Department of Technology Management and Economics Division of Innovation Engineering and Management CHALMERS UNIVERSITY OF TECHNOLOGY Göteborg, Sverige 2015 Report No. E 2015:036 Applying Gartner’s Hype Cycle on the Health Food Industry An analysis of growth opportunities Master of Science Thesis in the Management and Economics of Innovation Programme ANNA TEIWIK JOHANNA YDERGÅRD MASTER’S THESIS E 2015:036 Applying Gartner’s Hype Cycle on the Health Food Industry An analysis of growth opportunities ANNA TEIWIK JOHANNA YDERGÅRD Tutor, Chalmers: Bengt Järrehult Tutor, company: Undisclosed Department of Technology Management and Economics Division of Innovation Engineering and Management CHALMERS UNIVERSITY OF TECHNOLOGY Göteborg, Sweden 2015 Applying Gartner’s Hype Cycle on the Health Food Industry An analysis of growth opportunities ANNA TEIWIK and JOHANNA YDERGÅRD © ANNA TEIWIK and JOHANNA YDERGÅRD, 2015. Master’s Thesis E 2015:036 Department of Technology Management and Economics Division of Innovation Engineering and Management Chalmers University of Technology SE-412 96 Göteborg, Sweden Telephone: + 46 (0)31-772 1000 Chalmers Reproservice Göteborg, Sweden 2015 Abstract One of the largest trends among Swedish consumers today is the adoption of a healthy lifestyle, leading to consumers becoming increasingly interested in health food. This has in turn opened up for interesting opportunities for firms within the health food industry. However, this industry is highly influenced by consumers' shifting perceptions of what constitute health food, making the industry volatile in terms of trends that come and go, which increases the difficulty for companies in knowing where to focus their product development efforts and investments. The purpose of this master's thesis is twofold. Firstly, it aims to identify growth opportunities within the health food industry. Since this industry is highly volatile in nature due to trends succeeding each other, the theory of Gartner's hype cycle is used in order to identify characteristics of health food that provide short- and long-term potential. Gartner's hype cycle is however usually applied on new technologies, which leads to the second purpose of this master's thesis; to investigate how Gartner's hype cycle may be applied on the less technological health food industry. In order to fulfil the purpose of this master's thesis, this study applies a qualitative approach, which is further quantitatively supported by a survey. The research process is divided into four parts, that starts with developing a definition of the health food industry through identifying what currently characterises health food. The second part of the research process aims to position the health attributes in Gartner's hype cycle, through following the approach suggested in the hype cycle literature. This is followed by an analysis of how to interpret the hype cycle in the context of health food. Lastly, the research includes an analysis of how firms could manage hypes and trends to benefit from the identified growth opportunities. This report concludes that the definition of health food is dynamic due to shifts in scientific paradigms, high levels of media attention and the individual nature of health. What defines health food is therefore what attributes consumers perceive as healthy at a certain point in time. The health food characteristics positioned in Gartner's hype cycle indicate on several growth opportunities, both in the short- and long term. The inherent difficulty of predicting a fad or megatrend implies that it is beneficial to invest in hypes in a small scale through for example product extensions, gaining first mover advantages while reducing risk. Characteristics in the slope of enlightenment that are more likely to become megatrends may on the other hand be incorporated into the brand strategy. The application of Gartner’s hype cycle leaves some implications for the interpretation of the theory in this context. These mainly include that the health food industry is characterised by higher levels of positive and negative hype that affects the analysis, and low switching costs making the adoption pattern different from the traditional model. Acknowledgements We have had the opportunity to conduct this master's thesis for an investment company, and indirectly for one of their holdings operating within the health food industry. Thus, we would firstly like to extend a big thank you to our supervisors at the investment company, both for guidance throughout the process and also for giving us the opportunity to conduct this master's thesis. We would also like to thank all our interviewees for valuable input. Especially, we would like to thank all representatives from the health food company who enthusiastically took their time to tell us about their work and answer our questions. We would furthermore like to express our gratitude to our supervisor Bengt Järrehult for his encouragement, guidance and support during the process of writing this master's thesis. Throughout the whole process Bengt has been available to answer our questions as well as involve us in interesting discussions. Finally, this thesis is the final step of our time at Chalmers University of Technology, Industrial Engineering and Management and the MEI programme. We would therefore also like to extend a thank you to all friends, lecturers, family and others who have been a part of and encouraged us throughout these five years. Table&of&Contents& 1.#Introduction#......................................................................................................................................#1! 1.1#Background#................................................................................................................................................#1! 1.2#Purpose#.......................................................................................................................................................#2! 1.3#Research#Questions#.................................................................................................................................#3! 1.4#Delimitations#.............................................................................................................................................#4! 1.5#Outline#of#the#Thesis#...............................................................................................................................#4! 2.#Literature#Review#............................................................................................................................#6! 2.1#Scientific#Paradigms#................................................................................................................................#6! 2.2#Diffusion#Curve#.........................................................................................................................................#7! 2.3#Technology#SJcurve#.................................................................................................................................#9! 2.4#Gartner’s#Hype#Cycle#.............................................................................................................................#10! 2.5#Relating#the#Diffusion#Curve#to#Gartner's#Hype#Cycle#...............................................................#16! 3.#Method#..............................................................................................................................................#17! 3.1#Research#Strategy#..................................................................................................................................#17! 3.2#Research#Process#...................................................................................................................................#17! 3.2.1!Definition!of!the!Health!Food!Industry!....................................................................................................!18! 3.2.2!Positioning!of!Characteristics!in!Gartner's!Hype!Cycle!.....................................................................!18! 3.2.3!Interpreting!the!Hype!Cycle!in!the!setting!of!the!Health!Food!Industry!....................................!19! 3.2.4!Managing!Hypes!and!Megatrends!...............................................................................................................!19! 3.3#Data#Collection#........................................................................................................................................#19! 3.3.1!Interviews!.............................................................................................................................................................!20! 3.3.2!Survey!.....................................................................................................................................................................!21! 3.3.3!Secondary!Sources!............................................................................................................................................!22! 3.4#Data#Analysis#...........................................................................................................................................#24! 3.4.1!Interview!and!Secondary!Data!Analysis!..................................................................................................!24! 3.4.2!Survey!Analysis!...................................................................................................................................................!24! 3.5#Research#Quality#....................................................................................................................................#25! 4.#Empirical#Study#..............................................................................................................................#27! 4.1#Health#Food#Background#.....................................................................................................................#27! 4.2#A#Brief#History#of#Dietary#Advice#......................................................................................................#28! 4.3#The#Nordic#Nutrition#Recommendations#.......................................................................................#29! 4.4#Attributes#Currently#Defining#Health#Food#...................................................................................#30! 4.4.1!Consumers!Driving!Importance!of!Health!Attributes!........................................................................!30! 4.4.2!Current!Research!on!Classification!of!Health!Attributes!..................................................................!30! 4.4.3!Survey!on!Classification!of!Health!Attributes!........................................................................................!34! 4.4.4!Exposure,!Interest!and!Research!on!Health!Attributes!.....................................................................!40! 4.4.3.1!Natural!Ingredients!.....................................................................................................................................................!40! 4.4.3.2!Organic!..............................................................................................................................................................................!42! 4.4.3.3!Vegetarian!........................................................................................................................................................................!43! 4.4.3.4!Locally!Sourced!.............................................................................................................................................................!45! 4.4.3.5!Fat!Free/Low!Fat!..........................................................................................................................................................!46! 4.4.3.6!Sugar!Free/Low!Sugar!...............................................................................................................................................!48! 