Department of Technology Management and Economics CHALMERS UNIVERSITY OF TECHNOLOGY Gothenburg, Sweden 2018 Report No. E2017:120 Are startups doing what they believe is important? Master’s Thesis in the Master’s Programme Management and Economics of Innovation YLLDRIN HALILI ii iii Master’s Thesis E2017:120 Are startups doing what they believe is important? Master’s Thesis in the Master’s Programme Management and Economics of Innovation YLLDRIN HALILI Tutor, Chalmers: Bengt Järrehult Department of Technology Management and Economics Division of Entrepreneurship and Strategy CHALMERS UNIVERSITY OF TECHNOLOGY Gothenburg, Sweden 2018 iv Are startups doing what they believe is important? Ó YLLDRIN HALILI, 2018. Master’s Thesis E2017:120 Department of Technology Management and Economics Division of Entrepreneurship and Strategy CHALMERS UNIVERSITY OF TECHNOLOGY SE-412 96 Gothenburg, Sweden +46 31 772 1000 Chalmers Reproservice Gothenburg, Sweden 2018 v Acknowledgements This master’s thesis was conducted during the spring of 2018 and was the author’s final project at the Management and Economics of Innovation programme at Chalmers University of Technology in Gothenburg, Sweden. I would like to thank my supervisor Bengt Järrehult. Without his help, it would not have been possible to do this study. I would also like to express my gratitude to all the startups that made this study feasible by participating in the interviews and survey. Ylldrin Halili Gothenburg, 2018 vi Abstract The lean startup methodology has become popular amongst entrepreneurs, professionals and scholars, and some of the world’s leading business schools and startup accelerators are using it today for their entrepreneurship programs. However, these principles do not touch psychological factors that have been proven important for teams in order to be creative and innovative. This study was done since it was deemed interesting to see if there is a discrepancy in what of the above mentioned that startups believe is important for their future success, and to what degree they actually practice it. In order to investigate the discrepancies, a framework consisting of in total 25 innovation psychology and lean startup methodology theory groups was built. The relevant theories, were identified after a literature review and ten interviews with startup founders. In an iterative manner, 25 statements for a survey were created based on the theory groups. Survey respondents had to rate the importance and the agreement for every statement. The sampling for the startups participating in the survey was delimited to Swedish startups due to the short time limit for the study and the lack of personal network of startups abroad. The study was further limited by only having one startup accelerator participating; the other startup accelerator did not manage to reach out to its startups in time. However, due to the sample size, the wide array of fields for the startups participating, and the relatively low spread of the answers, the author is confident that the findings can be generalizable to Swedish startups in general. The findings show that the average discrepancy for startups is -9.80%, which can be considered low. This result indicates that startups are much better than large companies at practicing what they perceive is important for their future success. Furthermore, startups seem to have a propensity for practicing the factors that can be derived from innovation psychology. The factors derived from the lean startup methodology are not practiced to the same extent. Moreover, attempts have been made to create profiles regarding the startups surveyed. These are hypotheses based on the responses that have stuck out in the survey and should be tested in the future. To conclude, a future study should look into which of the IPLSM factors correlate to successful performances for startups. vii viii Definitions The most common words/phrases in this master’s thesis are described below. Startup: A temporary organization used to search for a repeatable and scalable business model (Blank and Dorf, 2012). Business model: A business model describes the rationale of how an organization creates, delivers, and captures value (Osterwalder, 2009). Osterwalder (2009) defines the business model in nine different blocks. Lean startup methodology: A methodology for business and product development developed by Eric Ries (2011). Aims to cut development cycles by adopting a hypothesis- driven approach, where hypotheses about the different business model blocks are tested as efficiently as possible. The goal for a startup is to get validated learnings and finally reach a perfect product-market fit. ix Table of content 1. Introduction ............................................................................................................................ 1 1.1 Background ............................................................................................................................ 1 1.2 Problem description ............................................................................................................... 1 1.3 Aim ......................................................................................................................................... 2 1.4 Research questions ................................................................................................................ 2 1.5 Report .................................................................................................................................... 2 2. Literature review ..................................................................................................................... 4 2.1 Complexity spurs innovation .................................................................................................. 4 2.2 Diversity spurs success for entrepreneurial teams ................................................................. 4 2.3 Individual motivation spurs innovation .................................................................................. 5 2.4 Expectations of innovation increases probability of innovation ............................................ 5 2.5 Developing human capital is beneficial for entrepreneurial endeavors ................................. 5 2.6 Team unity increases performance ........................................................................................ 6 2.7 Common vision spurs innovation ........................................................................................... 6 2.8 Low disparity of power increases innovation culture ............................................................. 7 2.9 Safe psychosocial climate to try new things .......................................................................... 7 2.10 Industrial experience facilitates innovation ........................................................................... 7 2.11 Role experience spurs creativity ............................................................................................. 8 2.12 Handling setbacks can be learned and is positive .................................................................. 8 2.13 Having a network to turn to spurs entrepreneurialism .......................................................... 8 2.14 Hypothesis-driven approach to develop value proposition is more efficient ......................... 9 2.15 Finding the perfect product-market fit ................................................................................... 9 2.16 Getting out of the building ..................................................................................................... 9 2.17 Overview of the business model ........................................................................................... 10 2.18 Not wasting time by developing the wrong things .............................................................. 10 2.19 Testing crucial first ............................................................................................................... 11 2.20 Efficient testing through MVPs ............................................................................................ 11 2.21 Regular and scheduled meetings to discuss changes to business model ............................. 11 2.22 Using ratios .......................................................................................................................... 12 2.23 Measuring what is actionable .............................................................................................. 12 2.24 Splitting customers into cohorts .......................................................................................... 12 2.25 Using validated learnings as measure of productivity ......................................................... 13 3. Method ................................................................................................................................. 14 3.1 Research method ................................................................................................................. 14 3.2 Literature review .................................................................................................................. 14 3.3 Data used for the study ........................................................................................................ 15 3.4 Development of IPLSM framework ...................................................................................... 15 3.5 Analysis of interviews ........................................................................................................... 17 3.6 Matching interviews with the theory ................................................................................... 19 3.7 Survey ................................................................................................................................... 58 3.8 Analysis method ................................................................................................................... 70 3.9 Scope .................................................................................................................................... 73 3.10 Trustworthiness .................................................................................................................... 74 4. Empirical findings .................................................................................................................. 77 4.1 Raw survey data ................................................................................................................... 77 4.2 Visualization of survey data ................................................................................................. 82 5. Analysis ................................................................................................................................. 95 5.1 Statistical analysis ................................................................................................................ 95 x 5.2 Analysis of results ................................................................................................................. 96 6. Discussion and future research ............................................................................................ 104 6.1 Analysis discussion ............................................................................................................. 104 6.2 Rearranging table 13 ......................................................................................................... 104 6.3 Filtering away soft factors ................................................................................................. 108 6.4 Comparison with large companies ..................................................................................... 109 6.5 Hypotheses regarding the different groups ....................................................................... 110 6.6 Future research .................................................................................................................. 