Copenhagen Business School
THE COMMERCIAL STRUGGLE OF
ONLINE GROCERY SHOPPING
A consumer behavioral study on why online grocery shopping has not had its
commercial breakthrough in Denmark
Written by
Dinah Malene Bjerregaard Thomsen (64805) Yasemin Seren (108286)
Supervision by Characters: 236,546
Christine Lenstrup Pages: 105
May 2018
Abstract
Objective: The objective of this study is to acquire an understanding to why online grocery shopping has not had its commercial breakthrough yet in Denmark. With a narrow focus on consumer behavioral studies, the study determines what underlying consumer behavioral factors that might explain why Danish consumers do not shop groceries online. Additionally, the study determines potential barriers to online grocery shopping that exist from a consumer perspective.
Methodology: The study will accomplish its research objective through a hermeneutic and social constructivist approach. The empirical data in this thesis originates from primary and secondary sources. The primary data consists of data collected through an online questionnaire. The online questionnaire consists of open and closed questions. The responses to the questions are used as the basis to understand why online grocery shopping has not had its commercial breakthrough in Denmark. The respondents consist of Danish women and men above the age of eighteen. The total sample size of the online questionnaire is 621.
Findings: The study found multiple consumer behavioral explanations that possibly explain why consumers do not shop groceries online. For example, grocery shopping covers more than a basic need, such as social needs, something that online grocery shopping is unable to cover. As grocery shopping covers more than a basic need, such as social needs, it requires more involvement. Online grocery shopping does not allow the same involvement, such as touching the products. Consumers believe that online grocery shopping requires too rigid a decision-making process. Additionally, many consumer characteristics clearly mismatch with the online grocery-shopping concept. Moreover, grocery shopping is highly related to habits. Changing habits is related to switching costs. Consumers express that switching from a physical to an online store has financial, relational and procedural switching costs. The study also found multiple consumer barriers to online grocery shopping. Examples include: negative attitude, perceived negative social consequences, lack of perceived quality control, and inability to touch items.
Conclusions: The study found that the answer to why online grocery shopping has not had its commercial breakthrough in Denmark is rooted in two main explanations: 1) There is a general mismatch between the concept of online grocery shopping and the consumer behavior of Danish consumers. 2) Consumers have concerns that act as barriers. This ultimately hinders them from shopping groceries online.
Key words: consumer behavior, grocery shopping, online grocery shopping, needs, involvement, habits, decision-making process, switching costs, consumer barriers.
Table of contents
Acknowledgements 4
1. Introduction 6
1.1. Research question 7
1.2. Motivation for choice of topic 7
1.3. Purpose 7
1.4. Delimitation 8
1.5. Thesis structure 9
1.6. Definitions 10
2. The Danish Grocery Industry 13
2.1. Grocery Shopping in Denmark 13
2.2. Online shopping in Denmark 13
2.3. Online Grocery shopping in Denmark 14
2.4. Online grocery shopping on a global scale 15
3. Methodology 17
3.1. Research philosophy 18
3.1.1. Hermeneutics 18
3.1.2. Social constructivism 20
3.2. Research approaches 21
3.3. Research strategy 22
3.3.1. Questionnaire 22
3.4. Choice of method 25
3.5. Time Horizons 26
3.6. Research Techniques & Procedures 26
3.6.1. Data Collection 27
3.6.2. Questionnaire Limitations and bias 27
3.6.3. Reliability and validity of research findings 29
4. Theoretical framework 32
4.1. Introduction to the field of consumer behavior 32
4.2. Needs 34
4.2.1. Maslow’s Hierarchy of Needs 34
4.3. High vs. low involvement 35
4.4. Introduction to consumer decision-making 36
4.4.1. The cognitive decision-making process 36
4.4.2. The theory of planned action by Fishbein and Ajzen 38
4.4.3. The habitual consumer decision-making process 40
4.5. The importance of switching costs 43
4.6. Decision-making in a household 45
4.7. Sproles & Kendall’s method for profiling consumers’ decision-making styles 47
4.8. Danish grocery shopping consumer types 49
5. Analysis 53
5.1. Data overview 53
5.2. Danish Consumer types 58
5.3. Needs, involvement and consumer decision-making 61
5.4. The habit of grocery shopping 74
5.7. Sub conclusion 89 5.8. Online grocery shopping and its lack of success in Denmark 90
6. Discussion 95
7. Conclusion 100
8. Future Research 103
9.Bibliography 107
10. Appendix 114
List of Figures
Fig. 1. Thesis Structure ...9
Fig. 2. Thesis structure inspired by Bloom’s Taxonomy, (Forehand, 2005) ...10
Fig. 3. Thesis research methodology in the context of the research onion ...17
Fig. 4. The Hermeneutic circle inspired by Bontekoe (1996) ...19
Fig. 5. Cognitive decision-making process inspired by Olshavsky & Granbois (1979) ...36
Fig. 6 The theory of planned action, inspired by Fishbein & Ajzen (1998) ...38
Fig. 7. Switching Cost, inspired by Burnham, Frels & Mahjan (2003) ...44
Fig. 8. Decision-making framework inspired by Lee & Levy (2014) ...47
Fig. 9. Consumer types and their involvement inspired by InsightGroup (2012) ...51
Fig. 10. Have you shopped groceries online before? ...53
Fig. 11. Do you shop online? (e.g. travels, clothing, electronics etc.) ...53
Fig. 12. Would you consider buying groceries online? ...54
Fig. 13. What is the likelihood of you buying the following categories online? ...54
Fig. 14. Responses: attitude towards online grocery shopping ...56
Fig. 15: Tendencies found in the responses to the open survey question ...57
Fig. 16: Overview of consumer types based on the questionnaire answers ...59
Fig. 17. Overview of consumer types and their willingness to buy groceries online...61
Fig. 18. The relationship between decision-making, needs and involvement ...61
Fig. 19: Why have you not shopped groceries online? ...66
Fig. 20 The classic decision-making process inspired by Solomon et. al. (2014) ...66
Fig. 21. Are you the one who is grocery shopping in your household? ...69
Fig. 22. What is your level of education vs. Would you shop groceries online ...70
Fig. 23. What statements describe you the most as a consumer. ...72
Fig. 24. Which description portray your grocery buying behavior ...73
Fig. 25. Where do you shop groceries? ...74
Fig. 26. How often do you shop groceries within the household?...76
Fig. 27. Why have you not shopped groceries online? ...80
Fig. 28. The likelihood that you would buy the following categories of groceries online?...81
Fig. 29. Would you consider buying groceries online? ...84
Fig. 30. Screenshot from Nemlig.com taken on March 27, 2018. ...97
Fig. 31. Screenshot from Nemlig.com taken on March 27, 2018. ...97
Acknowledgements
This research process has been a rewarding journey that brings our education to an end. We would not have made through this process without the help and support by others.
