IN
DEGREE PROJECT MECHANICAL ENGINEERING, SECOND CYCLE, 30 CREDITS
STOCKHOLM SWEDEN 2019,
Characteristics of Early Adopters and Early Majority Adopting a
Vertical Social Network
A case study of the applicability of the Diffusions of Innovations theory in a Vertical Social Network
SOPHIE HANTMAN KOLLÉN REBECCA MANHEM
KTH ROYAL INSTITUTE OF TECHNOLOGY
IN
DEGREE PROJECT DESIGN AND PRODUCT REALISATION, SECOND CYCLE, 30 CREDITS
STOCKHOLM SWEDEN 2019,
Characteristics of Early Adopters and Early Majority Adopting a
Vertical Social Network
A case study of the applicability of the Diffusions of Innovations theory in a Vertical Social Network
REBECCA MANHEM
SOPHIE HANTMAN KOLLÉN
KTH ROYAL INSTITUTE OF TECHNOLOGY
This page is intentionally left blank
Characteristics of Early Adopters and Early Majority Adopting a Vertical Social Network
A case study of the applicability of the Diffusion of Innovations theory in a Vertical Social Network
Hantman Kollén, Sophie Manhem, Rebecca
Master of Science Thesis TRITA-ITM-EX 2019:175 KTH Industrial Engineering and Management
Industrial Marketing and Entrepreneurship SE-100 44 STOCKHOLM
Karaktärsdrag av tidiga och sena användare av vertikala sociala nätverk
En fallstudie av lämpligheten att använda Diffusion of Innovations teorin i ett vertikalt socialt nätverk
Hantman Kollén, Sophie Manhem, Rebecca
Examensarbete TRITA-ITM-EX 2019:175 KTH Industriell teknik och management Industriell marknadsföring och entreprenörskap
SE-100 44 STOCKHOLM
Master of Science Thesis TRITA-ITM-EX 2019:175
Characteristics of Early Adopters and Early Majority Adopting a Vertical Social Network
A case study of the applicability of the Diffusion of Innovations theory in a Vertical Social Network
Hantman Kollén, Sophie
Manhem, Rebecca
Approved
2019-05-29
Examiner
Brown, Terrence
Supervisor
Blomgren, Henrik
TRITA-ITM-EX 2019:175
Commissioner
Fishbrain AB
Contact person
Kennelly, Lisa
Abstract
To remain relevant, even the successful social networks need to evolve. Vertical Social Networks (VSN) are the response of this, satisfying people who are seeking more niched and personalized content. However, this is a highly competitive environment where taking advantage of the first mover advantages is crucial for the future success. This study concerns diffusion of innovations and segmentation according to the technology adoption lifecycle to investigate how a VSN can reach user growth by understanding their current users better. We have built this study upon an explorative case study of a case company operating in the VSN landscape providing their users with an application. Building on extensive literature research, theory, surveys and conducting interviews, differences in user characteristics were identified. Theoretical and empirical evidence was further analyzed based on user behavior and level of satisfaction. Our recommendations suggest that considerations for different phases of the diffusion journey has to be made by targeting the appropriate users as this is vital for satisfying the existing users, and for reaching new ones.
Keywords: Vertical social network, user growth, diffusion of innovations, technology adoption lifecycle, user characteristics of adopters, the chasm, early adopter, early majority
Examensarbete TRITA-ITM-EX 2019:175
Karaktärsdrag av tidiga och sena användare av vertikala sociala nätverk
En fallstudie av lämpligheten att använda Diffusion of Innovations teorin i ett vertikalt socialt nätverk
Hantman Kollén, Sophie
Manhem, Rebecca
Godkänt
2019-05-29
Examinator
Brown, Terrence
Handledare
Blomgren, Henrik
TRITA-ITM-EX 2019:175
Uppdragsgivare
Fishbrain AB
Kontaktperson
Kennelly, Lisa
Sammanfattning
För att bibehålla relevans, måste även de mest framgångsrika sociala nätverken utvecklas.
Vertikala Sociala Nätverk (VSN) är svaret på detta, genom att tillfredsställa människor som söker mer nischade och personifierat innehåll. Detta landskap är dock präglat av hög konkurrens där vikten av att vara först på marknaden är avgörande för framtida framgång. Denna studie belyser Innovationsspridning och segmentering enligt Technology Adoption Lifecycle för att undersöka hur ett VSN kan nå användartillväxt genom att förstå sina nuvarande användare bättre. Vi har genomfört en undersökande fallstudie av ett företag inom VSN branschen som bistår sina användare med en applikation. Baserat på tidigare forskning, teori, enkäter och intervjuer identifierades olikheter i karaktärsdrag hos deras användare. Teoretiska och empiriska bevis beträffande olikheter i karaktärsdrag analyserades och diskuterades vidare baserat på användarbeteende och nivå av tillfredsställelse. De föreslagna rekommendationerna innefattar att ett övervägande för olika faser av diffusionsresan måste göras för att tilltala lämpliga målgrupper, då detta är avgörande för att tillfredsställa befintliga användare, och för att nå nya.