114 7. Conclusion .......................................................................................................................... 116 References .................................................................................................................................. 117 Appendix A .................................................................................................................................. 123 1. Interview template ................................................................................................................. 123 2. Constructs ............................................................................................................................... 125 3. Email template ....................................................................................................................... 127 4. Survey feedback ..................................................................................................................... 129 Appendix B .................................................................................................................................. 130 5. Company level - interesting graphs ....................................................................................... 130 Appendix C .................................................................................................................................. 147 6. Business model canvas ........................................................................................................... 147 7. Study done on large companies ............................................................................................. 148 1 1. Introduction The introduction will present the reader to the background of this thesis. The aim and goal with the report and the problem description will also be covered. The sections regarding literature, methodology, findings, analysis and discussion will be briefly described as well. 1.1 Background As Mansoori (2017) states, the lean startup methodology (Ries, 2011) has become popular amongst entrepreneurs, professionals and scholars (Eisenmann et al., 2013). Mansoori (2017) further notes that “A growing number of prominent entrepreneurship programmes (e.g. Stanford University, Harvard Business School, Berkeley, Columbia University) and accelerators (e.g. Techstars, 500 Startups, Y Combinator) have begun to favour the use of the lean startup methodology over business planning approaches [...]. These programmes explicitly encourage, and in some cases, require students and entrepreneurs to follow and adhere to the instructions of the lean startup methodology.” (p. 812-813). In brief, the Lean Startup Methodology (LSM) is applying the philosophy of traditional Lean, i.e. reducing waste in manufacturing, on entrepreneurial startups’ processes (Ries, 2011). As startups are formed to search for repeatable and scalable business models (Blank and Dorf, 2012) and work in environments with extreme uncertainties (Ries, 2011), the LSM aims to lower the uncertainty by involving target customers throughout all product development processes, which Blank (2007; Blank and Dorf; 2012) calls customer development. Ries (2011) means that “The goal of a startup is to figure out the right thing to build - the thing customers want and will pay for - as quickly as possible” (p. 20), hence a startup needs to avoid all the “[...] tremendous waste I saw all around me: startups that build products that nobody wanted [...]” (p.7). Ries (2011) proposes a set of processes in his methodology, that reduce the risk of building products that there is no demand for, hence saving time and money for new ventures. Most of the processes proposed are about getting real customers’ behaviors in order to validate or invalidate the assumptions the startup has about its target customers, and doing so in an as efficient manner as possible (Ries, 2011; Bieraugel, 2015) However, the principles of the LSM can be considered “hard” factors, since Ries (2011) is proposing processes to use, i.e. ways of working. The LSM (Ries, 2011) does not touch “soft” psychological factors that are important for teams in order to be creative and hence innovative, such as intrinsic motivation of the team members, or the importance of “togetherness” (Denti, 2012), nor does it touch the diversity of the team (Kakarika, 2013). Furthermore, many studies have looked into the importance of individual employees’ positive moods for workplace creativity and innovation (Amabile, 1996). This is also true on a group level (Shin, 2014; Meneghel et al., 2016), i.e. a positive, happy and excited group is more likely to be creative and hence innovative. 1.2 Problem description Seeing that the modus operandi of some of the world’s highest ranked business schools (The Economist, 2017) is more and more turning to the principles proposed by Ries (2011) in the Lean Startup (Mansoori, 2017), it would be interesting to see what startups think of the importance of the LSM, and to what degree they actually practice it. As previously mentioned in the background, it is clear that psychological factors also are important in order for teams 2 to be innovative and create new useful products, services or solutions. It would hence also be interesting to see startups’ perception of the factors that innovation psychology encompasses, i.e. to see what startups think of the importance of these innovation psychology factors for their future success, and to what degree the factors are actually practiced by startups. 1.3 Aim The main goal of this thesis is to better understand the discrepancies between how important startups believe innovation psychology and LSM (IPLSM) factors are for their future success, and to what degree said factors actually are practiced by the startups. In order to do this, a sub goal of this master’s thesis is to understand what factors current literature proposes to startups and teams, in order to be innovative and successful. A framework consisting of factors proposed by innovation psychology literature and the LSM will have to be developed for this, and will be called IPLSM henceforth in order to enhance the readability. The framework will hence be based on two pillars, comprising the hard and soft factors mentioned above. Another purpose of this thesis is to gather data for a future study which will look into which of the identified factors actually are correlated to a successful venture. 1.4 Research questions The study is divided into two parts in order to reach the main goal of this thesis. The first part comprises building the framework that current literature proposes for startups to use in order to be innovative, i.e. finding what IPLSM factors are interesting to analyze. The second part comprises analyzing if there are differences in what factors the startups believe are important for their future success and to what degree they actually follow them. The above mentioned can be summarized to the following questions: 1. What IPLSM factors are considered important for startups’ success? 2. Is there a discrepancy in what startups believe are important IPLSM factors and the degree these factors are actually practiced by the startups? 1.5 Report Introduction In the introduction, the background of this master thesis is described. The reader is introduced to the aim with the thesis and why it is important. Literature In the literature section the reader is introduced to the relevant literature in order to fulfill the goal of the thesis. The literature subjects are lean startup methodology as well various fields within innovation psychology. Method In the methodology section the research methods and design of the study are presented. The development of a framework containing 25 IPLSM statements, is described in closer detail. Moreover, the data gathering and data analyzing methods are described, as well as the quality, validity and the generalizability of the study. The scope and delimitations of the study are also discussed. 3 Empirical findings In the findings section the raw data gathered from 34 unique startups and in total 42 respondents is disclosed. The startups’ self-assessments using the IPLSM framework are synthesized in 25 graphs, illustrating the differences in what factors the startups believe to be important for their future success and to what degree the factors are practiced by the startups. Analysis In the analysis section the empirical findings are analyzed based on the literature from section 2. The analysis will be done on the 25 factors individually. Discussion and further studies In the discussion section the findings are discussed and potential reasons for the findings are proposed. Furthermore, the findings are compared to equivalent self-assessments done on larger firms, in order to compare startups and large firms. Finally, further research is proposed. Based on responses that are sticking out, some startups are further discussed in the appendix. Conclusion The research questions will be answered and contributions to theory presented. 4 2. Literature review This chapter will highlight theories that are recommended to organizations in order to be innovative and efficient. The theories are grouped together based on the topics they touch. This is done in order for the reader to get an introductory understanding of the subjects and to obtain a foundation of theory to follow the study. As described in the introduction, the IPLSM framework is supported by two pillars; one pillar with softer factors, such as team spirit, motivation and diversity, and a second pillar with harder factors, such as way of working, measuring and development drivers. The first pillar touches various topics within innovation psychology. The latter one is exclusively comprising of elements from the Lean Startup Methodology (LSM), which can be interpreted as the “new school” of product development, and can be explained as the Customer Development proposed by Steve Blank (2007) combined with agile engineering and business model design (Osterwalder, 2009). The contents of the subsections in this chapter are directly corresponding to the statements in the self-assessment used for this study. These are used in the method chapter (section 3), the following findings (section 4) and in the analysis (section 5). 2.1 Complexity spurs innovation Denti (2012) states that projects being complex and challenging activate inherent motivation, which is an important element of creativity. Hammond et al. (2011) also touch this subject and argue that the strongest relationship with creativity and innovation is job characteristics. A complex job, i.e. a job consisting of many different and connected parts (Google, 2017), may stimulate creativity and innovation as those jobs typically include differing activities and challenges (Hammond et al., 2011). Further on, Hammond et al. (2011) state that in order to further stimulate creativity and innovation, the job can be rebuilt to increase complexity and autonomy. The statement of complexity being positive to innovation is not something new, which Damanpour and Aravind (2012) points out in their meta-analysis of organizational innovation. They lift up that complexity, or functional differentiation, has been positive to innovation in both older studies as well as newer ones. Somech and Drach-Zahavy (2013) also argue that functional heterogeneity is positive to innovation. It affects team creativity, which might boost implementation of innovation. Denti and Hemlin (2012) conclude that heterogeneous teams working on complex tasks also have the highest capability for innovation. Denti and Hemlin (2016) further state that innovative work is usually done in unpredictable and complex environments, such as a changing industry or entering a new market. 