4.4.3.7!Low/No!Salt!....................................................................................................................................................................!49! 4.4.3.8!Low!Cholesterol!............................................................................................................................................................!50! 4.4.3.9!High!Protein!....................................................................................................................................................................!51! 4.4.3.10!High!Fiber!......................................................................................................................................................................!53! 4.4.3.11!Food!Fortification!(Vitamins)!...............................................................................................................................!54! 4.4.3.12!Gluten!Free!...................................................................................................................................................................!55! 4.4.3.13!Lactose!Free!.................................................................................................................................................................!57! 4.4.3.14!GMO!Free!.......................................................................................................................................................................!58! 4.5#Trends#Influencing#the#Health#Food#Industry#..............................................................................#59! 4.5.1!Macro!Trends!.......................................................................................................................................................!59! 4.5.2!Consumer!Behaviour!........................................................................................................................................!61! 5.#Analysis#.............................................................................................................................................#62! 5.1#Analysis#of#the#Health#Food#Industry#..............................................................................................#62! 5.1.1!Definition!of!Health!Food!...............................................................................................................................!62! 5.1.2!Survey!Analysis!...................................................................................................................................................!64! 5.2#Hype#Cycle#Analysis#...............................................................................................................................#66! 5.2.1!Analysis!of!Health!Attributes!........................................................................................................................!66! 5.2.1.1!Natural!Ingredients!.....................................................................................................................................................!66! 5.2.1.2!Organic!..............................................................................................................................................................................!68! 5.2.1.3!Vegetarian!........................................................................................................................................................................!69! 5.2.1.4!Locally!Sourced!.............................................................................................................................................................!71! 5.2.1.5!Fat!Free/!Low!Fat!.........................................................................................................................................................!73! 5.2.1.6!Sugar!Free/!Low!Sugar!..............................................................................................................................................!74! 5.2.1.7!Low/!No!Salt!...................................................................................................................................................................!75! 5.2.1.8!Low!Cholesterol!............................................................................................................................................................!76! 5.2.1.9!High!Protein!....................................................................................................................................................................!78! 5.2.1.10!High!Fiber!......................................................................................................................................................................!79! 5.2.1.11!Food!fortification!(Vitamins)!................................................................................................................................!80! 5.2.1.12!Gluten!Free!...................................................................................................................................................................!81! 5.2.1.13!Lactose!Free!.................................................................................................................................................................!82! 5.2.1.14!GMO!Free!.......................................................................................................................................................................!84! 5.2.2!ShortZ!and!LongZterm!Growth!Opportunities!........................................................................................!86! 5.3#Interpreting#the#Hype#Cycle#in#the#Setting#of#Health#Food#.......................................................#88! 5.4#Managing#Hypes#and#Megatrends#.....................................................................................................#90! 6.#Conclusion#........................................................................................................................................#93! 7.#References#........................................................................................................................................#95! 1 1. Introduction The opening chapter encompasses a background to why this study was conducted, followed by the purpose of this master's thesis. The purpose is then further broken down into four research questions that this study will answer. Furthermore, this chapter will also present the scope of this master's thesis through outlining the delimitations, and will then end by presenting the outline of the report. 1.1 Background The amount of people suffering from overweight or obesity is increasing around the globe, with Sweden not being any exception. According to Folkhälsomyndigheten (2014), approximately half of the Swedish population is either overweight or obese. These are shocking numbers, and they are still increasing (Folkhälsomyndigheten, 2014). To cope with the increasing obesity, it is critical that the Swedish consumers start changing diets towards more healthy food, which actually appears to be happening. ICA, Sweden's largest supermarket chain, stated in their 2014 annual report that one of the largest trends they will respond to in the coming years is health and consumers' willingness to live a long and healthy life. This is also supported by Nielsen (2015), a leading global information and measurement firm specialised on consumer behaviour, that further stated that health is a large trend in our society today, and for businesses this could provide interesting opportunities. Health food has previously been considered a niche market, separated from the mainstream food industry. This is currently changing, and businesses have started to realise that health is going to become mainstream (Nielsen, 2015), and that food actually is equal to health (Healthy Marketing Team, 2015). Traditional large food producers are now experiencing a volatile environment, where the competition has shifted from mass manufacturers with low prices towards new players with new business models and technologies better adapted to consumers demand for healthier food. Moreover, trends such as the sugar debate, gluten free diets and natural ingredients succeed each other, making it difficult for food producers to predict what consumers of tomorrow actually will demand (Healthy Marketing Team, 2015). The overall health trend as well as difficulties with this market was identified by a company, in this thesis called Investment AB, who therefore initiated this study. Investment AB is an investment firm that owns and develops companies on a long-term basis, with the goal of diversifying the operations of its parent company and contribute to increased value growth. One of their holdings is a Nordic health food company, in this thesis called Healthy Inc., that sells products within diet and health, superfood, sports nutrition, cold remedies and hygiene. More than half of the company’s products are sold through food retail stores, and Investments AB now wants to investigate opportunities for Healthy Inc. to increase growth through this sales channel. 2 Healthy Inc. operates within two main categories; the traditional self-care market and the newer health food market. The self-care market includes herbal and natural OTC medicines, dietary supplements, sports nutrition and weight-loss products (Svensk Egenvård, 2015). The health food market on the other hand is not as clearly defined but includes food that provide, or are perceived to provide, some kind of health benefit. The overall trend towards greater awareness of health and wellbeing has significantly increased the interest in health food, while the growth in the self-care market has stagnated over the last years (Svensk Egenvård, 2014). Healthy Inc. therefore believes that health food sold through food retail stores offer the greatest future growth opportunities. Healthy Inc. is a brand company, meaning that they manage brands within the health food category by either acquiring new brands, licensing external brands or developing new products in-house through their currently owned brands. Healthy Inc. has outsourced all production in order to focus on their core competencies that include product development, sales and distribution. The company’s customers include large food retail store chains, pharmacies and more specialised health stores. This necessary use of retail stores as an intermediary creates a gap between Healthy Inc. and their end consumers, which complicates the process of understanding consumer preferences and behaviour. Due to the large trend of interest in health and wellness among consumers, there are great opportunities for businesses to take advantage of that health is becoming mainstream, enabling actors to reach the mass market and not only a niche group of consumers. Healthy Inc. is well equipped to be a part of this market, taking advantage of their brands or capabilities to acquire or license external brands, as well as their distribution channel through food retail stores that enables them to reach the mass market. However, due to the volatile health food market in the form of different trends that succeed each other, the question is where on this market they should focus their efforts. As stated by Susan Dunn, executive vice president, Global Professional Services, Nielsen: “There is a tremendous opportunity for food manufacturers and retailers to lead a healthy movement by providing the products and services that consumers want and need. While diet fads come and go over time, innovative, back-to-basics food that taste good, are easy to prepare and provide healthful benefits will have staying power. The first step is knowing where to put your product development efforts.” (Nielsen, 2015, pp3) 1.2 Purpose The purpose of this thesis is twofold. Firstly, the increasing interest in health and wellness among consumers in Sweden as well as worldwide is, as previously mentioned, likely to provide growth opportunities for firms within the industry. However, due to the large amount of trends that come and go 3 within this industry, it is difficult for companies to know where to focus their efforts. This leads to the first purpose of this thesis: The thesis aims to identify growth opportunities within the health food industry. Secondly, the volatility of trends that come and go raises a question of whether there are any existing theories that may explain this phenomenon. The hype cycle is a theory developed by Gartner, aiming to describe the phases of expectation that a technology goes through when it is new to the market, illustrating a pattern of an initial irrationally positive reaction to a novelty. This phenomenon appears to be observed also in the less technological health food industry, where different characteristics in health food are hyped in media and succeed each other in a fast pace. As the hype cycle may be used in order to decide whether, how and when to invest in a new technology, this will be used in order to fulfil the first purpose. Thus, this also leads to a second purpose of this thesis as follows: The thesis aims to investigate how to apply Gartner's hype cycle on the health food industry. 