Nyckelord: Vertikala sociala nätverk, användartillväxt, innovationsspridning, technology adoption lifecycle, karaktärsdrag av användare, klyftan, tidiga användare, sena användare
Table of Contents
Abstract III
Sammanfattning IV
Acknowledgements VIII
1. Introduction 1
1.1 Background 1
1.2 The Case Company - Fishbrain 2
1.3 Problem Statement 2
1.4 Purpose 3
1.5 Research Questions 4
1.6 Expected Contributions 4
1.7 Delimitations 4
1.8 Chapter Outline 5
1.9 Chapter Summary 6
2. Theoretical Framework 7
2.1 Overview of Previous Research 7
2.1.1 Vertical Social Networks 7
2.1.3 Segmentation 8
2.2 Diffusion of Innovations 9
2.2.1 Technology Adoption Lifecycle and Adopter Categories 9
2.2.2 The Chasm 10
2.2.3 Characteristics of Early Adopters and Early Majority 11
2.3 Chapter Summary 12
3. Methodology 13
3.1 Research Process 13
3.2 Research Design 15
3.3 Literature Research 16
3.3.1 Sources and Search Terms 16
3.3.2 Source Criticism 18
3.4 Explanation of Structure 18
3.5 Empirical Data Collection Methodology 19
3.5.1 Interviews 20
3.5.2 Survey 20
3.5.3 Kano Survey 21
3.5.4 User Behavior 25
3.6 Analysis Methodology 27
3.6.1 User Segmentation 27
3.6.2 User Satisfaction 29
3.6.3 User Behavior 31
3.7 Ethics 32
3.8 Reliability, Validity and Generalizability 32
3.9 Chapter Summary 34
4. Empirical Findings 35
4.1 Pre-Study Results 35
4.2 User Satisfaction 37
4.3 User Behavior 39
4.4 Key Summation Points of Characteristics 40
4.5 Chapter Summary 41
5. Analysis and Discussion 42
5.1 Differences and Similarities between Early Adopters and Early Majority 42
5.1.1 User Satisfaction 42
5.1.2 User Behavior 44
5.2 Applicability of Rogers 47
5.3 Managerial Implications 48
5.5 Chapter Summary 49
6. Conclusions and Recommendations 50
6.1 Summation of Research 50
6.2 Recommendations for Future Research 52
Appendices 53
Appendix A. Interviews 53
References 54
List of Figures
Figure 1 Technology Adoption Lifecycle 9
Figure 2 The Chasm 10
Figure 3 Research Process 13
Figure 4 Checklist for Source Criticism 18
Figure 5 Explanation of Structure 18
Figure 6 Kano Survey Format 24
Figure 7 Kano Survey in Fishbrains Feed 25
Figure 8 Segmentation Components 27
Figure 9 Tam-Sam-Som 28
Figure 10 Adoption Percentage of the Potential Users in the US 29 Figure 11 Possible Answer Combinations in a Two-dimensional Plane 30 Figure 12 Share of Users in each Level of Skill, presented in the two each User Segment 35
Figure 13 Each Features Mean Value for Both User Segments 37
List of Tables
Table 1 Characteristics of the Early Adopters and the Early Majority 11
Table 2 Sources and Key Search Terms 17
Table 3 Targeted Features Used in the Kano survey 22
Table 4 Events Used in Amplitude 26
Table 5 Scoring of Possible Answers to Functional and Dysfunctional Features 30 Table 6 Share of Users Adopting a VSN, Prior Being a Member of Social Network 36 Table 7 Functional and Dysfunctional Answer Combinations Presented of Each Feature 37 Table 8 The Active % of Users Within Each User Segment Who Performed an Event 39
Table 9 Key Summation Points 40
Acronyms
DoI Diffusion of Innovations
SAM Serviceable Addressable Market TAM Total Addressable Market TAL Technology Adoption Lifecycle VSN Vertical Social Network
Acknowledgements
This master thesis was written with help from some extraordinary people, to whom we wish to express our special thanks and gratitude to.
Firstly, a very special thanks to our supervisors Henrik Blomgren and Lisa Kennelly, for the continuous guidance and support throughout this whole project.
Secondly, we wish to thank all the participants in the interviews, for taking their time and for sharing their knowledge of the investigated field.
Thirdly, we would like to thank Fishbrain, for giving us the golden opportunity of doing this study based on their users and based on their internal product analysis tool. In addition, to the employees at the Fishbrain who more than willingly helped us, but also for cheering us by being engaged and curios in our work.
Lastly, we like to thank our peers at KTH, who first and foremost gave us feedback and support, but also interesting (and long) discussions.
Once again, thanks!
Sophie Hantman Kollén Rebecca Manhem
June 2019, Stockholm June 2019, Stockholm
This page is intentionally left blank
1. Introduction
In this chapter, an introduction to the background of this research is presented together with a description of the case company Fishbrain. Included in the chapter is the problem statement, purpose and the research questions that seek to be answered. These are presented together with a hypothesis. Thereafter, expected contributions, the delimitations and chapter outline is presented.
1.1 Background
Facebook, LinkedIn and Twitter are some of the successful players and examples of social networks, but in order to remain relevant, even they need to evolve. Their rapid growth comes with challenges connected to fake news, data breaches and overcrowded marketplaces. Users are getting fed-up by the overload of irrelevant content exposure (Bridgwater 2014). However, their rampant growth is paving the way for other, more effective digital platforms.
People are seeking niche groups within social networks - Vertical Social Networks (VSN). These platforms are tailored specifically to certain groups of people or industries, offering a more personalized approach linking them together with others who share similar interests, and providing a curated content (Leadem 2018).
As the world has gone digital, this enables disruptive technological innovations to be explored and implemented frequently. However, technological innovation alone (such as mobile-commerce or a VSN) is insufficient. The business-model innovation has therefore become more important now than ever (Lindgardt et al. 2009). Discussions are held about whether the world-wide web drives market monopolization (Haucap & Heimeshoff 2014). Empirical evidence reveals that there are indeed standards on social network service markets (Baran et al. 2015). This insinuates a business model plagued by rapid scaling of users; and once a scale-up occupies the high ground in a specific market, both capital and talent floods are attracted whereas there is no room for other competitors (Hoffman & Yeh 2018). Unfortunately, there is no single success story in how to grow in users.
Although VSN still are in an emerging phase, it has a promising future (Solis 2017). Thus, uncertainties may stand in the way and potentially limit the growth of companies wanting to provide such offerings. As this kind of digital platform is relatively new and limited knowledge exists in how to transform it to a successful business strategy, it becomes essential to gain further understanding about the drivers behind such a platform; the users.
1.2 The Case Company - Fishbrain
The case company, Fishbrain, represents a small- and medium-sized enterprise (SME) and was established in 2009 in Sweden. In 2012, the company refocused its efforts toward what they are today: a smart phone application and vertical social network with multiple features targeted for a niche social hobby with passionate anglers. Today, they are the largest and fastest growing application in this specific industry with top user ratings on both the Android and iOS operating systems. The company has close to 8 million registered users. 80 % of the registered users are located in the US, but the company’s presence is widely spread throughout the world.
To expand, the vision of the company is to build a platform and ecosystem for the entire industry.
The opportunities in this originates from the lack of competition in the market, which is highly unexploited and underdeveloped in all aspects. The company can therefore become the driving force in changing the industry by combining its current scalability with more advanced technologies to bring the platform to new markets.