2.2 Diversity spurs success for entrepreneurial teams The teams that have the highest potential for innovation are the ones that also are heterogeneous (Denti and Hemlin, 2012). The mix of members from different disciplines and functions affects the team’s creativity and hence its innovation implementation (Somech and Drach-Zahavy, 2013). West (2002) further means that a great predictor of innovation is the diversity of knowledge and skills. Kakarika (2013) suggests two key factors regarding diversity that should be taken into consideration when aiming to build a successful entrepreneurial team; diversity of opinion and diversity of expertise. The prior type, refers to “[...] differences among team members in attitudes, values or beliefs [...]” (Kakarika, 2013, p.33) whereas the diversity of expertise is that “[...] members may differ in their level and specialization of education, functional background [...]” (Kakarika, 2013, p.33). Kakarika (2013) further proposes in order for a firm to be successful, the diversity of opinions ought to be moderate; i.e. “[...] some disagreement on hot issues but not to the extreme that create 5 polarization.” (Kakarika, 2013, p.36). This allows the team members to engage in constructive debates in order to find the best solution to problems they may face. Furthermore, the diversity of expertise ought to be high, in order for team members to obtain correct information to evaluate options and see problems from different angles (Klein and Harrison, 2007). Teams with high diversity of expertise get more legitimacy from stakeholders (investors, customers, suppliers, employees) who overall feel that a successful company should include people with complementary backgrounds (Kakarika, 2013). Startups are more homogeneous compared to other firms, which indicates that the costs related to workforce heterogeneity (e.g. coordination cost) may outweigh the benefits of heterogeneity (Kaiser and Müller, 2015). The startup’s heterogeneity rises with time, even though the increase is smaller than the one of other firms, as team members with other characteristics than the founding team are recruited. This is truer for knowledge-intensive startups than the non-knowledge-intensive ones (Kaisa and Müller, 2015). 2.3 Individual motivation spurs innovation Individuals need a driving force to support them in overcoming obstacles related to their creative and innovative work (Hammond et al., 2011). Therefore, individual motivation is a critical factor to consider when predicting creative performances (Hammond et al., 2011). Denti and Hemlin (2016) also found an association between personal initiative of a team member and the individual innovation. They conclude that when looking for new hires for R&D, the recruiters should premiere candidates that show initiative and motivational characteristics, besides the standard engineering or scientific skills (Denti and Hemlin, 2016). When an employee’s level of personal initiative is high, new ideas are more probable to become innovations (Denti, 2013). 2.4 Expectations of innovation increases probability of innovation Denti (2012) argues that expecting more creativity from an employee increases the chances of that person being more creative. This is in alignment with what is known as the Pygmalion effect; expecting a certain behavior from someone increases the probability of that behavior coming true (Rosenthal and Jacobson, 1968). Hence, as Denti (2012) argues, leaders should learn how to harness this “self-fulfilling prophecy”. To clearly encourage creativity and innovation on the job and expecting, or requiring, innovative behavior, can be used as a tool to promote creativity among employees (Hammond et al., 2011). This is further backed by West (2002), who says that clarified objectives lead to creativeness and innovation by contributing to a safe psychosocial climate. Gabarro and Harlan (1986) note that not stating goals, time limits or clarifying functions may lead to teams being inefficient, slow and frustrated. 2.5 Developing human capital is beneficial for entrepreneurial endeavors Investing in and developing human capital is connected to productivity, venture growth and innovation (Evans-Raoul, 2013; Holmberg-Wright and Hribar, 2016). Holmberg-Wright and Hribar (2016) define human capital as “[...] the value that employees provide through application of their skills, knowledge, and expertise which provides a necessary means for solving business problems, [...] seen as the cognitive skills, abilities, knowledge, personality, attitude, motivation, decision making, interests, and creativity that the worker provides in the workplace. These attributes and skills allow the workers to perform labor which will produce 6 economic value.” (p.12). For instance, to maintain a competitive advantage, ventures (entrepreneurial and small businesses alike) need to have a continuous increase of interpersonal skills (Holmberg-Wright and Hribar, 2016). Teixeira and Forte (2017) also recommend that entrepreneurial training should emphasize on factors related to a person’s intrinsic and entrepreneurial drive, e.g. stress and emotions management. In addition to studying theoretical subjects (Schwarz et al. 2009), a crucial part in entrepreneurship programs should involve “[...] a social learning process where the development of crucial life capacities should be the main target of all university faculties.” (Teixeira and Forte, 2017, p. 381). 2.6 Team unity increases performance Employees who feel that others are caring about them are also more probable engaging in innovative practices, since they experience bigger psychological safety and feel greater meaningfulness about their work (Hammond et al., 2011). In line with this, Denti (2012) means that a team will cooperate efficiently for their collective gain when they experience a feeling of “togetherness”. Both the team innovation performance as well as the individual performance is increased by this (Denti, 2012). Hammond et al. (2011) also support this view; positive relationships with coworkers may stimulate innovation by affecting motivation and psychological conditions. In order for a team to be creative and innovative, West (2002) states that “[...] there must be strong group integration processes and a high level of intra- group safety” (p. 380). Team conflict on the other hand, could be bad for a firm’s performance (de Jong et al., 2013). Relationship conflict, which de Jong et al. (2013) define as “When team members disagree about interpersonal styles and personal tastes or sociocultural norms and values and involve interpersonal clashes characterized by negative feelings and emotions, such as anger, hostility, and frustration” (p. 1828), lowers a firm’s performance and also lowers the positive side effects from task conflict. 2.7 Common vision spurs innovation One of the strongest predictors of innovation according to Hülsheger et al. (2009) is the vision of the firm and “[...] the extent to which team members have a common understanding of objectives” (p.1131). By taking the necessary time to state the goals and vision, a leader can promote a common understanding within the group. This makes it easier to cooperate and may enhance the innovative performance (Hülsheger et al., 2009). A crucial activity for a founding entrepreneur is to shape the team members’ awareness of means and ends so that the venture perception becomes collectively shared within the firm (Witt, 1998). Entrepreneurs can impact venture growth by mobilizing their team members’ passion by aligning the team members’ self-identity with the venture’s purpose (Yitshaki, 2012). It is beneficial for an organization to have team members that identify with the goals and values of the venture, since they are more likely to take risks and take innovative actions (Moriano et al., 2014). Denning (2014) means that in order to thrive in the new, creative economy of today, the communication within organizations has to change. He states, “A shift from top down- directives to multi-directional conversations. Instead of telling people what to do, leaders inspire people across organizational boundaries to work together on common goals” (Denning, 2014, p.4). 7 2.8 Low disparity of power increases innovation culture Kakarika (2013) identifies the distribution of power and resources within a team, as one of the key factors in order to build a successful venture. Minimizing the disparity of power is recommended (Kakarika, 2013) as it ensures that the venture is democratic, encouraging active participation for all team members. Bayraktar (2016) identifies the venture thought of as “the founder’s organization” rather than “our organization” as a barrier to building an innovative culture in an entrepreneurial venture. If the founder plays the central role and is perceived as a “hero” by the rest of the team, the venture risks ignoring negative aspects of ideas. “Yes people” may surround the founder, which can lead to groupthink and an inclination to agreeing with the leader’s ideas. Decision- making may then suffer, as well as creative thinking (Conger, 1990, Janis, 1971; Jaussi and Dionne, 2003; Bayraktar, 2016) Gabarro and Harlan (1986) states that a group is inefficient if someone, or parts of the group, has so much influence, that others’ ideas are dismissed out of hand. The asymmetry is especially dangerous when minority opinions are systematically rejected without sufficient exploration. Members who feel that they have had the possibility to influence a group discussion, are more committed to the decisions decided upon, disregarding if their own opinion has been accepted by the other members or not (Gabarro and Harlan, 1986). 2.9 Safe psychosocial climate to try new things For a team to be creative and innovative, West (2002) states that there has to be “[...] a high level of intra-group safety” (p. 380). Further, the context needs to be demanding (West, 2002). This requires the team to develop a safe psychosocial climate. Somech and Drach- Zahavy (2013) mean that “A climate in which it is safe to speak up and take risks is suggested to complement the adaptation and implementation of innovation” (p.702). They further conclude the importance of the team members’ right to feel safe when taking risks, such as proposing new ways to work or coming up with different ways of solving problems. Denti and Hemlin (2012) also acknowledge the importance of a good climate for creativity, suggesting that leaders promote emotional safety and respect in the organization. Having this emotional support from the environment may further stimulate team members to engage in innovative behavior (Hammond et al., 2011). Denti (2012) further suggests that leaders ought to recognize and reward creative efforts, and that they should tolerate a certain degree of experimentation. Risk is deep-rooted in innovation and team members should be allowed to fail. An individual in an environment where it is safe for to take risks, is more likely to engage in taking risks (Hammond et al., 2011). However, tolerating this risk, not annihilating it, is the best strategy (Denti, 2012). 2.10 Industrial experience facilitates innovation Delmar and Shane (2006) mean that the prior industry experience of the founding team increases the sales and the survival chances for a new venture. However, these effects are not linear and may differ with venture age (Delmar and Shane, 2006). Castrogiovanni and Ribeiro (2012) on the other hand mean that profitability and productivity have a positive relation to the owners’ industry-specific know-how from before starting ventures and the owners’ overall business know-how acquired after starting the ventures. However, the know- how resulting from having worked in a company in the same industry before starting the venture, is related to productivity alone and not related with profitability (Castrogiovanni and Ribeiro, 2012). Hurt et al. (2015) state that a better match between the entrepreneur’s characteristics (know-how, skills, abilities) and the characteristics of the opportunity, can lead to venture success. Further, Song et al. (2008) state that industry experience is positively 8 correlated to venture performance, and mean that entrepreneurial teams ought to acquire more industry experience in order to increase the venture performance. Lee and Tsang (2001) also mean that an entrepreneur’s industrial experience is of importance and state that it is one of the factors that has the greatest effect on venture growth. Startup managers with previous industrial experience can be beneficial for startups (Delmar and Shane, 2006). An example is the role they may play in the implementation of open innovation (Usman and Vanhaverbeke, 2017). This is due to their credibility amongst the managers in the larger counterparts in the innovation network, and their ability to efficiently handle the latter party. Usman and Vanhaverbeke (2017) explains this as a manager who “[...] knows to knock at the right door at the right time”. 