, 2015), which are available in Appendix 1, figure 1 ‘Comparison Netherlands & Turkey’. An exploration of these differences shows explicitly that the Netherlands compared to Turkey has a more individualistic culture, is less stronger in uncertainty avoidance, scores explicitly lower on masculinity and power distance, and is a more long-term oriented culture (Hofstede, 2015). Based on these results, the top three differences are respectively in individualism versus collectivism, uncertainty avoidance, and masculinity versus femininity.
Due to a lack of extensive research, the aim of this research paper will be to examine whether there is a difference in the customer experience between Dutch and Turkish customers, and if there is a difference, could it be explained by the cross cultural and economic situation differences between the Netherlands and Turkey. This results in the following problem statement:
‘‘How do price, assortment, convenience, and supermarket environment influence the customer
experience and subsequently word-of-mouth when comparing Dutch customers with Turkish customers?’’
To provide an accurate answer, the problem statement is divided into the following sub-research questions:
1. How does price influence the customer experience when comparing Dutch customers with Turkish customers?
2. How does assortment influence the customer experience when comparing Dutch customers with Turkish customers?
4. How does supermarket environment influence the customer experience when comparing Dutch customers with Turkish customers?
5. How does customer experience influence word-of-mouth when comparing Dutch customers with Turkish customers?
Based on the answer of the problem statement, the practical contribution of this research will provide a further look for Dutch retailers in case of establishing businesses in Turkey. Besides, this research will give Dutch retailers practical insights for attracting Turkish customers. Furthermore, this study will contribute with theoretical insights for further research of the comparison between Dutch and Turkish customers.
2. Theoretical framework
2.1 Customer experience and customer satisfaction
In the retail world customer experience can be seen as a crucial predictor of the behavior of customers. As stated earlier, customer experience is a very broad concept. There are several definitions explaining customer experience. Verhoef et al. (2009) refer customer experience to ‘‘the cognitive, affective, emotional and physical responses of a customer to the retailer’’. They explicitly indicate that the experience of a customer is combined with experiences established over time (Verhoef et al., 2009). In other words, customers may have several touch points with the retailer during the search for a specific product before possible purchases (Verhoef et al., 2009). Additionally, current retailers have the excellent opportunity to use an omni-channel retailing strategy for creating improved and superior interactions with customers. With omni-channel retailing a retailer can create an integrated sales experience, whereby there is an existence of both physical and digital benefits of shopping for customers (Rigby, 2011). Furthermore, Zomerdijk and Voss (2010) explain that
customer experience is the core of experience-centric service organizations. In such organizations establishing customer engagement is very important. Customer engagement occurs if a retailer has the capacity to create value for the firm and the customer through establishing an excellent customer experience which leads to emotional responses of the customer to the retailer (Sorescu et al., 2011).
As already mentioned, the focus in this research will be on customer satisfaction as an important outcome of customer experience (Chahal and Dutta, 2015; De Haan et al., 2015). Customer
satisfaction refers to ‘‘the pleasurable outcome of the fulfillment of the needs, goals, and desires of a customer’’ (Oliver, 1997; Oliver, 1999). Subsequently, due to the influence of customer satisfaction on the purchase intentions of customers, satisfied customers are more likely to buy a product compared to dissatisfied customers (Kumar, 2002). This is also in line with the fact that dissatisfaction has a negative impact on the retention of customers (Mittal et al., 1998), and with the finding that customer satisfaction has a positive significant influence on customer loyalty (Sutanto and Minantyo, 2014; Flint et al., 2011). Furthermore, customers make an overall judgement based on the consumption of
products, and subsequently the current customer satisfaction has also an influence on the future expectations of the customers (Verhoef et al., 2009).
2.2 Influence of customer satisfaction on WoM
more online information about products and services available, and customers have quicker and effortless access to this information. Providing opinions about specific products and services is easily done by online reviews. Furthermore, recommendation of the products through WoM is also a
possibility to exchange opinions and evaluations of products.
Customer satisfaction is an important driver of WoM. WoM is indicated as the customers’
communication with each other in which evaluation and recommendation about products and services take place (Anderson, 1998; Zhang et al., 2014; Su et al., 2015). According to Verhoef et al. (2009), the experience of a customer with a product or service has an influence on the experience of others. Consequently, a positive or negative recommendation of a product influences the purchase intentions of other customers. The findings of Confente and Russo (2015) indicate a positive significant
influence of customer satisfaction on WoM. Based on this finding the first hypothesis in this research has been outlined.
H1: Customer satisfaction has a positive significant relationship with word-of-mouth.
2.3 Antecedents
One important aim of this research is to examine the overall satisfaction of customers regarding their experience with any supermarket. Factors that have an explicit influence on customer satisfaction are price, assortment, convenience, and supermarket environment.
2.3.1 Price
According to Zeithaml (1988), price is defined as ‘‘what is given up or sacrificed to obtain a product’’. Looking from the perspective of a customer the price of a product is relevant to the positioning of a retailer. A high price corresponds with a high quality product (Verma and Gupta, 2004). Zeithaml (1988) indicates also that a perceived price is encoded by the customer, as a price could be noticed as expensive or cheap. Furthermore, a customer who is satisfied with a relevant product or service is inclined to pay a higher price (Consuegra, 2007). Besides this, the study of Consuegra (2007)
indicates that the perceived price fairness has a positive influence on customer satisfaction.
the findings of Varki and Colgate (2001) is that comparative price perceptions directly influence customer satisfaction.
Another important point is that price can be seen as the perceived value by customers regarding the received quality per valuta (Morgeson et al., 2015). In other words, a higher price of the product refers to a higher perceived value of price. Besides this, Morgeson et al. (2015) have examined developed and less developed countries regarding the differences in the customer satisfaction derived from the perceived value. They found a significant strong effect of perceived value on the customer
satisfaction comparing Turkey and the developed country United Kingdom (Morgeson et al., 2015).
