Improving customer metrics in a subscription based business A study of a company within the pre-packed grocery bag industry Master of Science Thesis in the Master Degree Programmes Entrepreneurship and Business Design and Management and Economics of Innovation DOUGLAS HAEGER GABRIELLA NILSBY Department of Technology Management and Economics CHALMERS UNIVERSITY OF TECHNOLOGY Gothenburg, Sweden, 2013 Report No. E2013:102 Improving customer metrics in a subscription based business A study of a company within the pre-packed grocery bag industry DOUGLAS HAEGER GABRIELLA NILSBY Supervisor: Magnus Eriksson         Department of Technology Management and Economics CHALMERS UNIVERSITY OF TECHNOLOGY Gothenburg, Sweden, 2013 Report     Acknowledgement  This thesis is the result from the work performed by the authors during the spring semester of 2013. First of all, we would like to express our gratitude to all interview respondents for their time and valuable insights they have provided to this report. Furthermore, we would like to thank the employees at GU Holding for contribution with data, comments and input. Finally, we would like to express a special thanks to our supervisor Magnus Erikssson at Gothenburg University for his encouragement, guidance and support throughout the process of writing this thesis. Douglas Haeger & Gabriella Nilsby Gothenburg, June 2013     1 Abstract  This report studies the Pre‐Packed Grocery Bag (PPGB) industry and how these e‐commerce companies can make use of as well as improve customer metrics. The majority of companies in this industry use a subscription‐based business model that enables the use of two key customer metrics namely, Churn and Customer Acquisition Cost (CAC). By not actively using these metrics the companies might be missing important information and make under informed marketing investment decisions. The purpose is to describe and analyse how an e‐commerce company, with a subscription based product and limited financial resources, can in a cost efficient way improve their CAC and Churn. Furthermore, give implications about how managing CAC and Churn can support decision‐ making processes regarding marketing investments. To improve CAC the first step is to map the company’s marketing processes to see which marketing media channels that is most cost‐efficient. The most efficient channels are online marketing as they target the customers that are in the buying process and provide the customer with a direct link to the buying place. To improve Churn it is important to improve customer loyalty and one of the drivers of loyalty is relationship commitment. This is in turn divided in calculative and affective commitment. Improving these commitments will improve customer retention and hence Churn. Calculative commitment is created through switching barriers and affective commitment is created through the customer/firm interaction points and aims to develop emotional ties and direct relationships. To support the marketing investment decisions a third metric, Customer Lifetime Value (CLV), is presented. This metric is based on CAC and Churn and shows the long‐term profitability of a customer. Incorporating these metrics into a recurring revenue model a firm can forecast their future financial performance. Therefore this model can be a useful tool for decision making‐processes regarding marketing investments. The report provides evaluations of marketing media channels for e‐commerce companies and suggestions of alternative channels that generate low CAC are provided. Online media channels are recommended as most suitable for a company to improve CAC, because they are cheaper and provide better targeting than alternative media. The qualitative content analysis provides an overview of the PPGB companies means to retain their customers. It is shown that the levels of calculative commitment differs among the PPGB companies and that Linas matkasse is the most successful company in this area. Affective commitment does not seem to be primary focus in their retention strategies and suggestions for measures to create this commitment is provided. Finally, the recurring revenue model is suggested as a good way to improve the reliability of marketing investments decisions, since it provides a good understanding of how the customer metrics are affected by these decisions. 2 Table of content 1. Introduction ..................................................................................................................... 4 1.1. Background ........................................................................................................................... 4 1.2. Problem formulation ......................................................................................................... 5 1.3. Purpose ................................................................................................................................... 7 1.4. Research questions ............................................................................................................. 8 1.5. Report structure .................................................................................................................. 8 1.6. Limitation ............................................................................................................................... 9 2. Method ............................................................................................................................ 10 2.1. Research methods ............................................................................................................. 10 2.1.1. First research question ......................................................................................................... 10 2.1.2. Second research question .................................................................................................... 10 2.1.3. Third research question ....................................................................................................... 11 2.2. Research procedure ......................................................................................................... 12 2.3. Ensuring quality ................................................................................................................. 14 3. Theoretical framework ............................................................................................. 16 3.1. Two key customer metrics ............................................................................................. 16 3.2. CAC .......................................................................................................................................... 17 3.3. Churn...................................................................................................................................... 19 3.4. Financial performance and decision‐making processes ..................................... 24 3.5. Summary of theoretical findings ................................................................................. 26 4. Empirics .......................................................................................................................... 28 4.1. CAC .......................................................................................................................................... 28 4.1.1. Matkasse AB .............................................................................................................................. 28 4.1.2. Marketing activities with low CAC ................................................................................... 31 4.2. Churn...................................................................................................................................... 35 4.2.1. Linas matkasse ......................................................................................................................... 36 4.2.2. Middagsfrid ................................................................................................................................ 37 4.2.3. Mathem.se .................................................................................................................................. 38 4.2.4. Matkasse AB .............................................................................................................................. 40 4.2.5. Members.com ........................................................................................................................... 41 4.2.6. Milkandhoney.se...................................................................................................................... 43 4.3. Recurring revenue model to support decision‐making processes .................. 43 4.3.1. Testing the Recurring Revenue Model: An example of Chef & Friend .............. 44 5. Analysis ........................................................................................................................... 48 5.1. CAC .......................................................................................................................................... 48 5.2. Churn...................................................................................................................................... 51 5.3. Recurring revenue model ............................................................................................... 55 6. Discussion ...................................................................................................................... 57 6.1. CAC .......................................................................................................................................... 57 6.2. Churn...................................................................................................................................... 62 6.3. Recurring revenue model ............................................................................................... 67 7. Conclusion ..................................................................................................................... 69 3 7.1. Suggestions for further research ................................................................................. 71 8. List of References ........................................................................................................ 72   4 1. Introduction  This section should provide an understanding of why this report has been written. First the background is presented to create a setting for the problem formulation. The problem formulation is followed by the purpose of the report and finally the research questions are presented. Concepts that may need to be clarified are marked with an asterisk (*) and the definition can be found in Appendix I. 1.1. Background  In this report the definition of a pre‐packed grocery bag (PPGB) is a pre‐ composed bag of groceries with matching recipes, sufficient for a certain number of dinners. Furthermore, the grocery bag is ordered online, from an e‐commerce* company in the PPGB industry, by entering a subscription for each or every other week and gets delivered to the customer’s home. The industry of pre‐packed grocery bags was founded around 2007 when Middagsfrid began their business. The industry is in an early stage of its lifecycle and the competitors are swarming the market. Since 2007 the business of pre‐ packed grocery bags has increased in popularity, especially in the larger cities. 