A key strength of the company, apart from being first to market, is how their data collection technology allows them to be up-to-date and have the highest quality data in the market. Instead of being a competitive force for existing services, they can become the channel for others to reach out to an ever-broader audience. However, while the industry of the vertical is huge, it is
traditional and slow to innovate.
There are few direct competitors as of today; the second largest has about 1 million users. Thus, there are several substitute competitors, for example general social media platforms, niche forums, e-commerce and YouTube.
With great expansion potential, the company today focuses on being customer-centric and developing the app to fit their customers’ needs. The primary focal point has been to understand and investigate their current users so that they can bring more value to the app by regularly releasing new features that provide the users with unique intelligence.
1.3 Problem Statement
Marketers have traditionally identified different kinds of technology users: Innovators, Early Adopters, Early Majority, Late Majority and Laggards. The traditional model named the Technology Adoption Lifecycle (TAL) implies a level of inevitability in the flow of one category to another. However, there are gaps in the model large enough to derail the promising startups as they transition from one category of users to the next. The reason for these gaps are that the different users are described to have distinctly different expectations and incentives for using technological innovations (Moore 2014).
Digital platforms are a relatively new phenomenon with a fast pace of change plagued by high competitiveness. This pressures companies competing in such landscape towards claiming large market share early and rapidly, especially in the niche markets (I3). Hence, the importance of being first to market and using the benefits of the first mover advantages (CFI 2019).
The characteristics defining the technology users are clearly presented by Rogers in the Diffusion of Innovations (DoI) theory in relation to technology adoption. However, digitalization, globalization, tech giants and disrupting startups marks the new era after the industrial age, possibly changing the way potential users adopt technology (Eden 2018). VSN are one of the emerging result out of this new era, where users strive for individuality and connection with people with the same interests (Wiese 2015). Due to the novelty of this phenomenon, there is limited knowledge about the users that can be used to predict its future. Today, there is no best practice or success-stories to be used as role-models in the VSN architecture.
The need for understanding the users is the primary focus for all customer-centric organizations.
If this is not done accurately, there is a high risk that new technology can stall and possibly fall to the bottom of the chasm and not to be heard from again (Wohlers Associates, Inc. 2002). How to define the characteristics and expectations for the technology users in today’s VSN landscape may provide a tool to properly reach the full potential of users.
1.4 Purpose
The purpose is to investigate how a Vertical Social Network can reach user growth by understanding their current users better. This, by looking into the adequacy of applying the TAL from DoI theory and testing the characteristics of two of its appurtenant segments; early adopters and early majority.
The objective is to investigate if such a segmentation is valid to increase the knowledge in who the next potential adopter is, the possibility of targeting the next potential adopter and influencing the diffusion rate.
1.5 Research Questions
To successfully answer the research question of this study, two sub-research questions have been formulated. The questions follow;
RQ: How to reach user growth by investigating the chasm?
Sub RQ 1: What behaviors of the two user segments early adopters and early majority in a VSN, correlates to their appurtenant characteristics stated by Rogers?
Sub RQ 2:What are the satisfaction level of in-application features for the two user segments early adopters and early majority of a VSN?
The hypothesis is that the characteristics of the early adopters and early majority presented by Rogers theory Diffusion of Innovations will be visible when segment the current users of a VSN by the TAL at a point in time before the product has reached the whole serviceable addressable market.
1.6 Expected Contributions
Even though it exists suggestions regarding how to segment users, applying the TAL in the context of VSN has not been done before. The expected contributions are both empirical contributions and contributions to the literature within the theory of DoI. The main empirical contribution will be provided by data collection of the two user segments early adopters and early majority, in the context of a VSN. While the contribution to the literature is focusing on the applicability of the theory by Rogers in a not yet explored context.
The outcome of this study can be used for a deeper understanding of the current users of a VSN, and serve as a foundation for future development, segmentation and user growth.
1.7 Delimitations
Geographical delimitation: The pre-study is focusing on getting to know users of VSN and is limited to Swedish forums and networks. The reason for this is that these forums are more available to the authors and can provide more responses during the limited period that is provided for this study. The geographical aspect is at this point not relevant, when the aim only is to get a wider perspective. Further on in the study, the investigation is limited to data from users in the US. This choice is based on the Fishbrain’s present user database, where data can be found.
User growth delimitation: The meaning of user growth, which is the primary focus of this study, may be interpreted in different ways. The chosen focus on this is how to diffuse an innovation in order to attract and reach new customer, leaving user retention outside of the scope.
Market delimitation: The market that is investigated is the specific hobby-related industry in which Fishbrain operates within, i.e. the recreational fishing industry. Other surrounding or similar actors will not be benchmarked or compared with.
1.8 Chapter Outline
Chapter 1 - Introduction
The first chapter provides an introduction to this study; covering the background, a description of the case company Fishbrain, and a problem statement. This is followed by a description of the purpose and the formulation of research question that this study aim to answer, together with a hypothesis. In the final part of this chapter, the expected contributions and delimitations are presented.
Chapter 2 - Theoretical Framework
This chapter provides a thorough literature review of the area of interest, including a description of previous research in this specific field. In addition, this chapter introduces theoretical frameworks and key concepts that are later used.
Chapter 3 - Methodology
This chapter covers the methodology used in this study. It includes; the research process, research design and a description of the literature research. This is followed by a clarification of the used structure coupled to the empirical data collection and the analysis methodology. Lastly, reliability, validity, generalizability as well as ethics is discussed.
Chapter 4 - Empirical Findings
This chapter presents the empirical findings. This by providing an overview of the pre-study results, citations from conducted interviews, and in-depth findings regarding user behavior and user satisfaction. To get a clear overview, this chapter is ended by key summation points of the empirical findings.
Chapter 5 - Analysis and Discussion
This chapter analyses and discusses the empirical findings. Firstly, by analyzing similarities and differences, both regarding user behavior and user satisfaction. Thereafter, by analyzing whether is the characteristics by Rogers is applicable in the context of a VSN or not, followed by managerial implications of the presented findings.