In addition, based on the findings of the studies of Hunneman et al. (2015), Al-Ali et al. (2015), and Akbar (2014), a higher perceived value or a higher price has a positive influence on the satisfaction of customers. The findings of the study of Dapkevičius and Melnikas (2009) is in line with this finding. They found a positive influence of price on customer satisfaction (Dapkevičius and Melnikas, 2009). To further test this relationship, the following hypothesis is been set up:
H2: Price has a positive significant influence on customer satisfaction.
2.3.2 Assortment
Currently, due to the rapid and continuous upcoming of new trends and the frequent changing customers’ needs and requirements, customers pay considerable attention to new and different products offered by retailers. Furthermore, customers become more and more selective and are not satisfied with every new or different product. Therefore, retailers should be able to provide an
attractive assortment for customers. Assortment refers to the number of items offered by a retailer (Dhar et al., 2001; Ailawadi and Keller, 2004). The assortment can be divided into cross category assortment and within category assortment, which is been defined as respectively the breadth and the depth of an assortment (Ailawadi and Keller, 2004).
depth assortment lead to the attraction of more customers, due to meeting customers’ needs and requirements with a large variety of products (Dhar et al., 2001). However, due to a choice overload, more assortment is not always attractive for customers. This is congruent with the findings of the study of Iyengar and Lepper (2000), which emphasize that a large assortment lowers both purchase intention and choice satisfaction of the customers.
Additionally, offering a large assortment is beneficial for customers for an important economic reason. Due to a variety of different products, customers have the opportunity to purchase more products in less trips to the retailer which results in transport costs savings (Akbar, 2014; Leszczyc et al., 2004). Furthermore, Akbar (2014) refers to the positive significant influence of assortment on the satisfaction of a customer. Taken into account this important finding, the following hypothesis of the influence of assortment on customer satisfaction is been set up.
H3: Assortment has a positive significant influence on customer satisfaction.
2.3.3 Convenience
In this research convenience is the next antecedent of customer satisfaction. It has been
conceptualized as the perceived overall costs and easiness for customers by going to a supermarket (Hunneman et al., 2015; Berry et al., 2002). In addition, Morganosky (1986) indicates that a customer who is convenience-oriented tries to reduce the required human energy and time to a minimum level. This is line with the study of Hunneman et al. (2015), whereby they indicate that if the required human energy and time reduces to visit a supermarket, this subsequently will lead to both higher value
perceptions and customer satisfaction. Thus, they argue the existence of a positive relationship between convenience and satisfaction (Hunneman et al., 2015). This leads to the fourth hypothesis in this research.
H4: Convenience has a positive significant influence on customer satisfaction.
2.3.4 Supermarket environment
The last antecedent of customer satisfaction is the supermarket environment, which is divided into physical aspects and personal interaction.
customers. Besides this, high perceptions of the layout or design of the supermarket illustrates high expected prices of the offered products or services (Baker et al., 2002). This is also in line with the findings of Ailawadi and Keller (2004), which designate that a pleasant atmosphere in the
supermarket leads to the creation of a unique image, and also to higher perceived quality of the offered products. In addition to this, Al-Ali et al. (2015) refer to the positive significant relationship between the perceptions of the quality image of a store or supermarket and the customer satisfaction. As stated earlier, the study of Beneke et al. (2012) have shown the significant influence of the
physical aspects in the supermarket on the customer satisfaction. A replication of this leads to the following hypothesis:
H5: Physical aspects have a positive significant influence on customer satisfaction.
Personal interaction is the other important factor in the supermarket environment. Bitner et al. (2008) suggest that the personal interaction with the employees occurs at the front stage of the supermarket, which is indicated as the line of visibility for the customers. Personal interaction is being defined as the visible actions of employees, with regard to helping customers and providing appropriate answers for their questions (Bitner et. al., 2008; Huang, 2009). Furthermore, personal interaction explicitly indicates the service quality of a supermarket, and therefore good personal interactions of employees lead to a higher perceived service quality (Huang, 2009). Beneke et al. (2012) found also a positive significant relationship between personal interaction and customer satisfaction. Therefore, to examine this relationship also in this research, the next hypothesis has been set up:
H6: Personal interaction has a positive significant influence on customer satisfaction.
2.4 Culture and economic situation
Within this study, one of the aims is to investigate the differences in the customer satisfaction
comparing Dutch customers with Turkish customers. As stated earlier, the expected differences in the customer satisfaction are based on cross cultural and economic situation differences between both groups.
2.4.1 Culture
Within a country, culture refers to the shared common values and beliefs (Peterson and Smith, 1997; Triandis, 1993). As stated before, culture can be explained with the five cultural dimensions.
is an interdependence between people (Hui and Triandis, 1986; Triandis,1993). Furthermore,
Jarvenpaa and Tractinsky (1999) refer to the fact that customers from an individualistic culture make less effort for information seeking about products or services. This might be a consequence of a stronger self interest of individualistic customers, which leads to more focus on their own ideas. Furthermore, Donthu and Yoo (1998) indicate that customers from an individualistic culture have stronger service quality expectations compared with customers from a collectivistic culture.
The second cultural dimension, is uncertainty avoidance. According to De Mooij and Hofstede (2010), uncertainty avoidance is indicated as a persons’ attempts to prevent specific circumstances which lead to the feeling of uncertainty. As mentioned before, Dutch customers are indicated as weaker in uncertainty avoidance than Turkish customers (Hofstede, 2015). Based on this, one can argue that customers from a stronger uncertainty avoidance culture take less risks or avoid risky situations. This is also in line with the findings of Hwang et al. (2006), which indicate that Turkish customers are sensitive regarding the security of information, and they do not trust online shopping.
Masculinity versus femininity, is also a cultural dimension that makes a remarkable differentiation in cultures. Masculinity compared to femininity implies the existence of a large differentiation between the roles that men and women have (De Mooij and Hofstede, 2010). Furthermore, De Mooij and Hofstede (2010) state that feminine cultures have a focus on the quality of life. In addition to this, a culture strong in femininity pays attention on an equal treatment of all customers (Wursten et al., 2009).
