Für neue Kunden:
Für bereits registrierte Kunden:
183 Seiten, Note: 1,7
List of tables
List of abbreviations
1.2 Problem formulation
1.3 Research questions
1.4 Intended Contribution
1.5 Structure of the thesis
2.1 Research approach
2.1.1 Research design
2.1.2 Research strategy
2.2 Data collection
2.3 Sampling approach
2.4 Data analysis
2.5 Ethical considerations
2.6 Potential weaknesses
3 Literature Review
3.1 Description of the process
3.2 Omnichannel retail practices
3.3 Customer journey
3.3.1 The customer j ourney concept
3.3.2 Customer journey phases
3.3.3 Customer journey touchpoints
3.3.4 Customer journey analysis techniques
3.4 Chapter summary
4 Empirical presentation
4.1 Preliminary framework
4.2 Findings and analysis
4.2.1 Requirements derived from the omnichannel literature
4.2.2 Requirements derived from the customer journey literature
4.2.3 Requirements derived from the customer journey touchpoints and phases literature
4.2.4 Requirements derived from the customer journey modelling tools literature
4.2.5 Additional requirements derived from the interviews
5.3 Data gathering
5.4 Data analysis
6 Concluding remarks
6.1 Reflection of the research approach
6.2 Conclusion regarding the research questions
6.3 Critical evaluation and recommendation for future research
List of references
The customer journey is a central topic for all retailers to consider to be able to successfully manage customer experience and to thereby ensure the survival of a business. Hence, managing and influencing customer journeys has been of importance since the emergence of retailing.
However, to successfully manage customer journeys has become increasingly complex with the development of omnichannel retail. Classical techniques to analyze and influence customer journeys are insufficient for today’s customers who are seamlessly using multiple devices and switching between different channels during their purchase journey while expecting a uniform and personalized customer experience.
This research paper studies the topic of customer journey analysis in omnichannel retail and discusses which requirements need to be fulfilled to successfully model, analyze and influence the respective customer journeys. By conducting a systematic literature review it was proven that a tool is needed to model and analyze individual customer journeys in omnichannel retail in real-time, making use of online and offline customer data. Through the method of semi-structured interviews, requirements that are related to developing and deploying such a tool were collected in an exploratory way.
This is from scientific importance as until now few research has addressed the topic of ways to model and analyze customer journeys in omnichannel retail. Additionally, with rising customer expectations and an increased competition between retailers, the importance of successfully managing customer journeys is indispensable in retailing.
The findings of this paper imply a high probability of an increasing need to deploy a customer journey modelling and analysis tool by omnichannel retailers. However, further in-depth analysis and specification of characteristics that this tool requires to fulfill are needed to successfully develop it.
Keywords: Customer journey, omnichannel retail, customer experience
Figure 1: Process steps of a CJMAT
Figure 2: The methodology of this thesis
Figure 3: Characteristics of omnichannel retail
Figure 4: The customer decision making as per the EKB model
Figure 5: Process model for the customer journey
Figure 6: Customer journey touchpoints
Figure 7: Current CJA techniques
Figure 8: Research rationale of this thesis
Figure 9: Preliminary framework
Figure 10: Matching of requirements and interviews
Figure 11: Process steps for a project to develop a CJMAT
List of tables
Table 1: Characteristics of the sample of this study
Table 2: Characteristics of current CJA techniques
Table 3: Quantity of articles discussing two literature streams
Abbildung in dieser Leseprobe nicht enthalten
This first chapter gives an insight into the author’s motivation and the scientific relevance to conduct research on the topic of customer journey modelling and analysis in omnichannel retail. The aim of this thesis will be pointed out and the structure of the paper will be outlined.
Digitalization is a major enabler that has changed the retail industry in recent decades. Nowadays, there are various retail companies operating in multiple channels, offline and online. A new form of retail has emerged, where all channels are seamlessly integrated; this is referred to as omnichannel retail (Bell et al., 2014). Omnichannel retail aims to make the customer experience as seamless as possible by providing the opportunity to choose from all available channels during all phases of the buying process according to the customers’ needs (Griffiths & Howard, 2008). In 2013, 86% of consumers used mobile devices and computers for shopping activities. Thereby consumers see the different channels as complementary services of information sources and not as separate ones (Deloitte, 2014). When consumers use multiple channels during the purchase process, retail companies must involve themselves in developing new ways to interact with their customers while evolving their business in an omnichannel environment.
How successfully retailers can provide services for their customers depends on their perceived service quality and their ability to deliver customer-centric services providing a unique customer experience (Halvorsrud et al., 2016). A description of customer experience with a retailer during the purchase cycle over time with multiple touchpoints is called a customer’s journey and has been widely studied by marketing and service management researchers. The total customer experience and thereby the customer journey is a dynamic process that ranges from pre-purchase, including search, to purchase and post-purchase (Lemon & Verhoef, 2016). At each stage, customers are influenced by their past experience and external factors while they interact with various touchpoints, each partly under a firm’s control. Taking into account the process stages and the touchpoints, guidance can be provided to systematically examine the customer experience during a customer journey and to model the effects of different touchpoints on the customer experience. During the customer journey analysis (CJA), firms investigate how their customers interact with the different touchpoints, aiming to first describe the journey and then to understand the options and choices for touchpoints during the multiple purchase phases (Verhoef et al., 2016).
Different methods can be applied to plan the customer journey. In service design, the method of service blue printing has been studied extensively, first being introduced by Shostack in 1982. Customer journey mapping (CJM) is another technique to display customer journeys within retail. Additionally, other process-oriented methods such as the critical incident technique (CIT), the sequential incident technique (SIT) or sequence-oriented problem identification (SOPI) can be deployed (Halvorsrud et al., 2016). Customer journey data can be collected and modeled to create value for a company. A common approach to do so is attribution modelling, which has the capability to determine the effectiveness of different marketing channels and to derive insight on the interplay between different channels across different data sets and conditions in a multichannel environment (Anderl et al., 2016). Nevertheless, most of these techniques focus on collecting data through qualitative methods such as interviewing or brainstorming, and are prone to different biases or focuses in the online world of retail. Tools to collect individual customer data in order to model customer journeys, while being able to instantly depict changes within these customer journeys, are needed to identify critical touchpoints and influence customer decision making in an optimum way. This thesis aims to study the requirements that such a modelling and analysis tool needs to meet from an omnichannel perspective.