2011 the product became Christmas present of the year, which is recognition of its popularity by the Swedish consumers (Svensk Handel, 2011). The concept was founded in Sweden and is now spreading across Europe and the U.S (Moskin, 2013). The online commerce of fast‐moving consumer goods (FMCG) represents circa 350 MSEK (0,5%) of the total and is expected to grow in the future. Mainly due to that younger generations are more comfortable and familiar with the online market (Svensk Handel, 2011). Matkasse AB was started in 2010. Today the company provides pre‐packed grocery bags to circa 150 subscribers in the Gothenburg area. The niche is to provide pre‐cooked sauces with the bags that Gustav Svensson has prepared by himself. Matkasse AB offer four different subscription products, either two or four meals for four days a week to be provided each week or every two weeks. The customers are able to end their subscription or pause it for any preferred amount of time without being charged additional fees. (Svensson, 2013) The most common business model in this industry is a subscription based business model. To further elaborate on how a subscription based business model is defined the term business model should be outlined. There are many authors who have tried this and several different models can be identified. Rappa (2000) identifies 29 different types of models, which he orders in nine categories. This demonstrates the diversity and complexity that can be * Find definition in Appendix 1 5 associated with this terminology. In this report the definition will rely on Osterwalder (2009, p. 14), namely “A business model describes the rationale of how an organization creates, delivers and captures value”. Osterwalder (2009) further presents that the business model contains nine building blocks; Customer segments, Value propositions, Channels, Customer Relationships, Revenue streams, Key resources, Key activities, Key partnerships and Cost structure. A subscription based business model is mainly distinguished in the fifth building block, Revenue streams. This part should state how the company will generate revenue from the customers and can be divided into two different parts; Transaction revenues that are generated from a one‐time purchase and Recurring revenues* that are generated from either repeated delivery of the product or post‐purchase customer support (Osterwalder, 2009). The subscription based business model is within recurring revenues and is generated by selling periodic use or access to products or services. The customer is entering into a formal agreement with the vendor for a period of time, which creates recurring revenues. This results in a reduced uncertainty due to that the vendor knows its current active members. It furthermore assures constant revenues during the customer’s subscription. The direct downside is that one might loose customers that only would like to pay a one‐time fee (Sundelin, 2008). For a company with a subscription based business model two key customer metrics are commonly used to measure the company’s performance. These metrics are Churn and Customer Acquisition Cost (CAC) and they are related to the company’s customer base* and sales. Churn is a metric that relates to the retention of customers and measures how many customers that leaves the firm during a specific period of time. CAC, as the name indicates, measures how much it costs for the firm to get a new customer. A company’s decisions affect these metrics and similarly these metrics affects the company’s decision‐making processes. Therefore, understanding and actively using the metrics is critical for the company’s managerial decisions and the performance of the firm. (Gupta et al, 2004) 1.2. Problem formulation  In the early phase for a company it is highly prioritized to increase the customer base as without customers, the firm has neither revenues nor profits. To attract new customers the company need to invest in marketing activities with the objective to create a good customer‐firm relationship. However, it is a challenge * Find definition in Appendix I 6 to understand which customers to acquire and how to retain them; hence, it is a challenge to manage the allocation of marketing expenses* and be able to justify these decisions. (Gupta et al, 2004) The fundamental goal of a firm is to increase profits, not just to obtain new customers, and therefore using effective marketing activities, meaning higher generated value per invested dollar, is critical. (Dacey, 1997) Understanding the link between customer‐firm relationship and financial performance* will support firms’ investment decisions. This link can be described as what a firm can do (marketing activities), what customers do, and how this affects a firm’s financial performance. Marketing activities and how customers respond can be understood by analysing key customer metrics. (Gupta et al, 2004) Customer metrics can be categorized as observable and unobservable metrics, where observable relates to purchase or consumption of a product or service and unobservable relates to customer perception, attitude and behavioural intentions. Economists mean that unobserved constructs are stated preferences while observed are revealed preferences. Unobserved metrics relate to observed metrics, which are directly linked to financial gains. For instance, if a customer’s perception of a product is positive it will most likely lead to a purchase, which translates into increased sales and profits. (Reinartz et al., 2005) As it is the observable metrics that generate financial improvement to a company these are the metrics that we will focus on in this report. Key observable metrics are Churn and Customer Acquisition Cost. (Gupta et al, 2004) * Find definition in Appendix I 7 Figure 1 Illustration of relationship between the firm, customers and the firm's financial performance. Compiled by the authors. CAC and Churn are metrics that measures a company’s financial performance and together they constitute a good benchmark to how well the company is performing against its competitors. There are furthermore different ways for the company to affect the metrics in order to enhance the financial performance; therefore, understanding the role of these metrics and how they can be improved becomes critical. The relationship between the firm, customers and the firm’s financial performance are illustrated in Figure 1. (Gupta et al, 2004) As Matkasse AB provides a subscription‐based product, managing CAC and Churn is important, but is currently not a part of their decision making processes regarding marketing activities. By not actively using these metrics it can be that they are missing important information and making under informed marketing decisions. Introducing these metrics to the company will be beneficial for the managerial decision‐making processes. Matkasse AB don’t know if they are over or under investing in customer acquisition or retention or whether these investments are effective. They need new customers to improve their financial performance and strengthen their position on the PPGB market. 1.3. Purpose  Describe and analyse how an e‐commerce company, with a subscription based product and limited financial resources, can in a cost efficient way improve their CAC and Churn. Furthermore, give implications about how managing CAC and Churn can support decision‐making processes regarding marketing investments. 8 1.4. Research questions  1. Which marketing media channels*, suitable for an e‐commerce company, with a subscription product and limited financial resources, can lower CAC? Studying the literature will provide an understanding of the meaning of CAC, why it is important and how companies theoretically can affect the metric. This will provide a framework that can be used as a basis for analysis. Next step will be to find marketing media channels that are cost efficient and suitable for a company in the mentioned situation by talking to an expert within the field. Furthermore, the authors will evaluate marketing media channels already used by a PPGB company and assess their impact on customer acquisition. Combining these empirical data will provide a comprehensive understanding of which marketing media channels that should be used in order to lower CAC. 2. How can an e‐commerce company, with a subscription product and limited financial resources, take measures to retain their existing customers, i.e. lower Churn? Studying the literature to understand why Churn is an important metric and how to improve it will provide a framework that can be used as a basis for analysis. By researching four PPGB company’s websites it can be shown how they take measures to retain customers. Interviewing an e‐commerce company from another line of business will ad insights and make up for good analysis and discussion regarding how to lower Churn. 3. How can a company increase reliability of marketing investment decisions by understanding the impact of CAC and Churn on the financial performance? Based on obtained knowledge from the theoretical framework and insights from interview with Investment manager, build a forecasting model to see future financial results from marketing investments in marketing activities with the purpose of improving CAC and Churn. The model will then be tested in a fictitious case to demonstrate how the model can be utilized as a tool for supporting marketing decisions. 1.5. Report structure  The next chapter, Method, describes the procedure to achieve the purpose of the report, i.e. how the work has been structured and which research methods that have been used. The methods validity, i.e. how far it measures what it is intended to measure, and reliability of the measurements, will be discussed. 9 Following chapter, Theoretical framework, is a review of the theories that are considered relevant to the report. These theories will create theoretical frameworks, which together serves as a starting point for the empirical and analytical chapters. The empirical chapter will present the empirical findings from the reality the study aims to investigate. This chapter is followed by the Analysis, where the theoretical and empirical findings are connected and analysed. In the next chapter Discussion, the research questions are answered sequentially by discussing previous analysis. Finally, Conclusions of the report will be presented. 1.6. Limitation  The thesis will focus on improving the two metrics CAC and Churn, as well as, describe how they can be managed to support decision‐making processes regarding marketing investments. This focus in chosen as they main priority at this stage is to increase their customer base, hence, investing in valuable marketing activities is relevant. CAC and Churn are observable metrics that are directly linked to a purchase of the product; therefore, these metrics have an impact on the financial performance of the company. As unobservable describes the perception, attitude and behavioural intentions of the customers, such metrics cannot be utilized as measurement of financial performance. Therefore, unobservable metrics will not be considered in this report. Unobservable metrics relates to the concept of the PPGB; for instance, the products offered by the PPGB company. As unobservable metrics are outside the scope of this report, Matkasse AB’s concept of the PPGB will not be considered. The reason for this limitation is that Matkasse AB already have taken several measures to improve the concept. When determining the impact on the firm’s financial performance the authors will use the firm’s profits as a measure of the financial performance. The product price is a critical financial parameter for the company and affects the company’s profits. However, this parameter will not be considered in the thesis as Matkasse AB decides on the price level of their products. The thesis will limit the geographical aspect to the Gothenburg region, as this is where the analysed company Matkasse AB is active. 10 2. Method  In this chapter of the report it is described how the work has been structured and which research methods that have been used. The chapter is divided in three parts, namely Research methods, Research procedure and Ensuring quality. The first section is further divided according to the three research questions and brings up the different research methods that have been applied. The second section contains a description of the research process and how this was built. The third and final section is a discussion where different aspects of the quality of the report are brought up in terms of credibility, transferability and dependability. 