The fourth cultural dimension is power distance. Power distance is defined as the acceptance of the unequal distribution of power (De Mooij and Hofstede, 2010). Subsequently, a culture low in power distance is indicated as low in accepting inequality and whereby power is fairly shared among people (Wursten et al., 2009).
Long term versus short term orientation is the last cultural dimension. It is defined as the perspective or view of a society in terms of long term (future) or short term orientation (De Mooij and Hofstede, 2010; Yoo et al., 2011). Furthermore, there are several main work values of both views. According to Hofstede and Minkov (2010, p. 497) ‘‘learning, honesty, self-discipline’’, and ‘‘freedom, rights,
2.4.2 Economic situation: disposable income
Another important variable in this research is the economic situation of a customer, with regard to the disposable income of a customer. The disposable income is the income after the deduction of taxes, thus it is the income that customers have on hand (Economic Development Winnipeg, 2014). A lower disposable income leads to less focus on quality and more focus on the price of a product (Morgeson et al., 2015). Furthermore, Topcu and Uzundumlu (2009) argue that customers with a high income and thus a high disposable income pay more attention on product satisfaction. Besides, a crucial point is that a low disposable income leads to more reluctance of spending money (Hunneman et al., 2015).
In the following paragraph, Dutch and Turkish customers will be compared with regard to the culture and economic situation differences. Also the relevant hypotheses will be drawn up.
2.4.3 Comparison Dutch and Turkish customers
As mentioned before, the finding of Morgeson et al. (2015) refers to a strong influence of perceived value on customer satisfaction in less developed countries compared to developed countries.
Literature indicates also that customers with a lower disposable income, focus less on the quality of a product and focus more on the price of it (Morgeson et al., 2015). This is also congruent with the fact that Turkish customers prefer a cheaper price (Hwang et al., 2006). Consequently, based on the fact that Turkey is a less developed country than the Netherlands, and on the fact that in general Turkish customers have a lower disposable income compared to Dutch customers, one can argue that Turkish customers will focus less on the quality of products and more on the price of products. Therefore, the following hypothesis is been formulated.
H7: Price has a stronger significant relationship with customer satisfaction for Turkish customers than for Dutch customers.
This also has to do with the fact that the Turkish culture is more collectivistic and strong in uncertainty avoidance compared to the Dutch culture. As Turkish customers are from a collectivistic culture, they have the priority to relate their choices for products to the opinions of other customers. Additionally, the strong uncertainty avoidance of Turkish customers leads to more avoidance of risky situations and risky information. Therefore, a large assortment enables and inclines Turkish customers to seek more information about a range of products and compare the opinions of other customers, and find a product which is more economical for them. Subsequently, this results in the following hypothesis:
H8: Assortment has a stronger significant relationship with customer satisfaction for Turkish customers than for Dutch customers.
Furthermore, the study of Hunneman et al. (2015) indicate that the ability to buy differs among customers, explaining that a higher income and thus also a higher disposable income provides a higher ability to buy products. Besides, the authors argue that the willingness to buy depends on the future expectations of the customers (Hunneman et al., 2015). Thus, if a customer has a low
disposable income then he/she is more reluctant to spend money, and tries to achieve more savings (Hunneman et al., 2015). Also, the findings of Morgeson et al. (2015) designate that developed countries focus more on quality. The fact that Dutch customers have in general a higher disposable income compared to Turkish customers shows that the quality of the products for Dutch customers is more important than the price of these products. In other words, Dutch customers aim more for value, and less for saving money. Subsequently, the convenience to visit a supermarket might play a less important role for customers with a higher income. Thus, one can argue that the ability to visit a supermarket and the travel costs have less importance for Dutch customers compared to Turkish customers. This leads to the following hypothesis:
H9: Convenience has a stronger significant relationship with customer satisfaction for Turkish customers than for Dutch customers.
Additionally, Topcu and Uzundumlu (2009) argue that customers with a high income and thus a high disposable income pay more attention to product satisfaction, and also to the image of the store or supermarket. One important reason they indicate is that people with a high income are more likely to show their social status (Topcu and Uzundumlu, 2009). Subsequently, based on this crucial point the expectation in this research is that the physical aspects, as the appearance and layout of the
H10: Physical aspects have a stronger significant relationship with customer satisfaction for Dutch customers than for Turkish customers.
Lastly, according to the study of Donthu and Yoo (1998), higher expectations of the service quality is being linked with customers who score low on power distance. As referred before, Donthu and Yoo (1998) found also that customers from an individualistic culture have stronger expectations of the service quality. Also as mentioned above, feminine cultures focus more on an equal treatment of all customers (Wursten et al., 2009), and thus also have more expectations about the service quality. Subsequently, one can argue that customers from an individualistic and feminine culture focus more on the quality of the service, thus on the personal interaction with an employee. Taken into account these points, the last hypothesis in this research is been formulated.
H11: Personal interaction has a stronger significant relationship with customer satisfaction for Dutch customers than for Turkish customers.
2.5 Conceptual model
Based on the provided literature and the consulted hypotheses, the relationship between the relevant constructs is presented in figure 2. This conceptual model is corrected with several control variables. These control variables could also have a possible influence, and subsequently they might influence the results of this research.
Gender is the first control variable. The findings of the study of Topcu and Uzundumlu (2009) refer to that females make more visits to a store or supermarket and also focus more on the service quality compared to men. Therefore, a possible influence of gender on customer satisfaction could be expected.
A possible influence of age on customer satisfaction is based on the expectation that young
customers have more energy to regularly visit a supermarket compared to customers of an age of 60 and older. Therefore, young customers could have more positive and negative experiences with a particular supermarket. This expectation corresponds also with the influence of the shopping frequency on customer satisfaction.