Chapter 6 - Conclusions and Recommendation
The final chapter of this study covers a summation of this research by providing a conclusion and recommendations for future research. The conclusion is made by answering the research question and verifying the hypothesis.
1.9 Chapter Summary
Facebook, LinkedIn and Twitter are two of the successful players and examples of social networks, but in order to remain relevant, even they need to evolve. Instead, people are seeking niche groups - Vertical Social Networks (VSN). Consequently, an there has been an increasing interest for more effective digital platforms than the ones existing today, which creates business opportunities.
However, since VSN is relatively new, there is limited knowledge in how to transform it to a successful business strategy, and it becomes essential to gain further understanding about the drivers behind such a platform - the users.
This study aims to investigate the characteristics of the two user segments early adopters and early majority, adopting a VSN. This, by looking into the adequacy of applying the Technology Adoption Lifecycle from Diffusion of Innovations theory and testing the characteristics of two of its appurtenant segments; early adopters and early majority.
2. Theoretical Framework
In this chapter, an overview of previous research from which this research originates is presented.
The targeted scope of research includes Vertical Social Networks, Spreading an Innovation, Segmentation and different aspects of Diffusion of Innovations which are briefly presented. A more in-depth description is presented about key concepts used from the Diffusion of Innovations theory, being the Technology Adoption Lifecycle, the chasm, the characteristics of early adopters and early majority. The theories presented in this chapter helps building the theoretical framework that is used in the analysis.
2.1 Overview of Previous Research
Previous research done within the field of interest, directly within the topic of VSN and surrounding areas, were investigated in order to gain further knowledge. The existing findings in literature and the possibility for future research of such studies is presented.
2.1.1 Vertical Social Networks
A vertical social network is described as the following: “Vertical networks (or vertical social networks) connect people who have very specific interests or passions. If your target audience is part of a vertical social network, you may be able to integrate that network into your inbound or content marketing strategy. This can also apply to niche groups within broader social networks like Facebook & LinkedIn groups” (Narrative Industries 2013).
The existing literature covering VSN are extremely scarce and is therefore a field that requires significant attention. Recent articles about the phenomena of emerging VSN, explains its rapid growth as a result from a large social networks offering, leading to a large and irrelevant content exposure for the users (Frasco 2013). From 2003 and onward, many new social network services was launched, promoting the social software analyst Clay Shirky (2003) to coin the term “Yet another Social Networking Service” (Boyd & Ellison 2007). VSN are therefore explained to derive as a substitute for users looking for a more vertical content, specialized in their individual interest, hobby, demographics or other factors (Leadem 2018).
Scholars from disparate fields have thus examined the roots from which VSN appear; the social network services. The practices, implications, culture, the meaning of the sites, as well as users’
engagement have been investigated in order to understand it better (Boyd & Ellison 2007).
Research has focused a wide range of topics including; impression management and friendship performance (Boyd 2004), network and network structure (Golder et al. 2007), the roles users play in the growth of networks (Kumar et al. 2010), the role of language typology (Herring et al. 2007), and the importance of geography (Liben-Nowell et al. 2005). Additional research of how a social
network evolves as its members attributes change has been the subject of several models (Sarkar
& Moore 2005); (Holme & Newman 2006). Mathematical models for group evolution and change have been proposed in a number of social science contexts (Backstrom et al. 2006), amongst them Boorman and Levitt (1980) applying an approach to this issue in terms of diffusion models (Boorman & Levitt 1980).
2.1.2 Spreading an innovation
There has been a new wave of research about user growth in relation to product innovations. A larger emphasis has been directed towards the underlying social network’s structure rather than on the aggregate level (Muller & Peres 2019).
When describing how social network structure influences the user growth of an innovation, the underlying theory has been described as; being a part of a social network plays a role in the network users adoption behaviors. The main attributes of this role is social contagion meaning how the users are impacted by each other in their adoption decisions. Contagion is enabled through multiple social network mechanisms in which the users gain information about the innovation and are persuaded to adopt it (Muller & Peres 2019). The characteristics presented in chapter 2.2.3 Characteristics of Early Adopters and Early Majority, will be viewed as mechanisms and used as theoretical backgrounds that will later in this study aim to describe some of the empirical findings.
A user growth process that is high-performing may intuitively be one in which 1) many people adopt the innovation 2) it spreads to the far ends of the social network 3) it does this within a short time frame 4) it does so with low marketing efforts. When describing it as such, this implies that the performance is a multidimensional construction. However, existing literature mainly chooses to focus on one of these dimensions. The diversifying metrics used imposes a challenge on empirically generalizing findings across papers. Possible future research suggestions that has been articulated, is therefore using a measure that captures more than one dimension, and for future research to strive for agreeing on one multidimensional metric (Muller & Peres 2019).
2.1.3 Segmentation
Segmentation refers to dividing a population into relatively homogenous groups of individuals in terms of their characteristics, beliefs, preferences, and/or behavior (Althuizen 2017).
Successfully segmenting a population depends on the presence of substantial similarities as well as meaningful differences between individuals (Althuizen 2017). No user is identical to another but are rather plagued by variation. Different users have different needs which cannot be satisfied by the same offerings. Therefore, it is important that a company does not approach all their users in the same way, when aiming for increasing their market share (Nilsson 2011).
A segment-focused analysis of the technology acceptance phenomenon has been advocated in literature, prominently by the work of Rogers (1994) and Moore (2002). In the DoI, Rogers discerns five segments based on the individual’s time of adoption of disruption (i.e., innovations that require significant behavioral change on the part of the user). In a similar way, Moore focuses on people’s attitudinal disposition towards new technologies (Althuizen 2017).
2.2 Diffusion of Innovations
Modes of transportation, daily habits, computer software, and smartphone applications are innovations that are explored and exploited every day. However, it is not until the innovation is adopted that it becomes an innovation for the user (Karakaya 2015). The diffusion of an innovation is defined as the process of how an innovation propagates in, and is communicated through, certain channels among potential adopters in a social system (Rogers 1983). The diffusion process is complex and involves incremental adjustments and related activities to make the different parts of the system fit (Rosenberg 1995). Even though an innovation has obvious advantages, it is difficult to understand why an innovation is adopted by users or not (Rogers 1995). How and why this occurs, are questions worth pondering for both governments and companies (Karakaya &
Lundberg 2016).