2.11 Role experience spurs creativity Many studies suggest that education and tenure reflect some sort of task or domain knowledge, either through explicit training or experience on the job (Oldham and Cummings, 1996; Kark and Carmeli, 2009; Tierney and Farmer, 2004). This relationship is often motivated by the authors through citing Amabile’s (1988) model of creativity. When a person becomes more experienced and gets more knowledge, s/he builds a greater and more integrated inner archive of ideas and facts that can be used as a response to different situations. This would allow her/him to come up with creative ideas to solve problems (Amabile, 1983) which in turn may lead to creative and innovative performance (Perkins, 1986). However, in their meta-analysis, Hammond et al. (2011) do not find that education and tenure consistently are related to creativity and performance. A possible reason for this inconsistency may be that “[...] the relationship between these factors and innovation may not be linear as creativity may develop and decline across the lifespan” (Hammond et al, 2011, p. 99). 2.12 Handling setbacks can be learned and is positive Blank and Dorf (2012) mean that failing is an essential part of creating a successful venture. Further, Gabrielsson and Politis (2009) believe that a positive attitude to failing can be important to entrepreneurs. It can help them deal with and learn from mistakes and move forward. Closing a previous business due to bad performance gives more learnings than closing a previous business due to other reasons, such as personal reasons (studies, family situations etc.) (Gabrielsson and Politis, 2009), something Stokes and Blackburn (2002) mean can be due to the rather concrete and obvious experience, which may result in more time for learning through personal reflection. The positive attitude towards failure can be learned through new experiences and information (Gabrielsson and Politis, 2009). Gabrielsson and Politis (2009) believe that it is more beneficial for firms to view failure in the steps of venture creation as something usual and inevitable. If dealing with failure is done right, it may present opportunities to learn and develop the venture (McGrath, 1999). However, critical failures in the process of creating a new venture are less important than having closed a previous business (Gabrielsson and Politis, 2009). 2.13 Having a network to turn to spurs entrepreneurialism Turning to an external network in order to access information and for relevant inputs is something Balodi and Prabhu (2014) mean can compensate for the lack of entrepreneurial orientation of the founders. Read (2017) also highlights the importance of a network, stating that “One of the greatest assets of an entrepreneur is the people s/he knows” (p. 78). This is supported by Sarasvathy (2001), who means that one component of what makes entrepreneurs entrepreneurial is whom they know, i.e. their social networks. Furthermore, communication with people outside the team is an important factor when stimulating innovation in a workplace (Hülsheger et al., 2009). Team members that keep social relations 9 with individuals outside their own team are more likely to get exposed to different perspectives and new information, hence being able to come up with new ideas (Hülsheger et al., 2009). This is further backed by Hsieh and Kelley (2016), who also mean that recurrent exposure to other professionals may give entrepreneurs different and up-to-date information sources, which may facilitate for spotting innovative opportunities. 2.14 Hypothesis-driven approach to develop value proposition is more efficient Eisenmann et al. (2013) propose a hypothesis-driven approach to entrepreneurship, which “[...] maximizes, per unit of resources expended, the amount of information gained for resolving such uncertainty [the uncertainty in the beginning]” (p.1). The hypotheses are the founders’ underlying assumptions about their business model, and should be validated or invalidated through efficient tests, which is proposed by Ries (2011) in the LSM. Validating or invalidating hypotheses guides entrepreneurs in finding the perfect business model (Blank, 2007). Eisenmann et al. (2013) mean that “[...] the lean startup approach evaluates an early stage startup’s entire business model, whereas intellectual antecedents focus more narrowly on a startup’s product.” (p. 12). Furthermore, Klofsten (2005) concludes that “The process of ideas development does not really get going until the founders become more receptive to the world around them and involve external partners in the process. One central actor is, naturally, the potential client who becomes involved in the development work.” (p.116). Klofsten (2005) means that usually, in the earliest processes of venture creation, the idea development processes are to a large extent driven by technology, with the technical knowledge of the founders being decisive (Klofsten, 2005). Subsequently, the soft parts of ideas development are underestimated. Startup founders usually lack important resources in combination with an uncertainty about the feasibility of their business model (Eisenmann et al. (2013). 2.15 Finding the perfect product-market fit A venture should only start the scaling process, going from startups to large, “real” companies, when they reach the product-market fit (Ries, 2011). The product-market fit is when a startup team have optimized its offering to fit the market; i.e. when the product, or solution, in a profitable way meets the needs of the customers on the targeted market (Blank, 2007; Maurya, 2016; Eisenmann et al. 2013). Ries (2011) quotes Marc Andreessen’s description of what the product-market fit means: “In a great market - a market with lots of real potential customers - the market pulls products out of the startup. This is the story of search keyword advertising, Internet auctions and TCP/IP routers. Conversely, in a terrible market, you can have the best product in the world and an absolutely killer team, and it doesn’t matter - you’re going to fail.” (p.219). Ellis (2009) means that a startup has reached the product-market fit when more than 40% of the customer base would be “very disappointed” if the product or solution would cease to exist. He further means that being above the 40% line is an indicator that the startup is building the right thing. 2.16 Getting out of the building Ries (2011) means that startups need to interact with potential customers in order to understand them. Furthermore, Blank (2007) argues that the information and facts needed about the potential customers, markets, partners and sales channels are all “outside the building” and hence have to be experienced by the entrepreneurs themselves. This may be 10 done by using tests, doing observations or by interviewing potential customers (Constable, 2014). Moreover, the LSM proposes the usage tests or metrics that measure the “requested”/good behaviors of customers, e.g. number of transactions per month, rather than using non-actionable metrics, such as visitors on website or (Ries, 2011). These types of metrics are further described in section 2.22 through 2.25. Moreover, both Ries (2011) and Blank (2007) are in alignment with Constable (2014), who means that “Being told your idea is cool is not useful; seeing behavior that validates your customer’s willingness to buy is very useful” (p.29). Based on the feedback from the interactions with customers (such as testing or metrics), entrepreneurs have to make decisions whether to continue with their current business model, to persevere, or if they have to pivot, changing some of the components in the business model (Ries, 2011; Eisenmann et al. 2013). The last alternative is to perish, which means completely abandoning the venture (Eisenmann, 2013). 2.17 Overview of the business model Blank and Dorf (2012) propose to use the flexible business model canvas (Osterwalder, 2009) as opposed to using the more static business plan, and mean that this could be the difference between having to close down the venture and success. Osterwalder (2009) describes the business model as “[...] the rationale of how an organization creates, delivers and captures value.” (p.14) and proposes that ventures should use the business model canvas in order to describe their business models. Osterwalder’s business model canvas (2005) contains nine building blocks which are further described in appendix 6. Furthermore, there are other variants of the business model canvas that have been developed; the most famous one is the lean canvas (Maurya, 2016), which focuses on the customers’ broader problems. In the customer development process, a startup can use the business model canvas to manage the different hypotheses regarding each component and making changes to it as they get more insights (Blank and Dorf, 2012). Based on real customer behaviors, the firm can then either accept the customer approval or, in the case of customer negatives, make pivots and change the business model to better fit the market (Ries, 2011; Blank and Dorf, 2012). Using the business model canvas as a tool facilitates the process of pivoting, since the canvas visualizes the venture’s different alternatives and helps them see possible changes. Each time the founders make a change to the business model, they should create a new canvas visualizing the changes (Blank and Dorf, 2012). 2.18 Not wasting time by developing the wrong things Paul Graham, co-founder of Y-Combinator (famous Silicon Valley accelerator), recommends startups to do things manually initially in order to not develop automatic solutions before knowing if there is a demand for them (Graham, 2013). This means startups initially should do things that are not possible to scale to a larger company (Graham, 2013). For instance, startups could be recruiting customers manually and steadily shift to more automatic methods. This is something nearly all startups have to do (Graham, 2013) and the method was also used by AirBnB initially. Furthermore, Graham (2013) also argues that startup team members initially can pretend to be their products for as long as possible, i.e. the team members do the back-end tasks manually, while the users think they are interacting with the actual product (Ries, 2011). As time goes, the team members could steadily automate the bottlenecks (Graham, 2013). Ries (2011) calls this approach Wizard of Oz testing, and means it would be highly inefficient if the product would work this way the whole time. However, the goal of the approach is not to permanently do things manually, but rather to see if there is a demand, hence avoiding putting effort into developing an automatic product with no demand (Ries, 2011; Graham, 2013). 11 2.19 Testing crucial first Ries (2011) calls the crucial assumptions for leap-of-faith assumptions; these are the assumptions on which everything depends and hence the riskiest elements for a startup. Should product features be based on leap-of-faith assumptions that are untrue, building the features perfectly and within the time frame does not matter. It is still a waste of time since there is no one willing to pay for it, and the startup might completely fail (Ries, 2011). Furthermore, Eisenmann et al. (2013) bring up the importance of prioritizing the testing of different hypotheses. Their general principle is that “[...] an entrepreneur should give priority to tests that can eliminate considerable risk at low cost.” (Eisenmann et al., 2013, p. 9). Should the hypotheses be serially dependent of each other, it makes most sense to try the first hypothesis. If the hypotheses are not serially dependent, the founders can parallel test them, which is beneficial especially in winner-takes-it-all markets (Eisenmann et al., 2013). 