First and foremost, we would like to thank our supervisor Christine Lenstrup for her supervision. We never had a dull moment with Christine, and most importantly, her feedback was always thorough, insightful and helpful, which is something we truly cherish.
We would also like to thank our family and friends, not only for their endless support during the research process, but also for believing in us these past five years when we did not believe in ourselves. Thank you from the bottom of our hearts.
A major thank you and a virtual high-five to Coop MAD for sponsoring a gift card of 500 DKK to one lucky respondent. We wish you and other online grocery stores the best of luck in the future.
CHAPTER 1
Introduction
1. Introduction
Grocery shopping has always been an important aspect of human beings lives.
As long as humans have lived, they have had the need for food. Later as humans have evolved, the needs for other commodities such as personal care products have emerged too. However, the ways in which we acquire these products have changed.
Previously, consumers had to buy their products in separate specialized stores, for example: meat from the local butcher, groceries from the local market, and dairy products from local farmers (Rosenwald, 2017).
Today, all products are located in one store: the grocery store. In the past few years a new version of grocery stores has appeared: the online grocery store. The online grocery store has the exact same products as the physical grocery store. The only difference is that you do not need to step out of your home to buy groceries. You can collect all your groceries with a few clicks and get it delivered to your home address.
Generally speaking, Danes are very willing to buy products online (Thorstensen &
Munkøe, 2017). In 2017, Danes shopped for 83.8 billion DKK online, which is 19%
more than in 2016 (DIBS by Nets, 2017). When comparing to 2013, it is a growth of 78% (DIBS by Nets, 2017). On the other hand, the overall consumption has actually decreased by 10% from 2015 to 2016 (DIBS by Nets, 2017). Thus, growth in online consumption is unrelated to an increase of consumption in general. Danish consumers seem to just have embraced e-commerce. The most popular product that Danes buy online is travels, strongly followed by shoes and accessories (DIBS by Nets, 2017). To fully understand how much Danes actually love to online shop, it is relevant to compare it to others EU countries. According to official statistics by the EU, 82% of Danes have purchased goods or services online in the past 12 months (Eurostat, 2017). This is equal to one out of eight Dane. It is the third highest number in the EU, only topped by the United Kingdom (86%) and Sweden (84%). The average percentage for the European Union is 68% (Eurostat, 2017). In comparison, the countries that like to shop the least online are: Romania, where only 23% have shopped online the past 12 months, Bulgaria (27%), and Cyprus (39%) (Eurostat, 2017).
It is thus safe to say that Danes really like to shop goods online. However, even though Danes like to shop online, it is not reflected in all categories. According to recent research made by The Danish Chamber of Commerce, online grocery shopping in Denmark has not had its commercial breakthrough yet (Thorstensen & Munkøe, 2017).
Online grocery shopping only accounts for 2.5 billion Danish kroner of the total e- commerce turnover (Thorstensen & Munkøe, 2017). 98 per cent of all grocery shopping is done in physical stores, which means that online grocery shopping is still in its infancy in Denmark (Thorstensen & Munkøe, 2017).
1.1. Research question
Analyses made by The Danish Chamber of Commerce suggest that online grocery shopping in Denmark has not experienced the same breakthrough as other categories of online shopping. We are wondering why Danes have not embraced online grocery shopping like other categories in e-commerce. Therefore, the following research question have been established:
Why has online grocery shopping not had its commercial breakthrough in Denmark?
To answer this question, we ask the following sub questions:
• What are the underlying consumer behavioral explanations to why Danes choose not to do their grocery shopping online?
• What barriers exist from a consumer perspective?
1.2. Motivation for choice of topic
As marketing students strongly interested in consumer behavior, we always knew that we wanted to do our thesis within the area of consumer behavior. Grocery shopping is something that we all have to do. Roughly speaking, if we did not do grocery shopping, we would die. As brutal as this sounds, we need to acknowledge that grocery shopping is something that is of great importance in our lives. We would not have the vital necessities we needed if we did not buy our groceries somewhere. Of course, there is always the option of stealing or producing everything yourself. That, however, are options for the very few.
As consumers ourselves we noticed that online grocery websites started advertising more and more for their services in Denmark. Our immediate thought was that this had to mean that online grocery shopping had become common and popular in Denmark. On the other hand, none of us knew anyone who actually did their grocery shopping online. We chose to dig deeper in to this. Soon we found out that despite online grocery shopping appearing to be more common in Denmark, it had actually not had its breakthrough yet. We, as researchers, found ourselves guilty too. We are not buying our groceries online either. But, we seem to buy everything else online. Shoes.
Clothes. Books. But not groceries for some reason. This made us wonder why and eventually led to our research question.
1.3. Purpose
The main purpose of this thesis is to acquire an understanding to why online grocery shopping has not had its commercial breakthrough yet in Denmark.
Furthermore, the purpose of this thesis is to provide new knowledge in the area of consumer behavior. We aim to provide a new understanding of Danish consumers as grocery shoppers with this thesis. In particular, we aim to provide an understanding to
why Danish consumers choose to do grocery shopping offline rather than online.
However, we do not aim to solve the problem of the lacking breakthrough of online grocery shopping.
In one way the topic of grocery shopping is timeless in the sense that consumers will always need groceries. Researching grocery shopping behavior will never not be relevant because it is always relevant in consumers’ lives. At the same time, the topic is also very contemporary because it deals with how consumers do grocery shopping right now. Or how consumers are not doing their grocery shopping.
By analyzing the reason why online grocery shopping has not had its breakthrough in Denmark through a consumer behavioral lens, we aim to provide an answer to a contemporary problem. We hope that our findings are useful for online grocery stores and that they are able to use our thesis as a basis for overcoming consumer behavioral barriers that possibly hinder online grocery shopping from becoming a success in Denmark. Finally, we hope that our findings are useful in future studies for reflections or comparisons.