Taking into account another finding of Topcu and Uzundumlu (2009) leads to the expectation
Antecedents customer experience - price - assortment - convenience - supermarket environment Customer experience customer satisfaction
Nationality
Dutch vs Turkish customers Control variables - gender - age - education - shopping frequency - household sizeWoM
with a high income focus more on the store layout and also on their social status (Topcu and Uzundumlu, 2009).
The last control variable in this research is the household size. The expectation is that customers with a greater household size might need to focus more on saving money, and therefore the price of the offered products could be more important than the quality of the products.
3. Methodology
3.1 Research designIn the previous chapter the independent and dependent variables have been defined. Also the
existing moderators have been indicated. Several hypotheses have been established to research the relationships with these variables.
To examine these relationships, a quantitative research in the form of an online questionnaire has been conducted. To make this online questionnaire understandable for both Dutch and Turkish customers, a questionnaire in English and in Turkish has been made, which can be found respectively in Appendix 2 and 3.
The online questionnaire consists of several questions. It starts with two screening questions. These are about the nationality of the customers and the frequency of visiting a supermarket. After these screening questions, some general questions follow about the demographic characteristics of the customers. Then the online questionnaire continues with specific questions regarding the constructs in this research. These questions are on a 7-point Likert scale, whereby 1 refers to ‘totally
disagree/very bad’ and 7 for ‘totally agree/excellent’. Also, the answer scale of these questions indicates the anchoring point of 4. The several scale items used to measure the constructs in this research are available in Appendix 4. The specific questions are related to a supermarket that the customers most frequently visit. Well-known supermarkets in the Netherlands are Albert Heijn, Jumbo, Aldi, and Plus. These four supermarkets are included in the questionnaire for the Dutch
customers. However, in Turkey these supermarkets do not exist. Therefore, well-known supermarkets in Turkey have also been chosen for the Turkish questionnaire: Carrefour SA, Kiler, Bim, and A101.
Furthermore, the data has been collected by use of a convenience sampling. A major benefit of a convenience sampling is the ability to find quicker and effortless participants for a research
(Pelsmacker and Kenhove, 2006). An appropriate amount of participants is 100 for each group. In addition, based on the provided literature both groups are different from each other, and therefore a between-participants design has been used in this research.
3.2 Target population
their origin domestic country and customers who live in a foreign country. Therefore, to avoid subcultures, the first screening question about the nationality is included in both online
questionnaires.
Furthermore, also a main point in this research is the examination of the customer satisfaction of a customer with regard to a particular supermarket. Therefore, participants should have experiences with a particular supermarket to participate in this research. Based on this, the second screening question about shopping frequency is included in the questionnaire. In this case, if a participant does not visit a supermarket and clicks on the answer scale ‘never’, then the questionnaire finishes for this participant. Subsequently, the data of these participants will be deleted for analyzing.
3.3 Analysis plan
After collecting the data of both groups, the data has been analyzed by using SPSS.
Firstly, the data has been imported in SPSS and then the data has been coded to explicitly
distinguish the specific scale items of the different constructs. Subsequently, the descriptive variables have been analyzed by comparing both groups with each other. In this research, the descriptive variables are the same as the control variables.
A factor analysis was a further step, whereby one item has been deleted. The internal correlation of all the specific scale items have been measured with the Cronbach's alpha of the reliability analysis. Based on the results of the factor analysis and the Cronbach's alpha, the scale items of the
constructs are combined into one variable.
After conducting the factor analysis and the reliability analysis, the specific relationships of the different constructs have been analyzed by using several tests. A regression analysis has been conducted for the first hypothesis regarding the relationship between customer satisfaction and WoM. Also, an independent T-test has been used to analyze possible differences between Dutch and
4. Results
Firstly, in this chapter the characteristics of the Dutch and Turkish customers will be described. Then the factor and reliability analysis of all the constructs in this research will follow. After that, all the defined hypotheses will be examined, and lastly the control variables will be researched.
4.1 Sample characteristics
As stated earlier, in this research there are two groups of participants, Dutch and Turkish customers. For both groups separate data collections have been conducted. In total 227 respondents participated in this research, which is 113 and 114 participants in the Dutch and Turkish group respectively.
4.1.1 Sample characteristics: Dutch participants
In the Dutch group 113 respondents have participated. A dropout rate of 2.0% resulted in 110
participants. The first screening question about the nationality of the participants indicated that there were 6 participants of the 110 who did not have the Dutch nationality. These non-Dutch participants did not continue with this survey after the first screening question. Thus, in total 104 Dutch
participants have participated in this research. However, also from these Dutch participants 4
participants have stopped before completing the whole questionnaire. The incomplete data of these participants have been removed from this research. Therefore, the definitive amount of Dutch participants was 100.
These 100 Dutch participants have been researched with regard to their gender, age, education, shopping frequency and household size. An overview of the results can be seen in table 1 ‘sample characteristics’. Respectively, 39.0% and 61.0% of the participants are males and females. The major part of these respondents is between 18 and 30 years old. 37 Dutch participants of the 100 have graduated from high school, and 47 Dutch participants have graduated from a university.
Furthermore, 23.0% of the Dutch participants share their household with one more person, and almost half of the participants do daily visits to a supermarket.
Figure 3: Dutch customers: most frequently visited supermarket
4.1.2 Sample characteristics: Turkish participants
The second group in this research are the Turkish customers who were 114 participants. Due to a minor dropout rate of 1.0%, this group resulted in 112 participants. Also, for this group there was a screening question about the nationality of the participants. Based on the analysis, 3 participants from these 112 participants did not have the Turkish nationality. Thus, in total 109 Turkish participants have participated in this research. Besides, 4 Turkish participants did not complete the survey, and the data of these 4 participants have been deleted from this research. Therefore, the definitive amount of Turkish participants was 105.