2.1. Research methods  The research methods in this report is of a qualitative manner and consists of two main methods for the collection and generation of data, namely semi‐ structured interviews and a qualitative content analysis. The following sections describe the means taken to answer each research question. 2.1.1. First research question  This part of the study aimed to investigate which marketing channels that were suitable for an e‐commerce company, with a subscription based product and limited resources, to lower CAC. This was done by studying what marketing channels are most useful to attain a low CAC and what measures that can be taken. Next step was to map the marketing channels already used by a PPGB company and assess the impact on CAC. These two findings formed the basis for understanding what media channels that can be used to lower CAC. To find the most useful marketing channels to attain a low CAC a semi‐structured interview were conducted with Anders Katz campaign planner at MediaView. Teodor Abreu, Sales Representative at Groupon, was also interviewed to get more information about the “deal‐of‐the‐day” marketing media channel. To map the marketing channels of a PPGB company a semi‐structured interview with the founder and CEO of Matkasse AB, were made. More about how these samples were chosen can be found in section 2.2 Research procedure. 2.1.2. Second research question  This part of the study aimed to investigate how the PPGB industry engages in activities to increase loyalty*, but more specific, how one can engage in activities to build relationship commitment with customers. The authors also examined how other e‐commerce companies have engaged in activities to increase loyalty and compared with the previous findings from the PPGB industry. By studying * Find definition in Appendix I 11 how the PPGB industry does this and comparing with other e‐commerce companies the authors hoped to point out improvement potential. To find how the PPGB industry retains their customers a qualitative content analysis was used by studying documents as sources of data. More specifically, virtual documents in the form of company webpages were studied to find information about how these companies engage in activities to build loyalty with their customers. Four different companies within this industry were studied. The three biggest actors, Linas matkasse, Middagsfrid and mathem.se, together with the Gothenburg based PPGB company, Matkasse AB, constituted the sample for the content analysis. More about how these samples were chosen can be found in section 2.2 Research procedure. The aim of the content analysis was to analyse switching barriers that the customers within the PPGB industry were subjected to and hence related to the second research question. The aim of the content analysis was set during the literature study of the second research question related to customer retention. The literature study showed that switching barriers were partly the improvement focus when trying to build customer loyalty and hence increase retention. Since all PPGB companies are e‐commerce businesses studying their websites gave the authors’ access to analyse the exact same situation that the customers in this industry had and hence the switching barriers subjected to them. To fulfil the aim of the content analysis a coding scheme was set up and structured according to the drivers of loyalty and switching barriers highlighted in chapter 3, section 3.2 Churn, all in accordance to Bryman & Bell (2011). This coding scheme were then used to conduct the content analysis and also served as a basis when structuring section 4.2. Churn. To find how other e‐commerce companies are improving loyalty two semi‐ structured interviews were conducted with two different companies. One with Albin Johansson, Product Manager at members.com, and one with Max Svärdh, CEO at milkandhoney.se. More about how these samples were chosen can be found in section 2.2 Research procedure. 2.1.3. Third research question  This part of the study aimed to investigate how a company can increase the understanding of CAC and Churns impact on the financial performance and how this can be linked to decision‐making processes regarding marketing investments. The initial step was to obtain an understanding of how to improve a firm’s profits by managing CAC and Churn. The next step was to build a tool in Excel that can demonstrate how the company’s profits are affected by improving CAC and Churn. 12 To obtain an understanding of how to improve a firm’s profits, the literature was studied. A semi‐structured interview with Gregory Carson, Investment manager at GU Holding, was also conducted. Carson has experience in forecasting recurring revenues and gave insights in how to build a forecasting model, suitable for a business with a subscription based business model. Based on the obtained knowledge from the theoretical framework and insights from Investment manager, a forecasting model that can be used as a tool for understanding the impact from CAC and Churn on a firm’s profits were built. The model was then tested with a fictitious case to demonstrate how the model can support marketing investment decisions. A fictional company called Chef & Friend is presented with a certain setting. Five scenarios are then demonstrated where CAC and Churn are altered to demonstrate how CLV is affected. Scenario 1, with no alterations of the metrics, Scenario 2, with a CAC decrease of 10 %, Scenario 3, with a Churn decrease of 10 %, Scenario 4, with a CAC decrease of 20 % and Scenario 5, with a Churn decrease of 20 %. These scenarios were chosen to give an understanding of how CAC and Churn relate to each other, how they relate to CLV and how this model can be used as a tool for evaluating marketing decisions. The different alterations of the metrics will affect CLV in different ways and hence provide the user of the model with information about the profitability, in terms of CLV, in each decision. These tests and the different scenarios are presented in section 4.3.1. Testing the Recurring Revenue Model: An example of Chef & Friend. 2.2. Research procedure  Some parts of the research procedures have been brought up in the previous section and this section aim to further elaborate on this subject. The purpose of the report has been formulated with the purpose of fulfilling the requests from Matkasse AB. The initial problem identified by the constituent was reformulated together with the authors, academic contacts and business advisors at Matkasse AB. This reformulation was also based on a substantial literature study conducted by the authors and resulted in the problem identification. This process resulted in the formulation of the purpose and the research questions, which has been the foundation and scope of the report. The theory has been studied in two different phases with two different purposes. In the initial phase the focus has been to get familiar with the problem and get a better understanding of the background to the issues. Next step was to divide the research according to the different research questions and provide theoretical frameworks to assist with structuring of the empirics and fulfilling the purpose of the report. Each research question served as the basis for this progress and the literature study has been organized accordingly. The sections in the literature chapter all results in a framework that served as a basis for the 13 analysis chapter (Chapter 5) described further down. The use of multiple sources to create the frameworks has prevented uncertain opinions from single authors. The empirics have been based on a series of semi‐structured interviews and a qualitative content analysis where virtual documents in the form of webpages were studied. The empiric’s chapter (Chapter 4) has been structured in accordance to the three research questions used in this report. The interviewees for the semi‐structured interviews served as the empirical basis for all research questions and were chosen as follows. Anders Katz, campaign planner at MediaView, was chosen because of his experience in the marketing business. He could provide the authors with the relevant information about marketing media channels that were needed to answer the first research question about customer acquisition. After this interview had been conducted the authors also tried to discuss the subject of customer retention (third research question). Anders Katz recommended, Max Svärdh and Albin Johansson as good references when studying customer retention since they both were actively working with this subject at their e‐commerce businesses. The interviewee Teodor Abreu was also contacted after recommendation from Anders Katz. Hence, these interviewees (Svärdh, Johansson and Abreu) were sampled through snowballing, in accordance to Bryman & Bell (2011). These interviews provided the authors with relevant information about creating loyalty with customers and hence relating to the second research question. The interview object Gregory Carson was chosen because of his involvement as business advisor to Matkasse AB, but also for his background in finance and experience from recurring revenue models. This interview provided the authors with relevant information about how to create a recurring revenue model. The sample chosen for the content analysis were the four PPGB companies Matkasse AB, Middagsfrid, Mathem.se and Linas matkasse. Matkasse AB was chosen because of their role as constituents for the subject of this report. Middagsfrid, Mathem.se and Linas matkasse were chosen because of their significant sizes in terms of revenue in the PPGB industry. All these three are the sole biggest actors on this market in terms of revenue. With this sample the authors’ got access to study the best practice within customer retention in the PPGB industry. More discussion about how the quality of the research methods used in this report has been ensured is presented in the next section of this chapter (2.3. Ensuring quality). The analysis has been based on the findings from the empirics and compared with the theoretical frameworks from the literature study. Each section in the analysis chapter has been structured according to the three research questions used in this report. Since chapter 3 Theoretical framework and chapter 4 Empirics has been structured the same way, the analysis were brought together by 14 combining the theoretical frameworks from each research question with the empirical findings from each research question. The theoretical findings has furthermore served as a basis for the process of generating the empirical data, hence the analysis were conducted in a structured manner providing a rich evaluation of the different results presented in the report. To ensure an analysis of high quality the authors’ own views, ideas and other input has been presented in the discussion. This was important to provide an objective and correct analysis comparing the theoretical findings with the empirical findings. The discussion has also been structured to fulfil the purpose and answer the research questions sequentially. The conclusions have been based on all progress from the theory, empirics, analysis and discussions. 