The Effect of Key Performance Indicators on Startup Growth Master of Science in Management, End of Studies Project 2017 Melinda Elmborg, Colombe Godeluck Supervisor: Anita Quas Abstract Substantial amounts of capital are yearly invested in venture capital in the digital age, even though 70 % of investments have historically been considered failures. Industry experts discuss the knowledge and use of key performance indicators (KPIs) to make startups succeed, but no research supporting this hypothesis has been advanced. Against this background, this study examines data collected by surveying 12 early stage digital startups, comparing different variables against revenue growth. Previous research has revealed the importance of experience, financial literacy, social qualifications and economic context to improve the success of a new venture. KPIs and management control systems (MCSs) are considered playing an active role in innovation and entrepreneurship. The data shows that KPIs are indeed important for the success and more precisely the growth of a startup as the knowledge and use of KPIs are correlated by 0.8 to the revenue growth, with a corresponding adjusted R2 of 0.5. This study contributes to the development of critical success factors for digital startups and can hopefully encourage stakeholders and entrepreneurs in the eco-system to learn more and further research KPIs. 1 Table of Contents Abstract ......................................................................................................................................................................... 1 1. Introduction ....................................................................................................................................3 1.1 Background ........................................................................................................................................................ 3 1.2 Problem Discussion .......................................................................................................................................... 4 2. Literature Review.............................................................................................................................6 2.1 Creation and Survival of New Ventures ......................................................................................................... 6 2.1.1 Creation ...................................................................................................................................................... 6 2.1.2 Survival ....................................................................................................................................................... 7 2.2 Key Performance Indicators ............................................................................................................................ 9 2.2.1 Definition of Key Performance Indicators ............................................................................................. 9 2.2.2 Types of KPIs ............................................................................................................................................ 10 2.2.3 The Balanced Scorecard ......................................................................................................................... 11 2.2.4 MCS, MAS and Performance Measurement in Startups ................................................................... 12 2.3 The Digital Industry......................................................................................................................................... 14 2.3.2. Overview of the industry ...................................................................................................................... 14 2.3.2. Critical Success factors of the industry ............................................................................................... 17 3. Methodology................................................................................................................................. 19 3.1 Sample .............................................................................................................................................................. 19 3.2 Procedure......................................................................................................................................................... 19 4. Results .......................................................................................................................................... 20 4.1 Background ...................................................................................................................................................... 20 4.2 Analysis ............................................................................................................................................................. 21 4.2.1 Descriptive Statistics ............................................................................................................................... 21 4.2.2 Correlations.............................................................................................................................................. 21 4.2.3 Regression Analysis ................................................................................................................................. 24 4.2.4 Specific KPIs ............................................................................................................................................. 26 4.3 Limitations ....................................................................................................................................................... 28 5. Conclusion and Discussion .............................................................................................................. 30 References .................................................................................................................................................................. 32 Appendix – Survey Questions .................................................................................................................................. 36 2 1. Introduction 1.1 Background As business students in a leading Business School focusing on entrepreneurship, we were very interested in researching information about successful entrepreneurs as we might become startup founders later in life. In 2016, €7.9 billion was invested in venture capital funds (Invest Europe, 2017). When looking at the success rate of investments in new ventures in Europe, the numbers are not encouraging. According to the European Investment Fund’s (EIF) latest report from 2017, only 30 % of investments generated a return. Out of the 3 600 seed and VC supported investments by EIF during 1996-2015 the multiples on cost was only 1,16x, with a median being at 0,12x. Furthermore, there was only 4 % of exits that returned more than five times the investment while 70 % of exited investments were either written off or sold for an amount below the initial investment (Prencipe, 2017). Why do most fail and a few succeed? Most people in the sector have theories and arguments, such as this example by Steve Jobs: “I’m convinced that about half of what separates the successful entrepreneurs from the non-successful ones is pure perseverance. It is so hard and you pour so much of your life into this thing, there are such rough moments in time that most people give up. And I don't blame them, it's really tough.” (Lie, 2010) Steve Hogan has devoted his life to trying to save failing startups and has with his years of experience interesting opinions to share. According to him, it is especially common for sole first-time founders to fail as they lack a partner and experience. The second common factor among failing startups, according to Hogan, is that the founders did not study potential buyers before developing the product and now lack customers. Indeed, the lack of interested customers seems to be a key factor also according to other studies on failed startups. CB Insights summaries every year the “failure post-mortems” written by the founders themselves of the latest startups that have closed, and they have identified certain reasons to which the founders refer. According to the study, the most common cause seen in 42 % of the post-mortems is that the product of the startup was not needed on the market, as seen in figure 1. The second most common reason is that the company ran out of cash, seen in 29 % of the cases and certainly closely linked to high costs, lack of revenues and investments (Griffith, 2014). 3 0% 10% 20% 30% No Market Need 29% Not the Right Team 23% Get Outcompeted 19% Pricing/Cost Issues 18% Poor Product 17% Need/Lack Business Model 17% Ignore Customers 14% Product Mis-Timed 13% Lose Focus 13% Disharmony on Team/Investors 13% Lack Passion Bad Location 10% 9% 9% No Financing/Investor Interest 8% Legal Challenges 8% Don't Use Network/Advisors 8% Burn Out 8% Failure to Pivot 50% 42% Ran Out of Cash Pivot Gone Bad 40% 7% Figure 1 Top 20 reasons startups fail (Griffith, 2014) The factors leading to a company running out of cash can be controlled when using tools to measure, analyze and act according to them. That is probably why people in the field of new ventures find KPIs crucial for success. According to venture capitalist Phil Nadel (2016), one of the ways to succeed is to know and use KPIs. He states in an article written for Techcrunch that there is a direct correlation between the knowledge and use of KPIs and the success of a startup. Furthermore, he explains that the KPIs allow the company to tweak the product and business model to achieve high performance. However, Nadel does not mention any proof, and the claimed correlation seems to be more of a theory, which is what gave us the idea to this End of Studies Project. 