2.2.1 Technology Adoption Lifecycle and Adopter Categories
The TAL is a model describing the diffusion and adoption rate of products, services and innovations. The model was defined by Rogers in year 1962, and provides insights in the current market for a product or service, by analyzing the adopters and the percentage of the population they represent (Wohlers Associates, Inc. 2002). The TAL is illustrated in Figure 1 below.
Figure 1 Technology Adoption Lifecycle (Rogers 2009)
The TAL is a normal distribution curve and defines the quantity of adopters over time. The normal frequency distribution is divided into five adopter categories; innovators, early adopters, early majority, later majority, and laggards. Whereas each adopter category is mutually exclusive and represents the ideal types of adopters. Ideal types are designed to make comparisons possible and to serve as a framework and is based on observations. However, those ideal types are classified after complete adoption of an innovation, and are therefore difficult to find for an innovation before reaching 100 percentage of potential users (Rogers 1995). Additionally, finding the current position for an innovation is difficult, since the adoption lifecycle is nonlinear and complex (Meade
& Rabelo 2004).
As illustrated in Figure 1 the innovators and the early adopters represents a mere 16% of the market, while the early majority and late majority represent 68% of the market. From these numbers, it can be seen that the adoption of an innovation will not reach substantial growth until the early majority welcomes the innovation (Wohlers Associates, Inc. 2002). The innovators and the early adopters are defined as representants of the early market, while the early majority and the late majority represents the mature and late market respectively (Moore 2014).
2.2.2 The Chasm
As previously illustrated, the TAL is a curve with smooth transitions between each adopter category. However, the curve ignores a critical element; a disconnection and disassociation between the user segment early adopters and the user segment early majority, which is referred to as the chasm (Moore 2014). The chasm is illustrated in Figure 2 below.
Figure 2 The Chasm
The chasm represents a chaos existing at this specific point of the TAL (Wohlers Associates, Inc.
2002). The existing chaos is due to the great differences between the user segments on each side of the chasm. If not properly bridged, the innovation falls into the chasm, risking to not being further adopted and to lose its’ market share and fail. However, the first users of an innovation can be a prerequisite for funding from venture capitalists (Lin & Hong 2011), which furthermore can lead to a chance to bridging the chasm (Popovic & Fahrni 2004).
2.2.3 Characteristics of Early Adopters and Early Majority
As previously described, there is a difference between the user segments on each side of the chasm;
the early adopters and the early majority. Each of those two user segments has different characteristics, reactions, criteria, and needs from the new innovation (Meade & Rabelo 2004), and adopts the innovation with different expectations and motivators (Li et al. 2017). These differences manifest themselves in preferences and behaviors such as the early adopters being more enthusiastic than the early majority towards new tech innovations, and are more careless if the product or service does not fulfill their expectations (Weaver 2019). Some of the additional characteristics of the early adopters and the early majority, described by Rogers (1995), are presented in Table 1 below.
Table 1 Characteristics of the Early Adopters and the Early Majority
Early Adopters Early Majority
● Visionary
● Technology focused
● Willing to take risks
● Willing to experiment
● Interact with others than their personal network
● The decision period of adopting an innovation or not is relatively short
● More willingly to spend money
● Pragmatic
● Not technically focused
● Averse to take risks
● Looking for proven applications
● Interact frequently with their peers
● The decision period of adopting an innovation or not is relatively long
● Less willingly to spend money
The early adopters are the respectable adopters (Rogers 1995). Furthermore, they are the visionary adopters who drive the high-tech industry, are willing to take risks to pursue a goal, and are willing to experiment. They are considered by the potential adopters to be the ones to ask advice from before adopting an innovation. The early adopters’ decision period of whether to adopt an innovation or not is relatively short in comparison to the potential adopters. However, since the early adopters are buying a dream and the business behind the technology rather than the technology itself, they are easy to sell to but hard to please. Therefore, succeeding among the early adopters leads to a reputation about the company of being a high flyer with an innovation that is hot (Moore 2014).
The early majority are the deliberate adopters, who adopts an innovation just before the average adopter in a social system (Rogers 1995). They are the pragmatic adopters who care about the reliability and infrastructure of supporting products and services of that innovation, rather than the technology itself. The early majority tends to communicate with peers within their own industry whereas relationships and references are important and will not adopt the innovation until it is established or when the company behind the innovation is proven to be a market leader. However, when a startup has earned its spurs; the early majority adopts the innovation, tend to stay loyal, and even help the innovation to be further adopted within the adopter’s social system. Therefore, this user segment makes a great market. It is not until the early majority adopts the innovation, that the innovation will reach substantial growth (Moore 2014).
Since the early adopters and the early majority have different criteria and needs, the innovation needs to evolve and improve over time. Initially, when the criteria and needs from the early adopters who tends to have superior technical skills dominate, the innovation is usually designed thereafter. These initial criteria and needs might have a disproportionate impact on the development of the innovation (Tidd & Bessant 2013).
There is a general agreement regarding the characteristics of the adopter segments, but there is no consensus in the relative importance (Tidd & Bessant 2013).
2.3 Chapter Summary
This chapter introduces relevant literature related to user segmentation and user characteristics.
Firstly, an overview of previous research was presented, whereas it was clear that the existing literature covering VSN is scarce. Secondly, the theory about Diffusions of Innovations, including;
the adoption lifecycle, adopter categories, the chasm between the adopter categories, and a further explanation of the characteristics of the two user segments on each side of the chasm - the early adopters and the early majority, are presented. The chasm exists because of the great differences in characteristics between the two user segments.
3. Methodology
This chapter presents the selected research methods and motivates the choices made regarding the course of action. The chapter is divided into subchapters that provides a walk-through of the research process, research design, literature research, empirical data collection and analysis methodology. Reliability, validity, generalizability and ethics are discussed at the end of this chapter.
3.1 Research Process
As described in chapter 1.4 Purpose, the purpose of this study was to investigate how a VSN can reach user growth by understanding the current users. This by looking into the characteristics of the early adopters and early majority. The term user growth may include several meanings in a business context, but as described in chapter 1.7 Delimitations, this study only focuses on the growth by new users. The approach of this project was structured so that the outcomes of Phase 1
& 2 served as a foundation for the following phases. An overview of the research process is presented in Figure 3 below. Due to a limited period, there were continuous actions throughout Phase 1-5: writing the academic report and having weekly meetings with supervisors.