As outlined above, changes in the social and technological environment have affected the retail industry, as well as consumer behavior, in many ways. This results in the fact that consumer interactions during customer journeys have diversified. While traditional customer journeys took place in an offline environment, nowadays also various forms of online interactions between consumers and firms occur, for example by social media. Since there is no sign of stagnation in technological development, it is expected that customer behavior will continue to evolve in upcoming years. Consumers nowadays have higher trust in other consumers and brands therefore no longer have the ability to steer their desired message and image of themselves (Grave, 2012). Hence, power has shifted from brands to consumers (Labrecque et al., 2013) and firms have limited control of the touchpoints that consumers experience during the purchase process (Lemon & Verhoef, 2016). Since more channel options are being introduced to the market, and seamlessly adopted by the omnichannel consumer, a more detailed understanding of customers’ shopping behavior is needed (Labrecque et al., 2013). Therefore, it is of interest to understand customer behavior during the customer journey.
To enable firms to understand their customers’ behavior, it is first necessary to collect relevant data from the purchasing process and to model different customer journeys while being able to depict changes in the journeys in real time. Currently, different methods of modelling and analyzing customer journeys are discussed in literature and in practice. However, there is a lack of tools for using real customer data collected during the purchasing process at different touchpoints while considering the characteristics of omnichannel retail.
The above-mentioned developments are of scientific interest since they lead to the need for an integrated awareness and understanding of a customer’s journey and influential factors across different business functions to enhance customer experience and sustain high customer satisfaction. This can be enabled by developing and deploying a tool that can first compile the relevant data and model customer journeys accordingly in real time, to then systematically assess customer buying behavior and to derive implications to successfully manage the customer journey. Hence, a customer journey modelling and analysis tool (CJMAT) needs to be developed. The process steps to be conducted by this tool are depicted in figure 1.
Figure 1: Process steps of a CJMAT
Abbildung in dieser Leseprobe nicht enthalten
Source: own graphic
To develop a tool combining these functionalities, it needs to be determined which requirements the tool needs to fulfill in order to first model customer journeys and to then analyze and influence them successfully. Hence, the aim of this paper is to examine which requirements a CJMAT needs to meet from an omnichannel perspective. This paper takes an exploratory approach and is focused on providing an overview of the requirements that different stakeholders place on a CJMAT. The aim is not to depict requirements in a way that requirement engineering proposes to collect and analyze requirements. This approach is justified since the topic of CJA in omnichannel retail has received limited attention until now and first exploratory insights need to be captured and developed.
From the above-mentioned motivation, the purpose of the study is developed and presented below:
The present study aims to generate new insight into the characteristics of the customer journey in omnichannel retail, into potential analysis techniques and into what requirements a CJMATfor an omnichannel retailer needs to fulfill.
More specifically, the following research aims to answer these questions:
1. What are the characteristics of the customer journey in omnichannel retail?
2. What are current techniques in use to analyze the customer journey in omnichannel retail?
3. Which requirements does a tool for an omnichannel retailer need to meet that can first model the customer journey and then analyze it to, ultimately, successfully manage the customer journey?
These research questions are outlined in an exploratory manner in order to allow rich insights to surface. The in-depth approach and the purpose of the paper aim to generate contributions that are discussed in the following chapter.
Previous research on CJA has mostly focused on deductive research approaches (Dennis et al., 2010; Emrich et al., 2015; Chiou et al., 2017). This might be due to the fact that the customer journey concept is originally based on the Engel-Kollat- Blackwell (EKB) decision-making model which was believed for a long time to be a generalizable and transferable model across different contexts and consumers and therefore applicable for conclusive research designs (Engel et al., 1995; Erasmus et al., 2001). However, several researchers have criticized this view and claimed that each customer experiences his own journey during the purchase process, hence claiming the need to take a constructionist approach in order to reveal insights into customer decision-making during the purchase journey (Harrel, 1990; Du Plessis et al, 1991; Bettmann, 1993; Erasmus et al., 2001). Nevertheless, some methodological exceptions exist within the customer journey literature. For example, Halvorsrud et al. (2016) analyzed particular customer journeys and Wolny & Charoensuksai. (2014) created customer journey typologies by mapping customer journeys in a multichannel environment by deploying qualitative data collection techniques.
Even though there are studies focusing on individual customer journeys, no research has been conducted in an omnichannel setting. There has been various research conducted in the field of decision-making and service management studying methods to analyze customer journeys (e.g. Botschen et al., 1996; Stauss & Weinlich, 1997; Kahnemann et al., 2004; Bitner et al., 2008; Karapanos et al., 2010; Li & Kannan, 2014; Boreham, 2014; De Haan, 2015; Lemon & Verhoef, 2016). However, with recent developments in technology, the need has arisen to deploy systematic CJA.
Hence, this research aims to study the requirements that a tool to model and analyze the customer journey on an individual level within omnichannel retail needs to fulfill. Therefore, the fields of omnichannel retail, the customer journey concept and analysis methods will be studied and synthesized in order to deduct the necessary requirements; semi-structured interviews will then be conducted to verify these requirements and to discover any additional ones. To fulfill this purpose, the study makes use of inductive and deductive approaches to research. The exploratory nature of the research will furthermore allow the participants of the interviews to spontaneously share ideas and thoughts when nearing a decision (Erasmus et al., 2001). Additionally, this paper focuses on a particular area of retail, namely omnichannel retail, with the aim of finding interesting insights into this new field of research. The findings of the paper will be of relevance for researchers in the fields of omnichannel retail and CJA. The paper also aims to provide the basis and directions for future research within the mentioned fields.
Furthermore, the thesis seeks to generate different practical implications. First, the insights created will help omnichannel retailers to understand the relevance of CJA within omnichannel retail practices. Second, the findings aim to provide the basis for the development of a CJMAT for omnichannel retailers. Third, the findings aim to challenge retailers to rethink their existing views on their customer journeys since the topic can be rather complex. These points are of practical interest because they can help to optimize resource allocation, measure channel attribution and successfully influence omnichannel customer experience in order to achieve high customer satisfaction (Wolny & Charoensuksai, 2014). Lastly, since the empirical focus of this paper is on omnichannel retail, these retailers will especially benefit from its findings.