1.4. Delimitation
This is a consumer behavior-based study. We have chosen to strictly look at consumer behavior. This means that we will not take macro-economic factors, the marketing of the companies (including the branding, advertising and promotion) into consideration. Although, these factors may possibly contribute to the answer to the question proposed in some way or another. However, we find it too extensive to include due to the scope of this particular thesis. Furthermore, the focus of this study is not to analyze which segments shop groceries the most online and why because it is out of scope in relation to our research question.
Moreover, this study is based on the Danish market. In order to keep the study as detailed and focused as possible, we do not draw any parallels to other markets, although this could be a relevant angle to the study. We focus on Danish men and women above the age of eighteen. These individuals are considered to have legal capacity in Denmark (Statsforvaltningen, u.d.) While, we focus mostly on consumers who have not tried online grocery shopping, we also look into those who have tried grocery shopping but have had negative experiences. The reason for including both consumers is that the answer to our research question may stem from multiple consumer behavioral explanations. Anything from preconceptions of online grocery shopping to possible negative experiences may contribute to an obstruction of a commercial breakthrough in Denmark.
This is not a cultural based study. We do not consider any cultural differences within Denmark, although we know these exist. This is simply too extensive for this study. To keep it as simple as possible, we look at the Danish population as a
literature to look at national cultures from a homogenous perspective (Hofstede, 2018) (Inglehart, 2006) However, when referring to respondents in the thesis, we do include the respondent’s gender, age and region. This is solely for reference purposes. This information can be found in Appendix 6: Masterark.
In most analyses on grocery shopping in Denmark, pre-packed food boxes with recipes, called måltidskasser in Danish, are considered a part of online grocery shopping. However, we have chosen not to examine these in this study, as pre-packed boxes cannot easily be compared to offline grocery shopping. A pre-packed box is a unique service and a product in itself. We believe this requires a separate study, hence the exclusion in this study.
1.5. Thesis structure
Fig. 1. Thesis Structure
This thesis consists of eight chapters. Each chapter has its own purpose. The first introductory chapter consists of an introduction to the topic and the overall purpose of the thesis. In the second chapter, a short overview and description of the grocery industry in Denmark will be provided. The purpose of this chapter is to provide general foundation of understanding of the Danish grocery industry in order to support and provide a better understanding the research problem. In the third chapter the philosophical science perspective of this thesis will be discussed. Furthermore, the methodology used is described and its limitations are discussed. The purpose of this chapter is to provide a foundation and understanding of how the research is to be conducted. In the fourth chapter, the theoretical framework is introduced. In this chapter each concept used in the thesis will be explained carefully. The relevancy of the concepts in relation to the research question will be mentioned. Finally, possible limitations to the concepts are discussed. In the fifth chapter, our data will be analyzed in relation to our proposed research question. In the sixth chapter, we will discuss our results. In the seventh chapter, an overall conclusion will be presented. In chapter eight, research implications and further research suggestions of online grocery shopping are discussed.
1. Introduction 2. Industry 3.
Methodo -logy
Theoretical 4.
framework 5. Analysis 6.
Discussion
Conclusio7.
n
8. Future Research
Fig. 2. Thesis structure inspired by Bloom’s Taxonomy, (Forehand, 2005)
Our thesis structure is based upon Bloom’s Taxonomy of Educational Objectives. Bloom’s Taxonomy is used to categorize the degree of complexity of learning. This is used in order for students to obtain realization and acquire knowledge that they can apply (Blooms Taksonomi, 2018). Bloom’s Taxonomy consists of six steps. Each step becomes more complex. It is not possible to skip a step. The steps need to be acquired in chronological order (Blooms Taksonomi, 2018). The steps start from the known and ends up with the unknown (Blooms Taksonomi, 2018). By structuring our thesis in an order that fulfills these steps we make sure that we reach the most complex learning as possible.
1.6. Definitions
To provide a foundation of common understanding, this section provides an overview of definitions of key terms:
Groceries: we define groceries as products that are purchased in a grocery store (Oxford Advanced Learner's Dictionary, 2017). Key products include: food items, personal care products such as deodorants, shampoos and the like, home necessities such as cleaning supplies, bath tissues and tinfoil.
Grocery store: “a large shop/store that sells food, drinks and goods used in the home. People choose what they want from the shelves and pay for it as they leave”
(Oxford Advanced Learner's Dictionary, 2017). Retail Institute Scandinavia divides grocery stores into three overall categories: discount chains (e.g. Netto, Føtex, Aldi), supermarkets (e.g. SuperBrugsen, Kvickly) and department stores (e.g. Bilka, Føtex) (FødevarerWatch, 2014).
Evaluate
Synthesis
Analysis
Applica- tion
Compre- hension
Know- ledge
Online grocery store: due to the fact that the online grocery store is still a relatively new phenomenon, the classic English dictionaries do not offer a definition. However, we define an online grocery store as a grocery store, which does not have any physical stores. Online grocery stores only exist online. Consumers select products and pay for them directly on the website, where after the products are delivered by the grocery store to the consumer’s address. Online grocery stores have the same selection of products as traditional grocery stores. Online grocery stores often require a minimum purchase amount before delivery and purchase is possible. This price is often between 400-500 DKK. Moreover, online grocery stores often have a delivery fee. The delivery fee depends on the number of purchased items and the specific grocery store. The consumer needs to choose a specific timeframe, when the consumer wants the goods to be delivered.
Family: “a group of two or more persons related by blood, marriage or adoption, and residing together as a household” (Lee & Levy, 2004).
Household: “all the people living together in a house or flat/ apartment” (Oxford Advanced Learner's Dictionary, 2017)
CHAPTER 2
Industry
2. The Danish Grocery Industry
In this chapter we will present an overview of the current Danish grocery tendencies that dominates the industry. The aim of this chapter is to provide a common understanding and a foundation for understanding context of the thesis.
2.1. Grocery Shopping in Denmark
The Danish grocery industry has a big selection. It has everything from budget friendly grocery stores to big food markets and inspiring delicatessens. Despite the big variety and selection of different stores, two retailers mainly dominate the Danish grocery- market: Dansk Supermarked and Coop Denmark (BIQ, 2017). According to a newly conducted survey, the Danes’ favorite grocery stores are Netto and Rema 1000 (Lund, 2017). More than one million Danish households state that Netto is their preferred grocery store (Lund, 2017). On the other hand, 900.000 households state that Rema1000 is their preferred place to buy groceries (Lund, 2017). Both grocery stores are considered discount stores (FødevarerWatch, 2014).