Furthermore, the Turkish customers have been researched based on the variables gender, age, education, shopping frequency and household size. The analysis has shown that 53.3% of the
Turkish participants are females and 46.7% are males. 43.8% of the Turkish participants have an age between the 31 and 40 years. From the 105 Turkish participants 63 have obtained a university
degree, and 29 Turkish participants of the 105 have finished high school. From the Turkish participants 34.3% share their household with 3 persons, and the major part of these Turkish
Table 1: sample characteristics
Variables NTotal %Total NDutch %Dutch NTurkish %Turkish
Gender Male Female 88 117 42.9 57.1 39 61 39.0 61.0 49 56 46.7 53.3 Age
Younger than 18 years Between 18-30 years Between 31-40 years Between 41-50 years Between 51-60 years Between 61-70 years 4 111 49 24 13 4 2.0 54.1 23.9 11.7 6.3 2.0 1 88 3 1 4 3 1.0 88.0 3.0 1.0 4.0 3.0 3 23 46 23 9 1 2.9 21.9 43.8 21.9 8.6 1.0 Education Primary school Secondary/high school College/university degree Master degree PhD or higher 3 66 110 23 3 1.5 32.2 53.7 11.2 1.5 1 37 47 13 2 1.0 37.0 47.0 13.0 2.0 2 29 63 10 1 1.9 27.6 60.0 9.5 1.0 Shopping frequency
Once per three months Once a month Once per two weeks
Once a week Two or three times a week
Daily 3 16 20 37 72 57 1.5 7.8 9.8 18.0 35.1 27.8 0 0 1 6 44 49 0.0 0.0 1.0 6.0 44.0 49.0 3 16 19 31 28 8 2.9 15.2 18.1 29.5 26.7 7.6 Household size No persons 1 person 2 persons 3 persons 4 persons More than 4 persons
Additionally, the two Turkish supermarkets Carrefour SA and Bim share the first place with regard to the most frequently visited supermarket by the Turkish participants. And respectively, A101 and Kiler are at the second and third place of the most frequently visited supermarkets.
Figure 4: Turkish customers: most frequently visited supermarket
4.2 Factor analysis: antecedents customer satisfaction
The antecedents of customer satisfaction in this research have been measured by means of specific statements. A factor analysis has been conducted to examine whether these statements load into a factor. The factor analysis has shown that the Kaiser-Meyer-Olkin Measure of Sampling Adequacy is high, namely 0.833. Also the Bartlett's Test of Sphericity is significant p= 0.000. An overview of the results of the factor analysis can be found in table 2.
Based on the factor analysis, the statements related to each antecedent load into one factor. Every statement has a communality value above 0.4. The statement ‘the overall appearance of this
Table 2: Factor analysis antecedents customer satisfaction
Construct Scale items F1 F2 F3 F4 F5 Comm unality Price - The products offered in this
supermarket are reasonably priced - The products offered in this supermarket offer value for money - The products offered in this supermarket are good products for the price
- The products offered in this supermarket would be economical
0.821 0.838 0.805 0.802 0.667 0.820 0.782 0.708
Assortment - Good supply of fresh products in this supermarket
- Large assortment in this supermarket
0.687 0.816
0.640 0.702
Convenience - This supermarket is nearby - Sufficient supply of other supermarkets close to this supermarket 0.561 0.894 0.410 0.567 Physical aspects
- The overall appearance of this supermarket
- This supermarket has a clean shopping environment and efficient running
- This supermarket has a design that enables me to move around with ease
- This supermarket has a layout that allows me to easily find the products I need 0.423 0.705 0.856 0.748 0.656 0.709 0.803 0.685 Personal interaction
- The staff of this supermarket have enough knowledge to answer my questions
- The staff of this supermarket are friendly and polite
- The staff of this supermarket are always willing to help me
4.2.1 Factor analysis: customer satisfaction and WoM
Again a factor analysis has been conducted for the constructs customer satisfaction and WoM. The Bartlett's Test of Sphericity shows a significant value of p= 0.000. Besides this, also the value of Kaiser-Meyer-Olkin Measure of Sampling Adequacy is high, namely 0.889. An overview of the results of the factor analysis can be found in table 3.
Table 3. Factor analysis: customer satisfaction and WoM
Construct Scale items Factor 1 Factor 2 Communality
Customer satisfaction
- My feelings towards this supermarket is very positive
0.798 0.686
- I feel good about coming to this supermarket for the
offerings I am looking for 0.788 0.687
- Overall I am satisfied with this supermarket and the
service they provide 0.863 0.773
- I feel satisfied that this
supermarket produce the best results that can be
achieved 0.726 0.567
- The extent to which the supermarket has produced
the best possible outcome for me is satisfying. 0.761 0.625
Word-of-Mouth
- Mention to others that you do purchases by this
supermarket 0.701 0.571
- Make sure that others know that you do purchases by
this supermarket 0.795 0.668
- Talk positively about the employee(s) of this
supermarket to others 0.761 0.609
- Recommended the supermarket to family members
0.881 0.807
- Talk positively about the supermarket to others
0.800 0.767
- Recommended the supermarket to acquaintances
0.893 0.829
- Recommended the supermarket to close personal
Based on the factor analysis, it can be concluded that the statements of both constructs load into a factor, and that all the loadings are higher than 0.5. Moreover, the communality values of all
statements are higher than 0.4, which means that no statement should be deleted. The eigen values of both factors are higher than 1, and the total variance explained by the two factors is above 60%, namely 69.655%. Therefore, the factor analysis shows that 2 factors should be extracted.
4.3 Reliability analysis
After the factor analysis, a reliability analysis has been conducted to guarantee that the specific statements of each construct can be summed up into one variable. The results of the reliability
analysis are available in table 4. The analysis is conducted with the combined data of both Dutch and Turkish customers. To get more insights, also the data of both groups have been researched
separately. In table 4 there is an overview of the important values of the reliability analysis.