Figure 3 Research Process
Phase 1 was the initiation of this project. This phase consisted of an iterative process between theoretical searches and internal interviews at Fishbrain. Both processes were held in an unstructured approach, aiming to gain understanding of Fishbrain’s needs and understandings, as well as achieving basic knowledge about VSN. The findings indicated a limited existence of research regarding VSN and understanding for user behavior in that context. The theoretical search process and internal interviews are described further in 3.3 Literature Research and 3.5.1 Interviews.
Phase 2 consisted of contacting external interviewees and formulating a pre-study survey. The external interviewees searched for were mainly candidates with substantial knowledge and new potential perspectives of the topics of spreading an innovation, user growth and segmentation. The pre-study survey was conducted with the purpose of validating certain theoretical standpoints, and to get a general understanding of the users of a VSN. The external interviews, and the formulating of the pre-study survey are presented in greater detail in chapter 3.5.1 Interviews and 3.5.2 Survey.
Phase 3 had a deductive approach, where the pre-study survey was analyzed quantitatively. The purpose was to investigate and detect overall patterns and behaviors. The outcome of the pre-study was compared to findings in theory, where a validation or rejection about certain aspects were made. These findings served as a direction to which aspects to investigate, and about how to continue with user segmentation. Further, internal interviews were conducted at this phase, being more specific and in depth with a semi-structured approach. The outcome of the pre-study survey and the internal interviews additionally served as a foundation when searching for user behavior in the internal product analytics tool Amplitude. The primary outcome of this phase was the understanding that the user segmentation could be based on year of registration. The analysis of the pre-study, the semi-structured interviews, Amplitude and segmentation procedure are presented in greater detail in 3.6 Analysis Methodology.
Phase 4 had an inductive approach. When all the interviews were completed, and data collected from Amplitude, the results were cross-checked with existing theoretical framework. Knowledge- gaps were investigated by both analyzing existing literature, but also by a second survey, referred to as the Kano survey. The Kano survey aimed to map user expectations in accordance to the Kano model, and was therefore carefully designed to simplify the analysis. The operationalization of the Kano survey was an iterative process with support from the supervisor at Fishbrain. The Kano survey and the analysis is presented in chapter 3.5.3 Kano Survey and chapter 3.6.2 User Satisfaction.
Phase 5 was the final phase, consisting of analysis and discussion. Similarities and differences between the investigated segments, and the applicability of DoI based on the findings were discussed to present conclusions, leading to a confirmation of the hypothesis. To ensure high
quality of the results, this phase also consisted of a feedback and a revision period. This was done by supervisors and peers.
3.2 Research Design
The choice of research purpose should be based on the nature of the objective, analysis and the research questions that the study seek to answer (Saunders et al. 2016). However, the research purpose may have more than one purpose whereas the classification of research design is not mutually exclusive. The purpose of this study could be divided into dual aspects; (1) to investigate the adequacy of applying the TAL in the context of a VSN, and (2) investigate similarities and differences in characteristics between the two user segments early adopters and early majority; a combination of an exploratory and explanatory research design was chosen.
Exploratory research design is the preferred research purpose when the aim is to gain deeper understanding of the research subject and the investigated area is unexplored (Blomkvist & Hallin 2015). Exploratory studies must be flexible as the direction of the study might change in accordance to discovery of new insights or data (Dudovskiy 2019). The study has therefore been designed with a broad initial focus that narrows and gets more specific as the research progresses.
As earlier described, VSN are a relatively new phenomenon and previous research is considered to be scarce. Consequently, the research context is immature. It therefore becomes favorable to use an exploratory approach as the purpose of the study and research questions can be reformulated when additional findings are made. However, the explanatory research design is suitable when the objective is to portray an accurate profile of persons or situations, which was executed at a later stage of the analysis; when patterns and observations had been made. Hence, the combination of the two designs made it possible to create visuals to present our findings (Saunders et al. 2016) with the product analytics tool Amplitude.
Furthermore, a mixed-model research approach was used in this study. This by combining quantitative and qualitative data collection techniques and analysis procedures, as well as combining the different approaches at different phases of the research. In this study, quantitative data was used, and then qualitative into narratives in order to analyze it qualitatively. Multiple methods provided solid opportunities of answering research questions, and to evaluate the extent to which the research can be trusted (Saunders et al. 2016). The mixed-model research was valuable as unstructured interviews could be conducted at an exploratory stage. This to get a better understanding of key issues - before using a pre-study survey to collect explanatory data. The quantitative data, collected from the interviews, was thereafter used as a foundation when searching for in-depth qualitative data (Blomkvist & Hallin 2015). Different data collection methods were used which will be presented in greater detail in the section chapter 3.5 Empirical Data Collection Methodology.
3.3 Literature Research
A literature review was carried out to gain an overview of previous studies, with the aim to achieve enhanced understanding of the research topic. The fields that the search focused on were:
1. Vertical Social Networks 2. Diffusion of Innovations
3. The Technology Adoption Lifecycle 4. Crossing the Chasm
5. Characteristics of Early Adopters and Early Majority
The literature review was initially conducted as a preliminary research method aiming to gain a general understanding on VSN and their corresponding industries and strategies. Additionally, it included overall research about DoI and the TAL. The second part of the literature review focused on critically targeting material that was increasingly detailed and relevant for the outcome of the study. Even though the literature review mainly occurred in the initial research phase, additional information was processed as the study progressed, where each subsequent search hade more precise focus on relevant literature.
3.3.1 Sources and Search Terms
The purpose of the literature review was to get a deeper understanding about how different technology users could be characterized and understood in the context of DoI when VSN aims for user growth. Both primary and secondary sources were used in the search and for this study. The primary sources were obtained from empirical material through interviews, surveys and documents provided by Fishbrain and are presented in detail in chapter 3.5 Empirical Data Collection Methodology. As new insights were gained, additional information within the scope of the study was searched for continuously. The literature was found using the online search engines; Google Scholar, Web of Science and the KTH Royal Institute of Technology’s own search engine KTH Primo.