1.3 Research Questions In order to identify growth opportunities within the health food industry, the industry firstly needs to be clearly defined. This will be done through exploring what attributes consumers believe characterise health food and what affects their perception. Due to the high amount of trends that come and go, it is furthermore interesting to identify if these characteristics defining health food today are short-term trends, i.e. hypes, or if they will persist in the long-term, i.e. if they are megatrends. Therefore, the next part of this study aims to identify the duration of these factors using the theory of Gartner's Hype Cycle. This will be followed by an analysis of how the hype cycle may be interpreted in this setting. Lastly, this master's thesis will also cover how a company could take advantage of these hypes or megatrends to enable further growth. Thereof, the below research questions have been developed in order to fulfil the purpose of this thesis. RQ 1: What characteristics currently define health food? RQ 2: Where in Gartner's hype cycle may these health food characteristics be positioned? RQ 3: How, based on the findings in research question 2, may Gartner's hype cycle be interpreted in the setting of the health food industry? RQ 4: How may hypes and trends be managed in order to create long-term growth within this industry? 4 1.4 Delimitations This study will only focus on identifying growth opportunities within the health food industry, and hence not include the traditional self-care market, even though Healthy Inc. operates in both. Furthermore, Healthy Inc. is only mentioned to provide an understanding of the background to this study, and no specific recommendations to them will be outlined in this report. This study will also be focused on primarily investigating the Swedish market. Furthermore, when identifying growth opportunities, different levels of analysis could be of interest. This study is focusing on health characteristics and whether they are hypes or megatrends, but it is possible to narrow it down even further and consider for instance different diets or even products that are considered healthy. However, due to time and resource constraints, it was deemed most appropriate to only focus on the level of what is called characteristics of healthy food, while the reader should bear in mind that other levels could be considered. Moreover, there are also various tools and frameworks for how to identify growth opportunities within an industry, but this study aims to find growth opportunities by using Gartner's hype cycle, why no other tools and frameworks are included. 1.5 Outline of the Thesis The thesis begins with a literature review that aims to provide an overview of relevant literature. Thereafter, the methods used to answer the research questions and fulfil the purpose in this thesis are thoroughly outlined. This is followed by a presentation of the empirical findings, covering secondary as well as primary data underlying the analysis. Thereafter, the findings are analysed to determine growth opportunities and how the hype cycle may be applied to this industry. Lastly, conclusions are presented. The outline is presented in Figure 1 on the next page. 5 Figure 1 - Outline of the thesis. 6. Conclusion Concludes and wraps up the analysis 5. Analysis Presents an analysis of growth opportunities, how to handle these and how the hype cycle may be applied to this industry 4. Empirical findings Presents the findings regarding overall perception of health food, specific health food attributes and macro trends affecting the industry 3. Method Outlines the research strategy, process, data collection and analysis 2. Literature review Presents literature relevant to answer the research questions 1. Introduction Explains the background to and purpose of the study 6 2. Literature Review In order to answer the first research question of this master’s thesis, i.e. what characteristics currently define health food, the Theory of Scientific Paradigms will firstly be presented in this literature review. This will provide an understanding of why consumers' perceptions may change over time. In order to further identify growth opportunities within this industry, the theories of the Diffusion Curve and the Technology S-Curve will firstly be described, in order to explain how an innovation is adopted among consumers and how it develops through its life cycle. This will be followed by a description of Gartner’s Hype Cycle, explaining the phases of expectation that a technology goes through when it is new to the market, which will be useful in order to identify and determine how to manage short- and long term growth opportunities within the health food industry. 2.1 Scientific Paradigms The notion of scientific paradigms was introduced by Kuhn (1962), and is based on the idea that a scientific community cannot function and progress without some set of common beliefs. If a scientist would not base his or her work on assumptions from previous research, he or she would never be able to provide anything new to society. This type of science, i.e. science based on accumulated accepted facts and theories, is according to Kuhn (1962) called normal science. Kuhn (1962) further argues that a paradigm starts with a random collection of mere facts. During the early stages of an inquiry, different researchers are tackling the same phenomenon in different ways, resulting in different opinions of how to interpret and describe it. After some time, pre-paradigmatic schools appear, each emphasising on a specific part of the collection of facts. Through competition between these different pre-paradigmatic schools, one particular paradigm eventually emerges. The paradigm is thus based on the most credible of the competing theories, but often it is not enough to explain all aspects of the particular phenomenon, making new research, i.e. normal science, possible. As the paradigm grows and new disciplines and professions form within it, the other pre-paradigmatic schools eventually fade away. During the normal science within a particular paradigm, no resources are allocated to call forth new types of phenomenon, and if anomalies occur, they are often discarded or ignored. Thus, in the early stages of a scientific paradigm, theoretical alternatives are easily invented, but as the paradigm gets entrenched, the alternatives become more and more resisted (Kuhn, 1962). However, in most cases, researchers find anomalies within the paradigm that at one point cannot be ignored any longer. This, according to Kuhn (1962) results in a crisis of the paradigm, which in turn is the beginning of a science revolution. If misstatements of the normal science is found during the period of revolutionary science, new paradigms will be formed, changing the rules of the game and guide science 7 into new directions. Thus, Kuhn (1962) suggests that the scientific paradigms are often cyclical, where periods with normal science will eventually be interrupted by revolutionary science. 2.2 Diffusion Curve The diffusion curve displays the market penetration of an innovation, through illustrating the types of consumers it attracts during its life cycle (Moore, 2001), as shown in Figure 2 below. It is a process dependent on how people are gradually becoming prone to adopt a new idea, meaning it is a social learning process resulting in change of attitudes and values (Brown, 1992). The rate of diffusion among consumers is determined by how information is spread within a social system over time (Rogers, 2003). The number of adopters over time is assumed to be normally distributed, which is illustrated in the adoption curve below. The different types of adopters are further categorised according to their propensity to adopt an innovation, beginning with the innovators and continuing on to the early adopters, early majority, late majority and laggards (Lindmark, 2006). Figure 2 - Diffusion of an innovation (Rogers, 2003), axes added by the authors of this master's thesis. Innovators have a gatekeeper role in the diffusion process since they are the first to adopt an innovation, i.e. they import the innovation from outside the system’s boundaries (Rogers, 2003). The next group of people to adopt are called early adopters. These people have the ability to easily imagine, understand and appreciate the benefits of a new technology in relation to their other concerns and they are willing to base their buying decision on their own intuition (Moore, 2001). Consequently, they are visionaries and expect to get first mover advantages by being the first to introduce new technologies to their industry. However, in order to capture any substantial profits and growth, it is essential to reach the next group, early majority, since this group constitutes approximately one third of the whole adoption population (Moore, 2001). These people are similar to early adopters as they possess some ability to relate to a new technology, but unlike the early adopters, they require well-established references when buying a new technology in order to not risk buying into a fad (Moore, 2001). Therefore, bridging the gap between the 8 early adopters and the early majority can be highly difficult (Moore, 2001). The subsequent group, late majority, is similar to the early majority, but they tend to require even more proof of usefulness in order to consider purchasing an innovation, resulting in that they will wait until the majority of the actors in their system have adopted (Moore, 2001). Lastly, there are also laggards, who despite proofs of benefits do not want anything to do with the innovation (Moore, 2001). The theory of diffusion of an innovation is also useful in order to understand why some products fail to reach a substantial customer base. Moore (2001) separates between early and mainstream markets: early referring to the consumer groups innovators and early adopters, and mainstream referring to consumers belonging to early and late majority. In order to leave the early market and enter the mainstream, firms need to move from the group early adopters to early majority, as illustrated in Figure 3 below. As previously mentioned, the characteristics of the consumers within these groups differ: early adopters have the ambitions to be first and early majority are pragmatists that prefer well-established references. This contrast results in a chasm that is difficult to cross, especially since early adopters are likely to lose interest in the technology if it reaches the mainstream market and since early majority does not consider early adopters to be relevant references but prefer references from within their own group (Moore, 2001). Figure 3 - The chasm between the early and the mainstream market (Moore, 2001). Moore (2001) further provides directions for how firms can overcome the chasm and successfully reach the mainstream market. Firstly, the author suggests to target the point of attack, meaning firms ought to target a specific niche market where all resources are to be focused in order to gain a dominant position in that particular segment. In this phase, firms are suggested to consider aspects such as who the target customer is, a compelling reason they should buy the product, what the whole product includes and what constitutes the competition. Secondly, the author suggests firms should "assemble an innovation force", meaning they should create the whole product based on the customers problems and possible solutions, 9 either in-house or through collaboration with other parties. The third step is, according to Moore (2001), to define the battle, i.e. to understand the competition and their relationships with the target customers and thereafter define their own position in the market to force the competition out of the target segment. Finally, firms ought to launch the invasion, which refers to distribution and pricing (Moore, 2001). When crossing the chasm, it is of essence to secure a channel into the mainstream market that is comfortable for the early majority to use. Since an established distribution channel is critical in order to reach the early majority, firms should, during the chasm period, also use a pricing strategy that motivates the channel, rather than to satisfy customers or investors (Moore, 2001). 