The reliability analysis indicates that the constructs price, assortment, physical aspects, personal interaction, customer satisfaction, and WoM have a higher Cronbach’s alpha than 0.6, which means a high reliability between the items. To increase the Cronbach’s alpha more, also the values of ‘if items deleted’ have been researched. The ‘if items deleted values’ of the scale items of each construct are below the current Cronbach’s alpha, and therefore no items have been deleted.
However, the construct convenience has a Cronbach’s alpha of 0.581, which is slightly lower than 0.6. There is also a low reliability between the two scale items for the group of Dutch customers, namely a Cronbach’s alpha of 0.231. In contrast, the Turkish group has a Cronbach’s alpha of 0.780. Due to an extremely low Cronbach’s alpha for the Dutch participants, the option is to choose the most appropriate scale item of convenience for both groups, and delete the other scale item. The two scale items of convenience are ‘the supermarket is nearby’ and ‘sufficient supply of other supermarkets close to this supermarket’.
The theory and assumptions are used to choose the best convenient scale item to analyze the construct convenience. As mentioned in the theory part, Turkish customers have in general a lower income compared to Dutch customers. Based on the findings of the study of Hunneman et al. (2015), customers with a low income are more reluctant to spend their money and thus have more focus on savings. Therefore, the Turkish customer will be more likely to save money by making overall
likely to visit a supermarket more often without considering the transport costs. Taking into account these assumptions, and also based on the theory, a conclusion has been made to choose the first scale item ‘the supermarket is nearby’ for further analysis of the construct convenience.
Furthermore, also the correlations between the scale items have been researched. The results have shown that the correlations for each construct are significant, namely p=0.000. Due to significant correlations and a good reliability between the scale items, all the scale items of price, assortment, physical aspects and personal interaction are combined into one sum variable.
Table 4: Results reliability analysis
4.3.1 Analysis of the new sum variables
As stated above, the corresponding scale items of the constructs are combined into one sum
variable. Therefore, a descriptive analysis of the constructs has been conducted to examine whether there are differences between both groups. An overview of the results can be seen in table 5.
As it can be concluded from table 5, the construct with the highest average in this research is
convenience with a mean of 6.00. The construct with the lowest average is WoM with a mean of 4.67. When comparing both groups the results show that the constructs price, assortment, convenience, physical aspects, personal interaction and customer satisfaction have slightly a higher average for the Dutch customers than for the Turkish customers. However, the construct WoM has a higher mean for Turkish customers than for Dutch customers, with respectively a mean of 4.98 and 4.34.
Table 5: Descriptive analysis of all the constructs: means and standard deviations
Construct MeanTotal SDTotal MeanDutch SDDutch MeanTurkish SDTurkish
Price 4.94 1.143 5.03 0.919 4.86 1.321 Assortment 5.40 0.884 5.68 0.740 5.12 0.927 Convenience 6.00 1.107 6.39 0.777 5.63 1.242 Physical aspects 5.32 1.124 5.43 0.903 5.23 1.297 Personal interaction 5.26 1.071 5.31 0.940 5.21 1.186 Customer satisfaction 5.31 0.867 5.36 0.824 5.26 0.908 WoM 4.67 1.244 4.34 1.111 4.98 1.288
An independent T-test has been used to test whether there are significant differences in the
Table 6. results independent T-test
Construct t-statistic p-value
Price 1.068 0.287 Assortment 4.732 0.000 Convenience 5.231 0.000 Physical aspects 1.283 0.201 Personal interaction 0.627 0.532 Customer satisfaction 0.769 0.443 WoM -3.815 0.000
4.4 Testing the hypotheses
In this research several hypotheses have been drawn up. Firstly, the influence of customer satisfaction on WoM will be analyzed. Then follows the examination of the influences of price, assortment, convenience, physical aspects and personal interaction on customer satisfaction.
Subsequently, the analysis of the comparison between Dutch and Turkish customers will take place. Lastly, the control variables in this research will be tested. Due to the fact of a small sample size in this research an alpha of 0.10 has been used for the examination of the significance of each analysis.
4.4.1 Influence of customer satisfaction on WoM
The first step in this research is to test the influence of customer satisfaction on WoM. As mentioned before, a positive significant relationship is expected between the variables. A regression analysis has been conducted to test this relationship. This regression analysis is significant R=0.510,
F(1,203)=71.31, p=0.00. Thus, customer satisfaction has a positive significant relationship with WoM. Therefore, based on the results, the first hypothesis in this research is supported.
Table 7: Regression analysis: influence customer satisfaction on WoM
Construct Beta p-value VIF
Customer satisfaction 0.732 0.000 1
R²=0.260, R² adjusted=0.256
satisfaction has a positive significant influence on WoM for both Dutch customers and Turkish
customers, respectively R=0.505, F(1,98)=33.516, p=0.000, and R=0.572, F(1,103)=50.027, p=0.000. Besides this, WoM is explained by 26% with customer satisfaction for the Dutch customers, and for the Turkish customers it is slightly more, namely 33%. This means that Turkish customers are more likely to recommend a supermarket when they are satisfied compared to Dutch customers.
Table 8: Influence customer satisfaction on WoM: comparing both groups
Construct Beta p-value VIF R2 R2 adjusted
Customer satisfaction, Dutch customers 0.732 0.000 1 0.255 0.247 Customer satisfaction, Turkish customers 0.811 0.000 1 0.327 0.320
4.4.2 Influence antecedents on customer satisfaction
The next step in this research is to test the influence of price, assortment, convenience, physical aspects and personal interaction on customer satisfaction. A multiple regression analysis have researched the influence of all the antecedents on customer satisfaction. Based on the analysis, it can be concluded that the total influence of all the antecedents on the customer satisfaction is
significant R=0.586, F(5,199)= 20.840, p=0.000. In table 9 the results of the analysis of the combined data of both groups can be found. A R² of 0.344 indicates that the total customer satisfaction is explained by 34% with the constructs price, assortment, convenience, physical aspects and personal interaction. Besides, for the examination of the regression analysis, the standardized Beta has been used. This is due to the fact that the scale item of the construct convenience has a different answer scale than the other constructs in this research. Furthermore, there is no multicollinearity because all the VIF values are below 4.