1.2 Problem Discussion The research question of this End of Studies Project is formulated to: What is the impact of knowledge and use of KPIs on revenue growth in digital startups? When formulating this question, we have decided to study digital startups, which we define as early stage ventures with a software as a product, such as a subscription software (SaaS), an online marketplace, an app or a game for example. In the definition of a startup, it is assumed that startups are fast growing (en.wikipedia.org, 2017a), which is why we have decided to study revenue growth as a definition of 4 success. Moreover, the European Startup Monitor of 2016 showed that 88 % of startups find that revenue growth is important to them which further argues for this measurement (Kollmann et al., 2016). KPIs are defined as Key Performance Indicators which are used to do a performance measurement. According to The American Institute of Certified Public Accountants (AICPA), performance mana gement is one of three areas of management accounting, sometimes also referred to management accounting systems (MAS) (en.wikipedia.org, 2017b). In turn, MAS is part of an organizations’ management control systems MCS (en.wikipedia.org, 2017c). MCS MAS KPI Figure 2 The relationship between MCS, MAS and KPI We have decided to focus on KPIs because it is the expression with the most established and well-known out of the three. When using Google’s search function, we find that “key performance indicators” has 3,5 million matches, while “management accounting systems” only has 300 000 and “management control systems” has 400 000. However, in our literature review, we have included MAS and MCS for new ventures to allow for a more multifaceted understanding of the subject. This research study will first explore critical success factors (CSF) of startups based on the “Top 20 reasons why startups fail”, and KPIs, as explained above. Then, it will investigate the quantitative relationship between these two components within the digital sector, based on an industry analysis and survey made among twelve startups. 5 2. Literature Review 2.1 Creation and Survival of New Ventures 2.1.1 Creation The creation of a company depends on two factors that need to be present, lucrative opportunities and enterprising individuals identifying the opportunity. Lucrative opportunities appear when there is a difference in price of resources and the price at which the product or service can be sold. The recognition of the opportunity is highly dependent on asymmetrical information which makes the recognition a subjective process, even though the real opportunity is objective but not known to all parties. The subjective information might be about underutilized resources, new technology, non-satisfied demand or regulatory changes. To increase the probability of recognition of an entrepreneurial opportunity, the individual needs to possess prior information and the necessary cognitive properties to value it (Shane & Venkataraman, 2000). On the other hand, entrepreneurs in the field of new technologies might experience different kinds of opportunities as the market often is not yet existent or at the beginning of its development which assumes high uncertainty. Rather than using a causation process as explained above, it might be more relevant to use an effectuation process to create a new technological-intense venture. The causation process follows the strategic management theory by focusing on market, firm and customer analysis. This process is more about managerial thinking as it consists of selecting between given means to achieve a pre-determined goal. In other words, it is when an entrepreneur has a given specific goal and acquires the means to achieve it. This process demands an environment with high predictability and access to many recourses. The effectuation process is rather about entrepreneurial thinking as it consists of imagining possible new ends using a given set of means. In other words, it is when an entrepreneur has a certain amount of means, such as knowledge or social network, to achieve different effects. This process is more value-driven and is easier to operate in an environment with little predictability. Effect 1 Mean 1 M1 Mean 2 Goal M2 Effect 2 M3 MN Mean 3 Effect 3 Mean N Effect N Selection and creation of means to reach a specific goal Discovering potential effects from a given set of means Figure 3 Causation and effectuation processes 6 It has even been shown that in many entrepreneurial success cases, the entrepreneur has not only used the effectuation process to recognize and pursue an opportunity but also to create it and the market as part of the implementation of the entrepreneurial process (Sarasvathy, 2001). 2.1.2 Survival Previous findings suggest the significance of numerous variables determining entrepreneurial success. We will consider each of it, point out what can we learn from it, and try to link these findings together. For this project, we decided to analyze five sets of Critical Success Factors (CSF), based on the 20 most relevant indicators for startups’ failure as mentioned in the introduction. For an organization to succeed, there are a few things that must go well; these are the Critical Success Factors. They are essential for achieving high performance (Freund, 1988). Business experience and knowledge The first set of variables that are the determent in new ventures’ success is experience and knowledge. There is a relationship between selected organizational predictors and entrepreneurial success. Notably, entrepreneurs with professional experience in company management, having unique knowledge, and with a network, achieve greater entrepreneurial success than persons with no such skills and knowledge. To test this hypothesis, the researcher examined new ventures settled between 2008 and 2012 in Poland; 294 entrepreneurs’ responses were analyzed. Among the respondents, the majority were the sole owners of the businesses they managed and overall, they were not experienced in either running their own company or in managing a company (Staniewski, 2016). There is a relation between the human capital of founders of new technology firms and new venture success. The human capital of founders includes the education and prior work experience of the team. Mainly, founders with a university background in economic, managerial or scientific fields tend to enjoy greater growth rates than entrepreneurs with an education in other areas. As well, the fact that a founder has experience in the same industry of his new venture is positively associated with growth. An d finally, a team where individuals have prior entrepreneurial experience also results in superior growth (Colombo & Grilli, 2005). Those criteria related to the characteristics of the entrepreneur or entrepreneurial team are also those used by venture capitalists when assessing new businesses to select the best to invest in (Hall & Hofer, 1993). Financial literacy The second set of variables we found interesting is the impact of financial literacy, which means the ability to read, analyze, manage and communicate about financial issues. Among a study including 509 young entrepreneurs in Canada, there was a 77% chance of surviving the first year and a 36% chance of surviving the first five years among those who were financially literate. Furthermore, it was discovered that frequent use of financial tools, ratios and frequent production of financial statements increased the likelihood of loan repayment and decreased the probability of venture failure. According to the research, poor financial literacy undermined the entrepreneurial activity. Indeed, knowledge about financial management contributes to entrepreneurship skills as it increases financial understanding and responsibility, it seems to be vital for an entrepreneur to understand the nature of money and the economic consequences and risks associated with decisions. Moreover, these 7 decisions should influence financial tasks such as budgeting, financial analysis and accounting (Wise, 2013). Cognitive and social factors The third set of variables affecting entrepreneurial success is cognitive and social factors. The ability to get along with others is more likely to contribute to success - the higher an entrepreneur’s social competence, the greater its financial success. Indeed, it is fundamental to project a positive image, build a competent business team, and a solid reputation, be a strong negotiator and follow-up with customers and partners to make the new venture grow and prosper. All these characteristics rely on social competencies. Several cognitive and social factors have been proven to be linked to entrepreneurs and entrepreneur success. One of the factors is overconfidence, as it has been shown that entrepreneurs tend to be overconfident in their personal judgment rather than non-entrepreneurs. Regarding risk taking, entrepreneurs tend to perceive higher potential gain in uncertain situations compared to nonentrepreneurs. A study has been done on entrepreneurs in the field of high-tech, which showed that social perception, meaning accuracy in perceiving others, was a significant predictor of financial success for these entrepreneurs (Baron, 2000). Economic Context The fourth set of variables we found interesting is the economic environment. Indeed, new venture creation is vital to a healthy economy, as an important driver of job creation, innovation and economic growth (Wise, 2013). These facts lead Devece et al. (2016) to research the fundamental entrepreneurial factors that drive the growth of new businesses under different economic