2.3 Technology S-curve Diffusion research usually disregard the fact that technology changes during the diffusion process, which implies that the diffusion does not solely depend on the speed of which information is spread. Lindmark (2006) points to that the technological change itself is one of the main factors for explaining diffusion. Thus, gradual improvements in technology can activate new market segments over time and can therefore also explain the diffusion process. How a technology develops can be explained by the technology s-curve. The curve illustrates how a new technology improves as a function of R&D expenditure, as shown in Figure 4 below. Emergent technologies develop rather slowly, resulting in performance improvements being achieved at relatively high cost in terms of time and money (Brown, 1992). However, the technology eventually reaches a critical stage whereby the performance improves rapidly, until it starts reaching its inherent limits where large expenditures only have marginal effects on performance (Lindmark, 2006). As illustrated in the s- curve, the highest return on investment will be achieved when investing in a developing technology, rather than emerging or mature technologies (Brown, 1992). At this phase, the performance will be improved rapidly at relatively low R&D costs. However, companies investing in emergent technologies might gain an advantage due to the learning effects it implies, which they then can utilize in the developing phase and thus outcompete their competitors (Brown, 1992). Hence, it can be more beneficial to invest in emergent technologies even if these do not initially yield the greatest returns (Brown, 1992). 10 Figure 4 - The Technology S-curve (Brown, 1992). 2.4 Gartner’s Hype Cycle Gartner’s hype cycle illustrates the phases of expectation a technology goes through when it is new to the market. A new technology is typically first received with over-enthusiasm from media and users, which is followed by a period of disappointment after which the technology either ends up as a fad or arrives to an understanding of how it is relevant to the marketplace (Fenn & Raskino, 2008). The model distinguishes a hype that becomes a fad from one that lasts and eventually becomes a megatrend (Järrehult, 2011). The model may be used in the process of deciding whether, how and when to invest in a new technology (Steinert & Leifer, 2010). The hype curve is formed by two separate curves. The first curve, sub-curve A, illustrates the level of hype, which initially is high due to an irrationally positive reaction to the introduction of a new technology (Steinert & Leifer, 2010). The second curve, sub-curve B, illustrates an s-curve, that is related to how the technology matures over time as described in Chapter 2.3 and also to rising expectations among the early majority. The two curves together form the hype cycle, as seen in Figure 5 below. The hype cycle therefore illustrates the mismatch between human expectations and the nature of innovation, where expectations rise quickly while the performance gains and adoption of an innovation develops more slowly (Fenn & Raskino, 2008). 11 Figure 5 - The Hype Cycle in a combination of the hype level and the technology S-curve (Steinert & Leifer, 2010). A technology that is new to the market goes through five distinct phases in the hype cycle. The stages include the innovation trigger, peak of inflated expectations and the trough of disillusionment, and if it does not turn out to be a fad, also the slope of enlightenment and plateau of productivity (Steinert & Leifer, 2010). These five phases have different characteristics, as illustrated in Figure 6 below. During the innovation trigger, a public announcement or breakthrough draws media attention to a new technology (Steinert & Leifer, 2010). When the innovation reaches the peak, publicity has generated over-enthusiasm and unrealistic expectations, while there is still very low adoption in the marketplace and the performance of a product is poor (Fenn & Time, 2007). The press however reinforce the need to take part in the technology or be left behind. This often results in companies investing quickly in the hyped technology without having a clear strategy, and thus attempting to apply it to many different settings. After some time, the hasty investments and over-enthusiasm result in products that fail to meet the high expectations of performance and adoption rates. These disappointments are then spread by media and the negative hype makes the technology slide into the trough of disillusionment. At this point, some early investors opt out of the technology as it becomes widely criticised and unfashionable. There is however not always a drop in the overall adoption in the trough, but instead the anticipated rapid growth may slow down. The lasting value of a new technology is rarely found before the disillusionment sets in (Fenn & Raskino, 2008). Therefore, as the technology eventually reaches the slope of enlightenment investors start to experience performance improvements and the technology becomes better understood. At this point, there is however still a very little part of the consumer base that has adopted the technology. After some additional time, the technology demonstrates to be successful in the marketplace and the adoption accelerates into the plateau of productivity (Steinert & Leifer, 2010). It is possible that a technology experiences more than one hype before reaching this plateau or that it does not reach the plateau at all (Fenn & Raskino, 2008). 12 Figure 6 - The 5 phases of the Gartner Hype Cycle (Steinert & Leifer, 2010). When deciding whether and when to invest in a new technology, companies should avoid to invest in an innovation simply because it is hyped, as the expectations of the value of the technology are often exaggerated. At the same time, companies should also avoid to dismiss an innovation based on that it is not yet living up to early over-expectations (Linden & Fenn, 2003), see Figure 7. A key consideration also includes that some innovations never make it through the entire hype cycle and fail to ever reach the plateau of productivity. This could for example happen when an innovation is replaced by another technology, or when it becomes embedded in another technology. In these cases, the actual capabilities of the innovation tend to not fall off the cycle, but a specific technique to reach the capability may fall off the cycle in favour of another technique (Fenn & Raskino, 2008). Another reason that innovations do not progress beyond the trough is because they are fads, that is, when there turns out to not be enough physical or intrinsic value in the innovation for it to endure (Fenn & Raskino, 2008). 13 Figure 7 - Avoid basing investment decisions on the level of positive or negative hype (Linden & Fenn, 2003). The scale of each technology's hype cycle depends on its overall perceived importance to businesses and society. Technologies that appeal to a large number of enterprises or consumers will attain much higher levels of exposure and therefore hype than those that affect only a limited amount of stakeholders (Linden & Fenn, 2003). Different technologies may also take different routes in the hype cycle, as illustrated in Figure 8 below. In addition, the time to go through the cycle may also vary between different technologies: some may take years and others take decades. Fenn and Time (2007) outline three different adoption speeds. Normal technologies usually take between five to eight years to go through the cycle. A fast-track technology goes through the cycle within two to four years because of a high value, ease to use, possibility to use an existing infrastructure and a quick transition from consumer to corporate use. An example of such a technology is Short Message Service that went through a negligible trough of disillusionment and reached maturity relatively quickly (Fenn & Time, 2007). A long-fuse technology on the other hand may take several decades, as it is often characterised by an unrealistic fascination that is far more advanced than the actual technology, high complexity, regulations of adoption and reliance on a new infrastructure and new business models. For example, object-oriented programmering took 10 to 15 years to become widely adopted because of barriers in the form of established development processes and skills (Fenn & Time, 2007). 14 Figure 8 - Different routes a technology may take, take special notice of the possibility of a hype ending up as a fad. In order to position an innovation on the hype cycle, the level of hype and maturity may be quantitatively analysed by studying the visibility and user interest. Visibility, or expectations, may be assessed by extracting the number of newspaper articles that have been published on the topic over time. The user interest may be analysed by the number of search requests that users have made on the topic (Fenn & Raskino, 2008). The position in the hype cycle may also be assessed from a qualitative point of view, studying the tone and nature of media currently published on the topic. In the beginning of the hype cycle, media tends to focus on the innovation itself its the future possibilities, and in the end of the cycle focus instead lies on the applications, practicalities and results. Stories also tend to be overly positive at the peak whereas they turn negative and cynical towards the trough (Fenn & Raskino, 2008). Table 1 on the next page outlines the main characteristics of each phase in the hype cycle and may therefore be used to position an innovation in the hype cycle. Table 1 - Characteristics of the different sections in the hype cycle (Fenn & Raskino, 2008). Innovation Trigger Peak of Inflated Expectations Trough of Disillusionment Slope of Enlightenment Plateau of Productivity Suppliers funded by venture capital Frequent positive stories in press Negative press Positive press, but more modest High-growth adoption phase Few sellers on the market Many sellers General cynicism is spread 2nd & 3rd generation products 20-30% adoption Considerable customization needed Phenomenon has a nick name Consolidations and buy-outs from larger companies Offer embedded into other tools or services High price Suppliers use early adopters as references Less than 5% of potential audience has adopted Consultants offer solutions on how to adopt Speculations of future impact Reliable data on cost and value 15 Based on the above, the dilemma lies in attempting to predict what hypes will become megatrends in order to gain a competitive advantage, or to manage the gap between the hype and productivity without taking too large risks. There are three ways companies deal with the gap between the hype and the point where an innovation provides value (Fenn & Raskino, 2008). So-called type A companies