This thesis is structured as follows: The first chapter outlines the background of the study and the rationale behind conducting the research. The problem is formulated and the purpose of the research, the research questions and the intended contributions are discussed. Chapter 2 outlines the chosen research design and strategy, the approach to data collection and the data collection techniques used. The second chapter additionally includes the possible limitations associated with the chosen research method. Chapter 3 reviews the relevant literature concerning definitions and past research in order to outline the theoretical background of the study and to identify relevant concepts from the literature. The basis of this research are studies with omnichannel retail characteristics, the customer journey and its relevant phases and touchpoints as well as the CJA techniques currently deployed in the field. The aim of this chapter is to derive a framework for developing the data collection in the following phases of the research. Chapter 4 presents and analyses the findings of the empirical study conducted in relation to the preliminary framework developed in the beginning of the chapter. Afterwards, chapter 5 gives a recommendation based on respective findings in relation to the purpose of the study and also previous research. The final chapter concludes this study by reflecting on the research approach and by answering the research questions. Additionally, the study’s main limitations will be discussed and suggestions for future research will be made.
This chapter describes the selected research technique of this paper, focusing on the research approach, including the research design and research strategy. The selected data collection technique will be explained, as well as the sampling approach and analysis methods that are deployed to analyze the data. Additionally, ethical considerations and the potential limitations and weaknesses of the selected research methodology will be discussed.
Before conducting the research, it needs to be explained what will be observed to achieve the goals that the research has set to pursue and how these observations will take place (Easterby-Smith et al., 2015). Hence, the research design and strategy will be explained in the following two sub-chapters.
As per Malhotra (2010), research designs are frameworks to conduct a research project that can be categorized into conclusive and exploratory. While conclusive research is more structured, exploratory research is more flexible and has the aim to create understanding and insight before an approach can be developed. To achieve this aim, exploratory research mainly makes use of qualitative data collection techniques since it does not want to test a theory, it remains rather unstructured in order to create new insights (Bryman & Bell, 2011). On the basis of the stated research questions and objectives of this paper, an exploratory research design was chosen since it enables to capture the small nuances of the requirements a CJMAT needs to meet by interviewing different parties and thus discovering new insights (Malhotra, 2010).
Additionally, this study, to some extent, makes use of inductive and deductive approaches during the research process. The inductive approach is also known as a bottom-up approach and first collects data to then draw conclusions based on the data (Bryman & Bell, 2011). The deductive approach is also called a top-down approach and starts with a hypothesis or theory that is tested during the research (ibid). Within this thesis, the deductive approach is used to develop a preliminary framework which is based on the existing literature and serves as basis to collect empirical data. This is carried out by conducting a structured literature review. Afterwards, an inductive approach is deployed to verify the requirements depicted by the preliminary framework, based on insights which are gathered from the interviews conducted. The pragmatic perspective of this paper allows us to combine elements from inductive as well as deductive views.
Thus, the present study differs from previous studies on CJA, since it focuses on omnichannel retail in which limited research has been conducted. Additionally, rather than studying customer j ourneys or analyzing a technique to study a customer journey, it aims to detect the requirements an optimal tool to first model and then analyze the customer journey in omnichannel retail needs to meet. The study additionally provides new insight into the field of customer journey in omnichannel retail, since it takes a pragmatic approach and most previous studies have been conclusive and mainly made use of a deductive approach as discussed in chapter 3. The selected research design defines the research strategy, which will be discussed next.
The research strategy of a study should be aligned with the research questions needing to be answered. De Vaus (2001) argues that the aim of the research strategy is to answer the research questions as univocal as possible. Alongside studying the characteristics of customer journeys in omnichannel retail and the techniques that can be used to analyze these journeys, this paper aims to examine the views and opinions different stakeholders of a potential CJMAT have about its requirements. The research was therefore based on a cross-sectional study. This strategy involves the collection of data from multiple cases at a single point in time, from which patterns of association can be drawn (Bryman & Bell, 2011). Cross-sectional studies are typically associated with quantitative research. However, it is also often employed as qualitative research by using semi-structured or unstructured interviews with multiple participants (ibid). There are no reasons why quantitative or qualitative approaches should be more suitable for cross-sectional studies (De Vaus, 2001). Among qualitative research articles with cross-sectional studies, the method of using semi-structured interviews is most prevalent (Bryman & Bell, 2011). A potential limitation of a cross-sectional study is the fact that it can be difficult to detect causalities, because the data is collected at a single point in time (De Vaus, 2001). However, this research takes an exploratory approach and does not aim to establish causal relationships or to depict changes over time. The study is instead interested in exploring current opinions and awareness. Considering the available time and resources of this research project, the chosen strategy of conducting semi- structured interviews was able to answer the research questions and to fulfill the purpose of the study. In conclusion, the research design and strategy selected helped to establish the premise of undertaking the process of data collection, which will be elaborated in the next chapter.
Qualitative data collection methods are used in this research to explore the views and opinions different stakeholders of a potential customer journey tool place on its requirements. Therefore, semi-structured interviews are conducted individually with the CJA stakeholders, as identified by the literature review in chapter 3.3.2. These stakeholders are considered candidates for the interview since they are likely to be in touch with the development of a potential CJMAT.
When planning this research paper, the benefits and drawbacks of triangulation were considered with the intent of increasing validity. Triangulation aims to improve the validity of qualitative research by using different methods to collect data on the same topic and thereby allowing to compare and question findings accordingly (Bryman & Bell, 2011; Opdenakker, 2006; Doody, 2012). This approach shows that the likelihood that the results are influenced by a particular chosen method decreases when there is a consistency of results between the different methods. Conducting a survey within the different omnichannel companies, based on the preliminary framework developed, was considered. However, given the limited time and resources of this research only semi-structured interviews were conducted, since contacting different omnichannel companies and getting the right contacts to send the survey to would have not been manageable within the timeframe of this thesis. Hence, given the aim of exploring different views about CJMATs, semi-structured interviews were considered as the most effective and appropriate method.