conditions. In countries with greater inequalities as emerging economies, entrepreneurship is more a necessity rather than an opportunity. But, when an entrepreneur is pushed to discover and exploit business opportunities because of the lack of viable alternatives such as unemployment, it is more likely to be less successful because of a lack of motivation. Nonetheless, countries facing economic crisis provide more opportunities for entrepreneurs than during periods of prosperity. The key for them to succeed is opportunity recognition and innovation. It is particularly the case for new ventures where the founder has outstanding knowledge of a sector and is thus capable of identifying opportunities. The researchers validated this hypothesis, by using data related to entrepreneurship in Spain from 2004 to 2010, including data from the financial crisis of 2008-2009. They further recognize the importance of opportunity driven entrepreneurship achieves more success over necessity-driven entrepreneurship. In conclusion, during an economic boom the number of entrepreneurs is higher than during a recession, but on the other hand, there is a larger part of ventures created during a recession that achieves success rather than during an economic boom. Women-owned businesses The fifth and last set of variables is related to gender. Research studies have examined the relationship that exists between the skills possessed by women entrepreneurs and their motivations, barriers, and performance. 8 Women and men tend to turn to different industries when becoming entrepreneurs, as women often go into retail and services while men are more oriented towards manufacturing. Furthermore, research has shown that there are no differences in survival rate between male and female founders. Moreover, studies have concluded the following: lack of education and managerial skills are two of the most important variables when it comes to understanding the motivations and difficulties women business owners face and factors for success – revenue, size, and growth. Women believe a prior business experience has a positive influence on their managerial approach as it makes them more self-confident, more flexible, more tolerant, and more careful as well as a positive implication on their business performance as it enables them to build relationships intuition and perspective (Huarng et al., 2012). Lee et al. (2009) also analyzed how business performance is affected by the CSF of women-owned business and made a comparison between the USA and South Korea. Three different CSFs were tested, family support and succession, communication ability and product/service competency and managerial ability to see how these correlated with business performance. For the success of businesses owned by women, the competitiveness of the type of business and the female business owner’s managerial ability are relatively more important to communication ability, family support, and the succession. 2.2 Key Performance Indicators 2.2.1 Definition of Key Performance Indicators Key Performance Indicators (KPI) are often seen in a business environment but has become known as a mainstream expression. To define what a KPI is, one should first define performance. Performance is the ability of an object to produce results in relation to a target (Laitinen, 2002). It is a relative concept that is more advanced than just observed and measured as it also implies future outcome. Many academics, authors and experts have tried to describe what a KPI is, and one of them is Parmenter (2007). He considers the KPI to be one of three types of performance measures: Key Result Indicators that tell you how you have done in the past, Performance Indicators that tell you what to do and then the KPIs that focus on what to do to increase the performance radically. He explains it as the different indicators are parts of layers, as described in below. According to Parmenter, KPIs focus on the parts of an organization that is the most crucial for its success both currently and in the future. He also suggests a set of rules to what a KPI should be and how it should be treated: non-financial, measured frequently, acted on by the top management team, a connection between the measure and an action should be understood, responsibility tied to a team, and it should have a significant and positive impact. 9 KPIs PIs Peel the core to find the KPIs KRIs Peel the skin to find the PIs Figure 4 Three types of performance measurement (Parmenter, 2007) Fitz-Gibbon (1990) is talking about performance indicators instead of key performance indicators which he ultimately gives a more general description: “A performance indicator can be defined as an item of information collected at regular intervals to track the performance of a system.” He is also talking more about the implied behavioral impact of a performance indicator, as they often cause logical and emotional reactions. When putting performance indicators in place, a manager or business owner needs to consider the behavioral implications and pay attention when they are put in practice. Neely et al. (1995) consider the measurement of performance as the process of quantifying action where they link the action to performance and quantification to measurement. Performance can either be of effectiveness or efficiency, where effectiveness is a non-economic indicator and efficiency is economic. Consequently, a performance measure is a metric used to quantify the efficiency or effectiveness of an action. As a performance measure is quantifying the result of an action, it ultimately stimulates a particular action and requires a strategy before being used. Smith (2013) is the owner of the website Made to Measure KPIs and helps businesses to develop relevant KPIs to support their decision making. In his book KPI Checklists, he explains Key Performance Indicators to: “show you how you are doing at a particular activity to achieve a particular level or outcome” 2.2.2 Types of KPIs There are diverse types of KPIs, usually what is mentioned are financial and non-financial indicators and lagging and leading indicators. Financial indicators are based on the financial performance and often 10 closely connected to revenue and profit. Moreover, this type of KPIs were the only ones used historically. A later development has been the non-financial KPIs which can be about a performance that is not directly connected to revenue or profit but is any way closely linked to the company’s strategy. These ca n be about customer satisfaction or R&D progress for instance (Neely et al. 1995). Lagging and leading indicators are referring to the chronological order of the event and the indicator, for example, revenue is measuring an event that has already happened while customer satisfaction can indicate a future change in revenue. A good dashboard of any company should always include both, financial and non-financial as well as both lagging and leading KPIs (Smith 2013). Performance metrics can also be divided into either quantifying the efficiency or the effectiveness. Efficiency measures the economic outcome of the firm compared to resources utilized while effectiveness is referring to what extent the customer requirements are met. Neely et al. (1995) refer to an example: “Take, for example, one of the quality-related dimensions of performance – product reliability. In terms of effectiveness, achieving a higher level of product reliability might lead to greater customer satisfaction. In terms of efficiency, it might reduce the costs incurred by the business through decreased field failure and warranty claims.” In this way, a performance metric’s raison d’être can either refer to internal targets regarding efficiency or external targets regarding effectiveness (Neely et al., 1995). 2.2.3 The Balanced Scorecard The balanced scorecard has been essential for the development of the performance measurement field, as that is the first-time non-financial indicators were considered. The balanced scorecard was an answer to the mentality of the Industrial Age where measurement systems were closely connected to a company’s finance function. The traditional systems had a control bias, to regulate employees’ actions. The balanced scorecard is a way to structure and link performance indicators. Kaplan and Norton created four categories of metrics to achieve a holistic view of the company as it includes both financial and nonfinancial KPIs. The categories are called perspectives: Financial Perspective, Internal Business Perspective, Customer Perspective and Innovation and Learning Perspective. Each category will consist of several performance indicators with associated goals. 11 The Balanced Scorecard Links Performance Measures Financial Perspective Goals Measures How do we look to shareholders? How do customers see us? What must we excel at? Internal Business Perspective Goals Measures Customer Perspective Goals Measures Innovation and Learning Perspective Goals Measures Can we continue to improve and create value? Figure 5 The Balanced Scorecard (Kaplan & Norton, 1992) When creating the Balanced Scorecard, Kaplan and Norton focused on strategy and vision, rather than control. They thought that well-chosen performance indicators with supplied targets would motivate employees to certain behaviors and pull them toward the vision of the company. According to the authors, the Balanced Scorecard makes the company look and move forward, instead of backward (Kaplan & Norton 1992). Nevertheless, the Balanced Scorecard has received its fair amount of criticism. It is developed to be an optimal solution for all types of companies. Unfortunately, it is most suited for large and stable businesses in a mature development stage. On top of that, the methodology that should be used to develop the Balanced Scorecard demands a substantial amount of human resources, something that is not possible for many small or medium sized enterprises (SMEs) to supply. In the end, the Balanced Scorecard is not suitable for most types of companies, which in turn have been seen developing tailored Balanced Scorecards appropriate for their particular organization (Rillo, 2004). 