are often lead innovators that deliberately adopt high-risk strategies for high potential rewards, thus invest in or before the peak. Type B companies are moderately willing to take risk and adopt new innovations in the through. These companies are both lead innovators and fast followers but tend to be relatively unsuccessful. Lastly, type C companies are neither willing nor have the capability to take large risks, and hence they invest at the slope of enlightenment. Type C companies are usually the most successful types of fast followers (Fenn & Raskino, 2008). Järrehult (2011) suggests that in order to cope with this, companies should start investing in hypes in small scale in the beginning of the cycle. This will enable firms to gain valuable experience that will lead to an advantage if the hype becomes a megatrend, while not taking too large risks if the hype becomes a fad. The hype cycle is specifically developed to explain the advancement of technological innovations, but the phenomenon may also be observed in other industries and on other levels (Fenn & Raskino, 2008). Linden and Fenn (2003) argue that most technologies conform to the hype cycle because what is common in all cases is the human factor, i.e. how people react to a novelty, regardless of the nature of the technology. Fenn and Raskino (2008) furthermore highlight three reasons related to the human nature for why hype driven expectations interfere with the maturity curve. Firstly, people have a novelty preference, meaning they are inclined to be positive towards novelties and use imagination to overestimate new phenomenon to some degree. Secondly, humans are subject to social contagion, meaning that they are sensitive to what other people are doing and saying which can create the self-reinforcing effect of a hype. Lastly, decision heuristics play a large role, meaning that when somebody has become positive towards a phenomenon they tend to remain positively biased to this idea. Due to the relation between the hype cycle and the nature of human behaviour, the hype cycle may be highly applicable to innovations within less technology-intensive fields, such as the health food industry. The health food industry is furthermore naturally relevant to a large number of people and thus subject to extensive media attention, which is likely to fuel positive and negative hypes. 16 2.5 Relating the Diffusion Curve to Gartner's Hype Cycle According to Fenn & Raskino (2008), the trough of disillusionment coincides with the chasm described by Moore (2001), as illustrated in Figure 9 below. As previously seen in Figure 6, the early adopters begin to investigate a new technology during the innovation trigger, indicating that both innovators and early adopters contribute to creating the hype. At the peak of inflated expectations, adopters beyond the early majority also begin to get involved. In the trough that follows, the adoption decreases in speed as previously mentioned, and in order to make it out of the trough the early majority needs to be reached, thus the chasm needs to be crossed. This is therefore where the chasm coincides with the hype cycle. Finally, given that the innovation successfully crosses the chasm between the early and the mainstream market, it also reaches early majority. As the adoption accelerates, the technology enters the plateau of productivity in the hype cycle. The percentage rates denoted in Figure 2 and Figure 6 separately illustrating the diffusion curve and the hype cycle do however not coincide in the below figure. This could be due to that the adoption figures are estimations rather than precise guidelines. In addition, it should be noted that the two theories are of very different age. Thus, the differences may also be due to a change in industry dynamics over the years. Figure 9 - The diffusion curve in relation to Gartner's hype cycle, figure created by the authors of this thesis. 17 3. Method This chapter aims to outline the method applied to answer the research questions of this master’s thesis, and thus to fulfil the purpose. The chapter will cover the research strategy, the process of the research and the data collection of the study. This will be followed by a description of how the data has been analysed and finally a discussion of the quality of this research. 3.1 Research Strategy The aim of this study is to identify growth opportunities within the health food industry, and in doing so, discuss how the hype cycle may be applied in this setting. Since the key variables of the purpose, in this case referring to growth opportunities, are not completely defined, the design of the study is considered to be exploratory, in accordance to Ghauri and Grønhaug (2010). This requires a rather flexible approach (Ghauri & Grønhaug, 2010). The research strategy is inspired by a deductive approach, which according to Bryman and Bell (2011) uses existing theory to develop hypotheses, which then are confirmed or rejected by the findings of the data collection. Similarly, the hypothesis of this thesis is that the hype cycle could be used to explain the pattern of expectations in the health food industry, which is tested through applying the theory and analysing differences between health food attributes and technological innovations. This study furthermore follows a mixed approach of qualitative and quantitative research, with a predominantly qualitative approach. According to Bryman and Bell (2011), a qualitative study emphasises words rather than quantification of data and includes the consideration that the reality is constantly shifting. Thus, this was chosen in order to be able to consider the complexity that follows from the undefined industry boundaries and complexity in defining growth opportunities. Qualitative research is also descriptive with focus on providing an understanding of the studied phenomena (Gillham, 2010), which was also essential in order to fulfil the purpose of this thesis. A quantitative approach in the form of a survey was also used to support and validate the analysis of the qualitative investigation. 3.2 Research Process The research process of this study consists of four main parts. Firstly, a definition of the health food industry is developed through identifying what currently characterises health food. Thereafter, growth opportunities in the form of hypes and trends of health food characteristics are identified. This is followed by an analysis of how the hype cycle may be interpreted in the setting of health food. Lastly, appropriate strategies to manage hypes and trends are identified. In order to do this, the research followed a process as illustrated in Figure 10. The four parts are further described in the following sub-chapters. 18 Figure 10 - Research process illustrated in four steps, where the two last steps are derived from the second. 3.2.1 Definition of the Health Food Industry In order to develop a broad definition of the health food industry, the history of dietary advice is outlined and analysed with the theory of scientific paradigms, to explain why consumers consider certain food more healthy than other today, and why their perceptions of healthy food changes over time. Thereafter, factors characterising health food are identified and investigated through secondary sources and a conducted survey, in order to set the scope of what characterises health food today. These factors are called characteristics or attributes throughout the report. All the identified attributes are related to contents in food, for example being low in some ingredient, free from some ingredient or high in some nutrient. The chosen attributes are presented in Chapter 4.4.3. 3.2.2 Positioning of Characteristics in Gartner's Hype Cycle As previously mentioned, the health food industry is highly influenced by hypes and trends that affect consumers’ perception of health food overt time. Thus, in order to identify growth opportunities within this industry, it is essential to analyse which of the identified characteristics of health food that are hypes versus megatrends, and which current hypes that are believed to result in megatrends. The identified characteristics are therefore positioned in and analysed using Gartner’s hype cycle. Usually, Gartner's hype cycle is used in a technological context, but the health food industry has been observed to also Identify Growth Opportunities through Applying the Hype Cycle on the Health Food Industry 1. Define the Health Food Industry Understanding what affects consumers’ perceptions of health foods • Scientific paradigms • History of dietary advice Current characteristics of health foods • Industry reports • Survey among Swedish consumers 2. Identify Hypes & Megatrends in terms of Health Food Attributes Positioning of characteristics in Gartner’s hype cycle • Visibility • Search interest • Qualitative analysis of content • Scientific research • Macro trends & consumer behaviour 3. Interpret the Hype Cycle in the Context of Health Food Industry Main differences between technology and health attribute • Implications on the interpretation of the hype cycle analysis 4. Identify Strategies to Manage Hypes & Trends in the Health Food Industry How to manage hypes • Short-term How to manage megatrends • Long-term 19 experience the phenomenon illustrated in the hype cycle. Due to the fact that it is the human behaviour that creates the hypes, as outlined in Chapter 2.4, it is likely that a similar phenomenon is occurring within the health food industry as well. Thus, Gartner's hype cycle theory was applied in the context of health food and this analysis resulted in an identification of characteristics that have the potential to provide long-term growth opportunities. In order to assess where the investigated attributes may be positioned in the hype cycle, they are addressed both quantitatively and qualitatively. From a quantitative point of view, the visibility is illustrated through the number of newspaper articles published over time and the user interest through the relative change in search requests made on the topic over time. Thereafter, from a qualitative point of view, the articles are analysed in terms of focus and tone as proposed in Chapter 2.4, in order to determine the current level of hype. An attempt is also made to determine whether, or to what degree, the certain characteristic has a scientifically proven positive effect on health in order to find indicators for whether the hype has a possibility to become a megatrend. Lastly, in order to further analyse the long-term potential of the identified attributes, the study also includes how macro trends and changes in consumer behaviour may influence these. 3.2.3 Interpreting the Hype Cycle in the setting of the Health Food Industry Due to the fact that the hype cycle is originally developed to explain a phenomenon related to technological innovation, an analysis of how the theory may be interpreted in this particular setting is conducted. More specifically, the main differences between a technology and a health food characteristic is discussed, followed by what implications these differences have for the application of the hype cycle on the health food industry. This is done based on the findings and what has been learned in the previous step of the research process. 3.2.4 Managing Hypes and Megatrends The final part of the study aims to analyse how a firm could manage the hypes or megatrends identified in through the hype cycle analysis in order to enable long-term growth. This is done using the theory of Gartner's hype cycle about how to manage hypes and megatrends as well as applying the literature about crossing the chasm between the early and the mainstream market. 