In 2017 the Danish population spent approx 100 billion DKK on groceries in physical grocery stores (Pedersen, 2016). According to Coop Denmark, the bestselling groceries in Danish grocery stores are: bananas, beers, apples, pears and milk (Sennov, 2016).
In the last couple of years, there has been a change in how often Danes do their grocery shopping. Danes do not visit grocery stores as often as they used to. In 2012 the typical Dane visited a physical supermarket 3.6 times a week, which in 2017 declined to 3.5 times a week ( Landbrug & Fødevarer, 2017). The reason for this lies in the growth of discount chains. Today, Danes shop large-scale, which then decreases the number of weekly visits to the grocery store ( Landbrug & Fødevarer, 2017).
2.2. Online shopping in Denmark
In an increasingly digital world retail is moving online. This is also reflected in the Danish population’s shopping habits. In 2017 the Danish population spent about 84 billion DKK online (DIBS by Nets, 2017). The Danish e-commerce market has increased with 78 per cent over the past four years, from 47.2 billion DKK in 2013 to 83.8 billion DKK in 2017. More than 89 per cent of the Danish population is shopping online (DIBS by Nets, 2017).
The online consumption is allocated in three main categories: travelling, goods and services (DIBS by Nets, 2017). In 2017, 47 per cent of the total amount of online shopping in Denmark was spent on travelling. Hence, Danes in total spent 39.1 billion DKK on different travelling activities.
Goods on the other hand represented 37 per cent or 31.2 billion of the total amount of online shopping in Denmark. The goods category contains products such as clothing, groceries and other utilities. According to an analysis conducted by Nets (2017) this category has grown with 35 per cent from 2016 to 2017, from 23 billion to 31,2 billion.
In terms of the last category, service, 16 per cent or 13.5 billion DKK of the total amount of online shopping in Denmark was spent on different services.
2.3. Online Grocery shopping in Denmark
Despite the continuous increase of online shopping, the online-grocery market has not experienced the same rising demand in Denmark. Despite the fact that there has been a rise in distribution of online goods in general, data from The Danish Chamber of Commerce says that Danish consumers only spent about 2.5 billion on online grocery shopping in total in 2017 (DIBS by Nets, 2017) (The Danish Chamber of Commerce, 2017). That is to say, online-grocery shopping only covers 2 per cent of the total sales of grocery shopping (The Danish Chamber of Commerce, 2017). However, the abovementioned numbers also contain the turnover from pre-packed meal boxes, such as Aarstiderne, which we do not consider within this study (The Danish Chamber of Commerce, 2017). Therefore, the actual sales number of online grocery shopping in Denmark is even lower, when the pre-packed meal boxes are excluded.
The first online grocery stores in Denmark were established in 2007. Osuma.dk is considered one the first online grocery stores in Denmark (Konkurrence- og Forbrugerstyrelsen, 2010). Since 2007 several online grocery stores have been established. However, most of the companies that were established in the beginning have been forced to close down. Today, only two of the companies that were founded in 2007 are still operating, one of them being Osuma.dk. However, despite being first- movers on the market, these companies are still extremely small on the online grocery market (Konkurrence- og Forbrugerstyrelsen, 2010).
Currently some of the leading online grocery retailers on the Danish market are Nemlig.com and COOP Mad (Pedersen, 2016). Nemlig.com is the only Danish online grocery-retailer that has focused completely on the online market (Pedersen, 2016). The company offers its customers delivery service 365 days a year in larger cities with prices comparable to those in Netto (Pedersen, 2016). COOP Mad offers delivery of groceries and pre-packed food boxes, which they deliver all over Denmark. Despite being the biggest retailers on the market, Nemlig and COOP Mad have yet to experience a positive turnover on the Danish market (Pedersen, 2016).
Consumers who buy groceries online are mainly characterized as well-educated families with small children located in larger cities (The Danish Chamber of Commerce,
grocery shopping online (The Danish Chamber of Commerce, 2017). According to Coop Denmark, those who buy groceries online mainly buy fresh fish, rice, pasta, baby articles and tissue paper (Sennov, 2016).
2.4. Online grocery shopping on a global scale
Denmark is not the only country where online grocery shopping has not had its breakthrough. With a few exceptions, Europe’s online-grocery market has been stuck with a slow demand, which in turn has fostered a poor supply of online grocery-retailers (Galante et. al., 2013). On a European level, Danes are shopping more groceries online than the average European: 16 per cent of the Danish population has brought groceries online, where the European average is 12 per cent (The Danish Chamber of Commerce, 2017).
The United Kingdom is considered the frontrunner in Europe. 7.5 per cent of the total amount of grocery shopping is spent online. Outside of Europe, these numbers are slightly different. Asia has experienced a tremendous growth in its online-grocery market. One of the countries that have experienced the greatest growth is South Korea, where 20 per cent of the total amount of grocery shopping was spent online in 2017 (The Danish Chamber of Commerce, 2017). The leading global online grocery-retailers is considered to be Carrefour, Kroger, Target, Tesco and Walmart (TechNavio, 2017).
CHAPTER 3
Methodology
3. Methodology
This chapter will determine the methodological considerations and research philosophy this thesis is based on. The purpose of this chapter is to illustrate how this thesis will accomplish its research objectives and to justify the choice of method.
The research methodology of this thesis will be structured after the research onion provided by Saunders et. al. (2016). The research onion illustrates six different stages of a research strategy. This approach has been chosen as it effectively illustrates the development through which a research methodology can be considered. Hence, the research onion is used to explain this study’s research methodology, starting from the outer layer to the inner as illustrated in figure 3 (Saunders et. al., 2016).
First the research philosophy in this thesis will be defined. The research philosophy establishes the foundation for the research approach, which is considered in the second step. Third the research strategy is adopted. Following the choice of method will be presented within the fourth step. Fifth the time horizon of the study will be considered, and the final step represents the stage at which the data collection method is evaluated (Saunders et. al., 2016).