2.3. Ensuring quality  This report has solely used qualitative research methods, such as semi‐ structured interviews and qualitative content analysis for the collection and generation of data. Bryman & Bell (2011) states that the criteria of reliability and validity are an evident quality issue for most research, but especially for quantitative approaches. The grounding of these criteria is based on quantitative research. Therefore some authors argue that validity and reliability are inapplicable or inappropriate for qualitative research. Instead other criteria should apply to qualitative research, such as credibility in parallel with internal validity (how believable are the findings), transferability in parallel with external validity (do the findings apply to other contexts) and dependability in parallel with reliability (are the findings likely to apply at other times) (Bryman & Bell, 2011). Bryman & Bell (2011) states that qualitative content analysis probably is the most widespread approach to the qualitative analysis of documents. They further argue that to ensure the credibility and dependability of the findings the extracted themes found by the researchers should be illustrated by quotations or pictures. In this report pictures of the findings are found in Appendix II. Bryman & Bell (2011) further states that this is especially important studying virtual documents since websites continuously are updated and changed. The interviews have mainly been conducted over the phone but also in person and all of them have been carried out in a semi‐structured approach in accordance to the description of Bryman & Bell (2011). Interview guides (see Appendix IV) were prepared and the questions were limited to only a few to ensure that the respondent could answer freely. The interview guide with questions was sent to the respondent three days in advance were a description the purpose of the interview also was included. This was done to ensure that the respondent was prepared to answer the questions correctly. The questions were not asked in any specific order but adapted to the answers given from the 15 interviewee. This was done to get the conversation flowing. After the interviews the interviewee was able to see the transcripts to ensure that the answers was in line with her or his actual opinion. All these measures were taken to ensure the credibility of the interviews in accordance with Bryman & Bell (2011). The credibility has been further enhanced by triangulation of multiple sources of data from the research methods used in the report and the studied literature. As qualitative research involves the deeper study of a smaller group qualitative research tend to be oriented towards a specific setting and context (Bryman & Bell, 2011). These arguments are applicable to this report as well and the findings presented should not be viewed as conclusions that can be generalized but are rather solely applicable to the PPGB industry and the companies included in the study. Hence, the transferability could be viewed as rather weak. The dependability criteria involve the documentation of the research process and that records are kept during all stages (Bryman & Bell, 2011). In this report all interview templates, transcripts of the interviews as well as the findings from the qualitative content analysis has been saved. The problem identification is specified and the selection of research participants are documented and presented. Hence, the dependability of the research has been considered.   16 3. Theoretical framework  In this chapter of the report the findings from the literature will be presented. CAC and Churn will be presented in two different sections where these metrics will be defined and their role as customer metrics will be described. Furthermore, theoretical suggestions about how to improve CAC and Churn will be provided. This will be followed by a section about financial performance and decision‐making processes. Here is Customer Lifetime Value (CLV) introduced, a metric that is based on both CAC and Churn and is a key indicator of a firm’s financial performance. 3.1. Two key customer metrics  For a company with a subscription based business model two key customer metrics are commonly used to measure the company’s performance. These metrics are Customer Acquisition Cost (CAC) and Churn and they are related to the company’s customer base and sales, as illustrated in Figure 2. CAC measures how much it costs for the firm to get a new customer. Churn is a metric that relates to the retention of customers and measures how many customers that leaves the firm during a specific period of time. Improving these metrics will increase the customer base, and hence sales. To attract new customers the company need to invest in marketing activities that generate many customers to a low cost and to retain existing customers the company need to invest in marketing activities that prevent customers from cancelling the subscription. (Gupta et al, 2004) Figure 2 CAC and Churn in relation to the customer base and sales. Compiled by the authors. A company’s decisions affect these metrics and similarly these metrics affects the company’s decision‐making processes. Deciding to invest in a specific marketing activity will affect the metrics and if the metrics are aimed to be a specific value, this will affect the decision‐making processes regarding marketing investments. Therefore, understanding and actively using the metrics is critical for the company’s managerial decisions and the performance of the firm. (Gupta et al, 2004) 17 3.2. CAC  CAC stands for Customer Acquisition Cost and measures the cost to get a new customer. Many different terms are used to describe this metric, such as Cost Per Action (CPA) and Cost Per Acquired Account (CPAA); however, in this report we have chosen the term CAC. To compute CAC the entire cost of sales and marketing over a given period is divided by the number of customers acquired during that period. CAC should, as all other costs, be kept low and weighed against the generated value. In a subscription based business this metrics must not be higher than the ability to monetize those customers, meaning the total value the customer will generate to the company. (Skok, 2009) CAC could further be seen as a means of controlling marketing expenses and ensure that the firm is not spending money on terms that are not driving the business (Larry, 2013). As CAC is a cost for the firm, successful investments in this metric will save future expenses and therefor improve the firm’s financial performance (Gupta & Zeithaml, 2004). With this said a firm must understand how to allocate expenses in order to drive down the CAC. According to Zohn (2003), mapping out the company’s marketing processes and ensure that the most cost‐effective resources are used is a first step to lower CAC. Most companies will achieve a more low cost structure only by shifting from high‐cost voice and print to low‐cost online communication. Following step should be to examine each marketing activity* from acquisition to retention to determine where equally effective but less expensive resource can replace the existing one. Marketing activities, in this sense, relates to marketing media channels for reaching out to targeted customer, which are the drivers of CAC (Zohn, 2003). How to allocate marketing expenses across different media channels is a classic problem in today’s multichannel environment. In order to understand this problem, according to Reinartz et al. (2005) the individual customer should be the measured objective, because persuasion is created at the individual level and because media increasingly can be targeted at the individual level. Furthermore, marketing media channels can be characterized as more or less interpersonal, where personal selling could be argued to be the most interpersonal since it offer the ability for message customization and allows for personal relationship building. Prior studies have shown that more interpersonal channels will have a greater positive impact on customer acquisition as well as customer retention. (Gupta & Zeithaml, 2004) 18 Marketing media channels can be divided into three categorize: paid, owned, and earned media, as illustrated in Figure 3. Paid media includes traditional advertising, such as print, television, direct mail. Owned media consists of properties or channels owned by the company, such as company webpage, Facebook fan page*, mobile apps. The third category, earned media, are generated when the quality or uniqueness of a company’s product makes customers promote the company at no cost through external or owned media; for instance, word of mouth*, blogs, Twitter* (Edelman et al., 2010). Each media offers distinctive advantages. Paid media immediately reaches out to a broad audience. An owned media gives the company concrete ways to engage customers. And earned media gives the company insight in how customers perceive them as well as their competitors (Smith, 2010). Furthermore, paid and owned media are cost related channels, while earned media is produced by consumes at no cost. To create paid media the marketing company* invest in advertising creation, placements, purchasing logistics and management, while to create owned media, investments are related to content, technology, response resources, and management. (Edelman et al., 2010). The choice of marketing media channel can have a huge impact on the outcome of a marketing activity, as marketing media channels have varying cost efficiency, different target groups and can create different effects of branding or promotion (Grönlund et al., 2010). Owned and earned media is becoming more and more commonly used because of the emergence of social media*. Owned is often less costly than traditional paid media and, as already mentioned, earned is produced at no cost; therefore, these channels well suitable alternatives for lowering CAC (Sussin et al., 2013). According to Smith (2010) the best way to work with marketing media channels is to use a mix of the three, where each channel supports the others (Smith, 2010). This positive effect is more and more achievable since the three media channels increasingly interact with each other, mainly due to social media. Paid media can serve as feeders into owned media, where companies can offer a more engaging experience and a more targeted stream of contact with customers. Furthermore, companies can, through both paid and owned media, encourage customer’s to e.g. “share” things they like and hence produce earned media. This fan promotion is sometimes spontaneous, however, it is often encourage by giving the customers an attractive offer, such as rewarding for acquiring another customer through recommendation. Thus, paid and owned media efforts can generate earned media to the company. (Edelman et al., 2010) * Find definition in Appendix I 19 Figure 3 Categorization of marketing media channels. Compiled by the authors. Online marketing targets consumers that are “in the moment”. Consumers who respond to an ad are already interested in the product and therefore are considered to be in the buying process. This is beneficial as those consumers have high potential of purchasing a product. (Baker, 2005) Paid online marketing dramatically improves the chance that potential customers find the company’s website. (Heaphy, 2011). When the customer is subjected to an ad on the Internet, that ad can lead the customer directly to the buying place, where the customer can order the product. This increases the likelihood for an actual purchase. The direct link between the advertisement and the buying place can only be achieved on the Internet. (Sauter, 2009) 3.3. Churn  Churn, sometimes also referred to as attrition, is the metric for measuring the rate of customers leaving their subscription. Hence, it is in a negative way directly correlated to the customer base and is usually calculated monthly by dividing the number of defecting customers that month by the number of total customers. The strategy for this metric is of course to keep it as low as possible to be able to retain the customer base and a stable recurring revenue stream (Wileman, 1999). Another term that is of importance when trying to understand Churn is retention. This is the term for preservation of customers and hence, Churn is a measure of the retention rate or the customer loyalty (Livne et al., 2011). Having a low rate of defecting customers is good proof that the customers are loyal to the company and the products and services they are offered regardless of how 20 small the customer base may be. With a new product on the market it is important to keep track of the customer loyalty. If the Churn rate starts to increase it is most likely that the customers are taking their money elsewhere (Gustavsson et al., 2005). Having a small customer base it would be more reasonable to