The keywords that were used was created by breaking down the research questions and the underlying concepts. They were used in multiple different combinations and with synonyms. The main sources and key search terms are presented in Table 2 below.
Table 2 Sources and Key Search Terms
Topic Main Sources Search terms
Vertical social networks Google Scholar, KTH Primo, Web of Science, Fishbrain material
“Vertical+social+network”,
“Niche-networks”, “VSN”,
“Social+networks”,
“Social+network+niche”
Diffusion of Innovations Google Scholar, KTH Primo, Web of Science
“Diffusion of Innovations”,
“Diffusion of Innovations + SME”, “Diffusion of Innovations + startup”,
“Diffusion of Innovations + Network Effect”
The Technology Adoption Lifecycle
Google Scholar, KTH Primo, Web of Science; Textbook:
Crossing the chasm
“Early Adopter”, “Early Majority”, “Technology Consumers”,
“Tech+Adopters”,
“Adoption rate +
Technology Innovation”
Crossing the Chasm Google Scholar, KTH Primo, Web of Science; Textbook:
Crossing the chasm
“How+to+cross+the+chasm”
, “Startup + chasm”,
“Startups+pitfalls”
Characteristics of early adopters and early majority
Google Scholar, KTH Primo, Web of Science; Textbook:
Crossing the chasm
“Early Adopter”, “Early Majority”, “Technology Consumers”,
“Characteristics”,
“Technology Adopters”
3.3.2 Source Criticism
All primary and secondary sources used in this study have been evaluated in accordance with the checklist for source criticism and evaluation presented by Blomkvist & Hallin (2015). This checklist consists of four sections - authenticy, proximity and dependency, tendency, and representativeness (Blomkvist & Hallin 2015). This checklist is presented and further explained in Figure 4 below. .
Figure 4 Checklist for Source Criticism
3.4 Explanation of Structure
To investigate the user characteristics stated by Rogers, two indicators were investigated - user behavior and level of user satisfaction. When investigating the characteristics, behavior is a prediction of it as it derives from personal disposition that remains stable over time and across situations (Chatman and Goncalo 2015). The satisfaction of a person, is further influenced by personal characteristics including; stage in life cycle, values and community norms (Gifford 2001). These two factors were therefore suitable to investigate for further understanding about the user characteristics. These two indicators are presented in this chapter. The explanation of the empirical data collection and analysis structure is presented in Figure 5.
User behavior was investigated by looking for behavioral patterns in Amplitude. In Amplitude, a
Figure 5 Explanation of Structure
behavior is called an event, and is defined as a distinct action performed by a single user (e.g. start a game, add to chart, or open a push notification). Each event and how it correlated to the characteristics by Rogers were chosen and defined by the authors and validated by employees at Fishbrain. Thereafter, the collected data was compared to the different user characteristics of the two investigated user segments, as presented in chapter 2.2.3 Characteristics of Early Adopters and Early Majority.
The level of satisfaction of the users was investigated by looking into data collected from a Kano survey. A Kano survey investigates the functional and dysfunctional values of targeted features.
The targeted features were chosen based on; the interest of Fishbrain, current user requests, and if the authors believed that the features could be translated into the characteristics by Rogers.
Thereafter, the collected data was compared to the different user characteristics for the two investigated user segments.
However, to be able to compare user behavior and level of user satisfaction of the early adopters and early majority, the current users needed to be segmented. The segmentation process is described in chapter 2.1.3 Segmentation.
3.5 Empirical Data Collection Methodology
As previously mentioned, this study had a mixed model research approach. This approach is favorable when there are different purposes in different phases of the research. Furthermore, this creates a deeper understanding to employ interviews at an exploratory stage, to grasp key issues before using quantitative studies to collect explanatory data (Saunders et al. 2016). This gave the authors assurance that important issues were being addressed.
A combination of primary and secondary data was collected. The secondary data consisted of previous research and existing frameworks. Additionally, non-text material was collected, which was found in Fishbrain’s internal presentations and meetings attended for a deeper understanding of the investigated subject. Other data sources were textbook-literature, blogs, articles, a pre-study survey and company material provided by Fishbrain. The latter consisted of documented user- reviews and user-interviews. The primary data collection was conducted through different kinds of literature, exploratory discussions, interviews, internet searches and a survey. Furthermore, Fishbrain had data about their current users such as user behavior, retention rate and detailed dates for registration that was used as primary data.
In the following section a description of the different data collection methods will be described, with focus on the interviews, the surveys and Amplitude.
3.5.1 Interviews
The interviews conducted in this study were both unstructured and semi-structured, and constituted, together with the Kano survey presented in chapter 3.5.3 Kano Survey, the largest part of the empirical data collection. 11 interviews were held throughout Phase 1 to 3, with different purposes.
Unstructured interviews are commonly used in qualitative studies and provides opportunities to make unexpected discoveries. This approach was appropriate in the beginning of the study, during Phase 1, when there was an unbiased desire and need to explore the subject area further (Blomkvist
& Hallin 2015). The unstructured interviews were of open nature where there was only an overarching topic for the interview, with no predetermined aims for the findings. These interviews made it possible to refine the objectives of the study, get a deeper understanding of the phenomenon, and to explore new dimensions of it.
Semi-structured interviews are commonly used when empirical material is gathered through interviews (Blomkvist & Hallin 2015). These interviews were held in Phase 2 and 3 when the topic of the study had been more thoroughly identified and delimited, and there was a substantial understanding for the investigated field. The purpose of these interviews was to learn more about a narrower field. The methodology used when executing these interviews was that the questions were created during the interview in accordance to how the conversation with the respondent floated rather than formulating them beforehand. In order to keep the interview flowing, various techniques such as asking introductory, probing, interpreting or specifying question was done.
The interviewees were carefully selected based on their knowledge within the area. All the interviews were carried out in face-to-face meetings. Notes were taken during the interviews by one of the authors while the other held the interview. In Appendix A. Interviews , the entire list of participants can be found among with the length of each interview, job title, date of the interview, and for what purpose the interview was made with and the phase they were held in. For confidentiality reasons, each respondent was assigned a code name.