Fig. 3. Thesis research methodology in the context of the research onion adapted from Saunders (2016)
3.1. Research philosophy
The outer layer of the model focuses on the research philosophy that this study will adapt. Understanding the research philosophy is essential for the quality of this thesis, because it is considered to be the foundation of how the research is conducted (Saunders et. al., 2016 p. 108-130).
Every research philosophy consists of two main components: the ontology and the epistemology (ibid, p.108). Each contains important differences that will influence the research process. The ontology is concerned with the nature of reality and raises questions about the researcher’s assumption about the way the world operates. There are two aspects of ontology: objectivism and subjectivism (ibid, p. 109). The epistemology, on the other hand, is concerned with how knowledge can be created, acquired and communicated. Different paradigms contain different ontological and epistemological views, which in return result in different assumptions of reality that underline the research approach. This in return is reflected in the methodology and methods (Scotland, 2012)
Within the last decade most recognized researchers in consumer behavior have changed their positivistic research perspective to be humanistic (Hirschman, 1986) and naturalistic (Belk et.al 1988). This change is often referred to as the interpretive turn and is associated with qualitative research methods (Saunders et. al. 2016).
In this study, we have adopted an interpretive research philosophy.
Interpretivism is not a single paradigm but entails several diverse paradigms. In this study we will focus on the hermeneutic and the social constructionism paradigm.
Hence, this study will rely on a relativistic epistemology and a subjective ontology. The relativistic epistemology considers knowledge relative to time, place, society, culture etc. The subjective ontology, on the other hand, considers reality dependent on the consciousness of the individual. These orientations are oppositions to the nature of sciences and the belief in the objective reality (Saunders et al., 2016 p. 111).
This study considers both the hermeneutic and the social constructivist paradigm essential to obtain the research objectives of the thesis. The paradigms will enable the authors to explore the subjective meanings individuals establish and interpret the gathered data in a subjective manner (ibid p. 119). Moreover, both paradigms offer valuable research methods, which will help us to understand consumers as individuals and consider their behavior in a social and cultural context (ibid p. 119). However, the hermeneutic paradigm will be used to a greater extent in this study. This will be further elaborated in section 3.1.1 and 3.1.2.
3.1.1. Hermeneutics
Hermeneutics is the methodology of interpretation, which includes interpretation
explanation. This paradigm deals with the interpretation of human behavior and understanding of how individuals make sense of the world guided by their meanings (Saunders et al., 2016).
The hermeneutic paradigm is often used to process qualitative data like interviews (Rienecker, 2005). However, within this thesis we will conduct a questionnaire survey. Questionnaires are often considered a method within positivism because answers are obtained through closed-ended questions and analyzed using quantitative methods (Saunders Et. Al, 2016). However, the questionnaire within this thesis will be a mixture of both quantitative and qualitative methods since it contains open- and close-ended questions, which will be interpreted hermeneutically by the authors. There will be some positivistic elements in the questionnaire, such as statistics entailing consumer’s opinion about online grocery shopping, but these will all be interpreted and analyzed by the authors from a hermeneutic approach. The hermeneutic method has also been used when obtaining and evaluating relevant literature for the thesis.
WHOLE
INTEGRATION CONTEXTUALIZTION
(Define) (Illuminate)
PARTS
Fig. 4. The Hermeneutic circle inspired by Bontekoe (1996)
According to Andersen (2012), researchers should consider and challenge their own pre-understanding of a subject when using the hermeneutic paradigm. The pre- understanding consists of the researcher’s previous experience and observations (Andersen, 2012). Hence, we will attempt to be aware of our own pre-understanding of consumer behavior and grocery shopping, both online and offline since this may have an influence on our interpretation.
To challenge our pre-understanding, we will draw on the hermeneutic circle (fig.4.). The hermeneutic circle describes that no observation or explanation is free from the effect of the observer’s experiences and pre-understandings (Bontekoe, 1996). The circle considers meaning as holistic, since one’s understanding of a whole is established by reference to the individual parts, and the understanding of each individual part by reference of the whole. Hence, neither a concept as whole nor
EXPERIENCE
EXPERIENCE
individual parts can be understood without reference to one another. Similarly, it stresses that the meaning must be found within the cultural, historical and literary context (ibid).
Throughout our education both of the authors have had courses in consumer behavior and read secondary literature deriving from this field. Moreover, we have gained knowledge about online grocery shopping through different news media and social media pages. Hence, we have established a pre-understanding of this thesis’
problem statement. However, to challenge our pre-understanding, we will add new parts to the whole. For example, by conducting a questionnaire survey and following interpret the respondents’ answers. This in return will result in a new understanding of the subject and hence a new interpretation. This process will be repeated continually throughout the thesis.
A challenge that can be encountered when using the hermeneutic paradigm is to study literature and data objectively. It is unavoidable to have some degree of subjectivity, as our pre-understanding, interpretation and new frame of understanding will influence the outcome of our study (fig.4). However, we are well aware of this and will consider it during our analysis.
Another challenge of using the hermeneutic paradigm is to enter the social world of the research subjects and understand their world from their point of view (Saunders et al., 2016). Similarly, the subjective nature of this approach leave room for researcher bias, because results is heavily impacted by individual viewpoints and values.
Despite these challenges, we consider the hermeneutic paradigm relevant for our study as we aim to establish an understanding of the respondents’ attitude towards online grocery shopping and challenge our own pre-understanding of the topic.
3.1.2. Social constructivism
This study will also include elements from the social constructivist paradigm because it can explain some of the fundamental concepts within consumer behavior.
The social constructivism is the theory of knowledge, which argues that human development is socially dependent, and knowledge is established through interactions with other individuals (Saunders et al., 2016). Similarly, it is argued that we are born into a world where meaning already has been established. Hence, this scientific paradigm examines the development of a jointly constructed understanding of the world that form the basis for shared assumption about reality (Saunders et al., 2016).
Saunders et al. (2016) consider subjectivism as an ontology where reality is seen as socially constructed and the individual’s opinion is perceived as the foundation of reality. Individuals may perceive a specific phenomenon differently based on the reality that they consider to be true. This follows the interpretivist philosophy that one has to
Within this study the social constructivist paradigm will be used when analyzing the data that have been conducted through the questionnaire. The paradigm will allow us to explore the subjective meanings that motivate individuals to continue to shop in physical grocery stores rather than online grocery stores. This may be explained by different interpretations, as customers will perceive the concept different depending on their view of the world. Hence, when examining the problem statement from a social constructivist perspective we aim to understand the subjective reality of the consumers in order to understand their motives, actions and intentions in a meaningful way (ibid p.