take actions to increase the customer base rather to retain it. In absolute terms the cost of having a Churn rate of two percent is of course less for a smaller customer base than for a bigger one. Although managing the customer Churn would make the growth of new customers more stable since fewer customers are leaving their subscriptions. Furthermore, many studies have shown that the cost of retaining customers is less than acquiring new ones (Claude, 2000)(Huson, 1996). Claude (2000) states that it could be as much as 10 times more expensive. To understand what measures that can be made to affect the Churn rate and keep it as low as possible it is important to understand the drivers of this metric. The retention of customers is depending on their loyalty towards the products or brand and the drivers of customer loyalty can be divided into two big segments, namely customer satisfaction and relationship commitment. The former are defined as the overall evaluation of the products or services built up over time and the latter can in turn be divided into two smaller parts, namely affective commitment and calculative commitment. Calculative commitment is the more rational one, a sort of economic‐based commitment, created by lack of other choices or different kinds of switching barriers. Affective commitment is the softer one that develops through the amount of exchange and personal involvement the customer has with the firm, which develops a higher level of trust and commitment in the relationship. These forms of commitment makes the customers stay with the company and repurchase products or services even when the satisfaction is low (Gustavsson et al., 2005). One important difference between the two drivers is that customer satisfaction is more backward looking and relationship commitment is more forward looking. Customer satisfaction is built up over time and measures the relationship up to date, whereas relationship commitment is a measure of the strength of the relationship and the commitment to proceed further (Gustavsson et al., 2005). 21 Figure 4 Demonstration of drivers of retention/loyalty and improvement focus to affect each driver. Compiled by the authors. To lower Churn and increase retention there are different activities to pursue depending on what drives loyalty. If customer satisfaction is the driver there should be measures taken to improve product or service quality or offer a better price. If however relational commitment is the driver, focus should be spent on building more direct relationships to your customers or creating switching barriers (Gustavsson et al. 2005). Figure 4 demonstrates the drivers of loyalty and the improvement focus to affect each driver. Menon & O’Connor (2007) argues that affective and calculative commitment are fundamentally linked. Since affective commitment is created and fostered during episodes of interpersonal interactions in the relationship between the customers and the firm it is not possible to have affective commitment without first ensuring some extent of calculative commitment. Therefore the presence of calculative commitment should be viewed as a platform on which to build affective commitment. Besides that calculative commitment is a prerequisite for affective commitment, it is also argued that the presence of affective commitment will increase customer calculative commitment (Menon & O’Connor, 2007). Effective interpersonal interactions are driven by the assertiveness and affiliation that is demonstrated during the interactions in the relationship between the customers and the firm. Hence, to build affective commitment with the customers it is necessary to look at the interaction points (i.e. direct relationships). These interpersonal communication channels must therefore be the focus when trying to improve the affective commitment with customers (Menon & O’Connor, 2007). 22 Calculative commitment is created and driven by the customers’ perception of high (contractual or financial) switching barriers, customer switching inertia or the lack of alternative choices on the market. Hence, it is the customers’ inconvenience or difficulty to switch that makes calculative commitment occur and to manage calculative commitment a firm must build switching barriers (Menon & O’Connor, 2007). When trying to improve Churn it is important to consider where the firm operates its business. For companies with an Internet based business the importance of switching barriers become even more evident and understanding the electronic environment surrounding the company is a vital. In contrast to other more traditional environments the Internet decrease switching barriers and there are foremost three reasons for this. First because of its transparent nature, which is further enhanced by customers’ use of smart price agents (e.g. pricerunner.se, etc.), second, because of the absence of physical distance between customers, suppliers and competitors and third because of the lack of personal customer‐firm relationships (Wirtz & Lihotzky, 2003). Interaction with customers is limited to the firms owned media such as webpage and Facebook fan page. The lack of interpersonal interaction in web based businesses means that it is difficult to build an affective commitment in the relationship and develop emotional ties with a customer. Hence, building customer switching barriers becomes an even more important part of retention management when having an Internet based business. It is also apparent that improving customer retention is more difficult to achieve in an environment where there is great transparency and competitors is only a click away (Wirtz & Lihotzky, 2003). There are five types of switching barriers that can be derived from an Internet based firm, as illustrated in Figure 5; transaction cost, learning cost, relationship costs, convenience costs, and artificial costs (or contractual costs). Transaction costs are the costs that arise from starting a new relationship with a firm and involve loss of financially quantifiable resources, such as onetime fees e.g. deposit or initiation fee. In addition, switching may involve replacement of co‐ assets (assets consumer have bought that is compatible with the specific product). Learning costs primarily involves expenses of time and effort. These costs occur when a customer needs to familiarise and create the same comfort and facility with the new e‐store as with the old one. Once consumers become familiar with an e‐store*, they might be unwilling to change to another, particularly if they have spent time and effort to customise the site to their specific requirements. According to Balabanis et al. (2004), online shoppers do 23 stick to the same e‐store to save time and effort to avoid learning how to use a new site with new layout. Relational costs means if a customer feels emotional discomfort when switching to a new provider. This emotional discomfort can appear when a customer don’t trust the provider or feel insecure when purchasing a product. Trust may be more important online than in physical stores as customers often have to pay in advance and there are uncertainties whether or when the product will be delivered. Furthermore, customers might feel insecure when providing personal and credit card information online. An effect of risk for Internet fraud* may be that customers have to rely on a small number of trusted providers. Convenience barrier are the costs associated to the customer’s loss or lack of flexibility with alternative suppliers. For example if a customer has certain needs or are used to specific routines that other suppliers cannot or do not offer. Artificial costs are created through deliberate actions by the firm, such as loyalty programs and contractual barriers. Such actions create economic benefits for staying with a provider, as consumers may lose accumulated points, discount, or benefits that are not afforded to new customers. Hence, creating incentives for the customer to stay with the provider. Similar switching barriers can be created by contracts, when a customer signs a contract committing him‐self to be liable for the company and buy from the firm for a period of time, e.g. contractually bound to the company until a predetermined date (Pei‐Yu & Hitt, 2002) (Balabanis et al., 2004) (Burnham, 2003). Figure 5 Demonstration of switching barriers to build calculative commitment A loyalty program aims to encourage a customer to stay with a firm, through the offering of hard and soft benefits. Hard benefits are those that incur financial * Find definition in Appendix I 24 costs to the firm, for example by offering merchandise, which help build credibility and customer attention. Soft benefits should try to build a sense of “special status” with the customer by offering special information and recognition. This leads to a deeper and more long‐term loyalty compared to the hard benefits (Balabanis et al., 2004). Implementing a loyalty program is a common way to encourage customers to come back and buy more. A successful loyalty program identifies, connects with and retains the best customers. According to Mosier (1999) there are three key elements of effective loyalty programs: personalization, emotional tie, and soft and hard benefits. Personalization means that the loyalty program should have a targeted message to the best customers that appeals to their specific needs. The customers must perceive a value of the reward and be something that they feel is useful in someway. With the right offer companies can acquire customers but that does not necessarily means that the customers are loyal. If a better offer comes they might switch to another supplier if the company have not connected with the customer emotionally. Therefore, it is important to create an emotional tie to the customers by making rewards interactive, timely, and exclusive. (Mosier, 1999) Hard benefits are tangible economic rewards, while soft benefits represent the emotional and recognition aspect. Mosier (1999) argues that hard benefits attract customers to join a loyalty program, while soft benefits keep them there. Example of hard benefits can be an introduction discount, premiums for accumulated points, or reduced price for buying more products. Soft benefits, on the other hand, could be to provide a sense of exclusivity for the customer or send notice in advance for events or special promotions. The most successful loyalty programs balance these two types of benefits. (Harris, 2000) Implementing a loyalty program can be perceived as expensive for the company, however, as it is more expensive to acquire new customers than retaining existing ones, this investment have a short pay back time. (Mosier, 1999) 3.4. Financial performance and decision‐making processes  Important financial information for a business is presented in the company’s financial statement, which typically includes a balance sheet, income statement, and cash flow statement. These statements are important when determining the financial performance of a company as well as when making strategic decisions. (Hillier et al, 2010) Another tool that is important for supporting decision‐making processes is financial forecasting. It helps management track the company’s financial performance; hence, it provides a base for decision‐making processes as management can see the impact from a decision on the firm’s future financial performance. (Whitfield, 2011) 25 The term revenue means income received from sales of goods or services over a specific period of time. Revenue can be found at the top of an income statement and revenue growth is a key performance indicator for investors and shareholders. (Whitfield, 2011) A subscription based business model generates recurring revenues as customers are contracted for a period of time. Besides reduced uncertainty as the supplier knows its customers, this also improves the predictability of the future revenue streams (Sundelin, 2008). The level of predictability can be further improved with financial forecasting that track recurring revenues. According