2.20 Efficient testing through MVPs Savoia (2011) means that a venture should make sure they are building the right product before building the product right. The manner in which a venture should do this, by pretotyping the product, something Savoia (2011) explains as “[...] testing the initial appeal and actual usage of a potential new product by simulating its core experience with the smallest possible investment of time and money.” (p. 21) and “Make sure - as quickly and as cheaply as you can - that you are building the right it before you build it right” (p. 21). This is aligned with what Ries (2011) argues: a venture can lose valuable time by building features or even whole products, that there is no demand for. The LSM recommends startups to use minimum viable products (MVPs) in order to start the learning process as quickly as possible (Ries, 2011). There is no need for an actual physical product or prototype for the MVP; it is simply the smallest set of activities needed to validate or disprove a hypothesis (Eisenmann et al., 2013). The MVP allows a startup to go through a so-called feedback loop, meaning that testing an essential business hypothesis, with the minimum amount of effort put into it (Ries, 2011; Eisenmann et al., 2013). Ries (2011) further states that one of the most disturbing aspects of the MVP for professionals is the quality challenge. Professionals usually aim at always building high- quality products, where all the features and the functions are perfectly done, since the business model is already set and the customers are known (Ries, 2011). However, since a startup targets early adopters before it can sell to the mass market, there are no issues with selling a product that is non-perfect (Ries, 2011; Moore, 1998). 2.21 Regular and scheduled meetings to discuss changes to business model The LSM recommends that a startup should consider a pivot when the effectiveness of the product experiments decreases and feeling there should be a more productive product development (Ries, 2011; Eisenmann et al., 2013). Since the decision of pivoting is emotionally loaded, Ries (2011) suggests that it should be done in a structured way. Startups should therefore have scheduled meetings with the sole purpose of reflecting over whether to pivot, persevere or perish, where the both the product development side and the business development side are participating (Blank, 2007; Ries, 2011). Ries (2011) furthermore recommends that the effect on product optimization is discussed (over time) and compared to the expectations, as well disclosures of conversations with actual and potential customers. The reason for having the scheduled pivot, persevere or perish meetings is to not postpone 12 the inevitable, i.e. that a startup can lose precious time by not dealing with the pivot question (Blank, 2007; Ries, 2011). 2.22 Using ratios Croll and Yoskovitz (2013) recommend startups to use metrics that are ratios or rates of something, as they are easier to act on and by nature are comparative (Ries, 2011). A good metric could be comparable over time, groups of users or competitors (Croll and Yoskovitz, 2013). 2.23 Measuring what is actionable Ries (2011) means that by using bad metrics, so called vanity metrics, a startup may believe it is improving when in reality, it is not, i.e. vanity metrics hide the fact that initiatives of today are not having any impact. These metrics would typically be cumulative numbers, where it is hard to draw any fair cause-and-effect inferences, i.e. making it hard to see if e.g. a feature actually affects customer behavior or not (Ries, 2011; Croll and Yoskovitz, 2013). Croll and Yoskovitz (2013) mention some metrics to avoid: • Number of hits • Number of page views • Number of visits • Number of unique visitors • Number of followers/friends/likes • Time on site • Emails collected • Number of downloads What these metrics have in common is that they do not tell a startup anything about how the customers are using the product or if they are engaged in using it. (Ries, 2011; Croll and Yoskovitz, 2013). Hence, it is not possible to take action on these, and e.g. being positively sure to say that feature XY lead to this many more number of page views (Ries, 2011). Ries (2011) mentions the actionable metrics as good metrics. These would typically be metrics where there is a clear cause-and-effect (Ries, 2011; Croll and Yoskovitz, 2013). It can change the behavior of a startup, e.g. making a startup stop building on feature YX as it is clearly not changing the customer engagement (Ries, 2011; Croll and Yoskovitz, 2013). 2.24 Splitting customers into cohorts Ries (2011) suggests that the use of cohorts-based metrics is one of the most important tools for a startup. Cohorts make data more accessible, as it makes the data more comprehensible for the team members. A cohort analysis says “[...] among the people who used our product in this period, here’s how many of them exhibited each of the behaviors we care about.” (Ries, 2011, p. 145). In addition to cohorts, split-testing is another type of metrics that the Lean Startup recommends (Ries, 2011). Split-test experimentation (AB-testing) is when “[...] different versions of a product are offered to customers at the same time. By observing the changes in behavior between the two groups, one can make inferences about the impact of different variations.” (Ries, 2011, p. 136). Many times, this can expose that features that engineers and designer believe are good have no impact on the behaviors of customers. By using this type of testing and metrics, a startup can save much time by not doing work that customers do not care about (Ries, 2011). 13 2.25 Using validated learnings as measure of productivity Ries (2011) proposes that startup teams use learning milestones as a measure of productivity. This means using the number of validated learnings as opposed to other measurements, such as features enhanced or added to the product or solution (Ries, 2011). The rationale behind learning milestones is “If you are building the wrong thing, optimizing the product or its marketing will not yield significant results” (Ries, 2011, p. 126). Ries (2011) means that many startups blame the lack of results on the engineering team for “not working hard enough” (p. 126) whereas in reality, the problem is that the venture is executing a plan that does not work, clearly in need of a change of direction. This method of working means that product development is pulled from the business model hypotheses (Ries, 2011) and leads to faster insights regarding the business model, which in turn leads to faster opportunities to pivot. 14 3. Method The method used in this study is described below. The development of the IPLSM framework is further motivated. 3.1 Research method The purpose of the study is twofold; (1) to create a framework containing IPLSM factors considered important for startups’ success and (2) to see if there is a discrepancy in which of these factors startups believe are important and the what degree these are actually practiced by the startups. Since both purpose (1) and (2) are derived from the existing literature, this study will use a deductive research approach (Bryman and Bell, 2015). In order to fulfill purpose (1), this study uses a descriptive comparative case study in its first part. This is proposed by (Easterby-Smith et al., 2015) as it allows for comparing the opinions of the different startups to find the IPLSM factors considered most important for success. For purpose (2) to be fulfilled, the survey uses a cross-sectional survey. This allows for describing what startups’ opinions are and how they vary across the startups (Easterby- Smith et al., 2015). 3.1.1 Research flow The research flow is as follows: A pre-study was done in order to develop the IPLSM framework necessary for answering the research questions. The pre-study comprised of a literature review and interviews with startups. With the help of the pre-study, 25 statements were created and a survey testing these 25 statements was sent out to startups affiliated to a Swedish startup accelerator. 3.2 Literature review A framework consisting of IPLSM factors had to be developed in order to answer the research questions. A literature review was hence necessary to be conducted. As described in section 1.1, the IPLSM factors could be split into two types; the ”soft” ones (team spirit, intrinsic motivation etc.) and the ”hard” ones (processes, frameworks etc.). In order to be consistent, the literature review was also split this way. The first part comprised of innovation psychology and the latter comprised of the popular lean startup methodology. It shall however be noted that both parts were equally much used as foundation for the statements, as described in section 3.6.1, used in the survey for the data gathering. 3.2.1 Innovation psychology As the field of innovation psychology is rather broad, the author found direction by interpreting articles by Dr. Leif Denti, a psychologist currently researching organizational innovation at the University of Gothenburg. Based on the works of Denti, a structured literature search was conducted. The Chalmers Library Database was used for this purpose. The keywords for the search were: innovation, knowledge, leadership, implementation, diversity, experience, innovation psychology. These keywords were combined with the following words: startup, success factor and success. 3.2.2 Lean startup methodology To understand the Lean Startup movement, the book Lean Startup (Ries, 2011) was read. In order to get a better picture of its meaning, and to see possible interpretations of the book, a structured literature search of the topics it touched was conducted. The Chalmers Library Database was used for this purpose. 15 The keywords for the search were: lean startup, customer development, agile engineering and design thinking in combination with the following words: startup, success factor and success. Further on, the author took an entrepreneurship course at Chalmers University of Technology lead by Henrik Berglund, in the spring of 2017. The recommended literature for the course was also read and used for this study, as it relates to the lean startup methodology. 3.3 Data used for the study The study has used both primary data and secondary data (Easterby-Smith et al., 2015). The primary data was collected by the author in order to answer the research questions. Easterby-Smith et al. (2015) mean that “Primary data can lead to new insights and greater confidence in the outcomes of the research.” (p.8). The secondary data collected by Googol was used in order to make the comparison between large multinational companies and startups. This is presented in section 6 and in appendix 7. Easterby-Smith et al. (2015) mean that using this type of data “[...] has value through exploring new relationships and patterns within these existing data [...]” (p.8). However, Easterby-Smith et al. (2015) also mean that the context and the purpose of the data has to be taken into consideration. 3.4 Development of IPLSM framework In order to build an IPLSM framework, it was necessary to validate the literature read and ensure its relevance. For this to be done, interviews were conducted with co-founders of startups. Interviews were used as they allowed to follow up on important questions and topics. The interviews had dual purposes; besides seeing if there was a need to widen the literature review, they were also used in order to choose the most relevant theories for the IPLSM framework. Once identified, the relevant theories were iteratively made into statements. The wording of the statements was also iteratively trimmed. 