‘physical aspects have a positive significant relationship with customer satisfaction’, and the sixth hypothesis ‘personal interaction has a positive significant relationship with customer satisfaction’ are supported.
Table 9: Multiple Regression: influence antecedents on customer satisfaction
Construct Beta p-value VIF
Price 0.059 0.379 1.363 Assortment 0.309 0.000 1.537 Convenience 0.105 0.119 1.368 Physical aspects 0.197 0.008 1.637 Personal interaction 0.248 0.000 1.378 R²=0.344, R² adjusted=0.327
As stated above, with the multiple regression no significant relationship of price and convenience on customer satisfaction has been found. To test both relationships more accurately, the influence of price and convenience on customer satisfaction has been analyzed separately. The regression analysis of price R=0.292, F(1,203)=18.952, p=0.000 and convenience R=0.166, F(1,203)=5.786, p=0.017 are significant. Thus, price and convenience have also a positive significant relationship with customer satisfaction. The results can be found in table 10.
Table 10: Regression analysis: Influence price on customer satisfaction
Construct Beta p-value VIF R2 R2 adjusted
Price 0.222 0.000 1 0.085 0.081
Convenience 0.166 0.017 1 0.028 0.023
An appropriate reason for the opposite results in the multiple regression analysis could be that the total influence of assortment, physical aspects and personal interaction on customer satisfaction is stronger than the influence of price and convenience on customer satisfaction. As it can be seen in table 10, customer satisfaction is explained only by 9% with price and by 3% with convenience . Therefore, in the total model the significant influence of price and convenience on customer satisfaction may be dropped out.
(p<0.10). However, as expected price (0.292) and convenience (0.166) have a lower significant correlation with customer satisfaction than assortment (0.468), physical aspects (0.435) and personal interaction (0.459). Therefore, it can be concluded that in this research assortment, physical aspects and personal interaction have a stronger relationship with customer satisfaction than price and convenience with customer satisfaction.
Table 11: Correlation analysis: antecedents and customer satisfaction
Construct Correlation with
customer satisfaction p-value Price 0.292 0.000 Assortment 0.468 0.000 Convenience 0.166 0.017 Physical aspects 0.435 0.000 Personal interaction 0.459 0.000
Another possible reason for the opposite results of price and convenience in the multiple regression could be the consequence of the differences between Dutch and Turkish customers. To test this possible reason, the main effects of the antecedents will be analyzed by means of a moderation analysis. This analysis will be explained in paragraph 4.4.4.
4.4.3 Influence antecedents on customer satisfaction: comparing both groups
Furthermore, a multiple regression analysis is conducted for both groups separately. The results of the Dutch group and Turkish group are shown respectively in table 12 and 13. For the Dutch group the regression analysis is significant R=0.758, F(5,94)=25.450, p=0.000. Also the regression analysis is significant for the Turkish group R=0.564, F(5,99)=9.248, p=0.000. For both regression analyses there is no multicollinearity. As it is visible from table 11, the analysis indicates that the antecedents explain for 58% the customer satisfaction of the Dutch customers. In addition, it can be concluded that price, assortment, physical aspects, and personal interaction have a positive significant influence on customer satisfaction for Dutch customers (p<0.10). In contrast, convenience has no significant relationship with customer satisfaction for Dutch customers (p>0.10).
The results of the analysis of the Turkish customers show that the antecedents explain for 32% the total customer satisfaction of the Turkish customers. Besides this, assortment and personal
price, convenience and physical aspects have a non-significant relationship with customer satisfaction for Turkish customers (p>0.10).
In conclusion, these differences in the results of both regression analyses show explicitly that there are differences between Dutch and Turkish customers.
Table 12: Multiple Regression: influence antecedents on customer satisfaction for Dutch group
Construct Beta p-value VIF
Price 0.341 0.000 1.089 Assortment 0.358 0.000 1.283 Convenience 0.071 0.303 1.042 Physical aspects 0.265 0.000 1.193 Personal interaction 0.182 0.016 1.222 R²=0.575, R² adjusted=0.553
Table 13: Multiple Regression: influence antecedents on customer satisfaction for Turkish group
Construct Beta p-value VIF
Price -0.176 0.101 1.635 Assortment 0.256 0.017 1.615 Convenience 0.065 0.542 1.651 Physical aspects 0.132 0.280 2.148 Personal interaction 0.356 0.001 1.554 R²=0.318, R² adjusted=0.284
4.4.4 Moderation analysis: Dutch and Turkish customers
A subsequent step in this research is the analysis of the differences in customer satisfaction for the Dutch and Turkish customers. As mentioned before, to test these differences hypothesis 7 till hypothesis 11 have been drawn up. Analyzing the interaction effect is done by using a regression analysis. A dummy variable is used for the nationality of the customers, whereby the Dutch nationality is the benchmark category and is coded with 0. The Turkish nationality is coded with 1.
model to analyze whether these control variables also have an influence on customer satisfaction. The examination of the control variables is in paragraph 4.4.5.
The moderation analysis has indicated that the full interaction model is significant, R= 0.658,
F(11,193)= 13.385, p=0.000. The important results of the moderation analysis can be found in table 14. The moderation analysis has shown that the customer satisfaction is explained for 43% by the antecedents price, assortment, convenience, physical aspects and personal interaction. Besides this, the dummy variable is positive and significant, which shows that Turkish customers are on average more satisfied than Dutch customers.
In addition, the results have indicated that here are positive significant main effects of the
antecedents price, assortment, physical aspects and personal interaction (p<0.10). Subsequently, the significant main effect of price explicitly indicates that due to differences between both groups, the influence of price on customer satisfaction was non-significant (B=0.059, p=0.379) in the main model in paragraph 4.4.2. Based on this finding, the hypothesis ‘price has a positive significant influence on customer satisfaction’ is supported. In contrast, the analysis has shown that convenience has no significant main effect on customer satisfaction (p>0.10). Therefore, it can be concluded that the hypothesis ‘convenience has a positive significant influence on customer satisfaction’ is not supported.