3.5.2 Survey
The survey methodology is a typical quantitative data collection method and have been widely discussed in research method contexts. The specifications on of how to precisely construct and use them have been discussed by Dillman (2007), who provided a tailored design method that contributes with helpful guidelines (Saunders et al. 2016). Two surveys were distributed in this study; a pre-study survey in Phase 2, and a secondary Kano survey in Phase 3. The methodology behind them were similar; the questions were based on the research questions and in line with the objectives of the study. The pre-study survey had the aim of validating common understandings of theory and testing some of the answers received during the preliminary interviews in Phase 1
and had more of an exploratory purpose. The Kano survey aimed to identify expectations and attributes after a segmentation had been made and had more of an explanatory nature.
The operationalizing of a questionnaire requires a solid understanding of theory and previous studies (Saunders et al. 2016). The process of formulating the survey was iterative, whereas a pre- testing phase was executed in several steps. The surveys were administered to potential respondents and colleagues, who were asked to review the questions critically. The two surveys were distributed online, whereas it was important to think about the design, and that the language was easily understood. Other general considerations were kept in mind during the operationalizing to reassure the validity and reliability (Harrison 2007); giving instructions to the survey taker, keeping the survey short, considering in what order the questions were asked, and having primary closed ended questions.
The pre-study survey was distributed in different forums focusing on niched hobbies. The members of the forums had a wide range of different attributes and demographics, but only targeted people living in Sweden. The incentive of targeting solely Sweden for that investigation was that it was easier to get access to different forums, and the demographics was not considered relevant at this stage when only searching for additional perspectives and insights. The pre-study survey was divided into three parts including each of the three variables; opinion, behavior and attribute (Saunders et al. 2016).
The Kano survey was distributed in Fishbrain’s application, targeting only members registered in the US. The reason for this was that the US market was mature and had the highest number of registered users in relation to inhabitants, therefore making the results viable. Furthermore, there was an interest to follow the users journey throughout the development of the application which was suitable in the US because of the higher level of activity in relation to the launch date of the application. The methodology behind this survey is more specifically formulated and is therefore described more thoroughly in chapter 3.3.2 Kano Survey.
3.5.3 Kano Survey
The purpose of the Kano survey was to investigate the level of user satisfaction by linking specific features to certain user characteristics correlated to the ones presented in chapter 2.2.3 Characteristics of Early Adopters and Early Majority.
The first step of the Kano model was to collect data of the targeted features and current users. In this case, targeted features were both features that already existed and features that did not exist in Fishbrain’s application. These features were tangible features chosen based on the interest of Fishbrain and current user request and are presented and explained further in Table 3 below.
Table 3 Targeted Features Used in the Kano survey
Targeted Feature
Currently exists in the app
Explanation
See what others follow
yes In the app, you can see what other users are following, e.g. waters, species, gear, people.
Display your skill level
no Imagine you could choose to display your perceived fishing skill level in your profile.
Automatic photograph editing
no Imagine you could automatically improve how your posts look by auto-editing the photos (e.g., improving contrast, lighting and blur).
Marketplace no Imagine you could buy gear, equipment and license from a shopping page within the app.
Private Profile yes Possibility to use the app as a private diary and logbook. where only the user can see it.
Minor Bugs yes Minor bugs are small issues and problems in the application that do not affect your overall ability to use the app. Examples of minor bugs include misaligned text, missing translations, missing animations and incorrectly sized buttons.
The reasoning behind each features correlation to theory followed:
Following
The Following feature was an existing feature and had the purpose of investigating the nature of the networking procedure in the application. It was of interest to see how important it was for the current users to be able to view who their friends followed, which could imply that friends were a source of inspiration. According to Rogers, the early majority are more easily influenced by their environment and looks for validation from their peers. The early adopters are characterized to be more influencers, not affected as much by their surroundings. Therefore, the assumption was that the feature “following” would be more important for the early majority segment than for the early adopters.
Skill Level
The Skill Level feature was a non-existing feature in the application and had the purpose of investigating how relevant this was in the users content. As a segmentation already had been made based on level of skill, there was an interest to investigate if the differences in perceived level of skills had diverging impact on what content they wanted to be exposed for. From the pre-study, there was an indication that VSN-users did not want to be exposed to content from users below their own perceived level of skill. The assumption was therefore that the early adopter would find
this feature more important as this would mean that they could narrow their content to people at the same or higher skill-level.
Automatic Photo Edit
The feature Automatic Photo Edit was a non-existing feature and had the dual purposes of investigating the users acceptance towards lower quality photographs, as well as the level of time and interest they had on editing the quality of their photos. According to Rogers, the early adopters are more tech-familiar whereas the probability that they would rather edit the photo quality by themselves was assumed higher. Furthermore, there was an interest to test and compare the tolerance of lower quality pictures in between the two segments. This, because the pre-study survey revealed that a common concern and large reason for user’s termination of online networks was low quality pictures.
Marketplace
The feature Marketplace was a non-existing feature in the application and had the purpose of investigating the willingness to spend money on the hobby. According to Rogers, early adopters traditionally have more money to spend, are more influential and typically have some sort of leader role. It was therefore of interest to test if a distinction regarding money spent had a correlation to the segments.
Private Profile
The Private Profile feature was an existing feature and had the purpose of investigating how users interacted with each other. According to Rogers, early adopters are more likely to interact with people outside of their close network of friends, whereas an innovation spreads through early majority based on peer-to-peer communication. Furthermore, the user’s intentions for using the application was investigated. It was therefore of interest to investigate why people were using the VSN; if it had a “brag” factor attached to it where satisfaction was obtained by letting other people view, support or comment on your posts, or if the users rather were looking for a platform to use as a private logbook.
Minor Bugs
The fact that Minor Bugs appear during the development of software products are close to impossible to avoid. During internal interviews at Fishbrain, it was stated that the early adopters were more tolerant towards these (Bouhnick 2019). This statement is in line with Rogers, saying that the early adopters are more aware and acceptant to the fact that the innovation might not fulfill the expectations. The purpose of including the minor bugs as a feature was therefore to test this statement.
The Kano model classifies features into four categories; A - Attractive, M - Must-be, P - Performance and I - Indifferent. The category A - Attractive is when a feature causes a positive