142).
3.2. Research approaches
Within the second layer of the research onion, the research approach applied in this study will be considered.
The thesis will take form as an explorative study. An explorative study intends to seek new insights about a phenomenon and not offer a conclusive solution to the problems, but instead help to obtain a better understanding of the problem (Saunders et al., 2016 p.132-145). In this study we aim at obtaining new knowledge to why Danes do not shop groceries online and hence why online grocery shopping have not become a success in Denmark. Thus, the explorative approach has been deemed relevant to use in this study. However, the thesis will also rely on descriptive research, as the study aim at casting light on a current issue. Therefore, it is essential to have a clear understanding of the phenomena that is studied. Hence, the descriptive approach will be used to portray an accurate picture of the online grocery shopping market in Denmark and to describe characteristics or behavior of the questionnaire respondents.
We will also make use of the descriptive strategy when reviewing existing literature related to consumer behavior and when we portray the Danish online grocery market. Likewise, our choice of implementing a questionnaire is also a popular data collection method used in descriptive studies. The exploratory research will be used by implementing open-ended questions in the questionnaire survey. This will give us the opportunity to gain insights on topics we have previously not thought of. Hence, this approach will be used to seek new insights and assess the phenomena in a new light (ibid).
Moreover, researchers often consider two types of research approaches:
Deduction and induction. Deduction examines the possibilities to reach a specific and logical conclusion. When using deductive reasoning researchers move from something general - the theory, to the specific, the data. Induction on the other hand, makes broad generalizations and describe an explanation or theory. Hence, researchers consider something specific and turn it into something general (ibid p.124-126).
Deduction is often applied when using a descriptive strategy whereas the inductive approach often is used in connection with an exploratory study. Similarly, it is
also rare that researchers only make use of one approach, as it is difficult to understand everything from one of the perspectives (ibid p.124-126). Hence, this study will make use of both the deductive and inductive methods.
A deductive study usually starts by establishing a hypothesis. However, in this study the deductive approach will be used when we consider existing literature within consumer behavior to understand why consumers do not buy groceries online. This approach will also be used in the formation of the questionnaire, as the questions will be established on pre-existing theory that will be tested. The inductive approach will be used when we explore our empirical material to make generalizations about the reasons to why the Danish online grocery market has not experienced a breakthrough yet.
3.3. Research strategy
Within the third layer of the research onion, we will consider the choice of research strategy used in this thesis. The research strategy presents how we intend to conduct the research. A research strategy can include several different approaches, such as case study research, interviews or experimental research (Saunders et al., 2016. However, in this thesis we have as previously mentioned implemented a questionnaire.
3.3.1. Questionnaire
To enhance the validity of this study and to conduct a generalized analysis of the Danish populations grocery-shopping habits, the authors have chosen to conduct a questionnaire survey. The questionnaire strategy was chosen in this study because it allows the authors to collect a large amount of data from a sizeable population in a highly efficient way. The challenge of a questionnaire survey is that it can be time consuming to ensure that the sample is representative for the whole population (Saunders et al., 2016).
The objective of the questionnaire has been to obtain an understanding of the Danish population’s grocery shopping habits, both online and offline. Moreover, we have intended to record how the respondents feel about online and offline grocery shopping, gather data on consumer behavioral patterns in terms of grocery shopping and collect data about the respondents’ individual characteristics. The collected data will be used in the analysis to establish an understanding of why online grocery shopping has not become a success in Denmark yet.
The target group of the questionnaire has been Danish men and women above the age of eighteen. These individuals are considered to have legal capacity in Denmark (Statsforvaltningen, u.d.). The questionnaire has mostly focused on consumers who have not tried online grocery shopping to obtain the research objectives and answer the
a 500 DKK gift card to their online grocery-market to one of the respondents in the survey, which may also have enhanced the target group’s interest in participating in the questionnaire.
The questionnaire reached a total of 621 respondents. All of the participants were located in Denmark, above the age of eighteen and have either had or not had a previous experience with online grocery shopping. The questionnaire mainly went in depth with those respondents who stated that they have not tried shopping groceries online since these respondents were considered the most relevant in terms of obtaining the research objectives of this study.
The questionnaire was shared with friends, families, acquaintances and published in different Facebook groups and online forums. The authors, who originate from the Region of Southern Denmark but study in Copenhagen, shared the questionnaire on their private Facebook and LinkedIn accounts. This may explain the great proportion of answers from these regions in Denmark. Moreover, the authors’
networks contain many students, and this may also explain the great proportion of highly educated respondents.
In order to distribute the questionnaire to a larger audience than our networks, the questionnaire was published in different Facebook groups, such as: “Hvad skal vi ha´
til aften idag?”, “Køb og salg Nordsjælland”, “Lolland-Falster … hjælp hinanden” etc. The Facebook groups we chose to share the questionnaire in either had a great proportion of members, discussed somehow related topics (food/groceries) or enabled us to reach an audience within the different Danish regions. An overview of the Facebook groups and other online forums we shared the questionnaire in can be found in Appendix 1.
3.3.1.1. Questionnaire Design
According to Saunders et. al. (2006), the design of the questionnaire will affect the response rate and reliability of the data one collects. They argue that this can be maximized by:
• “Careful design of individual questions”
• “Clear and pleasing layout of the questionnaire”
• “Lucid explanation of the purpose of the questionnaire”
• “Pilot testing”
• “Carefully planned and executed administration”
Hence, we aimed at constructing the questionnaire after these guidelines presented by Saunders et. al. (2016). The final result of the questionnaire is illustrated in Appendix 2.
The questionnaire has been designed to be self-administered by the respondent through an internet-mediated platform. This choice was made because of the needed size of the sample.
According to Dillman (2007), it is important to introduce the questionnaire properly to obtain a high response rate. He suggests that researchers should include the following aspects in the introduction since it is the only opportunity one has to convince the respondent to participate in the research (Dilllman, 2007):
• “Clear and unbiased banner or title that conveys the topic of the questionnaire in an interesting way”
• “Interesting and informational subtitle that conveys the nature of the research”
• “Include graphic illustrations or logos to add interest and set the questionnaire apart” (Dillman, 2007)
Hence, in the beginning of our questionnaire we explained clearly and concisely the purpose of the survey, what kind of questions the respondents could expect and how much time it would take the respondent to finish the survey. We did also incorporate a simple colorful design with graphical illustrations and pictures to make the questionnaire attractive to encourage the respondent to participate. Moreover, we also highlighted the possibility for the respondent to participate in a giveaway, which also may have been a convincing element for the respondent. The final design of the questionnaire is illustrated in Appendix 3.