3.4.1 Interviews Ten semi-structured interviews were conducted with founders/co-founders of startups. The initial two interviews were done in order to ensure the relevance of the template and to ensure the time limit of an hour for the interviews. All the interviews were conducted in Swedish, but the key takeaways have been translated into English and can be found in section 3.6. Translating from one language to another might be a possible source of error; this risk can however be considered to be very low as the author is fluent in both English and Swedish. However, there might be differences in the nuances of the quotes due to the translation of them. It shall be noted that the startups that participated in the interviews were guaranteed anonymity; in order to not disclose their identity, some of the key takeaways have been slightly altered. However, the changes have in all cases been about “hiding” the business ideas, industries or names; there has hence not been any tampering with the actual opinions of an interviewee. 3.4.1.1 Semi-structured interviews A semi-structured interview approach was chosen, as this allows working in a more flexible manner (Easterby-Smith et al., 2015). The benefits of using semi-structured interview questions is that they can “[...] often give a higher degree of confidentiality, as the replies of the interviewees tend to be more personal in nature.” (Easterby-Smith et al., 2015, p.140). It 16 shall however be noted that what Easterby-Smith et al. (2015) mentions as one benefit of interviews, that the interviewer has the possibility to identify non-verbal clues and hence can dive into further questions, was not applicable for this study to the same degree as described by Easterby-Smith et al. (2015). This was because all interviews were done through either phone calls or Skype; it was thus possible to notice changes in tone of voice, but not possible to identify facial expressions or other body language attributed information. This may have lead to some minor degree of misinformation. Another reason for why a semi-structured approach was chosen was that it made it possible to ladder up or ladder down; Easterby-Smith et al. (2015) means that this enables to see the interviewee’s value base for interesting questions or allowing the interviewee to exemplify questions. This enabled the author to be flexible in his approach. 3.4.1.2 Interview template Based on identified topics in the literature review, an interview template was designed (see appendix 1). The template was initially tested in the first two interviews in order to see if there were any topics that were irrelevant or did not fit the time limit of one hour; the startups that signed up for interviewing were told that the interview would take less than an hour and it was important to ensure this. It should be noted that the changes made to the template after the first two interviews can be considered negligible; the changes were either of clarifying character or to remove repeated questions. In order for the interview questions to be clear and easy to understand, theoretical concepts and jargon talk was aimed at being avoided throughout the interviews. However, in some cases the interviewees themselves used both theoretical concepts and jargon talk, which implied that the author could safely use the same. Furthermore, the questions were designed to spur open-ended answers, allowing the interviewees to reflect. Also, in alignment with Easterby-Smith et al. (2015), leading questions were avoided, as this risked giving the answers the author “wanted to hear”. 3.4.1.3 Interview sampling The sampling of the interviewed startups had two major limitations; the limited time frame and the dependency on the personal network. Since the interviews were done in order to create the IPLSM framework intended for the actual data gathering, the time frame was limited to less than six weeks; reading the literature, setting up an interview template, reaching out to startups, setting up an IPLSM framework and creating a survey, all had to be done within this limited time. Furthermore, the one hour long interviews can be considered an effort for startups to “sacrifice” to an unknown person; having a personal contact was key in order to get the startup people to participate. In light of this, the sampling done for the interviews can be considered to be a mixture of convenience sampling and ad-hoc sampling, as the startups were chosen based on the ease of access (personal relationship) and availability (time) (Easterby-Smith et al., 2015). Easterby-Smith et al. (2015) mean that this is the most correct method when speed of collecting the data is the priority and with a difficult access. It should though be noted that this way of sampling might be biased, and that Easterby-Smith et al. (2015) mean that the researcher using these types of sampling cannot be confident that the findings are generalizable. An attempt to increase the generalizability was made by trying to talk to startups in completely different fields; see table 1 below. The aim of interviewing the startups was however to see if other fields of literature were needed to be studied, and to choose the most relevant factors for the framework. 17 Furthermore, in all cases except one, the interviewee was either a co-founder or the sole founder. This ensured that the opinions unfolded in the interviews were representative of the opinions of the startups in the sample, assuming that the founders “know” the organization the best. The possibility that the opinions proposed by the founder/co-founders were not representative to the venture they represent exists, but can be assumed to be lower than for other types of companies. This is assumed since startups usually are much smaller and hence their members’ opinions are much more coherent than in large companies, where the interactions between the members are less intensive. Furthermore, there is a possibility that the opinions voiced by the people interviewed do not reflect the reality; some of the interviewees were personal acquaintances to the author and it is reasonable to assume that these would be ashamed if they made their teams look bad, hence not telling the whole truth or even giving erroneous statements. However, this risk was mitigated to a high extent since interviewees were promised complete anonymity in combination with being given the choice to not respond to a question if they were not comfortable with answering it. Further on, the business developer interviewed from Echo 51 was assessed to have sufficient knowledge about the venture and to represent its opinions, even though s/he was not one of the original co-founders. This was largely based on that the business developer was asked to become a partner in the venture, which indicates that the co-founders confide in her/him. Startup Type of startup Interviewee Relationship (degree) Alpha 7 Advertising Co-founder 1st Charlie 11 Education Co-founder 1st Bravo 19 Medical, Health and wellness, IT Co-founder 2nd Echo 49 Logistics Sole founder 2nd Alpha 27 Logistics, Waste management Co-founder 2nd Bravo 32 Platform Co-founder 1st Alpha 28 IT, Social Co-founder 2nd Charlie 2 Advertising, IT Co-founder 1st Charlie 24 Retail, Life style Co-founder 2nd Echo 51 Financial Business developer 1st Table 1. Showing the interviewees and their relationship to the author. 3.4.1.4 Transcription of interviews All the interviews were recorded and transcribed. It should be noted that the parts prior to the interview, i.e. the “ice-breaker”, and the parts after the interview, were not transcribed as they were not relevant to the study. Transcribing the interviews allowed for a more precise analysis of the interviews, described below. It should be noted that transcribing interviews like this might mean that non-verbal clues are ignored (Easterby-Smith et al., 2015). 3.5 Analysis of interviews The transcribed interviews were coded and later matched to the literature. This is described below. 3.5.1 Ensuring a relevant literature review The author highlighted every paragraph in the interview where the interviewee expressed either (1) that something had been positive or negative for the venture, or (2) that something was perceived to be important or bad for the venture’s success. 18 The parts highlighted were later on labeled; the label described the meaning of the paragraph. A total of 503 labels were created for this purpose. A selection from the interview with the serial entrepreneur from Alpha 7 is listed below. It should be noted that the meaning of the paragraphs are interpretations made by the author. • Everyone has to contribute equally much • People in the team are the most important • The will to work a lot, towards common goals. Does not suffice with “just enough” It should also be noted that these labels have been “cleaned up” or masked in order to maintain the anonymity of the interviewees and their ventures. In order to see if the literature review was sufficient or had to be revised, constructs were created based on the labels (see appendix 2). The labels were first organized so they covered the same type of topic, and based on the topics’ contents, the constructs were created. The constructs created comprised of between one to sixteen labels, and were categorized as either positive (perceived as important by the startups) or negative (perceived as negative by the ventures, or directly contrasting the literature). Later on, the constructs were arranged to further see overarching themes. The themes are shown in table 2 below. Themes Business model canvas Goal/vision Lean startup methodology Idea Support Processes Team Tests Experience Traits Metrics Table 2. Shows the themes identified in the interviews. These ensured the relevance of the literature. The themes and constructs indicated that in general, the literature was relevant. The constructs regarding experience and traits lead to a deeper study in the fields of diversity and previous experience of team members. Furthermore, the few constructs for the themes of growth and finance indicated that the literature was abundant; this literature was hence not used and removed from the study. 3.5.2 Ensuring a relevant IPLSM framework Once the literature was deemed relevant, quotes from the interviewees were matched to the theory groups identified in the literature. A theory group was considered to be relevant for the framework if there were at least 3 positive or negative matches, i.e. 3 matches in total, between an interview quote and the theory group. The criterion for a positive match was when either of the following was true: (1) the venture actively practices or agrees to a theory group or (2) perceives a theory group as important or positive. Furthermore, the criterion for a negative match was when either of the following was true: (1) the venture is not practicing the theory group but believes it is important, (2) the interviewee perceives the theory group as negative to practice or (3) the theory group is directly contradicted by the venture. It should be noted that the negative matches criteria were included in order to fulfill the purpose 19 of collecting data for the future study which will further research which ones of the identified factors actually are correlated to successful ventures (performance-wise). The underlying quotes have been included in section 3.6 in order to further illustrate the matches. They are structured to follow the logic of the literature review. 3.6 Matching interviews with the theory Quotes from the interviews were linked together to the theories based on the criterion in 3.5.2. This can be seen in section 3.6.1. Note that this section is structured to follow the logic of the literature review. The table in section 3.6.2 further illustrates which interviews are linked to what theory. 3.6.1 Key takeaways from interviews matched to theory In this section, the opinions of the interviewees have been compared to the literature. The opinions are summarized in the end of this section. 