2.2.4 MCS, MAS and Performance Measurement in Startups Startups are working in settings with high volatility and unpredictability, as they often disrupt current markets or create new ones. To survive and create change, they need to be flexible to adapt to unexpected events and quickly absorb novelty (Davila et al. 2009). 12 When structuring companies into Defenders and Prospectors, startups belong in the category of Prospectors. It is because Prospectors develop markets by being innovative, while Defenders rather operate with fewer products differentiating themselves by cost, quality or service. Prospectors repetitively seek new markets and are therefore competing through new product development, while Defenders spend few recourses on this. Regarding management control systems, Prospectors attach great importance to them as they forecast data, set tight budget goals and monitor the outputs carefully while putting less focus on cost control. Defenders, on the other hand, use control systems less frequently and rarely change them. By this, it has even been observed that the companies with the most innovation of markets, models or products also have the tightest control. Similarly, companies operating in environments with high uncertainty use control systems that demand more interaction within the firm and more frequent attention of managers (Simons, 1987). While studying research done on the subject, we find very different results that are both contradicting and supporting the assumptions mentioned above. After considering new ventures, researchers arrive at different conclusions, some such as Granlund and Taipaleenmäki (2005) argues that startups spend little time on performance measurement while Endenich (2017) claims that they do to a considerable extent. In a study done on New Economy Firms (NEF) in the beginning of 2000, it was discovered that due to lack of time and resources these firms did not consider performance measurement, strategic planning and sometimes even internal financial analysis. The reason seemed to be the engineering oriented culture that was typical for the firms participating in the research project. The reasons to not do performance measurement were many, such as the high intensity of R&D, the uncertainty of the environment and the future as well as unusual roles and revenue models (Granlund & Taipaleenmäki 2005). It can be argued that entrepreneurs of new ventures do not need management control systems since they rely on personal controls, characteristic for small organizations. Moreover, some think that the innovative culture typical for new ventures might be at risk if adopting a management control system, which creates another reason not to implement a system (Chenhall & Moers 2015). When looking at a study done very recently, we find a behavior opposite to the one discussed above. The study looked at Information Technology driven earliest stage startups and found that they all used a formal version of a management control system that was dynamic, technology-driven and multi-faceted. The startups studied were part of an ecosystem of other startups, web agencies and software developers, which they used to outsource a significant part of their operations. In contrast to previous studies, it was shown that the startups need to control other than themselves, limiting them from using informal faceto-face communication and personal supervision. To assure the communication and success of both inhouse and outsourced tasks, MCSs were crucial. The characteristics of the operations demanded MCSs that were available worldwide from any device to allow a flexible and collaborative use of the systems. In a startup, the MCS has a complex and important role for survival and development of the business. The systems are not only used internally to support employees and founders but also to legitimate the startup to external stakeholders, for example, investors or incubators (Endenich, 2017). Having formal procedures support employees in achieving efficiency and increase their commitment to the organization (Adler & Borys, 1996). 13 This new role of management control systems found among startups is part of the New Control Paradigm, where control is positive and necessary for innovation and entrepreneurship. Since the characteristics of an MCS in the New Control Paradigm is so different from a traditional MCS, the system should be redefined as Smart Control Systems to emphasize the positive effect of control (Endenich, 2017). The use of control or management accounting systems seems to be in a dual relationship with the lifecycle stage of the startup (Granlund & Taipaleenmäki, 2005). The formality of control or management accounting systems change over the different stages of the lifecycle as the organization changes, but it is also firms with high growth that are also the fastest on increasing the formality of their system (Moores & Yuen, 2001). From a study of 78 early-stage startup companies, researching multiple variables with the rate of adoption of MCSs, clear associations were found to support the idea that MCSs are of high relevance to efficiently grow early-stage startups. The variables that are positively associated with the adoption of MCSs are the number of employees, the presence of venture capital, international operations, and time to revenue. Moreover, the study also implies that the rate of adoption affects the company size and the CEO turnover. Indeed, it was proven that CEOs that implemented fewer MCSs had shorter tenures (Davila & Foster, 2007). MCSs also have a significant role in signaling firm quality and future growth potential to external financiers as they believe these will lead to improved decision making. Effectively, it has been proven that MCS intensity is positively correlated with firm value and this goes for all industries and around the globe. Cross sectional analyses imply that it is particularly the case for startups operating in competitive sectors with high growth-rates (Davila et al., 2015). 2.3 The Digital Industry 2.3.2. Overview of the industry During the last couple of decades, the information technology revolution has developed to allow for the digital industry to appear. The revolution is considered a major historical event in line with the 18thcentury industrial revolution, with the IT today having the same central role as new sources of energy had then. At that time, the key element was the generation and distribution of energy; today, it is the technologies of information processing and communication. The exception of the IT revolution though is the speed at which it is applied globally, all thanks to the contact between civilizations in the 21st century (Castells, 2011). As of the recent development, a new economy of information- and knowledge-based assets has emerged. The digital economy transforms all sectors by digitizing information enabled by new technology. The transformation is still far from being completed, which functions as a generator of new businesses, creating opportunities for modern corporations to grow and new to emerge (Brynjolfsson & Kahin, 2000). The IT industry, including hardware, software, services, and telecommunications, was in 2016 estimated to be worth $3.8 trillion, which represents about 5 % of the world GDP. About one-third of this originated from the United States, while most of the growth lately comes from Asia and especially China (CompTIA, 2016). When concentrating on the software market, which is the focus of this project, we find that it represents $469 billion, so about 12 % of the total IT industry. The industry is led by giants such as Microsoft, IBM, and Oracle (Shields, 2014). On the other hand, it is hard to accurately categorize the market for digitized goods as new business models and industries emerge rapidly. We should mention 14 companies operating on the Internet such as Alphabet (parent company of Google) with a revenue of $90 billion, Facebook of $28 billion and Chinese Tencent of $22 billion (En.wikipedia.org, 2017d). From the digital economy, the term digital entrepreneurship has emerged as a subcategory of entrepreneurship. Digital entrepreneurship means that the entrepreneur sells something in a digital form, that earlier was physical. Since the goods sold are no longer physical, many of the rules and learnings of business are no longer valid, which opens for new business models, organizational structures, and products. The distinctions between traditional and digital entrepreneurship, and those among the three types of digital entrepreneurship, can be seen in the context of ease of entry, ease of manufacturing and storing, ease of distribution in the digital marketplace, digital workplace, digital goods, digital service, and digital commitment (Hull et al., 2011). For the second year in a row, the European Startup Monitor was released in 2016 by the German Startup Association and the European Startup Network. It is based on 2 500 startups from 17 countries responding to a survey. The survey is used to map European startups to gain unique and authentic insights into the emerging startup ecosystems. According to the ESM, startups are defined by three characteristics: • • • Startups are younger than 10 years. Startups feature (highly) innovative technologies and/or business models. Startups have (strive for) significant