Conducting interviews is the most widely-used method to obtain qualitative data within a study (Malhotra, 2010; Bryman & Bell, 2011). The method allows a conversational style and is an unstructured, flexible and direct data collection technique (Malhotra, 2010). The interviews in this study were conducted on a one- to-one basis and were therefore synchronous in space and time (Opdenakker, 2006), with the exception of three interviews conducted via telephone. According to Bryman & Bell (2011) interviews can either be structured, semi-structured or unstructured. For quantitative research, structured interviews are often most suitable while semi- and unstructured interviews, which include open-ended questions, are typically used in qualitative research such as the present study (ibid). Semi-structured interviews leave room for the respondents’ perspectives since the aim is to acquire rich and detailed answers. Researchers can make use of an outline that lists the topics to be covered during a semi-structured interview, even though the direction will be determined by the answers of the person that is being interviewed (Easterby-Smith et al., 2015). According to Doody & Noonan (2013), developing an interview guide as an outline for an interview can help to create comfortable interactions with a participant because the researcher is already familiar with the topics to be addressed and can therefore concentrate on listening and responding to the given answers. For the present study, an interview guideline was drafted, taking into account the preliminary framework from chapter 4.1, considering the areas of interest to be covered within each interview.
Hence, an interview guide was developed for this study based on the literature review in chapter 3. Therefore, the requirements pointed out by the concepts discussed within the literature review, namely omnichannel retail characteristics, the customer journey concept, customer journey touchpoints and phases as well as CJA techniques, were collected. These requirements were coded and questions were developed to validate the requirements already depicted and to get insights into any additional requirements when conducting the interviews with different stakeholders. To secure a reliable analysis of the interviews, the questions were also coded.
To gain elaboration on certain answers, slightly different wording was used and further questions were asked while conducting the interviews, depending on the participants’ replies to the questions. After asking a question, participants were asked to elaborate further on certain topics, depending on their answer. This is a motivational technique, known as probing, which helps to clarify, enlarge and further explain the answers given (Malhotra, 2010; Easterby-Smith et al., 2015). Even though most of the participants were willing to provide the necessary information, they required guidance on the details they were expected to give. Therefore, probing was a key technique deployed during the interviews.
As per Malhotra (2010), the length of an interview generally varies but is usually between 30 and 60 minutes. Nevertheless, Bryman & Bell (2011) argue that short interviews can also be revealing. Within this study, the interviews ranged from 27 minutes to 46 minutes. Best practice shows that interviews should be recorded and transcribed since it helps to correct the limitations of memory, allows for the possibility of reviewing how and what exactly was said and for the reusing of data gathered by other researchers (Malhotra, 2010; Bryman & Bell, 2011). Hence, all the interviews in this study were recorded and transcribed (see appendix). Additionally, interviews should be conducted in the language of the study itself because translating can lead to linguistic, sociocultural and methodological problems (Bryman & Bell, 2011). Within this study, two languages were used to conduct the interviews, namely English and German. The reason was to encourage participants to share as much information as possible and to provide results that were as reliable as possible. This was in some cases easier in the participants’ native language. The drawbacks associated with translation were mitigated by careful procedures within this study’s data analysis stage. These measures are described in subchapter 2.4.
The method of using semi-structured interviews to collect data for a study has certain limitations. One potential limitation is the skill of the interviewer, since the completeness and the quality of the results heavily depends on these skills (Malhotra, 2010). A completely objective approach is not possible within the interviews and is based on the interviewer’s own habits in moderations and interpretation, since the aim is to gain in-depth insights (Easterby-Smith et al., 2015). Therefore, the researcher could affect a participant’s answer by, for example, unintentionally showing disapproval or surprise (Doody & Noonan, 2013).
Additionally, the fact that people tend to rationalize their actions when they are asked to recall their previous experiences might lead to incorrect results (Easterby- Smith et al., 2015). To mitigate these limitations, different stakeholders were interviewed within this study with the aim of detecting similar opinions and acquiring consent on the requirements that need to be fulfilled when developing a CJMAT for omnichannel retail.
The following graphic sums up the different research steps of this thesis and points out the underlying methodological concepts.
Figure 2: The methodology of this thesis
Abbildung in dieser Leseprobe nicht enthalten
Source: own graphic
As displayed in figure 2, deductive and inductive reasoning will be used to derive a recommendation on which requirements to consider when developing a CJMAT for omnichannel retail.
Since the purpose of this study is to conceptually examine the requirements that different stakeholders place on a CJMAT in omnichannel retail, it is important to collect the data from participants that are able to provide the relevant insight into such requirements (Bryman & Bell, 2011). Hence, the criteria for choosing the participants of the semi-structured interviews was that they should be people in contact with a potential CJMAT in omnichannel retail, also called stakeholders, and people who have experience in developing CJMATs for companies.
Regarding the empirical focus of this study and the limited timeframe available, only stakeholders representing the internal opinions of omnichannel retailers and potential developers who have experience in developing and deploying a CJMAT, were interviewed. Nevertheless, the aim when selecting these people was to gather requirements in a holistic way, depicting the requirements in their full extent within this paper. Hence, people were interviewed who represent an area that is directly in touch with the customer journey concept and has the ability to influence customer journeys and the potential development of a CJMAT accordingly. These include the business functions IT, cross-channel, customer service, marketing and corporate development departments of an omnichannel retailer. Moreover, to collect information from more people than just stakeholders employed by an omnichannel retailer, two consultants were interviewed who have been working for various customers on CJA topics. By selecting these participants, the research makes use of purpose sampling, since the participants were selected in a strategic way in order to fit a certain criteria description (Malhotra, 2010; Bryman & Bell, 2011). Keeping in mind that the study does not aim to generate conclusive results, purposive sampling is an appropriate method to be employed.
In practice, the participants for the interviews were recruited by identifying individual stakeholders of a customer journey within an omnichannel retail company. Additionally, potential candidates from consulting and software development companies, who are working in the area of CJA, were contacted. The aim was to not only generate insights from internal stakeholders of an omnichannel retailer but also to capture insights from candidates who have been working in an external and therefore neutral role within the field and have experience in working on the topic of CJA with different omnichannel retailers in practice.
The period of data collection lasted three weeks, during which the interviews were conducted. In total, the final sample consisted of seven participants who were interviewed. It is important to consider the different characteristics of a sample in relation to the larger context from which the sample is drawn (Easterby-Smith et al., 2015), despite the fact that purpose sampling does not allow for the generalizing of the results of a study to the wider population (Bryman & Bell, 2011). The participants of this study varied in their key characteristics, e.g. gender, employer, position and department (Table 1), although it needs to be pointed out that more men than women volunteered for the study. However, a sufficient number of perspectives were included in this study, due to the different characteristics mentioned above. This is an important criterion when evaluating the validity of a constructionist study as this one (Eastery-Smith et al., 2015).