to Swanson (2008), the best forecasts are based on a good understanding of historical trends combined with future scenarios derived from a detailed analysis of known factors that can affect revenues and costs. Financial forecasting can predict the inflow of revenues over a specific period of time with respect to how many new customers the company acquires as well as how many existing customers that Churn (Saas Optics, 2012). The forecast provides a clear visibility of future revenues and hence clear picture of the business. Measuring recurring revenues gives a reflection of the state of the business and decisions can be made based on how to maximize the revenue growths and profits of the firm. It is possible to track annual recurring revenues; however, it is more suitable to track monthly since the numbers can change dramatically from month to month. (Cowan, 2012) CAC and Churn, together with profit generated from a customer, will form a third metric relevant for measuring a company’s financial performance, Customer lifetime value (CLV). CLV is the long‐run profitability of a customer and is defined as the discounted cash flows from a customer over its relationship with a firm (Gupta & Zeithaml, 2004). In other words, to calculate the CLV it is necessary to project the net cash flows generated from a customer over time and then calculate the present value of those cash flows (Berger & Nasr, 1998). As seen in equation 1, the lifetime (1) of a customer is one over Churn resulting in the average months that a customer stays with the company. Then using the sum of all discounted profits from a customer and subtracting the CAC will result in CLV (2). (Berger & Nasr, 1998) Equation 1 Formulas for the lifetime of a customer (LT) and customer lifetime value (CLV). Since CLV focuses on the long‐term profits maximizing this metric will maximize the long‐term profits and financial performance of the firm. Knowing what 26 profits your customers are generating will help better decision making about customer selection. In general different customers are generating different amount of profits to the firm and often the 80‐20 rule is applicable (80% of a company’s profits are generated by 20% of its customers) and furthermore that some customers destroy value. (Gupta & Zeithaml, 2004) Measuring CLV should be an important part of marketing investment decisions, such as budgeting a customer acquisition program and understanding how to maximize the metric will provide a better support for the allocation of marketing investments, as the company will understand which marketing activities to invest in. (Berger & Nasr, 1998) 3.5. Summary of theoretical findings  CAC and Churn can be improved by investing in marketing activities that can improve the drivers of the metrics. Figure 6 illustrates an overview of the drivers of respective metric as well as improvement focus for the drivers. Investing in marketing activities that affect the drivers in a positive way will improve CAC and Churn and hence CLV. As CAC is a cost, lowering this will have a positive impact on the firm’s profit and hence financial performance. Mapping out a company’s marketing processes to see which marketing media channel that is most cost‐efficient is a first step to improve CAC. Marketing media channels can be categorized as paid, owned, and earned media. Each media channel offers distinctive advantages, as they have varying cost efficiency, different target groups and can create different effects of branding or promotion. Online marketing targets customers that are in the buying process; hence, it targets customers that have high potential of purchasing a product. Furthermore, online marketing dramatically improves the chance that potential customers find the company’s website as it enables a direct link between the advertisement and the buying place, which increases the likelihood for a purchase. To lower Churn and increase retention building relational commitment is important. This could be done by building more direct relationships to customers (affective commitment) or creating switching barriers (calculative commitment). Creating switching barriers is an important part of retention management when having an Internet based business, as there is lack of interpersonal interaction in web‐based businesses. It is therefore difficult to build an affective commitment in the relationship and develop emotional ties with a customer. There are five types of switching barriers that can be derived from an Internet based firm; transaction cost, learning cost, relationship costs, convenience costs, and artificial costs (or contractual costs). 27 Figure 6 Illustration of the drivers of CAC and Churn, as well as improvement focus for the drivers. Compiled by the authors. Financial forecasting can support decision‐making processes as it provides a clear picture of the company’s revenue growth and profits as a result of a specific decision. Maximizing CLV can improve a firm’s profit and as CLV is dependent on the customer metrics CAC and Churn, these metrics are critical to manage. Understanding how to maximize CLV will also support decision‐making processes, as the company will understand which marketing activities to invest in, in order to improve CAC and Churn.   28 4. Empirics  This chapter of the report will present the empirical findings that are divided into separate sections for CAC, Churn and Recurring revenue model. The CAC section is introduced with a description of Matkasse AB’s already performed marketing activities. This is followed by suggestions from a media planner expert at Media View of cost efficient marketing media channels. The Churn section presents the findings from the qualitative content analysis of the three largest PPGB companies as well as Matkasse AB. Then measures to retain customers are presented based on interviews with the e‐commerce companies Members.com and Milkandhoney.se. The section about the Recurring revenue model presents how the model has been built followed by a section where the model is tested to show how it can support a company’s marketing investment decisions. 4.1. CAC  The first part of this section will describe which marketing media channels that Matkasse AB already have used and the second part will present low cost marketing media channels in general and how these can be tracked and traced. 4.1.1. Matkasse AB  This section is based on an interview with Gustav Svensson, the founder and CEO of Matkasse AB, 130403. So far, Matkasse AB has performed several marketing activities, mainly with the purpose of attracting new customers to the business but also to retain existing customers. They have used both paid, earned and owned media channels for fulfilling these two purposes. For acquiring new customers following paid media channels have been utilized (marketing media channels are summarized in Figure 7): Direct mail marketing*: Svensson himself has posted promotional messages in customers’ mailboxes and even given free grocery bags for customers to try. He have targeted streets where his current customers live, as neighbours might have heard of Matkasse AB and hence are interested in trying the bag. This marketing activity has more or less not generated any new customers. Google advertising: Matkasse AB purchases the service search engine optimization* (SEO) from another company. Svensson believes that this is the main traffic source that generates most new customers. He further believes that be placed on top among search results generates more customers than advertising in the top bar or side bar, hence, it is a better marketing media * Find definition in Appendix I 29 channel than Google AdWords (described in section 4.1.2.2. Google AdWords). Therefor, Matkasse AB only spends a small amount of money (around 100 SEK) per month on Google AdWords. Deal‐of‐the‐day*: The most recent marketing activity Matkasse AB have performed is a Let’s Deal campaign. Let’s Deal offer deals of the day to consumers, where suppliers gives minimum 50% discount on their products. Matkasse AB offered 100 pre‐packed grocery bags with dinner for 2 persons. The reason for the limited number of bags is that delivering too many bags during the same period can affect the quality of current customers’ bags. The deal meant 50% discount of ordinary price. Of the remaining gross income Let’s Deal charges 40 %. The deal sold out in 3h and 34 minutes. Svensson estimates that 5‐ 6 customers will be definitive subscribers, however, this result will be more accurately estimated after four weeks when deadline for ordering has passed. He further argues that generating 15 new subscribers from the deal would be a good outcome. Social media advertising: Matkasse AB does not spend much money on Facebook* advertising (note difference between Facebook advertising and Facebook fan page) as he believes that it does not generate any customers at all. However, he argues that it is a good channel for generating brand awareness. For acquiring new customers following owned media channels have been utilized: Social media: Matkasse AB utilized Facebook and Instagram* as marketing media channels for reaching out to customers as it provides viral spread*. They have a Facebook fan page where they display the menu for the coming week, including images, and announce the last day for ordering. After a week has past customers can vote for the best dish on the fan page. Furthermore, the Facebook fan page enables the possibility for Facebook users to “like”* and “share” * the fan page; as a result, the fan page will appear in the user’s newsfeed and be visible to his/her friends. At the time of writing, there is a contest on their fan page encouraging customers to make their own pizza and take pictures to upload on Facebook. The picture with most “likes” wins a dinner cooked by Gustav Svensson worth 5000 SEK. The purpose with this contest is to increase the viral spread of Matkasse AB. Furthermore, Matkasse AB have an Intagram account, where they upload images of dishes, ingredients, events etc. Svensson believes that people have not understood Instagram well enough for it to have any effect. * Find definition in Appendix I 30 Figure 7 Marketing media channels utilized by Matkasse AB. Compiled by the authors. Matkasse AB believes that earned media channels1, such as word of mouth, have generated the majority of customers so far. As already mentioned, Matkasse AB has a Facebook fan page and this enables “share” and “like” features, which can be considered as earned media. Matkasse AB also offers discounts to customers generates another customer by recommending the product. Independent webpages rating pre‐packed grocery bags show high scores for Matkasse AB compared to competitors; they are ranked as the best PPGB in Gothenburg at www.matkasse.se and www.matkasse.com. Matkasse AB track their marketing with a Stats counter, a program for visitor analysis on the websites. They have also installed Google analytics but do not use it yet. Matkasse AB define a good customer as a customer staying at least 16 weeks and purchase minimum 9 grocery bags. The critical point is when the customer buys their second bag, as that is when the customer decides if they want to continue or not. Overall, Svensson considers his customers being very loyal; nevertheless, they perform activities for keeping existing customers as well as winning Churners back. Svensson keeps track of his best customers and sometimes visits them with a gift or sends a gift with their grocery bag. Furthermore, Matkasse AB has considered implementing a loyalty program, however, they believe that it is not appropriate until they reach a certain number of customers. 31 1 Earned media channels are not illustrated in the picture, as it is difficult to categorize into specific media channels. Customers that have paused their subscription receive an e‐mail every week, including the coming menu, with the purpose to lure them to start subscribing again. For customers that have Churned, Matkasse AB calls or sends an e‐mail offering the same offer as for new subscribers. Matkasse AB mainly communicates with their customers via SMS. At delivery day, Matkasse AB sends SMSs about delivery time notifying which hour the bag will be delivered. They want to be able to send out these messages earlier to increase convenience for the customer. 