3.6.1.1 Complex solutions Hammond et al. (2011) mean that complex jobs might stimulate innovation (complex as in many factors depending on each other, see the full definition in section 3.7.4.1). Note that when asked what prior knowledge the team members had before joining the venture, the co- founder in Bravo 19 implies that the solution was complex since there were so many fields needed for it to work: “The problem with our solution is that it comprises so damn many fields. It is a hardware used in health care, thus it needs to follow so many standards and rules [...] a software is needed to run it, requiring servers etc.[...] I would say that talking about industry knowledge, we had, let’s say 20 %. So we had to learn extremely much the hard way” Denti and Hemlin (2012) conclude that teams working on complex tasks that are heterogeneous have the highest capability for innovation. Further, Damanpour and Aravind (2012) mean that functional differentiation is positive to innovation. Note that the Bravo 19 co-founder implies that the team members’ different expertise in different fields are important. When asked what makes her/his venture have a possibility to succeed compared to others, s/he says: “[...] and we are an interdisciplinary team. Technicians, developers, doctors.” Somech and Drach-Zahavy (2013) means that functional heterogeneity is positive to innovation. Note that the business developer interviewed in Echo 51 means that different competences are needed for their solution to work: “We have extremely different competences - that is what makes our team so important. We need to help one another for our idea to work” Further note that the co-founder of Charlie 24 responds that complementing each other with different competences has been beneficial for the venture. When asked what characteristics of people that s/he believes has been to their advantage, s/he says: “I would say we complement each other competence-wise. I am the technical one, working with product development, [...] X is the creative advertiser, [...] Y is super good in sales and has worked in that. [...]. We all have educations in our fields.” 20 Further on, note that, when the co-founder of Bravo 32 is asked what team members s/he would have chosen for a new venture in order for it to be successful, s/he implies that the venture would need several areas of expertise. This is aligned with Hammond et al. (2011) and Denti and Hemlin (2012). “I would have chosen a few, but very skilled people. A good team with cutting-edge competence in their areas” 3.6.1.2 Different people Klein and Harrison (2007) mean that team members need to be different to a certain degree, to have the right amount of information to evaluate options and see problems from different angles. Note that the co-founder of Charlie 24, when asked what team he would have handpicked for a new venture, means that it is not good to have a too similar group: “Number one, choose the right people. They are the people that are not too similar to you. [...] I think it’s good to have people of different ages. If you surround yourself with people that are too similar to you, you won’t get challenged and won’t develop” However, note that the same co-founder found that the team feeling was harder to keep when the team has grown, especially since some have been from abroad. It can hence be assumed that they have a different culture than the Swedish venture. This could imply a high diversity of opinion (Kakarika, 2013). The co-founder states: “[...] It has become harder to keep the team feeling, as our team has grown. And we have recruited from abroad as well, so we have different cultures.” Kakarika (2013) suggests that there should be moderate diversity of opinion and high diversity of expertise in an entrepreneurial team for it to be successful. Note that the co- founder of Bravo 19 believes it is important with people complimenting each other; this would be the advice given to herself/himself were s/he to create a new startup tomorrow: “To find people that you can work with and that compliment you, both competence- wise but also personality-wise.” Note that the business developer in Echo 51 believes that having too different people in the team, negatively affected the team unity: “There wasn’t always a team unity since some people were so different from each other and had very differing views on certain things” Note that the Bravo 19 co-founder implies that the team members’ different educations and functional backgrounds are important. This is in alignment with Kakarika (2013). It shall be added that the person saying the quote below herself/himself has an education and background within business administration. When asked what makes her/his venture have a possibility to succeed compared to others, s/he says: “[...] and we are an interdisciplinary team. Technicians, developers, doctors.” Kakarika (2013) proposes that the diversity of opinion should be moderate and not high. Note that the co-founder of Alpha 7 implies that there has been more friction in the venture where 21 s/he did not know the others from before. This could be interpreted as a result of a too high diversity of opinion. When asked about importance of the team members, s/he says: “It has actually been worse in the two startups that I was a part of when the school matched us, based on personality and background. It was way worse than when I started up firms with friends [...]. And besides, Y Combinator actually says that the big startups hired from their friends first because it is hard enough running a startup; if you get friction because you don’t go along well… I mean, you’ll work together so much, if you don’t get along well, it will cost more than it tastes.” Furthermore, note that when the same co-founder is asked about the value of team diversity, s/he implies that diversity is important but hard to achieve, and that it is easy to be fooled by “external signs” of diversity. S/he states: “Diversity sounds good and I think it is very good if you get someone that thinks and is in a different way. You can fool yourself that you are diverse if for example six guys take in a girl in the team, but if she thinks the same way as they do and doesn’t come with new point of views, it won’t lead to anything.” Kakarika (2013) means that diversity of expertise ought to be high for an entrepreneurial venture to be successful. When asked how the co-founder of Charlie 11 would go about creating a new venture today, note that s/he seems to be unhappy with having recruited without looking at what skills the team members had: “If I were to start a new venture today I wouldn’t just take a constellation that happens to exist, like we did then. We sort of had the mentality that “anyone that wants to join us can join us”. I would look more on what industry to go into and what we know.” 3.6.1.3 Dedicated team members Hammond et al. (2011) mean that individuals need a driving force to support them to overcome obstacles related to their innovative work. They therefore mean that individual motivation is a critical factor to consider when predicting creative performances. When the co-founder of Charlie 11 is asked about the important individual attributes of the team members, note that s/he means that it is important to have a “don’t give up”-mentality: “Perseverance is the most important attribute. That you don’t quit. [...] No real fails, it’s not like failure happens all of a sudden. What happens is that the founders decide that they do not want to continue any more. [...] If you lower your salary to 0 SEK and don’t have any debts, no one can make you shut down. [...] It doesn’t have to be that you give all you’ve got for a very long time - it’s just that you don’t quit when it gets tough.” Further, the same co-founder exemplifies the perseverance mentioned above later on in the interview, when asked about the venture’s finances: “The first year we had a salary of 3000 SEK a month. [...]. I ate porridge most of the meals and had to say no to many fun activities. It sounds bad now, and gave us a huge anxiety then. Some in our team were a bit older and left good salaries. Imagine going from about 50 000 to 3000 SEK a month. Haha!” Further, when asked if they ever thought of giving up, note that the co-founder of Bravo 19 also highly implies that perseverance is important: 22 “Initially we had some of that [thoughts of giving up]. We did some financial rounds where we could get half a million SEK in grants. And we didn’t get it. That made us depressed for a few days. [...] You’re not that used to getting those setbacks when you’re that young. Now we are more used to it. We have got 25 NOs in a row but we have money now. [...] I have one post-it note on my computer. It says “Perseverance” on it. That’s the most important thing, endurance and keep on fighting” - Further, note that when asked about the venture’s finances, the business developer in Echo 51 also implies that the co-founders were willing to go through a hard time in order to continue with the venture: “They have probably never put their own money in the venture but have rather lived on a very low or non-existing salary” Further, the co-founder of Bravo 19 also touches the topic of going through hard times. Note that s/he implies that s/he has had to give up much in social life in order to work with the startup. This can be interpreted as a sign of having the driving force that Hammond et al. (2015) mention. S/he states: “We bootstrapped for a very long time. So we’ve had more ownership but didn’t have it good financially on a private level. It has been quite tough to be honest. On average we’ve lived on 12 000 a month for 2.5 years. It’s alright when you’re a student, but it gets strained when your friends make about 3-4 more than you do. You miss out a lot of the social life - you don’t think of it but much of the social is about eating or having drinks out. I can’t do that, since I only have 1500 SEK after I’ve paid for rent and food. This means you lose a lot of your social life. You cannot join your other friends for an after-work. I think that’s a thing that many people forget about. [...] My partner has been the one supporting our relationship financially. I haven’t been able to save money for an apartment. Let’s say that our firm doesn’t go well - I’d be really broke and possibly on the street” Further note that the same co-founder’s experience is that the startups that are not committed to their startup, do not live that long. This could be interpreted as a lack of the driving force presented by Hammond et al. (2011). The co-founder states: “But I think our struggle is good. Some critique to the startup community [...], there are many “latte-entrepreneurs” that spend their time in coffee shops. They don’t have the motivation to cope with this commitment. They notice that they don’t make any money and then take in a million or so, giving away 40 % to a stupid angel. That way they finance their latte sessions another year. And then they notice that things weren’t that good. I’ve seen many of those, that aren’t fully committed. It’s more of an alternative to being a freelancer.” Further note that, when the co-founder of Charlie 24 is asked if they have thought of giving up the venture, s/he implies that it is important to focus the positive things and ignore the negative things. This can be interpreted as having the driving force mentioned by Hammond et al. (2011). S/he states: “Yes, many times. There have been times when things have been tough, like, the money will only last for two-three more months. What will we do then? We’ve worked our asses off and it still isn’t enough. And then you get a no, and things aren’t going as well as you thought. And every time, something new comes up. It hasn’t necessarily been anything big or much or so, but like getting a new distributor or opening a new key account or so, and you live on that for a while. But at times it gets 23 hard; it’s when things are hard you feel like “What the hell am I doing? I’m working 60-80 hours a week and can’t see any results of it”. And it’s like