employee and/or sales growth. Most participating startups generate revenue through enterprise customers, or B2B as it is often called, in opposite to generating revenue through consumer customers, B2C. Over half of respondents answered that they were mostly or only generating revenue by B2B while only about 5 % did B2C . Only B2B 9% 5% Mainly B2B 9% 38% Mainly B2B with some B2C B2B and B2C in equal measures 9% Mainly B2C with some B2B 12% 18% Mainly B2C Only B2C Figure 6 Customers through which startups generate revenue (Kollmann et al., 2016) 15 ESM 2016 has identified many different areas at which the startups operate, of which the most common were IT, Software Development and Software as a Service. However, the group of startups in the study has a high diversity regarding categories, see the graph below for details. IT, Software development Software as a Service (Saas) Industrial Technology, Production, Hardware Consumer Mobile, Web Application E-Commerce Bio- Nano-, and Medical Technology Finance Technology (FinTech) Online Marketplace Education Consultning Company, Agency Online Service Portal Green Technology Food Media and Creative Industries Offline Services Games Stationary Wholesale and Retail Other 377 307 209 171 166 146 131 123 121 116 106 101 86 83 33 33 15 199 0 50 100 150 200 250 300 350 400 Figure 7 Number of startups operating in each category (Kollmann et al., 2016) Founders are dominantly male, with only 15 % of founders being women. This minority position holds true in all countries, as the most equality was discovered in the United Kingdom where one-third of participating founders were female. The smallest part of female founders was found in Austria, where only 7 % were women. Most startups were founded by teams, as only one in four startups were founded by sole founders. Almost half of founders have previous experience of past created ventures, yet, only a third of these previous ventures’ operations have been discontinued as the other two-thirds of previous ventures still exist. The startups in the study are characterized by an international workforce, as one-third of employees are international. The study shows a good diversity regarding revenues, as one in five startups don’t yet have revenues while about the same amount had over €500 000 euros during the last fiscal year. However, when looking at a country level, there is a clear trend that startups in the Northern and Central Europe tend to have more revenues while those in Southern Europe tend to have lower revenues. As an example, out of the studied startups in Finland, 95 % had an annual revenue above €150 000. 16 > €500 000 393 €150 000 - €500 000 407 €50 000 - €150 000 374 €25 000 - €50 000 244 €1 - €25 000 568 No revenue 528 0 100 200 300 400 500 600 Figure 8 Annual revenue in the last fiscal year per number of startups (Kollmann et al., 2016) On average, 68,1 % of startups have received external funding, and here, as well, we see a trend that Northern and Western European startups have a higher chance of being funded and with larger amounts rather than the Southern Europeans. As an example, out of the studied startups in Finland, 95,7 % had received external funding while that number was 52,4 % in Cyprus. Regarding the development of the startup, most founders, 87,7 %, agree that revenue growth is necessary, while only 29,2 % are satisfied with their actual growth. When looking at the overall strategy of the startup, 77,7 % of founders find product development to be important or very important, while profitability only was considered important or very important for about 60 %. However, only 44,7 % of founders considered developing a strong company culture to be important or very important (Kollmann et al., 2016). 2.3.2. Critical Success factors of the industry As web technology increases throughout the industry, education, and government, the CSF has been discussed widely. The first set of variables is related to the ability to efficiently attract and maintain customers as it is the base of success within the digital industry - success being the capacity to achieve goals for which it was designed. Hence, e-commerce transactions have been classified into three phases of marketing: pre, online, and after sales. The first phase includes advertising, public relations, and other related services to raise brand awareness. The second phase is the one that will transform “visitors” into “buyers.” Kotler (1994) stressed that trustworthy, dependable and reliable characteristics are essential to trigger business transactions. Finally, the third phase is about problem resolution and customer services to ensure customers’ loyalty. In parallel of these three phases, a necessary condition is a security for online transactions. Customers would not pay for products or services if financial information could not have been transmitted securely. In a study where webmasters among Fortune 1000 companies were surveyed, where they focused on web design to identify success factors of an online business. They identified four essential elements of 17 success: quality of information and service, system use, playfulness, and system design service. Web designers and online business owners should consider these factors at all stages of their interaction with the customer, from presale to after-sale. It is important to control the online transaction process of the customer to make sure that it has high quality and promotes customer excitement. In conclusion, it was verified that system design quality leads to success for online businesses (Liu & Arnett, 1999). The second set of variables is related to the company itself: its behavior such as attitude and passion, its management team, its ability to attract quality investors, and to get a high-quality product to the market rapidly (Preston, 1997). Eid et al. (2002) noted two sets of variables as CSF. A strong marketing strategy incl uding advertising and customer relationships and support from the top management team is essential to digital business success. External and website related factors as trust, security, customer acceptance, ease-of-use and website design are also demonstrated as factors of success in the digital industry. When looking solely at software as a service (SaaS) businesses, it has been found that performance is the most significant success factor whereas cost saving is another crucial factor. Furthermore, it is the organizational level of a company that experiences the highest added value of the SaaS. To run a software company with success, it is considered necessary to have high system quality (Walther et al., 2012). 18 3. Methodology 3.1 Sample The scope of this study was set to startups with a digital service or product having experienced revenues for about 1-2 years. The startups were all based in Europe (Sweden, Finland, France, and Belgium) and receive revenues either in the form of subscriptions or commissions. One of the founders of the startup participated in the study, usually being the CEO (Chief Executive Officer), and more rarely the CMO (Chief Marketing Officer), or a responsible of business development. 63 entrepreneurs were asked to participate in the study. In total 24 founders participated in the study, ultimately yielding a response rate of 38,1 %. Out of these, 12 submitted sufficient information. The most common reason to why the information submitted was considered insufficient was an unwillingness to provide details on revenues due to confidentiality issues or having an irregular revenue stream which made it impossible to measure month on month revenue change. 3.2 Procedure The respondents participated in the study by submitting an online survey of 23 questions, a copy of the survey can be found in appendix 1. The survey was split up into three themes, 1. Background: We asked about gender, educational and entrepreneurial background. Samples of questions include “What is your entrepreneurial experience?”, “How often do you perform financial statements?”. 2. Knowledge and use of KPIs: We tested the knowledge by listing 20 frequently seen KPIs among startup blogs (for example David Slok’s website forentrepreneurs.com) without any explanation and often as an abbreviation. The entrepreneur was asked to tick the KPIs he or she could explain in a sentence what they mean. The entrepreneur was then requested to answer what type of revenue model they applied, to receive questions on which of the nine important KPIs of their revenue model that they use weekly and monthly. 