3.3 Data Collection The data collection mainly consisted of qualitative methods, due to the exploratory and rather unstructured nature of this study. However, as previously mentioned, a quantitative survey has been used in addition to this in order to further support the qualitative analysis. According to Ghauri & Grønhaug 20 (2010), qualitative data is useful in order to provide insights and an understanding of concepts. Collection of data in qualitative research is usually conducted simultaneously as the analysis, often in an interactive manner, in order to enable new questions and further collection of data. This is often necessary due to the unstructured research problem that becomes more clarified as more data is collected, which in turn clarifies what new data is needed to collect (Ghauri & Grønhaug, 2010). This was the case in this master’s thesis, especially since the data collection for the first part, i.e. the definition of the health food industry, set the scope of the following parts of the research. The data collection consisted of interviews, secondary sources as well as a survey, all described in further detail in the below chapters. 3.3.1 Interviews Interviews are usually considered appropriate when conducting an exploratory study (Ghauri & Grønhaug, 2010), and were also used in this study in order to gain a deeper understanding of Healthy Inc. as well as insight into the health food industry and what characterises health food. Hence, both internal and external interviews have been conducted. The internal interviews were conducted at Healthy Inc. in order to gain a basic understanding of the company and its market. Six employees were interviewed, who had positions as CEO, Head of Regulatory and New Product Development, Market Director and Brand Managers of the three largest brand categories. The interviews were all unstructured, which Ghauri and Grønhaug (2010) describe as interviews where the interviewee is given almost full liberty to discuss the particular topic, based on a few lead questions from the interviewer. Unstructured interviews are advantageous in the context of discovery (Ghauri & Grønhaug, 2010), as in this case. To gain an understanding of health food and the health food market, several external interviews were also conducted. These followed a semi-structured approach, which according to Bryman and Bell (2011) is useful as it allows for deviation from the interview guide and enables follow-up questions when interesting answers have been given. Furthermore, the questions used were open, which is also useful when the aim is to get insights into an area where the researcher has limited knowledge (Bryman & Bell, 2011), as in this case. Firstly, an associate professor and lecturer in the area of nutritional sciences at the University of Gothenburg was interviewed regarding the evolution of dietary advice and preferences among consumers today. However, it was concluded that this type of expert did not have much to add to the information already collected from secondary sources, and when asking questions regarding preferences among consumers today, it became clear that nutritionists working more closely with people would be more appropriate. Therefore, three additional semi-structured interviews with nutritionists were conducted, due to the fact that they are working more closely to consumers. However, these interviews 21 only revealed general advices about healthy eating, such as eating a varied diet, and did not reveal much about specific characteristics of health food or health trends. Consequently, this was followed by a semi- structured interview with a senior brand manager at ICA. ICA is currently investing in expanding their product range within health food (ICA Gruppen, 2014), and it was therefore considered interesting to understand how they characterise health food. Moreover, experts within the area of consumer behaviour were also contacted, which provided reports in the subject of health food consumption. All interviews were conducted over the phone. The interview guides may be found in Appendix 1. 3.3.2 Survey A survey has been used to support the definition of the industry in its current state through looking at what characteristics Swedish millennials currently associate with health food. Millennials have in this case been defined as people that are currently between 21-34 years of age, i.e. born between 1980-1995, similar to the definitions made by Kavounis (2008) and Fromm & Garton (2013). The focus on this age group is due to the fact that millennials are said to set the trends in the consumption of health food today (Cuppett, 2014), which is further described in chapter 4.4.1. As suggested by Ghauri & Grønhaug (2010), a review of the current knowledge in terms of previous similar investigations was performed before constructing the survey, and an understanding of different health attributes was formed through studying reports covering the worldwide and the US market. However, as Nielsen (2015) states, it would be misleading for firms to assume that consumers in different regions have the same preferences, and thus, a similar descriptive survey of the perception of health attributes among some Swedish consumers has been performed. The 14 investigated attributes were chosen qualitatively as suggested by Ghauri & Grønhaug (2010), through studying and qualitatively analysing reports and previous research on the subject. The researched population includes millennials currently living in Sweden. A survey should aim to access a representative and random sample of the researched population (Ghauri & Grønhaug, 2010). It is however resource- and time consuming to find a means that guaranteed the sample to be completely representative while at the same time managing to collect enough answers. Thus, this survey was conducted using convenience sampling. Convenience sampling is a non-probability sampling method used where entities in the population are included according to their ease of access rather than randomly (Easterby Smith, Horpe and Jackson, 2013). When using this type of sampling, inferences from the sample on the population cannot be drawn with the same level of confidence as for probability sampling methods (Easterby Smith et al., 2013). The results will therefore be analysed with particular attention drawn to the bias that this sampling method is likely to introduce. Bias is in this case introduced primarily in terms of geographic location and level of education, which is further discussed in chapter 3.5 Research Quality. The sample was collected through posting the survey on the Internet through Facebook, Twitter 22 and blogs. These channels were considered appropriate as they enable reaching a wide audience of millennials. The target minimum size of the sample was set to 100, and 147 people in total answered the survey out of which 137 were millennials. The nature of the questions allowed for a respondent-completed questionnaire, which means respondents fill in the questionnaire without interference from the interviewee which allows for it to be distributed online (Easterby Smith et al., 2013). The survey was created through the online service SurveyMonkey and consisted of closed questions. In order to obtain as many answers as possible, the survey was kept short with only five questions covering the 14 characteristics, which was enough for the purpose of this investigation. The first two questions were background or qualifying questions, determining the gender and age group of the respondent. The third question aimed to determine to what degree the respondent made active decisions to buy and consume healthy food today, in order to be able to analyse the difference between different respondent groups. The two remaining questions asked the respondent to firstly grade the health importance of 14 attributes on an ordinal scale, and then answer whether they would be willing to pay more for the same 14 attributes. On an ordinal scale, the importance lies in the order of the possible answers, but the exact distance between every answer is not quantifiable but rather subjective (Cunningham & Aldrich, 2012), as in this case where the possible answers ranged from “very important” to “not at all important”. The 14 characteristics were programmed to appear in a randomised order for every respondent in order to avoid bias. The respondent also had the possibility to add additional comments on what food attributes they considered important. It was furthermore deemed necessary to offer respondents a possibility to choose “no opinion”, as they may not know what they think of a certain attribute. Forcing the respondents to take a position would likely result in random and misleading answers when a respondent does not have an opinion. After constructing the questionnaire it was sent out to three people that were allowed to provide feedback, whereupon the questionnaire was revised. See questionnaire in appendix 2. 3.3.3 Secondary Sources In order to answer the research questions of this master’s thesis, the data collection also involved collecting data from secondary sources. In order to define the health food industry and what characterises health food, secondary sources in the form of market reports have been used in combination with books and articles. In addition, information from websites of health food organisations and nutritionists have also been used for this purpose. Thereafter, in order to identify macro trends and how consumers’ behaviour has changed, trend reports from several consultancy firms as well as from market research organisations have been used. 23 Secondary sources have also been used to position attributes in the hype cycle. When assessing visibility in terms of the number of newspaper articles published over the last years, the Chalmers library database Summon has been used. The search made includes newspaper articles published between 1994 and 2014. There is an option to include results beyond the library’s collection, which has been chosen in order to minimise bias in terms of the type of databases Chalmers provides. Newspaper articles have been chosen as they most accurately display the level of media attention at a certain point in time, whereas for example journal articles are often published with some delay. Different search phrases have been used in the searches in order to find the search producing most accurate results. It should be noted that the visibility does not necessarily directly reflect the level of hype, as the data include both positive and negative articles published on the particular subject. Thus, it is possible that a peak in the number of published newspaper articles actually represents the trough of disillusionment in the hype cycle, since the peak might reflect a high amount of negative articles getting published in the subject. This has been taken into consideration in the analysis through also qualitatively analysing the specific content and tone of published articles. Thus, also newspaper articles, both published internationally and in Sweden, have been analysed in order to position attributes in the hype cycle. These have been found through Google's search feature called News, where the most recently published news on the particular health food characteristics were found. Google News was chosen instead of Summon since this enabled searching for Swedish articles, which was not possible in Summon. Furthermore, the selection of the articles for the qualitative analysis was based on the relevance of the content, the publication date as well as the availability of the newspaper for consumers. Furthermore, Google trends has been used in the assessment of user interest. Google trends include a feature that illustrates the relative search interest of a search term from 2004 to the present. The absolute number of searches is not available, but is not needed in this analysis as the interest primarily lies in how the search numbers have changed over the past years. Google trends also enables searching on topics, which means looking at the search interest on not only one phrase but for all related phrases. When such a topic has been available and appropriate, it has been used as it is deemed to be more comprehensive than only one specific search term. All graphs that are dashed indicate that a topic has been used instead of a specific search phrase. 