Abbildung in dieser Leseprobe nicht enthalten
Table 1: Characteristics of the sample of this study
In conclusion, it is important to mention that the sample size of this study is small and non-representative. This is often the case in exploratory research, especially when conducting interviews. However, the aim of generating maximum insight rather than coming to conclusive findings can still be fulfilled (Malhotra, 2010).
The data gathered from the selected sample was rich and provided the necessary grounds for fulfilling the purpose of this thesis. The next subchapter focuses on how the collected data was analyzed.
It is important to carefully follow procedures during the analysis of data in regards to the purpose of a study, since qualitative data collection techniques generate a large set of data (Bryman & Bell, 2011). Before any actual analysis of the collected data took place, the data was systematically filed and organized, as Easterby-Smith et al. (2015) recommend.
The organization of the results of the interview was conducted in the following way: Firstly, each interview was transcribed word for word. After that, the transcripts were checked to avoid losing any data. This was followed by a translation if needed and another proofread. Additionally, the notes taken during the interviews by the interviewer were checked to make sure any additional information gathered was included. This process was conducted directly after each interview to ensure that the social nuances and settings of the interview were still in the mind of the interviewer (Bryman & Bell, 2011).
After all the interviews were transcribed and formatted, the transcripts were read multiple times. To identify categories from the non-standardized data (Saunders et al., 2009), the relevant concepts were identified with the help of the preliminary framework developed in chapter 4.1. The framework was derived from the literature review before the data collection took place. This allowed to group and distinguish themes related to the purpose of the research (ibid). Additionally, it provided the chance to bridge the gap between the theory of this paper, the empirical data and the purpose of the study (Bryman & Bell, 2011). In practice, all sentences related to the requirements defined in the framework were highlighted and underlined. Insights that could not directly be connected to the requirements already identified were retained in order to identify patterns within them at a later stage. Additionally, marginal notes were written on the printed transcripts and gradually refined into codes to group them accordingly (Bryman & Bell, 2011). Afterwards, notes were compared and quotes grouped according to the requirements pointed out in the framework. Since the data was sorted systematically according to the predefined concepts, a deductive process of analysis took place. However, evaluating themes that did not fit into the preliminary framework was an inductive approach.
In conclusion, the explained analysis procedures of this study were consistent with the research design and strategy. Additionally, they allowed for the deployment of a well-structured approach to pursue the analysis, answer the research questions and finally, to fulfill the purpose of this study (Saunders et al., 2009).
The principals of ethical considerations within the field of research aim to protect research subjects as well as the research community (Easterby-Smith et al., 2015). These principles can be broken down into four different categories, namely to prevent harm to participants, a lack of informed consent, an invasion of privacy and deception (Bryman & Bell, 2011).
First, the research should not cause harm to the research subjects. The main possible harm within this study that the research subjects could have experienced was stress during the data collection in form of interviews (Malhotra, 2010). To prevent this, participants were informed that there were no right or wrong answers to the questions asked and that their right of anonymity would be respected, keeping their identities confidential during the research process (Bryman & Bell, 2011). Additionally, the participants were free to end their participation if wanted. These aspects secured the privacy of the participants during the data collection for this study.
The category of informed consensus relates to the fact that participants should receive detailed information about the purpose of the study before taking part in it (Burns & Burns, 2008; Bryman & Bell, 2011). Therefore, the interviewer explained the topic of this study when approaching the participants, as well as at the beginning of the interviews. Additionally, the participants were told to feel free to ask any questions at any time before, during and after the phase of data collection. Before the interview, the participants were also informed that the interview would be recorded. These procedures reduced the risk of deception - the possibility that the study could be presented in the wrong way in order to obtain more natural responses by participants (Bryman & Bell, 2011).
There is also a political risk that needs to be considered when conducting research. There are four key stakeholders of this study: the researcher and the client, the respondent and the public (Malhotra, 2010). To mitigate any political risk, one should be aware that these stakeholders might want to influence the direction of the research as well as the results of the study (Easterby-Smith et al., 2015). Additionally, ethical issues can arise when the stakeholders’ interests are in conflict (Malhotra, 2010). Moreover, any research’s sources of support and funding should be declared to increase transparency (Bryman & Bell, 2011). This study was conducted as part of the research project SURTRADE at Kuhne Logistics University in Hamburg. The participants of the interviews mainly work for companies which are involved in the research project group. However, none of them determined the topic of this study or influenced the questions asked during the interviews, since this might have resulted in biased results (Easterby-Smith et al., 2015).
Another possible ethical risk was posed within this study while designing the interview guide. To prevent this, biased statements and wordings were avoided when developing the questions of the interview guide. For example, no omnichannel retail company was named within the guide because this could have resulted in comparisons to a competitor or could have affected the given answers. Additionally, collecting sensitive data such as the participants’ political opinions, ethics, sexual orientation, physical or mental health and religious views was avoided (Burns & Burns, 2008). Since it is considered ethical to inform the participants of a study about the outcome of the research (ibid), the final version was delivered to all participants via e-mail.
Overall, this study takes a universalist stance, hence meaning that throughout the study, all ethical principals were respected. To conclude this chapter, the potential weaknesses of the methodology employed within this research will be discussed.
There are four common criteria used to assess the quality of business and management research, namely the concepts of credibility, transferability, dependability and confirmability (Bryman & Bell, 2011). These criteria are the equivalents of reliability, replicability and validity which are used to evaluate quantitative research and will therefore be used within this chapter to reflect any potential weaknesses and the trustworthiness of this study.
The criterion of credibility refers to the step of evaluating how believable the findings of a research are (Bryman & Bell, 2011). To ensure plausible findings within this paper, interviews with different stakeholders were conducted. However, as pointed out in subchapter 2.3, the sampling approach deployed did not allow to generate generalizable findings. Nevertheless, to create such results was not within the scope of this research and therefore cannot be considered a main limitation.
Transferability relates to the fact that the findings of a study should be applicable in other contexts (Bryman & Bell, 2011). Concerning this criterion, the present research is limited for the same reasons as mentioned above. Because the data was only collected from a limited number of people, the findings are unique within the context and can therefore not directly be generalized to other markets or industries (Bryman & Bell, 2011). Transferability is considered a responsibility of the researcher. Hence, the research assumptions and methodologies are thoroughly described within this paper to enable other researchers to transfer the findings of the research to other environments.