Furthermore, there is a negative significant interaction effect between price and nationality on customer satisfaction (p=0.000). Given the fact of a positive significant main effect of price in this model, this means that the positive effect of price is smaller for Turkish customers than for Dutch customers. Therefore, Dutch customers value price more than Turkish customers, which means that price has a stronger significant relationship with customer satisfaction for Dutch customers than for Turkish customers. Subsequently, the hypothesis ‘Price has a stronger significant relationship with customer satisfaction for Turkish customers than for Dutch customers’ is not supported.
Moreover, the moderation analysis has shown a non-significant interaction effect between the constructs assortment, convenience and nationality on customer satisfaction (p>0.10). This means that assortment and convenience do not have a stronger significant relationship with customer
Furthermore, there is a non-significant interaction effect between physical aspects, personal
interaction and nationality on customer satisfaction (p>0.10). This means that physical aspects and personal interaction do not have a stronger significant relationship with customer satisfaction for Dutch customers compared to Turkish customers. Therefore, hypothesis 10 and 11 are also not supported.
Table 14: Results moderation analysis
Construct Beta p-value
Constant -0.116 0.885 Dummy 1.451 0.006 Price 0.403 0.000 Assortment 0.406 0.000 Convenience 0.096 0.397 Physical aspects 0.313 0.003 Personal interaction 0.197 0.046 Price x dummy -1.283 0.000 Assortment x dummy -0.452 0.283 Convenience x dummy -0.419 0.273
Physical aspects x dummy -0.477 0.179 Personal interaction x dummy 0.358 0.284
R²=0.433, R² adjusted=0.400
4.4.5 Moderation analysis including the control variables
The last step in this research is analyzing possible influences of the control variables on customer satisfaction. Therefore, the control variables are also included in the full model of the moderation analysis. Subsequently, the dummy variable is the nationality of the customers, whereby the Dutch nationality is the benchmark category and is coded with 0 and the Turkish nationality is coded with 1.
Furthermore, also in this full model there is a negative significant interaction effect between price and nationality on customer satisfaction (p<0.10). In contrast, the other antecedents of customer
satisfaction do not have a significant interaction effect (p>0.10).
Taken into account the control variables in this research show that gender, education, shopping frequency, and household size do not have a significant influence on customer satisfaction (p>0.10). However, according to the results ‘age’ has a significant influence on customer satisfaction (p<0.10). Furthermore, the results have shown that customer satisfaction is only explained by 46%. This means that possible differences in the customer satisfaction might also be a consequence of other variables which are not included in this research.
Table 15: Results moderation analysis: including the control variables
Construct Beta p-value
Constant 1.133 0.235 Gender 0.006 0.922 Age -0.137 0.029 Education -0.058 0.302 Shopping frequency -0.100 0.152 Household size -0.020 0.739 Dummy 1.274 0.019 Price 0.401 0.000 Assortment 0.375 0.001 Convenience 0.129 0.258 Physical aspects 0.316 0.004 Personal interaction 0.197 0.049 Price x dummy -1.311 0.000 Assortment x dummy -0.396 0.360 Convenience x dummy -0.514 0.184
Physical aspects x dummy -0.462 0.192 Personal interaction x dummy 0.383 0.251
4.4.6 Overview hypotheses
In table 16 there is a concise overview of the results of this research. As it can be seen, there is an indication whether the hypothesis is supported or not.
Table 16: overview hypotheses
Hypothesis Supported or
not supported
1. Customer satisfaction has a positive significant relationship with word-of-mouth.
Supported
2. Price has a positive significant influence on customer satisfaction. Supported 3. Assortment has a positive significant influence on customer satisfaction. Supported 4. Convenience has a positive significant influence on customer
satisfaction.
Not supported
5. Physical aspects have a positive significant influence on customer satisfaction.
Supported
6. Personal interaction has a positive significant influence on customer satisfaction.
Supported
7. Price has a stronger significant relationship with customer satisfaction for Turkish customers than for Dutch customers.
Not supported
8. Assortment has a stronger relationship with customer satisfaction for Turkish customers than for Dutch customers.
Not supported
9.Convenience has a stronger significant relationship with customer satisfaction for Turkish customers than for Dutch customers.
Not supported
10. Physical aspects have a stronger significant relationship with customer satisfaction for Dutch customers than for Turkish customers.
Not supported
11. Personal interaction has a stronger significant relationship with customer satisfaction for Dutch customers than for Turkish customers.
5. Conclusion
The results of this research lead to several concrete conclusions. As mentioned in the introduction chapter, the main research question of the study is ‘how do price, assortment, convenience, and supermarket environment influence the customer experience and subsequently word-of-mouth when comparing Dutch customers with Turkish customers?’ To provide an accurate answer, this main research question is divided into 5 sub-research questions. Furthermore, the control variables gender, age, education, shopping frequency, and household size are also included in the research model. However, the results have shown that only the control variable age has a significant influence on customer satisfaction.
Considering the first sub-research question regarding the influence of price on customer satisfaction, the conclusion could be made that price has a positive influence on the satisfaction of customers. However, a negative significant interaction effect has shown that there are explicit differences between Dutch and Turkish customers. Given the result of the positive main effect of price, the negative interaction effect means that the effect of price is smaller for Turkish customers than for Dutch customers. Subsequently, Dutch customers value ‘price’ more compared to Turkish customers.
In addition, the second sub-research question has researched the influence of assortment on
customer satisfaction. The analysis has shown that the provided assortment in a supermarket has a positive significant influence on customer satisfaction. This means that a good evaluated assortment leads to a higher customer satisfaction by customers. Besides this, the moderation analysis has shown no significant interaction effect between assortment and customer satisfaction, which means that there is no significant difference in how Dutch and Turkish customers value assortment.