24 3.4 Data Analysis Below follows a description of the methods used to analyse the qualitative interview data and data from secondary sources as well as the quantitative survey data. 3.4.1 Interview and Secondary Data Analysis As stated by Bryman & Bell (2011) one of the main difficulties with qualitative research is that it easily generates large amounts of data that is cumbersome to analyse. In order to facilitate this, the method of qualitative data analysis has been inspired by the Grounded Theory framework. This approach firstly includes theoretical sampling, meaning an iterative approach to data collection and analysis, where the emerging data constantly directs the process of data collection (Bryman & Bell, 2011). This enabled an iterative approach to data collection through interviews, where the findings directed the development of the research in terms of what further interviews to perform as described in Chapter 3.3.1. Secondly, grounded theory encompasses iterative and continuous coding of data, which entails breaking down the data into sections and give these specific names (Bryman & Bell, 2011). This has been done both when analysing macro trends, as seen in Table 5 and Appendix 3, as well as when analysing health characteristics, as seen in Table 2. The data has been coded along with its collection in line with this theory (Bryman & Bell, 2011). Grounded theory results in categories and properties of these categories (Bryman & Bell, 2011), where the properties in this case are the health attributes or characteristics as well as specific macro trends. As previously mentioned, secondary data was also collected for every health characteristic in the form of visibility, search interest, content and tone of newspaper articles as well as scientific research. The analysis of these data followed the method proposed in the theory of Gartner's hype cycle, presented in Chapter 2.4. 3.4.2 Survey Analysis The survey data has been analysed both in terms of all respondents and through separating the results between two different respondent groups. As the innovators and early adopters are those that adopt new trends earlier and therefore may be able to predict what the majority will adopt in the future, these formed a separate group from the rest of the respondents. The innovators and early adopters are in this instance defined as those who have claimed that they always make active choices to buy and consume healthy food. This group will through the rest of the report be referred to as health enthusiasts or simply enthusiasts. The rest of the respondents are considered as early and late majority or laggards. By dividing the results of these two groups, any large differences will be analysed and discussed in order to see whether some attributes are more important to one group than the other. However, as the survey has been 25 carried out using convenience sampling, no statistical methods evaluating whether the differences are statistically significant or not may be used (Cunningham & Aldrich, 2012). Thus, the result from the survey is only used to find indications of the preferences of Swedish consumers. 3.5 Research Quality The research may firstly be evaluated in terms of its reliability and validity. According to Bryman & Bell (2011), reliability concerns whether a study is repeatable and the measures are stable, and therefore creates consistent results. However, qualitative research generally does not generate consistent results. This is due to a number of factors such as the specific situation the research is conducted in and the interpretation of the results. Thus, reliability is generally less applicable to qualitative research (Bryman & Bell, 2011). Consistently with this, the unstructured and semi-structured interviews and qualitative analysis in this study make it difficult to replicate precisely. The reliability has however been increased by thoroughly describing the research method to enable replication of the study to an as large degree as possible. Furthermore, there are different types of validity, out of which ecological validity is mostly applicable to qualitative research (Bryman & Bell, 2011). Ecological validity concerns whether the results of a study are not only technically viable but also hold in people’s everyday actual social setting (Bryman & Bell, 2011). In this study, all interviews have been carried out in people’s own social setting which implies high ecological validity. However, as Bryman and Bell (2011) states, surveys tend to have low ecological validity as answering a questionnaire is in itself an unnatural task. There are several additional aspects of the quality of this study that are interesting to reflect upon. First of all, the studied characteristics of health food have been chosen based on two previously conducted reports. It is possible to argue that more references could be beneficial in this case, in order to get more trustworthy results. However, these were the only two comprehensive and recent reports found that were comparable and analysed characteristics on the same level of analysis. In addition, there are evidently more than the 14 chosen health attributes describing health food today. The 14 in this report are qualitatively chosen as most important, but can not be argued to represent a fully comprehensive view of health food. The time and resource constraints of this study however motivates choosing a limited number of attributes, and the methods used when analysing these may be used to analyse additional attributes of interest. The use of the Chalmers library database may also not perfectly reflect all articles published in the subjects, even though including results beyond the library's collection. It however gives an indication of the change in interest in the topic over time, which is what is important for the analysis. Furthermore, the searches on Google Trend would preferably had been on searches in Sweden only, but unfortunately 26 many of the attributes had insufficient search quantities in Sweden and could therefore not be plotted. Thus, worldwide search interest has been used instead. This could reduce the quality of the findings as there may be cultural differences between countries, but the risk of this has been minimised through a qualitative analysis of Swedish articles as well. In the conducted survey, a complete sample frame of the millennial population of Sweden was not possible to obtain due to time and resource constraints, and thus a random sampling method was not possible. The convenience sampling method used is however likely to have introduced bias. Most survey answers are likely to be from people within the networks of the two authors, and the result is therefore not a perfectly representative sample of the Swedish millennial population. The sample is expected to be skewed towards a specific geographic area as well as to a higher than average level and specific type of education. In addition to this, respondents may be prone to answer that they make active choices to be healthy to a larger extent than they actually do, as people tend to want to perceive themselves as healthy. Hence the measurement dividing enthusiasts and others is quite subjective. The measurement is however in line with AlixPartner’s (2013) findings stating that people who spend more than 40% of their food budget on health food are 43% more likely to rate themselves as health-conscious customers. It is furthermore a simplified assumption that the ones stating that they always make active choices to eat healthy are necessarily innovators or early adopters for all types of health food. The assumption is however made based on that people that make active choices to eat healthy are more likely to be interested in finding new information and research on the topic and therefore also most likely to try new things early on. 27 4. Empirical Study The empirical study will firstly outline a background of health food, which will be followed by a brief history of dietary advice, showing one example of how it may change over the years. Thereafter, the current dietary advice in Sweden will be presented in the form of the Nordic Nutrition Recommendations. The next part of the empirical study aims to firstly identify what attributes consumers actually perceive as healthy today, and these will then be further investigated in terms of their media exposure, search interest and research evidence. The final part of the empirical study will outline macro trends affecting the health food industry, and thereafter changes in consumer behaviour. 4.1 Health Food Background Health is, according to World Health Organisation, defined as: "Health is a state of complete physical, mental and social well-being and not merely the absence of illness or infirmity" (World Health Organisation, 2015). In order to determine physical health, the following key areas are often considered: physical activity, alcohol and drugs, rest and sleep, medical self care and nutrition and diet (Koshuta, 2015). The diet, i.e. what we eat and how much of the different components we eat, is considered one of the most important determinants of health (Santich, 2005). A well-balanced diet should consist of carbohydrates, proteins, fats, vitamins and minerals (Koshuta, 2015). However, the proportions of these components depend on the particular individual's needs (Hälsosidorna, 2015), allowing society and media to continually discuss what food is beneficial in order to live a healthy life. Nevertheless, the interest in healthy eating is increasing among consumers. In a report published by AlixPartners (2013) the large interest in healthy eating among American consumers is clearly shown, where 92 % of the respondents claimed healthy eating was very important or somewhat important to them as seen in Figure 11 below. However, it could be argued that people are likely to claim healthy eating is important to them without actually incorporating it into their lifestyles. Figure 11 - How important is eating healthy? (AlixPartners, 2013) 28 The primary difficulty in defining health food is the fact that health is individual, meaning what food will be perceived as healthy depends on the individuals’ own needs, definition and perspectives of healthy living. Furthermore, consumers’ perception of what is healthy to consume is also largely dependent on what society and media believes, i.e. the dietary advice they provide to consumers (Santich, 2005). According to Santich’s (2005) study of dietary advice in Australia from the middle of 1800s until now, it is clear that the dietary advice has changed significantly during these years, which in turn also changes the consumers’ perception of healthy eating. Even if the dietary advice is based on recommendations from medical and nutrition experts, which in turn base their recommendations on scientific studies, the understanding of what is healthy changes as new nutritional knowledge emerge and as the understanding of metabolic processes improves (Santich, 2005). Santich (2005) also states another important aspect in the article: even if the advice is based on scientific studies, the knowledge is nevertheless socially and culturally influenced, which is why dietary advices should be evaluated within the particular context, taking into account social, cultural, economic and practical aspects besides nutrition. 