To help other researchers in evaluating the trustworthiness of a study by explaining its context is called dependability (Bryman & Bell, 2011). Therefore, evidence from all steps of this study was documented, stored and is available on request. This enables a critical audit of the process later on. Earlier versions of this study, e.g. the literature review or problem formulation were retained in order to increase the transparency of the research process (Easterby-Smith et al., 2015).
Confirmability refers to evaluating whether the values of the researchers could influence a study to a high degree (Bryman & Bell, 2011). This criterion is difficult to estimate. However, to minimize the bias that could undermine the findings of this study, efforts were made, for example by standardizing the mode of data collection and by conducting a practice interview. Despite the fact that English is not the first language of the researcher, it was used to collect data since this study is in English. Before the data collection took place, it was made certain that each participant had a sufficient command of English to verify that no insights could be lost due to linguistic limitations. If participants were not able to answer a question in English, or felt more comfortable in answering in German, they were free to do so. The answers were than translated. Overall, these measures enabled to accurately depict the individuals’ views while minimizing the researchers’ bias. Therefore, the research provides access to the different experiences and opinions of the people who were part of the research (Easterby-Smith et al., 2015).
In conclusion, chapter 2 described the approach and techniques deployed in order to conduct this study in detail. Additionally, any possible limitations were critically pointed out. On top of these methodological requirements, it is important for a study to understand the theoretical framework it is built on. Hence, the next chapter seeks to give an introduction into the theoretical foundations of customer journeys in omnichannel retail and related analysis techniques.
In the following, a literature review will be conducted. According to Easterby- Smith et al. (2015) and Durach et al. (2017) a literature review provides context in addition to refining the topic of a study. Within this chapter, the theoretical background of this study will be discussed. First, how the literature review was conducted will be described. Second, the phenomena of omnichannel retail and its characteristics will be elaborated, as it provides the context for the following discussion. Third, the customer journey in omnichannel retail and the respective analysis tools in use will be pointed out. The aim is to deduct the relevant concepts from the literature in order to combine them within a preliminary framework to be used for data collection and analysis.
A systematic literature search was conducted in order to find relevant articles published in the period from 2000 to 2017 with a cut-off date on the 31st of August 2017. Relevant articles which relate to the topic of omnichannel retail, CJA in omnichannel retail and respective planning and modelling techniques were identified. The database of WebOfScience (https://webobknowledge.com/) was deployed as a first step. Separate searches were conducted for the above-mentioned categories, including a combination of the keywords “omnichannel retail”, “customer journey” and “analysis techniques”. The keyword search within the WebOfScience database yielded around 15 articles for the subject of omnichannel retail, 141 articles for the subject of customer journeys and, within these, 30 articles on the analysis of customer journeys. In the next step, the focus was narrowed down to articles published in peer-reviewed journals, excluding conference papers and similar material. Afterwards, all articles were screened by reading the abstracts and the irrelevant ones were filtered out. The resulting articles were analyzed in detail and references in the relevant parts of the articles were again checked by deploying WebOfScience’s data database along with Google Scholar.
In recent years new retail channels have emerged. In a retail environment, channels can be described as customer contact points or mediums, which enable a firm and a customer to interact (Verhoef et al., 2015). Retailers have the possibility to integrate all information these different channels provide by using new technologies. This phenomenon is known as omnichannel retailing (Brynjolfsson et al., 2013). The latin word “omnis” stands for “universal” or “all”, hence meaning all channels together. Therefore, while traditional multichannel retail makes a distinction between online and physical stores, in omnichannel retail, customers move freely across different channels, such as online, mobile devices or physical stores, within a single transaction (Melero & Verhoef, 2016).
New technologies, for example smartphones and tablets, as well as related software such as apps, mobile payments, digital flyers and location-based services, have driven these changes within the retail industry. Additionally, price optimization and personalization are enabled by changes in IT provision, access to technology, such as big data and cloud computing, and reduced costs. Furthermore, in-store- technologies have developed, for example virtual mirrors, digital signage, vending machines, intelligent self-service kiosks and virtual screens. The deployment of emerging technologies such as Google Glasses and 3D printing are expected to further push changes. There is also a growing usage of social media on the customer side. Hence, the connected mobile customer has the possibility to access information and to buy anything at anytime and anywhere. This leads to the fact that retailers have to remove barriers within the different channels and provide services across channels such as “order in-store, deliver home”, “click and collect” or “order online, return to store” (Cuthbertson & Piotrowicz, 2014). The natural boundaries that once separated the different channels begin to disappear, leading to interchangeable and seamless use during a customer’s search, purchase and postpurchase process of a product. For firms, it is difficult or virtually impossible to control this usage (Verhoef et al., 2015).
However, regardless of the channel that customers use, they expect a uniform, integrated and consistent service experience (Cuthbertson & Piotrowicz, 2014). They want to use their own devices to search and compare products, to ask for advice or to look for alternatives during their purchasing journey in order to take advantage of the benefits offered by different channels and retailing companies (Rippe et al., 2015). Because of the seamless usage of different channels by omnichannel customers, a traditional division between interactive and one-way communication channels becomes less obvious. Hence, it is important to broaden the scope of channels via customer touchpoints. Touchpoints can be one-way or two-way interactions between customers and firms, which can be extensive or superficial and also involve customer-to-customer interactions, for example via social media or peer-to-peer communication (Verhoef et al., 2015). To meet customer expectations, a main characteristic of omnichannel retail is its customer- centered strategy. The aim is to offer the shopper a holistic experience (Gupta et al., 2004). To ensure a satisfying customer experience, the interplay between channels and brands is of special importance. Omnichannel retail therefore does not only broaden the scope of channels, but also integrates the consideration of customer- brand retail channel interactions. Research in the field of omnichannel retail is therefore especially focused on how each customer touchpoint influences brand and retail performance (Verhoef et al., 2015).