a ketchup effect, all of a sudden everything comes at the same time. You get 3-4 wins in a row. And everything is perfect again. And you forget the old, bad things” Further, when asked about the relevance of passion and drive amongst team members, note that the co-founder of Alpha 7 implies that team members need to do more than “just enough”, which could be interpreted as having the driving force mentioned by Hammond et al. (2011). S/he states: “People are A and O. Some spend time on Google searching for their next vacation places, and then complain that they have a lot to do. I’ve been in those startups too. The important is that you have the motivation to work towards the common goal. That you want to move forward and don’t just say “I’ve done bare minimum now”.” Further, note that the same co-founder also exemplifies that the team members need to have an intrinsic will to work on the common goal: “I’ve tried doing other stuff as well and you can’t try to push people to do something. Everybody has to contribute and want to build this together. That was what worked.” Also note that the intrinsic will is touched in the interview with the co-founder of Alpha 28. Note that when asked about important individual characteristics of team members, s/he means that it is important that people are driven: “[...] the drive as well. The more time you are willing to put on this, the more probable you are to succeed. Of course, you have to work efficiently as well, but if you’re prepared to put the time on it you will succeed within something. Inch by inch, even if you’re not the brightest person, you will learn how to do it.” Further, the co-founder of Charlie 2 also touches the importance of personal drive; note that drive is the first thing he mentions. This could be interpreted as having the driving force that Hammond et al. (2011) mention. When asked what he would look for when recruiting people to his “startup dream team”: “Recruit people that are extremely driven, smart and humble. That’s my main advice” Further, this is opinion is further supported by the business developer at Echo 51. When asked what the best individual characteristics of the team members have been, s/he says: “That people are driven” Further, note that the founder of Echo 49 implies that it has become even more important to motivate people the more the venture has grown. This can be interpreted as that s/he needs to stimulate the intrinsic driving force of the team members (Hammond et al., 2011). When s/he is asked what factors have been the most important for the venture this far, s/he says: “When I was alone I didn’t need to motivate people, I just had to work hard for myself. But now when more people have joined, I have needed to lead by example. If the others see that you’re motivated even though it’s tough and that you keep working - it has been super important that I am the one that believes in this when times look bad.” 24 Further note that the same founder also highlights that people need to feel so much for the venture so it feels like their own: “Even if people aren’t owning the firm, it is important that they are part of driving it. It says explicitly in our “culture book”. Run things like it was your own company. People here should always think like it’s their own firm” Furthermore, note that passion is the first thing mentioned by the co-founder of Charlie 2. This can be interpreted as the team members have a driving force which helps their innovative work (Hammond et al., 2011). When asked for what individual characteristics have been to their advantage, s/he says: “Our people is extremely passionate in different ways; curious, thirsting for knowledge - listening to podcasts, reading books [...]” 3.6.1.4 Team members aware of responsibilities Denti (2012) argues that expecting more creativity from an employee increases the chances of that employee being more creative. Note that the co-founder of Bravo 19 implies that the team members are aware of what responsibilities they have. When s/he is asked about the atmosphere in the team, s/he says: “One of our core values is “Dare and Do”, it is premiered to do stuff you haven’t done before, but you are also accountable if things screw up” Further note that the co-founder of Bravo 32 implies that team members know their responsibilities. This is in alignment with West (2002) and Denti (2012). When asked about the leadership in the team, s/he says: “And I think it is important to challenge and put demands on the one that’s responsible, it’s not OK to be sloppy. You have to set some type of principle to come forward in the process. I think it is easy to have a picture of startups being sloppy and that things are allowed to go wrong, but if things go wrong and we lose our biggest client, we lose all of our capital. So you have to challenge people and make them feel responsible.” S/he further exemplifies this. Note that s/he points out that if people do not learn from their failures, it gets frustrating: “You have to be allowed to do mistakes but let’s say someone has created the module for our website, it is important that it doesn’t break. But if it breaks for the third time, one gets frustrated and wonders how we cannot control this.” Furthermore, note that when asked what factors have been important, the Echo 49 founder talks about empowerment. S/he implies that as long as the team members know the overall direction, they are free to work however they want. This is in alignment with Rosenthal and Jacobson (1968). S/he states: “Ownership can be in form of stocks, but in our firm people really own their thing. [...] People really own their own processes. I think that is super important, that I don’t get too much into details in their work. It’s the big brushstrokes that are important.” - 25 Further note that the co-founder of Bravo 19 finds it important that the venture trusts its team members know what is right to do. This can be interpreted as a safe psychosocial climate (West, 2002). When asked of characteristics of team members that have been important, s/he says: “We work a lot with autonomy, we have faith in that people work in the right direction and know what is important.” Also, note that the co-founder of Alpha 27 believes it is important that team members are trusted to run their field, indicating of a safe psychosocial climate (West, 2002): “When we had the developer in the team, he focused on his thing. We complete each other but we also let people do their thing/work, which is nice to do and also important I think.” Furthermore, note that the co-founder of Charlie 2 believes it is important that team members are aware of that they can take action on their own. This can be a variant of the Pygmalion effect (Denti, 2012). The co-founder of Charlie 2 says: “It’s important that people are independent and can make their own decisions, without being dependent on someone else. If you’re good, which you are if you are in our team, and if you think that what you’re doing is the right thing for the firm, we trust you on that.” Furthermore, when asked what traits have been important for the team, note that the co- founder of Alpha 27 implies that the venture lets team members know what is expected from them. This can be interpreted as the Pygmalion effect (Rosenthal and Jacobson, 1968; Denti, 2012). S/he states: “We are not afraid to question each other, set demands for each other and double check afterwards” Furthermore, note that the co-founder of Charlie 11 mentions that, as long as s/he knows it is within the explicit frames of a problem, s/he can do whatever s/he wants. This is clearly a team member that knows what is expected of her/him from the rest of the team. S/he says: “The ball comes to me and X especially. And we break it down to smaller pieces. We then start working much more independently, and I don’t ask the others in the group about what mini-problems to focus on - I choose them for myself. And as long as I know that they are within the frames of our main problem, I just go for it.” 3.6.1.5 Team members aiming to learn new things Investing in and developing human capital is connected to productivity, venture growth and innovation (Evans-Raoul, 2013; Holmberg-Wright and Hribar, 2016). Note that the co-founder of Charlie 24 believes it is important for the venture’s success that the team members continuously develops their capabilities. When asked what the venture does to learn how and what to do in the future, s/he says: “We study quite a lot. Trying to get everyone at work to listen to some type of podcast related to their job, for at least 30 minutes a day. That is equivalent to reading 18 books a year. It’s hard to take the time to sit down and actually read, but listening is easy. If we can get all our coworkers to do so, I think we have a lot to win.” 26 Furthermore, note that the co-founder of Charlie 2 also believes that team members continuously developing their capabilities is important. This is in alignment with Evans-Raoul (2013) and Holmberg-Wright and Hribar (2016). The co-founder of Charlie 2 states: “Our people is extremely passionate in different ways; curious, thirsting for knowledge - listening to podcasts, reading books [...]. If you are smart and thirsting for knowledge, you can beat the ones that know a lot already.” This opinion is also shared by the co-founder of Charlie 11. Note that s/he believes that being updated with the latest research distinguishes her/his venture from many other ventures. Staying up-to-date with the latest research is aligned with Holmberg-Wright and Hribar (2016). When asked what the venture has done to learn how to run the venture, s/he replies: “We’ve put a lot of effort on that. That’s probably the thing we’ve achieved as a team that I’m the most proud of. That’s where I find the biggest differences on what we’ve done and other teams have done [...]. We are in the industry of education and behavioral change. We didn’t get it from Y-Combinator but they phrase it very well. “If you’re going to operate in a field, you should be the leaders in that field. You should know more than the others do in the field. What you do is, you read books about the field, and to really get that edge, you have to read research about it.” That’s something that very few people do. You simply download scientific research papers from the web and just read them.” Furthermore, when asked about the knowledge prior the startup, note that the co-founder of Alpha 28 implies that the co-founders has had to learn many new things along the way. This is in alignment with Holmberg-Wright and Hribar (2016): “Most has been to learn. He’s [the other co-founder, developer] learned a lot on the way. It’s been a lot about commitment. He has put a tremendous amount of time to wrap his head around this” Furthermore, note that the co-founder of Bravo 19, means that that team members have learned new things has been important for them. This is in alignment with Evans-Raoul (2013) and Holmberg-Wright and Hribar (2016). S/he says: “I would say that talking about industry knowledge, we had, let’s say 20 %. So we had to learn an extremely lot the hard way. But we had other knowledge, like self- awareness that we don’t know that much but being self-confident in that we can learn, or find someone that can teach us. And like, usurp the knowledge in some way.” However, note that the founder of Echo 49 believes that not having offered people a workplace where they can develop and learn has been negative for the venture. This can be interpreted as that Echo 49 has not invested in human capital, which is not in alignment with Evans-Raoul (2013) and Holmberg-Wright and Hribar (2016). When asked what has been bad for the venture, the founder of Echo 49 says: “Some people have left our firm through the years [...] I should have coached them more and talked more about their future roles. Now, it’s more been like that they’ve said they want to go do an internship somewhere else to “learn and develop themselves”. So, working with the personal development, we didn’t do that earlier. We didn’t sh