4.2 A Brief History of Dietary Advice In order to understand what food is considered to be healthy today, it is interesting to understand how dietary advice has changed over the years. In the 1960s, health concerns from malnutrition had been overcome and instead the problem with over-nutrition arose (Santich, 2005). During the 1970s, heart diseases became a primary concern in the US, why the government decided to provide dietary recommendations to the public in order to prevent this type of disease (Harcombe, Baker, Cooper, Davies, Sculthorpe, Di Nicolantonio & Grace, 2015). Scientists developed two competing explanations to this phenomenon based on nutrition: one emphasised on that saturated fat was the cause of the problem, while the other theory stressed the effects of sugar (Eenfeldt, 2011). However, despite that the theory of saturated fat as the cause of the problem was lacking in evidence, it still got accepted by both the American, the British and the Australian government due to a strong consensus of opinion (Harcombe et al., 2015; Santich, 2005), while the latter theory was rejected (Eenfeldt, 2011). Based on this, in 1977 the US government released recommendations for a healthy diet, which involved reducing overall fat consumption to 30 % of total energy intake and saturated fat to 10 % (Harcombe, et al., 2015; Taubes, 2001). This was followed by the UK, which in 1983 provided the same recommendations to the British people (Harcombe, et al., 2015). However, by only focusing on reducing fat consumption, the public health team did not pay enough attention to other risks, such as increased intake of carbohydrates, which today is believed to be causing the increasing obesity and diabetes in the Western countries (The Guardian, 2015; Eenfeldt, 2011). Thus, experts are now rather encouraging consumers to reduce the intake of sugar and carbohydrates in order to improve their health (Hite, Goldstein Berkowitz & Berkowitz, 2011; Hu, 2010), which in turn has made governments change their dietary advice to the 29 public. The history shows that there is no clear definition of what constitute healthy food, and instead implies that the beliefs at any one time is dependent on the current scientific knowledge as well as the current circumstances in terms of political, economic and sociocultural values (Santich, 2005). 4.3 The Nordic Nutrition Recommendations The Nordic Nutrition Recommendations (2012) is a report released every 8th year that provides guidelines for the nutritional composition of a diet in order to provide a basis for good health in the Nordic countries. It is developed by a Nordic expert group and published by the Nordic Council of Ministers (Folkhälsomyndigheten, 2015). The report states that Western-type dietary patterns are characterised by a high consumption of processed meats and red meats (i.e. beef, pork, and lamb), a high consumption of food products low in essential nutrients but high in added sugar and fat, and a high consumption of food products with high level of salt. This type of diet is, according to the Nordic Nutrition Recommendations (2012), associated with adverse health effects and chronic diseases. In the report, it is suggested that dietary patterns with high level of vegetables, including dark green leaves, fresh peas and beans, cabbage, onion, root vegetables, fruiting vegetables (e.g., tomatoes, peppers, avocados, and olives), pulses, fruits and berries, nuts and seeds, whole grains, fish and seafood, vegetable oils and vegetable oil-based fat spreads and low-fat dairy products reduce the risk of most chronic diseases. Thus, the Nordic Nutrition Recommendations (2012) are suggesting the consumption of food patterns illustrated in Figure 12 in order to eat healthy. It is however highlighted in the report that the recommendations should be reassessed when appropriate due to the emergence of new scientific knowledge. Thus, these recommendations are not definitive (Nordic Nutrition Recommendations, 2012) but represents what the government of Sweden currently believes constitutes health food (Folkhälsomyndigheten, 2015). These recommendations therefore set the foundation of what is perceived as healthy food today, which also the interview with the senior brand manager at ICA revealed, since she stated that they follow these recommendations and use them as a basis for which products they consider to be good for the health. Figure 12 - The Nordic Nutrition Recommendations. 30 4.4 Attributes Currently Defining Health Food This part of the empirical study aims to provide a more specific understanding of what characteristics consumers in Sweden currently associate with health food. This is done by firstly present and further describe the generation called millennials, often claimed to be driving trends within the health food industry. This is followed by a presentation of previous research on consumer preferences in terms of health food, focusing on two large investigations of consumers worldwide and superusers in the US. This will be followed by the results of the conducted survey of this study, with the aim to provide indications on which health food attributes Swedish consumers perceive as most important and which they are willing to pay a premium for. Thereafter, these attributes are investigated in terms of their media exposure, search interest and support by scientific research, to provide the underlying data for the forthcoming analysis. 4.4.1 Consumers Driving Importance of Health Attributes The consumers currently driving trends within the food industry are the millennials (Cuppett, 2014): a generation that refers to people born roughly between 1980 and 1995 (Kavounis, 2008; Fromm & Garton, 2013). Millennials make up the largest generation of young people in history and are therefore a highly influential consumer group (Nielsen, 2015; French, 2014; Euromonitor, 2015). This generation is well educated but has scarce resources of capital (Food & Friends, 2014), and believe in meaningfulness in both the work they do as well as in the products they buy (Euromonitor, 2015). Hence, the millennials are savvy shoppers, especially since they have limited cash and are reluctant to spend what they do have (Euromonitor, 2015). They have also grown up with technologies enabling price and product comparisons as well as free services, games and media (Euromonitor, 2015). In addition, they use digital communication technologies to manage their lives and work (Mendelson, 2013). Millennials are also driving development of the health and wellness industry (Food & Friends, 2014; Cuppett, 2014). It is clear that they are aware of environmental concerns, and they therefore value for instance recyclable packaging of the food they consume, or in the best scenario, no packaging at all (Food & Friends, 2014). They want to purchase products that contribute to a better world, but they are not prepared to pay much for it (Food & Friends, 2014). For instance, many of the millennials instead chose to eat less meat and processed food (Food & Friends, 2014; Cuppett, 2014). 4.4.2 Current Research on Classification of Health Attributes Nielsen (2015) recently investigated what health attributes consumers in 60 countries worldwide are looking for and whether they are willing to pay more for these. Similarly, AlixPartners (2013) have investigated the health and wellness superuser in the food and beverage sector in the US, looking into what these consumers consider to be healthy food and to what degree they are willing to pay a premium 31 for certain features. The report by Nielsen (2015) was chosen due to the fact that Nielsen is a leading market research firm with access to comprehensive consumer data worldwide. The report by AlixPartners, investigating superusers in the US, was also of interest in terms of this study, since trends in the US often is adopted in Sweden some years later. According to both reports, the rapidly growing health awareness among consumers is going to lead to new opportunities and a larger market for consumer product firms. Nielsen (2015) outline four major categories of factors characterising healthy food. Firstly, consumers worldwide go back-to-basics, meaning they want to consume natural food with certain health benefits. Such food include natural, fresh and minimally processed food with no GMO. Secondly, consumers look for functional food providing benefits that either reduce the risk of disease or promote good health. Functional food are for example high in fiber, protein, whole grain, or fortified with calcium, vitamins and minerals. A third category, less is more, includes food that are low in cholesterol, sugar and fat, and free from caffeine, gluten and high fructose corn syrup. The last category includes environmental and socioeconomic concerns, that is, whether the food is sustainably sourced, organic and produced with local herbs/ingredients. The global average ranking of attributes and willingness to pay for these according to the Nielsen (2015) report is displayed in Figure 13 below. Considering how different age groups have rated importance, millennials (aged 21-34) rate health attributes highest, and percentages are lowest among the respondents aged 65 and older. When investigating whether consumers are willing to pay a premium for healthy attributes, Nielsen (2015) concludes that the percentage of respondents that consider an attribute very important when buying is consistently higher than the percentage that are very willing to pay a premium for the same feature. The only exception is organic food, having an equal percentage claiming it is very important as being very willing to pay a premium. Furthermore, the ones that are willing to pay most for health attributes are millennials and generation Z (aged under 20). Nielsen (2015) means that as the purchasing power of millennials increases over the coming years, companies within the health food industry making an effort to connect with this generation can increase their chances of success. 32 Figure 13 - Percentage of respondents worldwide ranking an attribute very important and the percentage that are very willing to pay a premium according to Nielsen (2015). Chart compiled by the authors of this thesis. AlixPartners (2013) similarly outline the preferences of US consumers, focusing on the health and wellness superusers. Superusers are defined as those who spend more than 40% of their yearly food and beverage budgets on health products. These types of consumers are generally more willing to pay a premium for certain health and wellness features, and also have clear preferences in terms of retail channels that may be used by firms to target these customers. This type of consumer is also more likely to have a higher income, a graduate degree and are more likely to live in the west of the US (AlixPartners, 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% Gluten-free Caffeine-free Low/no carbon hydrates Rich in unsaturated fats Mincronutritient-fortified No high fructose corn syrup Local herbs/ingredients Portion control Low/no calories Minerals-fortified Low/no fat Whole grain Calcium-fortified Vitamin-fortified Low sugar/sugar free High in protein Low salt/sodium Organic Ingredients sourced sustainably Natural flavours High in fiber Low/no cholesterol Made from vegetables/fruits No aritficial flavours No artificial colours All natural GMO-free Importance and willingness to pay % rating factor very important % very willing to pay more 33 2013). Superusers constitutes 26% of the US population, but accounts for 61% of the national food and beverage spending related to health which gives them a disproportionate impact on the market. This makes superusers a highly relevant consumer group to investigate for firms within the health food industry. Figure 14 below illustrates the importance and willingness to pay among this type of user in the US market. Figure 14 - The perceived importance of certain attributes and the willingness to pay a premium for these. Chart by AlixPartners (2013). The importance of certain features commands for significant price premiums that may be utilized by consumer product firms. However, AlixPartners (2013) note that importance is not the only factor driving purchase decisions, but price and performance has a significant impact that cannot be overlooked by health food companies. AlixPartners (2013) therefore conclude that superusers are an attractive but complex target group, with high awareness of price and high demands on quality and convenience despite a willingness to pay more for certain health characteristics. AlixPartners (2013) have also found significant confusion when it comes to health claims on product labels. For example, there is no consensus on what criterion define a product as “natural”, and 31% of 34 survey respondents do not know what it means that a product is genetically modified. Consumer confusion is also created from the use of scientific names on labels, where for example 97% of consumers