Following the above-mentioned discussion, omnichannel retail can be defined as a, from new technologies emerging, form of retail whereby customers interact with a firm and each other during a purchasing process through numerous available channels and touchpoints. Omnichannel retailing companies aim to maximize customer experience during this process in order to optimize performance seamlessly over these different channels. In order to do so, it is vital to align different business functions, such as information technology (IT), service operations, logistics, marketing, human resources and external partners (Lemon & Verhoef, 2016). These functions need to be aware of the influence they have on the touchpoints a customer has with a firm in order to use these moments of influence to reach consumers with the right messages at the right time in the right place. The following graphic depicts the main characteristics of omnichannel retail as explained above.
Figure 3: Characteristics of omnichannel retail
Abbildung in dieser Leseprobe nicht enthalten
However, to understand how these characteristics should be aligned, we need to investigate how customers choose to use the different channels and how these choices may influence their buying decisions. Hence, customer behavior during the purchase process needs to be systematically analyzed by the usage of relevant data. To be able to identify the relevant data, it is necessary to understand the characteristics of a customer journey in omnichannel retail and to model the customer journey with real time customer data.
The following chapter discusses the theory of customer journeys and the respective analysis tools currently used in practice. Having pointed out the characteristics of omnichannel retail and its development in chapter 2.2, this chapter provides the basis for understanding how a customer interacts with a firm during a purchasing process in an omnichannel context. First, it will be elaborated what customer journeys are and how the concept has developed. It will be presented which special characteristics customer journeys in omnichannel retail have. This will be followed by a discussion of the different touchpoints of customer journeys in omnichannel retail and a depiction of the techniques that are deployed to study customer journeys. This chapter and the following chapter (2.4) serve as the basis to develop a framework for conducting interviews to study the requirements that a CJMAT needs to meet.
The customer journey concept has evolved from initial research in the field of customer experience management and from decision making models. Customers’ channel choices and buying decisions are strongly influenced by their experiences. According to Lemon & Verhoef (2016) customer experience is “a customer’s journey with a firm over time during the purchase cycle across multiple touchpoints”. The customer journey is considered the treatment of customer experience by Pucinelli et al. (2009) and Verhoef et al. (2009). The Marketing Science Institute (2014, 2016) states that customer experience is one of the most important research topics for upcoming years. This is due to the fact that previous research has shown that there is an increasing number and complexity of customer touchpoints and that creating positive, strong experiences within customer journeys can improve a company’s financial results by accelerating performance within the customer journey at different touchpoints, positive word of mouth and higher customer loyalty (Court et al., 2009; Homburg et al., 2015; Edelmann, 2010). Additionally, studies have shown that a positive customer experience can support a brand and create emotional bonds or conversely emotional scarring (Pullman & Gross, 2004; Berry & Carbon, 2007).
Customer experience has widely been studied within marketing literature since the 1960s, with early works by Kotler (1967) and Howard & Sheth (1969) focusing on initial theories on marketing and consumer behavior. For this research paper, the theories of customer buying behavior process models and service quality are especially important, since they study techniques that depict and analyze customer journeys and will further be discussed in chapter 2.3.2. To successfully manage customer experience, an understanding of the customer decision-making process is needed.
The most popular model to explain the consumer decision-making process is from 1968 and was developed by Engel, Kollatt and Blackwell, resulting in the so-named EKB model. It consists of the five steps depicted in the figure below.
Figure 4: The customer decision making as per the EKB model
Abbildung in dieser Leseprobe nicht enthalten
Source: Own graphic based on Ashman et al., 2015
The EKB model was frequently used to display the steps that a consumer takes while preparing to buy a product or service. Additionally, the assumption was that the consumer goes through the different steps in the same order every time a purchase takes place (Wolny & Charoensuksai, 2014). Furthermore, it was assumed that the information search and the evaluation of alternatives before an actual purchase take place inside a store (Lofgren, 2005). However, there have been several changes in the world of retail since the development of the EKB model. As explained in the previous chapter, in an omnichannel context, consumers have nearly unlimited access to information and use different devices simultaneously (Dias, 2016; Saxon, 2017). Additionally, new retail concepts have developed and the complexity for marketers to understand consumer choices is high (Gullstrand et al., 2016). Hence, it is important to understand that events during a contemporary consumer purchase differ from those the EKB model proposes; researchers need models to capture the actual consumer path in detail (Wolny & Charoensuksai, 2014). Consumer decision making in the light of activities and considerations during a purchase journey remains complex, despite the efforts made by previous research Eramus et al., (2001, p. 89), which claims that consumer decision making “cannot be generalized over the wider spectrum of consumer goods”. Hence, the decision-making process needs to be studied in detail in reality.
While previous customer experience studies and decision-making models help in understanding the purchase process and related customer feelings in a broader context, the customer journey literature concentrates on describing the phenomenon in more detail. Thereby, the customer journey literature still makes use of the five steps that the EKB models proposes by portraying the overarching purchasing process and its general components (Erasmus et al., 2001).
Customer journey research mostly has a practical approach, with its main argument being that actual customer behavior in most cases is not linear, as is portrayed in decision-making models (Wolny & Charoensuksai, 2014). Kolsky (2016) states that instead each consumer has their own personalized journey, and that the purchase journey of a consumer from the process of need recognition to purchase cannot simply be illustrated as a universal and linear funnel (Court et al., 2009). Nowadays, customers change their paths regularly since the barriers to accessing information have nearly disappeared (Kolsky, 2016) and consumers interact with firms in multiple channels leading to more complex customer journeys (Lemon & Verhoef, 2016).
In order to enable an optimal resource allocation, to measure channel attribution and ultimately to manage the customer experience over different channels simultaneously in an omnichannel context, it is important to capture the actual customer journey in full detail (Wolny & Charoensuksai, 2014). Additionally, when firms understand the sequence of a journey from a customer’s point of view, it can help them to define the role each channel plays and to illustrate the different journeys consumers can undergo when navigating through the multiple channels available (ibid). Wolny & Charoensuksai (2014) argue that even customer journeys with similar end decisions can result in different customer experiences, pointing out the need to study the journeys on an individual level. Furthermore, customer journeys take emotional and behavioral drives into account while decision-making models assume rational consumer choices according to cognitive abilities (ibid). Thereby, customer experience is also considered in customer journey studies. In conclusion, it can be said that “mapping the relationship between the customer journey and company strategy process provides a coherent structure for the business model” (Norton & Pine, 2013, p. 14).