3. Revenue growth: We asked about the revenue per month over the last 12 months. We used the results to calculate an average month on month growth, to use to test correlations, according to the following formula: πππ£ Σ ( πππ£π+1 − 1) π πππ πΊπππ€π‘β = Σ ππππ‘βπ − 1 To be able to use the experience as a numerical independent variable in the analysis, we created a weighted score to test in relation to the average growth per month. Since the experience of having started or worked in a startup should be considered more important than if a family member did, or entrepreneurship was a part of the education, we decided to use these weights. No startup had both claimed to have an entrepreneur in the family and studied entrepreneurship when a startup founder has received 1 or more as a score; it is because he or she has earlier founded or worked in a startup. πΈπ₯ππππππππ πππππ = ππ‘πππ‘π’π ∗ 1 + πΉπππππ¦ ∗ 0,5 + πΈππ’πππ‘πππ ∗ 0,5 19 4. Results 4.1 Background Most of the startups we surveyed were led by only men (92%). In general, they have a master degree (83%), and all have a business education. Regarding their entrepreneurial experience, 75% have worked in or created a startup before. They all received funding from a third party, could it be relatives, banks, VC or angel investors. Looking back to Hall and Hofer’s article (2013), 67% received funds from a VC meaning those last ones perceived a high potential in the new ventures thanks to the educational or entrepreneurial background, which is verified by our results. The impact of financial literacy (Wise, 2013) cannot be verified explicitly, as 50% of the respondent perform a financial statement every month, 25% once a year, 17% every quarter, the rest every semester. Besides, the average of KPIs known is nine over a list of 20 KPIs. We cannot conclude anything regarding the management (male or female), business context and social competence since we do not have enough data about it. Table 1 represents the characteristics of the respondents (industry, gender, education, experience). Gender Number Percentage 11 0 1 92% 0% 8% Diploma High School Bachelor Master PhD 1 1 10 0 8% 8% 83% 0% Education Business Engineer Law Web Development Other 12 5 2 2 1 100% 42% 17% 17% 8% Only male Only female Female and male 20 Entrepreneurial experience Previously in a startup Relatives Education None 9 4 3 2 75% 33% 25% 17% Table 1 Characteristics of the respondents 4.2 Analysis To test our research question, we used averages, analysis of correlation, multiple regression and individual regressions including ANOVA to analyze the data provided by the entrepreneurs. 4.2.1 Descriptive Statistics The dependent variable is the average month-on-month growth in the analysis. The independent variables, which we tested to see the impact on the dependent variable, were: entrepreneurial experience, the number of KPIs known, the frequency of financial statements generation and the number of KPIs used. We have chosen to test more variables than the KPIs to be able to see if KPIs have a superior effect on growth over experience or frequent creation of financial statements. Initially, we have anticipated that any increase in the independent variables has a positive impact on business performance. Experience KPIs Known Statements KPIs Used Growth Average Standard Deviation 1.042 0.582 9.833 3.326 7.083 5.230 5.500 1.784 17.92% 9.96% Maximum Minimum 1.5 0 16 4 12 1 9 3 32% 4% Possible Maximum 2 20 12 9 Table 2 Descriptive statistics of the variables In table 2, we illustrate descriptive statistics of the explained variables. The table shows mean, standard deviation, maximum, minimum and possible maximum for each variable. When separating the startups between subscription and commission based, we see that subscriptions have a higher average monthly growth at 22 %, while it is 15 % for commissions. 4.2.2 Correlations To discover any relationship between the independent variables and the growth rates, we test the individual correlations. 21 Experience 35% 30% Growth 25% y = 0.066x + 0.111 R² = 0.148 20% 15% 10% 5% 0% 0 0.5 1 1.5 2 Experience score Figure 9 Relationship between experience and growth In figure 9, we can see the relationship between the founders’ experience and the growth rates, with a correlation of 0.385 and an R 2 of 0.148. Ultimately, the experience does not seem to very well predict the average growth of a startup. However, it is interesting to see that the startups performing above average had all earlier experience of founding or working in a startup, even if it does not seem to be a guarantee for high performance. KPI Knowledge 35% 30% Growth 25% y = 0.018x + 0.003 R² = 0.358 20% 15% 10% 5% 0% 0 5 10 15 20 KPIs known Figure 10 Relationship between knowledge of KPIs and growth In figure 10, we can see the relationship between the founders’ knowledge about KPIs and the growth rates, with a correlation of 0.598 and an R 2 of 0.358. Thus, the knowledge of KPIs seems to be fairly good at predicting the revenue growth of a startup. We interpret the knowledge of KPIs not only as a founder who knows KPIs by heart but as a founder who is curious about metrics, that has therefore researched the subject, and it probably is an important part of the culture of the company. 22 Financial Statements 35% 30% y = -0.005x + 0.216 R² = 0.073 Growth 25% 20% 15% 10% 5% 0% 0 2 4 6 8 10 12 14 Statements per year Figure 11 Relationship between financial statements per year and growth In figure 11, we can see the relationship between the number of financial statements produced per year and the growth rates, with a correlation of -0.270 and an R2 of 0.073. It might seem interesting to see the negative correlation, but the R2 is very low which indicates that there is no significant relationship between the frequency of financial statements and growth. KPI Use 35% 30% Growth 25% y = 0.026x + 0.035 R² = 0.219 20% 15% 10% 5% 0% 0 2 4 6 8 KPIs used regularly Figure 12 Relationship between KPIs regularly used and growth In figure 16, we can see the relationship between the number of KPIs measured at least monthly per and the growth rates, with a correlation of 0.468 and an R2 of 0.219. The relationship is not very significant, even though it is better at predicting the growth than experience and the number of financial statements per year. We interpret the use of KPIs as to the extent the startup is using formal processes to measure their performance. 23 Experience KPIs Known Statements KPIs Used Growth Experience 1 0.145 -0.106 0.591 0.385 KPIs Known Statements KPIs Used 1 0.393 -0.092 0.598 1 -0.745 -0.270 Growth 1 0.468 1 Table 3 Correlations between variables We anticipated some strong correlations between the independent variables; this can be because of multiple reasons such as that the small sample, omitted variables bias, multicollinearity or measurement error. We would still like to mention a couple of correlations that are particularly high. The strongest correlation of -0.745 is found between the use of KPIs and the frequency of production of financial statements. It seems like startups make a choice between measuring KPIs and produce financial statements. Another relatively high correlation is found between experience and the use of KPIs as the correlation is 0.591. Those founders who have previous experience from entrepreneurship seem also to measure KPIs to a greater extent. 4.2.3 Regression Analysis To better understand how the different independent variables, affect the growth rate of a startup, a multiple regression has been performed. The multi-regression statistics presented in table 4 indicates, at first sight, a good fit of the model as the multiple R is 0.838 and the Adjusted R 2 is 0.532. Regression Statistics Multiple R R2 Adjusted R2 Standard Error Observations 0.838 0.702 0.532 0.068 12 Table 4 Regression statistics for the variables experience, KPI knowledge, financial statements and KPI use Regression Residual Total Intercept Experience KPIs Known Statements KPIs Used Coefficients -0.015 0.036 0.024 -0.011 0.000 df 4 7 11 SS 0.077 0.033 0.109 Std Error 0.159 0.056 0.007 0.008 0.028 MS 0.019 0.005 T Stat -0.096 0.645 3.298 -1.262 0.002 F 4.124 P-value 0.926 0.539 0.013 0.248 0.998 Sign. F 0.050 Lower 95% -0.392 -0.096 0.007 -0.030 -0.067 Upper 95% 0.361 0.169 0.040 0.009 0.067 Table 5 ANOVA for the variables experience, KPI knowledge, financial statements and KPI use The Analysis of Variance (ANOVA) shows a significant effect of the four variables on business performance since the significance F is 0.05. It means that there is a less than a 5 % chance that the relationship between the variables is due to random chance. Analyzing in more detail the different variables, the 24 number of KPIs known appears to play a significant role in the revenue growth of a startup. This variable is the sole to have a satisfying P-value at 0.013. Therefore, it is a 98.7 % chance that there is an actual relationship between the number of KPIs known and the growth of the company. The entrepreneurial experience, the frequency of financial statements generation and the number of KPIs used are not significant since the value is many times more than 0.05. In the chapter on correlation above, we saw that knowledge and use of KPIs, in fact, had the most significant correlations, and therefore we would also like to test the multiple regression and ANOVA of only knowledge and use. Regression Statistics Multiple R R2 Adjusted R2 Standard Error Observations 0.796 0.634 0.552 0.067 12 Table 6 Regression statistics for the variables KPI knowledge and use Regression Residual Total Intercept KPIs Known KPIs Used df SS MS F Sign. F 2 9 11 0.069 0.040 0.109 0.035 0.004 7.782 0.011 Coefficients Std Error T Stat P-value Lower 95% Upper 95% -0.173 0.019 0.029 0.092 0.006 0.011 -1.881 3.190 2.604 0.093 0.011 0.029 -0.382 0.006 0.004 0.035 0.033 0.055 Table 7 ANOVA for the variables KPI knowledge and KPI use In this multiple regression, we have achieved a very strong regression of 0.796, supported by an adjusted R2 of 0.552. The close relationship showed here is further supported by the low Significance F of 0.011 which shows there is a less than a 1.1 % chance that the relationship between the KPIs and growth is due to random chance. The P-values for both the variables are low, as they are below 0.05 which supports that KPI knowledge and KPI use predict the growth of a company. This analysis confirms that the combination of knowledge and use is better at foreseeing the growth, in deference to the sole analysis of KPI knowledge and use, as the R 2 are 0.358 and 0.219 respectively. According to this linear regression analysis, the monthly average growth of a startup can be predicted according to this formula: πΊπππ€π‘β = 0.019 ∗ πΎππΌπ πΎπππ€π + 0.029 ∗ πΎππΌπ ππ ππ − 0.173 In other words, for every new KPI that the entrepreneur learns, the startup will improve the monthly average growth by 1.9 percentage points. For every KPI that the entrepreneur starts to analyze regularly, the startup will improve the average monthly growth by 2.9 percentage points. It is hard to present this graphically since it is a regression with two independent variables, the closest graphical presentation 25 would look like the figure 13. We have turned the absolute numbers into percentages to be able to add them together. 