4.1.2. Marketing activities with low CAC  This section is based on an interview with Anders Katz, 130410, except for ”Deal‐of‐ the‐day”, which is based on an interview with Teodor Abreu, 130502. Recommended media channels are summarized in Figure 8. 4.1.2.1. Online marketing is most cost efficient  As Matkasse AB is in an early phase it is highly prioritized to increase their customer base. This objective can be achieved by marketing the company through different media channels, but as the financial resources are limited they need to find low cost media channels. According to Anders Katz, Campaign Planner at MediaView, the most cost efficient marketing options are online marketing, where you pay less than for traditional media, such as newspaper, television, direct mail marketing, etc. and reaching more customers. Furthermore, as all orders of pre packed grocery bags are placed online (e‐ commerce), Katz suggests that Internet is the place for advertising. There is a risk that a company markets is competitors when, for instance, placing an ad in a newspaper, since the potential customer might only perceive a message saying that the concept pre‐packed grocery bags are attractive, but the company brand is ignored. It is easier to capture the persons exposed to the ad with an online ad as s/he can click on it in the exact moment it is exposed and then be directly linked to the website and the buying place. Katz believe that you can have higher control over marketing investments in online marketing as it is easy to monitor and track which media generated which visitor/potential customer. For a company as Matkasse AB, Anders Katz recommends two media channels to use for marketing, Google AdWords and online banner advertising, as these are both examples of paid media that are cost efficient and reach out to many potential customers.   4.1.2.2. Google AdWords  Google AdWords enables creation of own ads together with a keyword portfolio including key words and phrases that relates to the business. When a person “googles” on those specific words the ad will appear next to the search results. 32 According to Katz, this is a good alternative since the majority of people nowadays know about the concept pre‐packed grocery bags so it is not necessary to market the product concept, however, a company with this product need to create brand awareness so potential customers choose that specific supplier over another. The marketing company also has the possibility to create an ad with a specific campaign or something else they want to communicate. It is further beneficial since it actually targets persons that have actively done research on Google about the specific product and are most likely considering being a customer, which increases the chance for turning a visitor into a customer. When using Google AdWords one can control the ad so it targets a specific customer, by choosing keywords, region, time for exposing the ad (days and hours), etc. For a Google Ad companies pay per click, meaning paying a specific amount each time a visitor clicks on the add. Google AdWords can be expensive for a company in the pre‐packed grocery industry since the words related to the business are competitive keywords and will therefor cost more to include in the portfolio. However, they could choose words that are more related to the uniqueness of the company, for instance sauces for Matkasse AB, which might be cheaper but still reaches potential customers to the business. The marketing company can decide exactly how much they want to pay each month, i.e. maximum clicks, and how the amount should be distributed over days. Anders Katz, estimates that it will cost about 20 000 SEK per month for Google AdWords to get a fair share of clicks and exposure as well as attractive keywords. 4.1.2.3. Online banner advertising   The meaning of the second media channel, online banner advertising, is self‐ explanatory. Basically there are two ways to buy an online banner ad. One way is to directly contact a company, e.g. a newspaper, and request advertising space on their webpage. The most attractive ones have a lot of visitors on their webpage and can therefor charge an expensive price for the ad. The other alternative is to utilize a so called advertising network (annons nätverk), where the marketing company can buy unsold advertising space from publishers, such as Aftonbladet, Göteborgs‐Posten etc., in the network for a cheaper price. As for Google AdWords it is possible to choose where and when the banner should be exposed, however, it cannot be controlled on which website it will appear. Utilizing an advertising network is, as already mentioned a cheaper alternative than buying advertising space from a specific publisher. The company pay for number of exposures of the ad, using a metric called “cost per mille”, meaning the cost of thousand exposures. The CPM for a banner is normally within the range 20 – 100 SEK. 33 Banner advertising also enables a feature called retargeting*, meaning that the ad is exposed to previous visitors. When a person visits a company’s webpage, do some research, and then leave the site, this person’ s computer ID will be registered and remembered. Later if this same person visits a webpage within the advertising network s/he will be exposed to an ad from the owner of the first webpage. The intention is that this person will be reminded of the company and its product and return to the webpage and purchase the product. This is very effective marketing since the company knows that the targeted customer has previously visited its webpage, hence, done research about the products but without placing an order. As the customer is retargeted they will get a reminder of buying the product. Retargeting is cheaper than advertising on specific publishers’ websites; however, it is not possible to have full control over where the ad will be exposed. Even if lacking control of the advertising space the company will know that the ad will target a customer that is interested in its product as s/he has visited the website before. Retargeting costs around 10 to 20 SEK per exposure. Anders Katz recommends a combination of Google AdWords and banner advertising with retargeting. Then the company can get customer to visit your webpage with AdWords and then retarget them via the advertising network. 4.1.2.4. Facebook  In addition to AdWords and banners, Facebook is a cost efficient media channel, which could be argued is a mixture of the two. In a Facebook ad even more parameters than for a Google ad can be set along with what is desired to achieve with the ad; for instance, where the person live, age, interests. The Facebook features “like” and “share”* reveals a lot of information about the users interest, which is great information for marketing decisions. As for Google AdWords the company pay per click for a Facebook ad. It is also possible to pay extra for exposing a “status update” in specific customers’ news feed, even if they have not chosen to follow the specific page. For instance, the company can write something about a product or offer in their status field and then pay for this status to appear in specific person’s news feed. Companies can also create a Facebook fan page, which is categorized as owned media, where they can update their fans with information. This is also beneficial because it enables fans to “like” or “share” information from the page, which results in exposure to friends of the fan. This type of earned marketing is very effective marketing as words from a friend is more relevant and reliable than an ad and it is provided to no cost. * Find definition in Appendix I 34 Figure 8 Recommendation of marketing media channels suitable for Matkasse AB. Compiled by the authors. 4.1.2.5. Effectiveness of the marketing media channels  It is difficult to measure the effectiveness of a marketing media channel, which is important when comparing them with each other. Media agencies use a metric called Click Free Rate* (CFR), which measures the proportion of people exposed to the ad that clicks on the ad. For Google AdWords the CFR is usually 3‐5 %, while banners only have a CFR of 0,5%. The main reason for this is that banners are placed next to other interesting media such as news, videos, etc. However, a banner is exposed more frequently to a lower price so even if the CFR is low many people have probably seen the ad, which is important for building brand awareness. One could argue that Google AdWords generate more Internet traffic to the website, while banners increase brand awareness. The two medias also differ in which potential customers they target. A Google ad targets a person whom you are aware have actively has done research on a specific product/service by searching on specific keywords. This enhances the change of getting the potential customer to visit the company’s website and hopefully purchase a product. Using banner advertising it is impossible to know how far a person, the ad is exposed to, is in the buying process. Facebook allows for huge amount of exposure, possibly even more targeted marketing than Google AdWords and to a lower price; this is mainly because the ad can be customize to a very specific target group and hence avoid irrelevant customers. With that said, * Find definition in Appendix I 35 Google AdWords and Facebook provides a more accurate targeting, while banners increase brand awareness and with retargeting can remind potential customers to visit the website and find out about e.g. current deals. 4.1.2.6. Track and trace media channels  Online media is to track and trace, which is something Katz highly recommends that companies do. Understanding which media that generates the most new customers or visitors to the website enables optimized allocation of marketing expenses. For instance, let say that spending 10000 SEK on Facebook and 10000 SEK on Google AdWords one month results in 100 new visitors traced to derive from Facebook while only 10 new comes from the Google AdWords. Shifting part of the Google AdWords budget to Facebook advertising then should optimize the distribution of marketing allocation. Google Analytics is Google’s own statistical program that generates detailed statistics about a website’s traffic and traffic sources, and measures conversations and sales. The program provides reports showing how visitor use the site, how they arrived (from what source), which part of the site are performing well, which parts are most popular, and so on. Together all this information will provide a very solid decision‐base for marketing strategies, e.g. allocation of expenses, customization of marketing campaigns, improvement of advertise performance, etc. Furthermore, Google Analytics provides Return on Investment (ROI) analysis showing how much profit you have made from marketing compare to marketing expenses related to the marketing campaign. (Google Analytics, 2013) 4.1.2.7. Deal‐of‐the‐day  According to Abreu, previous employee at groupon.com, it is important to emphasize that groupon.com is not a sales channel; it is a marketing media channel. Their business model works as following: the marketing company offers their product to consumers with a discount (groupon.com have a minimum discount level of 60%). Of the remaining gross income groupon.com, or other site, charges a specific percentage (groupon.com charges 50%) for utilizing their marketing media channel. In the end such deal does not generate much income, Abreu argues. However, marketing on groupon.com is a good way to get customers to try the product. It is remarkable that deal‐of‐the‐day does not enables targeting a specific customer 4.2. Churn  The first four parts of this section (“Linas matkasse”, “Middagsfrid”, “Mathem.se”, “Matkasse AB”) will provide the information collected from the qualitative content analysis of virtual documents. Results from the analysis are summarized in Table 1 and pictures of the findings can be found in Appendix II. The final two 36 parts of this section (“Members.com”, “Milkandhoney.se”) presents how two e‐ commerce companies strive to increase loyalty with their customers. 