Despite the fact that every customer has their own purchase journey (Kolsky, 2016), researchers have attempted to identify similarities and to create different customer journey typologies. Using an inductive research approach, Wolny & Charoensuksai (2014) categorized customer journeys into impulsive, balanced and considered. With impulsive journeys, less time is spent on the search for information and the consumer relies on previous experiences while making rapid decisions. These journeys occur when making repeat purchases, e.g. in a grocery store, relying on simplifying heuristics (Valenzuela & Raghubir, 2015; Spanjaard & Freeman, 2012). Balanced customer journeys occur due to the influence of reference or aspirational groups, for example friends, celebrities or bloggers. They contain an extended information search as well as an evaluation on different channels and platforms (Wolny & Charoensuksai, 2014). The considered journey involves extensive information search and evaluation. There are no real buying intentions but the information is used when purchasing the product becomes relevant at a later point of time (ibid). However, more research is needed to identify the reasons why these differences in the journeys occur.
Overall, the customer journey concept has developed from decision-making models, aiming to include the rationale of customer experience to capture how consumers act during the purchase process in a more detailed way. This proposes also taking a process-oriented view when engaging in CJA. Thereby, the EKB model is used to place consumer decisions on a broad timeline, which is why in the following (chapter 3.3.2) the model will be used to study the phases and touchpoints of the customer journey in omnichannel retail in detail.
Having outlined the concept of the customer journey in chapter 3.3.1, it is of importance to study the phases and touchpoints within omnichannel customer journeys in more detail in order to investigate the requirements a CJMAT needs to meet. As explained above, customer journey theories still make use of the EKB model when conceptualizing the customer experience during the purchase process. Hence, Lemon & Verhoef (2016) propose a process model for the customer journey including the stages pre-purchase, purchase and post-purchase, to make the analysis of the customer journey more feasible. In the process model, they also highlight customer behaviors during the different phases and touchpoints. Figure 5 depicts the process model and in the following the different elements of the customer journey process model will be discussed. 1
As the graphic above shows, the customer journey is an ongoing process, with its different stages being constantly influenced by each other. Thereby, the current cus-
1 Note that the aim of this overview is not to provide an exhaustive literature review here, since this is not the focus of this study. There has been an extensive amount of research on customer experience and service design focusing on the three different stages of the customer journey. The aim of this thesis is to point out characteristics and developments that bear the need for systematic CJA, hence for detecting the requirements such an analysis tool needs to meet. For a more exhaustive literature review it can be referred to Hoyer & MacInnis (2007).
tomer experience at point t in time is shaped by previous experiences (t-1) and previous experiences as well as current experiences shape the future experience of a customer at point t+1.
In the pre-purchase phase, the first stage of the customer journey, all aspects of the interaction between a customer and a brand, category and environment before a purchase transaction takes places, are considered. Related behaviors have been characterized by traditional marketing literature as need recognition and building awareness, consideration and thereby search. This stage of the purchase process could, in theory, include all customer experiences before a purchase. However, this stage practically deals with the customer experience from the beginning of the recognition of a need, an impulse or a goal and goes over to consideration and evaluation of satisfying this need, impulse or goal by conducting the purchase (Pieters et al., 1995; Lemon & Verhoef, 2016). Over the years, the factors influencing need recognition have changed. In earlier days, it was described as form of unrecognized want or need that is triggered by advertisement or promotion (Bruner & Pamazal, 1988; Solomon et al., 2006). However, nowadays Ashman et al. (2015) argue that consumers are additionally affected by the interaction with peers, magazines, catalogues, store displays and online content. Every day, consumers are exposed to multiple promotional messages leading to the fact that traditional advertising is not as effective on its own anymore (Marshall, 2015; Saxon, 2017). Additionally, the way consumers seek for information regarding their purchase has changed. Traditionally, consumers searched and evaluated their alternatives in a retail store, at the same time (Ashman et al., 2015). Nowadays, consumers are inspired by social media services, search engines and blogs. They share experiences with a product or a brand through product descriptions, recommendations, reviews and ratings and they listen to more informal voices in the online environment (Park & Cho, 2012; Wolny & Mueller, 2013). Especially in an omnichannel environment, consumers show a tendency to simultaneously search for information on different devices (Ashman et al., 2015). At the end of the information search, consumers narrow down the purchasing options to several considered alternatives, which are called the consideration set and the evaluation of the alternatives starts (ibid). Consumers today use different online tools such as comparison sites, polls and reviews to narrow down their consideration set (Breugelmans et al., 2012). It can be said that consumers’ needs have remained rather stable, but new technologies lead to the fact that consumers recognize a higher quantity of needs and have new ways of satisfying them via the increased amount of information available. Due to these changes, it is important for firms to understand how customers’ needs are initiated and through which channels consumers interact with their brands. By studying these factors, firms can be enabled to positively influence customer satisfaction. Hence, different touchpoints in the customer journey will be discussed in the next chapter after analyzing the other two stages of the customer journey.
The second stage of the customer journey is the purchase phase. It involves all customer interactions with a brand and the environment during the event of the purchase itself. The behaviors of choice, ordering and payment characterize this phase (Lemon & Verhoef, 2016). Marketing literature has studied this phase significantly, even though it is the most “temporally compressed” of the three stages (ibid). Studies have focused on how marketing activities, for example the marketing mix (Ko- tler & Keller, 2015), and atmospherics and the environment (Bitner, 1990) as well as the service environment (Berry et al., 2002) influence the purchase decision, hence the behavior “choice”. In consumer products and retail research, there is an emphasis on the shopping experience (Baker et al., 2002; Ofir & Simson, 2007). New technologies have evolved which also influence the customer journey in the purchase phase, for example by providing multiple touchpoints and ways of ordering and paying for products. Therefore, concepts such as decision satisfaction, purchase confidence and choice overload are studied (Lemon & Verhoef, 2016). Due to the developments mentioned, customers might stop searching and start completing or deferring a purchase for different reasons or at a different point in time, which has been studied in assortment research, for example by Broniarczyk et al. (1998) and Iyengar & Lepper (2000). Additionally, because of the changes within the retail environment and especially because of the arise of omnichannel retail, researchers have extended studies of the purchase stage into digital environments (Manchanda et al., 2006; Elberse, 2010).
The third and final stage of the purchase process, according to the EKB model, is called post-purchase. It encompasses all interactions a customer has with a brand and the respective environment after an actual purchase has taken place.