35% 30% Growth 25% 20% 15% 10% 5% 0% KPI Knowledge and KPI Use Figure 13 The relationship between KPI knowledge and use In Figure 13 above, we see how the participating startups are spread around the line ar regression in a more aligned way, than when looking at the KPI knowledge and KPI use separately. There seems to be that both knowing KPIs and using KPIs are important to predict the growth, but it also means that a weaker ability in one can be compensated by the other. The two top performers having over 30 % of monthly growth are very good examples of how one can compensate the other. One of the top startups is very good at using KPIs since it scored 9 out of 9, but on knowledge, they scored close to the average at 10 out of 20. The second one, on the other hand, scored close to the average on the use of KPI with 6 out of 9, while being very good at the knowledge of KPIs by claiming to know 16 out of 20. It might seem puzzling that this analysis assumes the possibility of not knowing KPIs but still using them. Nevertheless, based on how the survey was structured, we have seen examples of startups declaring that they do not know certain KPIs but that they later claimed to use regularly. Out of 42 possible overlaps between use and knowledge, only 37 appeared. There were in total 3 startups out of the 12 that claimed they used certain KPIs but that they did not know in the previous question. It is probably due to the fact that to test the knowledge we only listed abbreviations and less prominent names, while when asking for KPIs that they regularly analyze we stated definitions to make sure that they recognize the KPI even if they use a variant. 4.2.4 Specific KPIs The respondents to the survey filled out the specific KPIs that they knew and used. We have then analyzed these and looked at the effect of using each one of them. In table 8, we list the KPIs and look at the number of startups that claimed to know these terms and abbreviations without seeing the definition, and the average month on month growth of these startups. 26 KPI Definition Frequency 1 Share of total 8% Average Growth 32% Cost per loyal user Cost per install DAU/MAU 3 25% 30% Cash/burn rate 4 33% 25% NPS Net promoter score 5 42% 23% MAU Monthly active users 6 50% 22% ARPA Average revenue per account 4 33% 21% 4 33% 21% % visitors visiting more than one web page Customer acquisition cost 9 75% 21% 9 75% 20% Total costs per month 10 83% 20% Average revenue per user 6 50% 20% % users that are active after a certain period Monthly recurring revenue 9 75% 19% 7 58% 19% 11 92% 18% 12 100% 18% DAU % customers that leave during a certain period % users that convert to paying customers Daily active users 3 25% 17% LTV Life time value of a customer 10 83% 17% AOV Average order value 1 8% 16% GMV Gross merchandise value 4 33% 15% EPMV Earnings per monthly visitor 0 0% CPLU CPI Sticky factor Runway Bounce rate CAC Burn rate ARPU Retention rate MRR Churn rate Conversion rate Table 8 KPIs known by the respondents and the average effect on month-on-month growth The most commonly known KPI was conversion rate, that was known by all the respondents, followed by churn rate known by all respondents except one. When looking at the top of the list at the KPIs that were known by the highest performing startups, we see a particular trend. None of the top KPIs are closely linked to revenues, rather costs as cost per install (CPI) or runway, otherwise user satisfaction as the sticky factor or NPS. In the bottom, we find KPIs closely linked to revenue, as GMV, AOV, and LTV. These metrics are on the other hand closely linked to the commission revenue model and more widely recognized by the commission startups. The average growth among these startups is lower, which decreases these metrics’ effect according to this list. We also asked about the relevant KPIs that the startups measure and analyze monthly. To make the KPIs relevant, they have been split up depending on the revenue model: subscription or commission. In table 9, we list the KPIs for the subscription model and study the number of startups claiming that they analyze this metric and their average month-on-month growth. 27 Subscription KPI LTV/CAC ratio Frequency 1 Share 20% Average Growth 32% Customer Lifetime 3 60% 24% Churn rate 4 80% 24% CAC 3 60% 23% Conversion rate 5 100% 22% LTV 2 40% 22% ARPA 3 60% 21% MRR 4 80% 20% App ranking 0 0% Table 9 KPIs used by the subscription respondents and the average effect on month-on-month growth All 5 startups claimed to regularly analyze the conversion rate, followed by the monthly recurring revenue (MRR) and churn rate which was used by 4 startups. Among these startups we see a clear trend, it seems like customer lifetime oriented metrics increase the performance the most. It appears that when a startup controls their churn rate and make sure customers stay customers as long as possible, it has an apparent positive effect on performance. In table 10, we list the KPIs for the commission model and study the number of startups claiming that they analyze this metric and their average month-on-month growth. Commission KPI Frequency Share AOV Average commission 2 2 29% 29% Average Growth 22% 22% Conversion rate Renewals CAC Number of transactions GMV Average listing price 6 4 7 7 5 2 86% 57% 100% 100% 71% 29% 17% 16% 15% 15% 12% 11% Buyer-to-seller ratio 0 0% Table 10 KPIs used by the commission respondents and the average effect on month-on-month growth All 7 startups claimed to analyze the customer acquisition cost (CAC) regularly, and the number of transactions and all by one periodically analyzed the conversion rate. It is difficult to draw any conclusions based on this data, but we see that startups analyzing average order value (AOV) and the average commission seem to achieve the highest growth rate. 4.3 Limitations Unfortunately, these results have been obtained on a limited sample of startups since there were only 12 startups who were willing to share their information with us and had the right type of revenue stream. Another limitation is that we were unable to test the effect of gender, sources of financing and education since all respondents except one consisted of a team of only men, they all had received external financing, and all but two had studied a master. Furthermore, the study is limited to two revenue models and has not been considering startups depending on advertising, one-time purchase or in-app purchases. 28 Another limitation is that we have not been able to see the results of KPIs in a long-term perspective, it would be of importance to know how KPIs affect the survival rate of startups. 29 5. Conclusion and Discussion The findings from this research has broadened and deepen our understanding of how KPIs affect business performance in startups. According to our study, it seems like key performance indicators influence the growth of startups during the first few years. Indeed, our correlation report supports our hypothesis. We found that the strongest correlation is found when combining knowledge and use of KPIs as we obtained a correlation of 0.8 and an adjusted R 2 of 0.5. Ultimately, it is not only important to know KPIs but also to use them regularly. Our results suggest that understanding an additional KPI increases the average monthly growth by 1.9 percentage points while analyzing an additional KPI regularly increases it by 2.9 percentage points. Though, the study indicates that a sound knowledge can compensate a bit lower use of KPIs and vice versa. It seems like KPIs do not only have to be a part of a startup’s formal process but can be compensated by founders being curious and conscious about them during the decision-making process. In other words, KPIs should rather be a part of both a company’s culture and the formal processes, even though strength in one can to some extent compensate the other. Our study confirms the finding of Endenich (2017) that the MCS has a complex and important role for survival and development of a startup. In line with Adler and Borys (1996), it appears that formal procedures seem to achieve efficiency, resulting in higher growth rates. External financiers appear to be right to consider that MCSs signal firm quality which ultimately leads to a higher valuation, according to research made by Davila et al. (2015). Opposite to what Chenhall & Moers (2015) and Granlund & Taipaleenmäki consider, it seems not to be sufficient to only rely on personal controls even though the organization is small. When innovating markets, models or products, startups appear to profit on structured systems according to the results of Simons (1987). The study cannot statistically confirm that previous entrepreneurial experience affects the growth of early stage startups, according to the study of Colombo & Grilli (2005), even though all respondents performing above average had previous experience. Furthermore, it appears that the regular production of financial statements does not have the same effect on growth as KPIs, and even has a weak negative correlation, which questions the traditional role of finance and accounting in a startup. It could, therefore, be interesting to examine the study of Wise (2013), where he claimed that a high frequency of financial statements improved the survival rate of the new ventures. A hypothesis might be that highly innovative technological startups might not profit from traditional financial accounting since they differ drastically in organizational structure and business model compared to traditional companies. Based on the result, several recommendations can be advanced. Understanding of KPIs should be a priority in the ecosystem, and the knowledge should be passed on to new entrepreneurs, as it today appears to be the serial entrepreneurs that are the best at analyzing KPIs. Implementation of KPIs should be of importance to entrepreneurs and investors; hopefully, this study can lead to that entrepreneurs are motivated to learn more about and measure KPIs regularly. To achieve this, stakeholders within the ecosystem of startups should take responsibility, might it be investors, incubators or governmental organizations, to train and motivate entrepreneurs in performance improvement. 