When constructing the questionnaire, we considered the order and flow of the questions, so it would be logical and easy for the respondent to fill in the survey.
The questionnaire has been based on relevant consumer behavior theories and the design of each question has been determined by the data we needed to collect. As mentioned earlier, we included a combination of open and closed questions in the questionnaire to obtain our survey objectives.
Some of the questions were also established on the basis of the Net Promotor Score (NPS) scale (Net Promotor Score, 2017). The NPS measures the consumers experience with a brand or product based on their rating of it on a scale from 1-10. The respondents’ answers are divided into three groups: Promotors, Passive and Detractors. Promotors are considered loyal consumers who will be likely to buy or refer others to use the product. These score between 9-10 on the NPS 1-10 scale. Passives are satisfied consumers that not engage with the same enthusiasm as the Promotors.
This group score between 7-8 on the NPS scale. Finally, the detractors are considered unhappy customers, who are more likely to grow negative word-of-mouth. Customers who score 0-6 are most likely to be detractors (Net Promotor Score, 2017).
3.3.1.2. Questionnaire Analysis
The data collected through the questionnaire have been processed and analyzed to make them useful in this thesis’ analysis. The results have been analyzed
1. “Survey data should be considered perceptions not facts”
2. “Results should be interpreted with other data sources”
3. “Understand what is behind the results to draw policy conclusions”
4. “Take into account the number and the way respondents were selected in the interpretation of the results”
5. “Take into account the response rate in the integration of the results”
6. “Document steps 1 to 6 and report transparently how the survey was conducted to assist users to interpret the results”
The results will be presented in graphs and tables within the fifth section of the thesis. The choice of statistics is influenced by this study’s research question (Saunders et. al. 2016). In this thesis we have illustrated our data in frequency tables, bar charts, pie charts and cross tables to present the results in an easy understandable manner.
This study has relied on inductive statics as we have aimed at generalizing our conducted data.
We have evaluated each of respondents’ answers to the open question and grouped them into different categories. This has been done to be able to clarify the respondents’ opinions and draw tendencies of our empirical data. Within the fifth chapter, we have used several of the respondents’ statements to support our analysis.
When referring to the respondents we refer to the number of the respondent, gender, age range and region. An overview of this can be found in Appendix 6. However, because of this study’s research question, we do not consider the demographic differences within the analysis but view the respondents as homogenous. The result of the questionnaire can be found in Appendix 4 and the analysis of the open question from the survey in Appendix 5.
3.4. Choice of method
Within the fourth layer of the research onion, we will consider the choice of method used in this thesis.
The use of either a quantitative or qualitative method introduces different data collection techniques and data analysis procedures. The quantitative method is often used when handling numeric data, whereas the qualitative method is used when using non-numeric data (Saunders et al., 2016 p. 151). Researchers can choose to use a mono method, mixed method or a multiple method. Mono method refers to only using a single data collection technique and analysis procedure. The mixed method refers to the use of two or more methods and using both a quantitative and qualitative methodology. While the multiple method refers to using more than one data collection technique and analysis procedure to answer the research question (ibid p.151)
In this thesis both qualitative and quantitative methods have been applied in order to obtain the research objectives and answering the research question. We have
used a quantitative data collection technique in this study, as we have chosen to implement a questionnaire survey (Saunders et al. 2016). The main intention of conducting a questionnaire is to be able to generalize. Hence, this method has been considered a valuable tool for collecting data in this thesis.
In contrast, the qualitative method has been used in our data analysis procedure. With the hermeneutic and social constructionism foundation, the results deriving from the questionnaire have required an interpretation of the answers.
Hence, this thesis will rely on both qualitative and quantitative techniques and procedures to answers the research question. Thus, this study has used a mixed method, as we have combined a single quantitative data collection technique with a single qualitative analysis procedure (ibid p. 151).
3.5. Time Horizons
Within the fifth layer of the research onion the time horizon of the study is considered. The time horizon is important to consider during a research, as the intended time horizon for completion has an influence on the research design (Saunders et al 2016). Saunders et. al. (2016) specifies two time-horizons in his research onion:
The cross sectional and longitudinal. The longitudinal time horizon refers to research that intent to examine the change over time. Hence, data is collected repeatedly over a period. The cross-sectional time horizon, on the other hand, refers to a study that examines data collected from a particular period (Saunders et al 2016).
Within this thesis we will use the cross-sectional time horizon, which examines data from a certain period. This approach has been deemed relevant, as this thesis aim at establishing an analysis of the current trends on the Danish online grocery market, year 2018. Similarly, this choice has been considered relevant in terms of the given scope and scale of the study. This study has been conducted in a period of three month, which also have had an influence on the collected data. Similarly, due to the constrained time it has not been possible to establish a longitudinal research process.
Moreover, the research question of this study is implying a cross-sectional time horizon. We aim at establishing an understanding of the current lack of success for Danish online grocery suppliers by considering potential consumer explanations for this matter. We also believe that this approach will make this study more relevant because the field of consumer behavior and digital trends is in constant development.
3.6. Research Techniques & Procedures
The data collection and data analysis are the inner layer of the research onion (fig. 3). Within this final layer we will consider the potential bias of the collected data, as well as the reliability and validity of the research findings.
3.6.1. Data Collection
The empirical data used in this study stem from both primary and secondary sources. The primary data consists of the data collected by the authors, while secondary data can be categorized as scientific data that have been collected from books, articles, academic research papers etc. (Saunders Et. Al., 2016). The secondary data will be presented in the theoretical framework whereas the primary data will be used within the analysis of this thesis. However, the secondary data was not conducted as a part of this thesis and may contain bias, as it was collected for another particular purpose. The potential limitations and biases have been considered in our literary review. We have also encountered problems in terms of gaining full access to some of the market research reports, which we have used in this study. Nevertheless, the advantage of using secondary data, such as saving resources, were considered an important aspect and hence relevant to use. The limitations associated with the primary data will be considered within the next section 3.6.2.