4.2.1. Linas matkasse  Linas matkasse started in 2008 and their turnover 2011 was 312 million SEK, which makes them the largest PPGB company in Sweden. The grocery bags can be delivered throughout Sweden. Linas matkasse offers three products; original, inspirational, and child friendly grocery bag and some add‐on products like fruit bag and snack bags. The price per portion ranges between 39–48 SEK. Three words that best describe the concept: tasty, healthy and quick. 4.2.1.1. Main page  On the front page of Linas Matkasse’s website visitors find five tabs in the top menu “About us”, “About the bags”, “Order”, “Contact us” and the final tab “Log in” is for present customers, where they can find personal information regarding their subscription of “Linas Matkasse”. In the middle of the page there is a banner, which states that Linas matkasse is the most sold PPGB in Sweden (Appendix II – 1). 4.2.1.2. Membership & subscription terms  When a customer is registered a subscription has started, which means that the new customer have to terminate the subscription him‐self if s/he wishes. After having registered the “Log in” page are available. Linas Matkasse requires 8 days notice in advance for booking or cancellation of an order (Appendix II – 2). 4.2.1.3. Delivery  They deliver the bag each other week on Mondays or Sundays, depending on where you live, between 17.00‐22.00. Residents in Stockholm, Gothenburg, and Malmö have the option to receive the bag every week. Customers cannot choose if they want deliveries even or odd week due to logistical issues. Furthermore, Linas Matkasse does not offer the possibility for customers to pick up their bag themselves (Appendix II – 3). 4.2.1.4. The menu  The menu and recipes are sent together with the food bag, but can also be found online, for registered customers only, 4‐5 days before delivery. A customer can only see the recipes from their previous orders, not from all grocery bags. For potential customers that want to try a Linas matkasse recipe, examples of recipes can be found online (Appendix II ‐ 4). 4.2.1.5. Contact  Contact channels to Linas Matkasse are phone and email. Their customer service is open during office hours with extended opening hours until 22.00 on delivery days Sundays and Mondays (Appendix II – 5). 37 4.2.1.6. Member’s page/club  Becoming a subscriber of Linas matkasse you are automatically a member of their customer club “Klubb Lina”. Customers collect points for every order they make and extra points if they recommend to someone who becomes a subscriber. The accumulated points can then be converted in to premiums, e.g. kitchen equipment, hotel nights, magazine subscriptions, etc. The original bag is worth 10 points and the first premium starts at 20 points. Recommendations generating a new customer generate 30 points. The validity of the points is two years. Customers buying more than 16 bags a year becomes a premium customer, a so‐called “Topp kund”. This gives the customers additional benefits, such as initial 20 points, menus with recipes from all grocery bags, special deals on selected products and a free membership of “Linas Kom I Form” (Appendix II – 6). As a member of “Klubb Lina” you can also join “Linas Kom i form” where customers get access to videos with instructions for different gym sessions for improving strength and cardio (Appendix II – 7). Members also have an opportunity to pay more for the bags and "extra” money will contribute to Linas matkasse’s project of giving food to poor families in Kenya (Appendix II – 8). Linas Matkasse furthermore provides a mobile software application* (here by referred to as an App*) for subscribers. There you can find recipes to ordered grocery bags and nutritional content of the dishes. Furthermore, the accumulated points in Klubb Lina are showed in the app and a feature where customers can “check in” on Facebook when receiving a bag. This means that a message will appear on the customers Facebook page and hence be displayed to friends of the customers. This feature can generate more point to the Klubb Lina account, as each share is worth 1 point. 4.2.2. Middagsfrid  Middagsfrid was first to enter the PPGB industry in the summer of 2007. 2011 they had a turnover of 85 MSEK. Their distribution reaches most of Sweden and the business has also expanded to Norway, Denmark, Germany and Belgium. They offer four main PPGB’s one large (4 persons for 5 days), one medium (4 persons for 3 days), one small (2 persons for 4 days) and one children friendly alternative (2 persons for four days). Aside from these offering there are add‐on products such as a breakfast bag and snacks. The average price per portion of * Find definition in Appendix I  38 food is between 45 and 72 SEK. Three words that best describe the concept: tasty, healthy and quality. 4.2.2.1. Main page  The first page when entering Middagsfrid’s website shows five main categories of navigation all connected to the PPGB service “How it works”, “The bags”, “Menu”, “About us” and “Order”. Through these tabs one could find information about their concept, how and where deliveries are made and order the products online. Certification in safe e‐commerce (Trygg E‐handel), which is a certificate from third party and “Best‐in‐test” awards that they have won are presented on the top of the page along with links to visit their social media (Appendix II – 9). Clicking on the link “About us” reveals information about the company and presents different prices and nominations that Middagsfrid has received during the years (Appendix II – 10). 4.2.2.2. Membership & subscription terms  There is no binding time when entering into a subscription and to pause or end the subscription one must notify the customer support nine days in advance (Appendix II – 11). 4.2.2.3. Delivery  Deliveries these can be done Sunday and Monday between 17‐22 in Gothenburg and Stockholm, and the smaller cities only have Monday between 17‐22. No other deliveries are possible (Appendix II – 12). 4.2.2.4. Menu  Their recipes and menu’s are publicly available on the website for 3 weeks in advance. Old recipes are also displayed for non‐members as well. All recipes are printed with pictures and instructions that come along with the bag (Appendix – 13). 4.2.2.5. Contact  Customer services are available on the telephone, chat and email and open during regular office hours. The ambition is to answer all emails from customers within 24 hours sorts of questions are welcome (Appendix – 14). 4.2.2.6. Member’s page/club  When entering into a subscription one gets access to a member’s page. From here it is possible log in and manage the account, choose delivery times and pause or terminate the subscription. All recipes are also available from this page. 4.2.3. Mathem.se  Mathem.se was started 2009 as an online grocery store and later came to offer PPGB’s the following year. In 2012 they had a turnover of 65 MSEK. Their 39 distribution range includes Stockholm, Uppsala, Gothenburg, Helsingborg, Lund and Malmö. They offer a range of four different products (all for four persons), one original bag for 3 or 5 days and one ecological bag for 3 or 5 days. Aside from this they offer all products that are available in a regular grocery store. The average price per portion of food is between 38 and 57 SEK. Three words that best describe the concept: healthy, quick and easy. 4.2.3.1. Main page  On the first page when reaching mathem.se the information is related to the general grocery store that they operate. Clicking on “Matkasse” gives the first information about their PPGB offerings. Displayed in the up‐right corner there is certification for safe e‐commerce (clicking on the icon for safe e‐commerce displays the whole certificate from third party) and a reward of “Best food site 2012” from Internetworld. Also customer reviews are displayed in the first view (Appendix II – 15). 4.2.3.2. Membership & subscription terms  Ordering a product can be done both by phone and online. The order will automatically start a subscription at mathem.se with PPGBs being delivered every week. A subscription with mathem.se can be terminated or paused at any time, but this must be done eight days before delivery by logging in to “My site” (Appendix II – 16). 4.2.3.3. Delivery   Deliveries are available from Monday to Friday 12.00‐22.00 and the customer can choose which hour between these times that they want their products delivered (Appendix II – 17). 4.2.3.4. The menu  Menu’s and recipes are available and displayed on the webpage for non‐ members as well. All recipes are printed with pictures and instructions that come along with the bag (Appendix II – 18). 4.2.3.5. Contact  Mathem.se provides two different direct phone numbers, one to order and the other to reach customer services. The opening hours are Monday to Friday 08.00‐22.00 and 12.00‐22.00 on Sundays (Appendix II – 19). 4.2.3.6. Member’s page/club  When ordering from mathem.se you create an account that one could manage through “My site” (Min sida). On this member page it is possible to look at current orders, create shopping lists and administer different account settings. There is also a bonus system connected to “My site” where the customer receives one point for each krona that they spend. Collecting 2500 points will result in a 40 bonus check of 25 SEK that only can be spent buying products on the website (Appendix II – 20). Through My site it is possible to buy delivery checks, which enables free deliveries for a time period of either three or six months. The checks also have three different denominations base, silver and gold where silver and base checks are restricted to certain delivery times and gold checks are not (Appendix II – 21). The services that mathem.se offers are also available in their free of charge mobile App, which is accessible for both Android and iPhone. In the app it is possible to both pay and order PPGB’s and access information that is stored in My site (Appendix II – 22). 4.2.4. Matkasse AB  Matkasse AB was started in 2010 by the chef Gustav Svensson and 2012 they had a turnover of 1,3 MSEK. They can deliver in the Gothenburg area and most of its suburbs. The products offered are one original bag and one that is free from gluten. Both can be ordered for either two or four persons for four days a week. Aside from these products it is also possible to order a snack bag. The price per portion is between 55‐73 SEK. Their niche is to provide pre‐cooked sauces that the chef has prepared himself. Three words that best describe the concept: nutritious, quality and easy. 4.2.4.1. Main page  There are eight tabs presented on the main page “Home”, “About Matkasse AB”, “About the bags”, “Order”, “Delivery”, “Contact us” and “My pages” where the customer can log in to the member page. In the top section of the page there are links to their fan pages in social media. There is also information about why the customers choose their PPGB according to a customer survey by Intermetra, 2013 (Appendix II – 23). 4.2.4.2. Membership & subscription terms  When ordering a PPGB it is possible to choose whether to buy one PPGB or enter into a subscription. The subscriptions can either be every week or every other week. There is no binging time in the agreements, but to end or pause your subscription this must be made 5 days in advance on the member’s page (Appendix II – 24). 4.2.4.3. Delivery  Deliveries are made on Mondays and can be delivered to workplaces during 14.00 to 16.00 and at home 16.00‐22.00 (Appendix II – 25). 41 4.2.4.4. The menu  Menu’s and recipes for the PPGB’s that are still offered for order are available on the homepage. More recipes can be found on the member’s page. All recip