30 According to the European Startup Monitor, 88 % of startup founders consider revenue growth to be important, but less than one third are happy with their growth (Kollmann, 2016). For the remaining startups who want to improve their growth, they might be motivated to overlook their understanding of KPIs and corresponding formal processes to hopefully enhance their growth to a satisfying level, according to the findings of this study. As we saw earlier, the European Investment Fund’s average yield on investment was only 1.16x as a multiple to cost, with a failure rate of 70 %, it should be of interest of investors to learn more about the critical success factors, such as KPIs, of their investments. In 2016, €7.9 billion was invested in venture capital funds (Invest Europe, 2017), but if it keeps on yielding low returns, investors might start to look elsewhere for investments which will be hurtful for the financing of future venture capital funds and startups. Knowledge about critical success factors of startups, as KPIs as in this report, could potentially improve the returns on the venture capital investment market which in turn would benefit innovation and entrepreneurship. 31 References Adler P., Borys B. (1996). Two types of bureaucracy: enabling and coercive. Administrative Science Quarterly, Vol. 41, pp. 61-89. Baron, R.A. (2000). Psychological Perspectives on Entrepreneurship. Current Directions in Psychological Science, Vol. 9, Issue 1, pp. 15 – 18. Brynjolfsson, E. and Kahin, B. (2000). Understanding the digital economy. Cambridge, Massachusetts, USA: MIT Press, pp. 1-12. Castells M. (2011). The Rise of the Network Society. West Sussex, United Kingdom: John Wiley & Sons, pp. 1-27. Chenhall R.H., Moers F. (2015). The role of innovation in the evolution of management accounting and its integration into management control. Accounting, Organizations and Society, Vol. 47, pp. 1- 13. Colombo M. G., Grilli L. (2005). Founders’ human capital and the growth of new technologybased firms: A competence-based view. Research Policy Vol. 34, Issue 6, pp. 795-816. CompTIA (2016). IT Industry Outlook 2016. [online] Available at: https://www.comptia.org/resources/it-industry-outlook-2016-final [Accessed 25 Jun. 2017]. Davila A., Foster G. (2007). Management Control Systems in Early-Stage Startup Companies. The Accounting Review, Vol. 82, No. 4, pp. 907–937. Davila A., Foster G., Oyon D. (2009). Accounting and control, entrepreneurship and innovation: venturing into new research opportunities. European Accounting Review, Vol. 18, No. 2, pp. 281-311. Devece C., Peris-Ortiz M., Rueda-Armengot C. (2016). Entrepreneurship during economic crisis: success factors and paths to failure. Journal of Business Research, Vol. 69, pp. 5366-5370. Eid R., Trueman M., Ahmed A.M. (2002). A cross-industry review of B2B critical success factors. Internet Research: Electronic Networking Applications and Policy, Vol. 12, No. 2, pp 110-123. En.wikipedia.org. (2017). List of largest Internet companies. [online] Available at: https://en.wikipedia.org/wiki/List_of_largest_Internet_companies [Accessed 25 Jun. 2017]. En.wikipedia.org. (2017a). Startup company. [online] Available at: https://en.wikipedia.org/wiki/Startup_company [Accessed 25 Jun. 2017]. En.wikipedia.org. (2017b). Management accounting . [online] Available at: https://en.wikipedia.org/wiki/Management_accounting [Accessed 25 Jun. 2017]. 32 En.wikipedia.org. (2017c). Management control system. [online] Available at: https://en.wikipedia.org/wiki/Management_control_system [Accessed 25 Jun. 2017]. Endenich C. (2017). Management control systems in the entrepreneurial arena: refining the new control paradigm. Working Paper. Fitz-Gibbon, C.T. (1990). Performance indicators. Clevedon, Avon, England: Multilingual Matters, pp. 1-4. Freund Y. P. (1988). Critical success factors. Planning Review, Vol. 16 Issue: 4, pp.20-23. Granlund M., Taipaleenmäki J. (2005). Management control and controllership in new economy firms: a life cycle perspective. Accounting, Organizations and Society, Vol. 16, pp. 21-57. Griffith, E. (2017). Why startups fail, according to their founders. [online] Fortune.com. Available at: http://fortune.com/2014/09/25/why-startups-fail-according-to-theirfounders/ [Accessed 22 Jun. 2017]. Hall J., Hofer C.W. (1993). Venture capitalists’ decision criteria in new venture evaluation . Journal of Business Venturing, Vol. 8, pp. 25-42. Huarng K-H., Mas-Tur A., Yu T. H-K. (2012). Factors affecting the success of women entrepreneurs. International Entrepreneurship and Management Journal 8, pp. 487497. Hull, C. E. K., Hung, Y. T. C., Hair, N., Perotti, V. (2007). Taking advantage of digital opportunities: a typology of digital entrepreneurship. International Journal of Networking and Virtual Organisations, Vol. 4, No. 3, pp. 290-303. Invest Europe (2017). 2016 European Private Equity Activity. [online] pp.10. Available at: https://www.investeurope.eu/research/activity-data/annual-activity-statistics/ [Accessed 25 Jun. 2017]. Kaplan R., Norton D.P. (1992). The balanced scorecard: measures that drives performance . Harvard Business Review, January-February 1992, pp. 71-79. Kollman T., Stöckmann C., Hensellek S., Kensbock J. (2016). European Startup Monitor. German Startups Association. Kotler P. (1994). Marketing Management: Analysis, Planning, Implementation, and Control, 8th Edition., Englewood Cliffs, New Jersey, USA: Prentice-Hall. Laitinen E.K. (2002). A dynamic performance measurement system: evidence from small Finnish technology companies. Scandinavian Journal of Management, Vol. 18, Issue 1, pp. 65-99. Lee S.S, Stearns T.M., Osteryoung J.S., Stephenson H.B. (2009). A comparison of the critical success factors in women-owned business between the United States and Korea. Int. Entrep. Manag. Journal 5, pp. 259-270. 33 Lie, E. (2010). INSTANT MBA: Passion And Perseverance Are What Separate Successful Entrepreneurs From Failures. [online] Business Insider. Available at: http://www.businessinsider.com/instant-mba-passion-and-perseverance-are-whatseparates-successful-entrepreneurs-from-failures?IR=T [Accessed 22 Jun. 2017]. Moores K., Yuen S. (2001). Management accounting systems and organizational configuration: a life-cycle perspective. Accounting, Organizations and Society, Vol. 26, pp. 351-389. Nadel, P. (2016). 11 reasons we didn’t invest in your company. [online] TechCrunch. Available at: https://techcrunch.com/2016/09/05/11-reasons-we-didnt-invest-in-yourcompany/ [Accessed 22 Jun. 2017]. Neely A., Gregory M., Platts K. (1995). Performance measurement system design. International Journal of Operations & Production Management, Vol. 25, No. 12, pp. 1228-1263. Parmenter, D. (2007). Key Performance Indicators: Developing, Implementing, and Using Winning KPIs. Hoboken, New Jersey, USA: John Wiley & Sons, pp.1-7. Prencipe D., (2017). The European venture capital landscape and EIF perspective. Volume III: liquidity events and returns of EIF-backed VC investments. EIF Research & Markets Analysis, working paper 2017/41. Preston, J. (1997). Success Factors in Technology-Based Entrepreneurship. Massachusetts Institute of Technology, MIT Entrepreneurship Center. Rillo M. (2004). Limitations of Balanced Scorecard. Proceedings of the 2nd Scientific and Educational Conference, Business Administration: Business in a Globalizing Economy, Parnu, 30-31 January 2004, 155-161. Sarasvathy S. D. (2001). Causation and Effectuation: Toward a Theoretical Shift from Economic Inevitability to Entrepreneurial Contingency. The Academy of Management Review, Vol. 26, No. 2, pp. 243-263. Shane S. & Venkataraman S. (2000). “The Promise of Entrepreneurship as a Field of Research”. The Academy of Management Review , Vol. 25, No. 1, pp. 217-226. Shields, A. (2014). Must-know: An overview of the software industry. [online] Marketrealist.com. Available at: http://marketrealist.com/2014/07/must-knowoverview-software-industry-2/ [Accessed 25 Jun. 2017]. Simons R. (1987). Accounting control systems and business strategy: an empirical analysis. Accounting, Organization and Society, Vol. 12, No. 4, pp. 357-374. Slok, D. (2013). SaaS Metrics 2.0 – A Guide to Measuring and Improving What Matters. Retrieved from: http://www.forentrepreneurs.com/saas-metrics-2/ Smith B. (2013). KPI Checklists. Sheffield, England: Metric Press, pp. 7-8. 34 Staniewski M.W. (2016). The contribution of business experience and knowledge to successful entrepreneurship. Journal of Business Research, Vol. 69, pp 5147-5152. Walther S., Plank A., Eymann, T., Singh N., Phadke G. (2012). Success Factors and Value Propositions of Software as a Service Providers – A Literature Review and Classification. Proceedings of the Eighteenth Americas Conference on Information Systems, Seattle, Washington, August 9-12, 2012. Wise S. (2013). The impact of financial literacy on new venture survival. International Journal of business and management, Vol. 8, No. 23. 35 Appendix – Survey Questions 36 37 38 39 40 41 42 43 44