3.6.2. Questionnaire Limitations and bias
The primary data in this study derive from a questionnaire survey. Before publishing the questionnaire, the authors conducted a pilot test of the survey. The aim of the pilot test was to identify any mistakes or weaknesses in the questionnaire design.
This strategy was chosen as we as market researchers considered ourselves too close to the survey work to discover ambiguous questions or wordings. The pilot test consisted of a smaller sample of respondents, who read through the questionnaire and responded to the questions just as they were taking the survey. After the pilot sample we made some final adjustments and then published the survey.
However, mistakes are inevitable. After conducting a great number of replies, we detected some bias within the questionnaire. Our questionnaire contained some misguiding wordings, which could make it difficult for the respondent to clarify the meaning of the questions. For example, in the questionnaire we ask the respondents the following question:” On a scale from 1-5 how great is the probability that you will shop the following grocery categories online? 1 is no probability and 5 is great probability.” The question may guide the respondent’s opinion as we ask, “How great is the probability”. Instead we could have phrased the sentence more neutrally by asking
“What is the probability”. Hence, there might be some wordings, which may confuse or guide the respondents’ opinions. Conversely, we are well aware of these biases and limitations in our interpretation of the results.
Moreover, since we based our questions on theories within consumer behavior, which tend to contain generalizing elements, consumers might not find the options to fit their individual shopping behavior. However, we aimed at implementing neutral positions and to enable the respondents to elaborate on their standings. Similarly, to avoid uncertain answers in the questionnaire, we allowed the respondents to tick of
multiple choices or choose the opportunity “Other” to give them the opportunity to elaborate on their response (Appendix 2).
In the questionnaire we did not state that pre-packed food boxes are not considered a part online grocery shopping in our particular study. Since, these normally are considered a part of online grocery shopping the respondents in this questionnaire may also have perceived these to be a part of online grocery shopping. This is especially relevant to consider when handling the empirical data from those respondents who stated that they have bought groceries online. However, these respondents are, as mentioned earlier, considered less important in this survey.
We had some concerns with sharing the questionnaire in different Facebook groups; we worried that we would receive fake replies. Thus, we carefully selected which Facebook groups we wanted to share the questionnaire in. Our selection of Facebook groups was based on: the number of followers the group had and their topic relevance. However, we also observed and evaluated recent group discussions to determine how actively members engaged and their general tone-of-voice. Most of the Facebook groups either had restrictions or member rules, which also ensured us that the group members would reply in an appropriate manner.
Since, we had implemented a gift card in our questionnaire, we also considered the risk of receiving less constructive answers due to our sponsored giveaway. The gift card could potentially attract individuals who solely wanted to participate to win the gift card and not be interested in providing a true picture of their grocery-shopping habits.
However, we do not consider these concerns to be accurate, as we believe that the answers of the questionnaire seem to be reliable and thorough. Nevertheless, it is important to note that we have no control over this and we cannot know for sure whether the findings are true.
We chose Facebook as our main distribution channel because more than 3.7 million Danes use Facebook actively and therefore, it is possible to reach a wide audience (Digital kommunikation, 2016). However, our questionnaire lacks respondents in the age group 68+, which may be explained by the choice of distribution channel.
Even though the questionnaire was distributed to a diverse audience, 86 per cent of the respondents were women. We cannot explain this result as several different explanations to this outcome can be found: women may in general be more responsible for grocery-shopping, women may be more active on social media, women may be more eager to help, the authors networks may consist of more women or we simply engaged in groups with more active women than men etc. Of course, explanations are only conjectures, and we do not have a researched answer to this. The explanations may be endless, and we can only make predictions about the outcome.
These potential limitations have been considered throughout the entire study and we as researchers have tried to take an objective stance, when designing the
3.6.3. Reliability and validity of research findings
It is hard to ensure that research findings are accurate or correct. Hence, to reduce the possibility of incorrect answers we will in the following consider the reliability and validity of this study (Saunders et al., 2016).
According to Boolsen & Linderman (2008), the method of examination influences the importance of reliability and validity because it depends on whether a quantitative or qualitative approach has been used. Boolsen & Linderman (2008) argue that validity is most important in connection with qualitative research methods. Hence, the researcher should be able to ask questions during the interview that the respondent is able to understand and to be able to interpret the interview afterwards. On the other hand, reliability is considered most important when conducting quantitative research because the way the data is evaluated should be consistent and provide the same results if the study should be repeated (Boolsen & Linderman, 2008).
Since, this study contains both quantitative and qualitative elements, which is further explained in section 3.4, the reliability and validity of our results will be considered in the following.
Reliability is concerned with the extent to which the data collection techniques or procedures will provide consistent results (Boolsen & Lindermann, 2008). Threats to reliability could be technical issues, as the questionnaire was distributed through the internet-based system Survey Xact. Moreover, the pre-interpretations of the authors could also to a certain degree influence the understanding of the conducted data.
Hence, our results may contain observer bias, which is the tendency to see what one expects to see due to prior knowledge and subjective feelings (Statistics How To, 2016).
The reliability of the research also indicates whether the responses are evaluated consistently. All the information gained through the questionnaire has been evaluated objectively. However, the data will still have to be subject to some generalization to produce comparable results. Likewise, within this study we have also had a hermeneutically approach. Hence, some level of subjectivity will not be entirely avoidable.
Validity on the other hand, refers to whether the findings are what they appear to be (Saunders et al., 2016). Validity also considers the control of the research design, which can be influenced by external variables. According to Robson (2002) several issues such as history, testing, instrumentation, mortality and maturation have an influence on the validity of a study.
Before and during the research we have considered the influence of these factors and controlled for potential confounding variables. However, we have not been able to detect any variables that may have had a significant effect on our results.
To further enhance the reliability and validity of the data we could have assessed in-depth interviews with some of the respondents of the questionnaire. This method is also known as parallel forms reliability, which is a measure of reliability obtained by administering different versions of an assessment tool to the same group of individuals (Boolsen & Lindermann, 2008).
Given the non-controversial subject at hand, none of these prospects for bias were estimated sufficiently significant to invalid the empirical data. The data may still have some uncertainties, but these are considered minor given the research question and the intended use of the data. Hence, we consider the data sufficiently reliable and valid, since the likelihood of significant errors are minimal, and the use of the data will not lead to an incorrect result.
