Determinants of the crowdfunding campaign success in the Areas of Music and Sports Graça Trindade Tiago Miguel Teixeira Campos Silva Instituto Universitário de Lisboa (ISCTE-IUL), Lisboa, Portugal Maria da Conceição Santos Business Research Unit (BRU-IUL), Lisboa, Portugal Instituto Universitário de Lisboa (ISCTE-IUL), Lisboa, Portugal Av. das Forças Armadas, Lisboa Av. das Forças Armadas, Lisboa 1649-026 Lisboa, Portugal 1649-026 Lisboa, Portugal email@example.com firstname.lastname@example.org and email@example.com Abstract —This study aims to contribute to a better understanding about the specificities of different crowdfunding areas in attracting the supporters’ attention. Using data from the largest Portuguese platform, PPL, the study compares the determinants of success of the crowdfunding campaigns in two different areas: Music and Sports. The conceptual model is based on a multiple regression specified in a multiplicative way, which is estimated under the assumption that the project is financed. The results show that the success determinants are different according to the crowdfunding area. In the area of Music, the key factors that contribute positively to the success of financing a project are the number of views per project and the level of requested amount per project. In the area of Sports, the factors that explain the success are the number of images per project, the runtime project and being connected online by Facebook and the factors that explain its failure are the time period to launch a project: failure occurs more often for projects launched by March/April or September/October. These results show the importance of specifying properly the management of a campaign and the network management by crowdfunding area and should give rise to new studies for different industries and in different geographic contexts. technology that made it easier to reach a large number of people . Web 2.0 is based on the principle that collective actions by online consumer communities can contribute to new forms of business organization in which consumer leisure activities can result in productive work being tapped as a free resource for businesses. The development of Web 2.0 is therefore identified as a prerequisite for the evolution of crowdsourcing and crowdfunding, as its structure has easily reached consumer networks. The term crowdfunding only emerged in 2006 and gained more prominence in 2012 with the law that regulates crowdfunding to encourage financing of small US companies, the JOBS Act (Jumpstart Our Business Startup) in the USA. It was initially developed in the arts industry, specifically in music, films and games, and was controlled by a single Platform, Sellaband, founded in 2006. Keywords - crowdfunding, PPL platform, entrepreneurs, supporters, gap between requested amount and raised amount. Kickstarter became a reference for crowdfunding and is considered the dominant platform with the largest impact in this market . In Portugal, there are already about half a dozen platforms, where PPL is the largest one, having raised about 1.3 million Euros by the end of 2015 . The first platforms emerged in 2011 and not all disclose data on amounts raised, number of campaigns supported (successful or unsuccessful) or number of investors/supporters. I. INTRODUCTION The crowdfunding term derives from the crowdsourcing concept since it is allowed to use collective knowledge of a group of people through a network of contacts to solve a specific problem. Building upon increasing internet access, crowdfunding became a rising alternative financing model to the current financial system, at least for small projects, triggering the attention of academic researchers . For small projects, it is very hard to get bank loans or equity investors, such as venture capitalists and business angels. Thus, the entrepreneurs are more likely to get funds from a large crowd of individuals through the use of a platform . Crowdfunding provides access to the financing of a new product or idea, to hold an event or to record a music album, but also allows to get feedback on that idea or product. From the marketing point of view, validating the idea of the entrepreneur or testing the impact of the idea or product on the market are extra benefits to the use of this type of financing methodology . Even in campaigns that fail to achieve their funding goal, the entrepreneurs obtain important feedback on the reasons why their products did not have an impact on the public or even figure out how they can improve the services or products presented. According to Massolution Crowdfunding Industry 2015 Report, globally, crowdfunding platforms raised $34.4 billion, up 18.2 billion dollars from 2014. The use of crowdfunding platforms, which act as intermediaries between entrepreneurs and investors, increases the potential of obtaining both financing and awareness of the projects. These platforms provide assistance to the launch of a campaign and to the collection of funds, usually in exchange for a commission. The use of this tool is in fact very simple, but, that does not guarantee the success in the pledge to obtain financing. In Portugal, this type of financing is still expanding. There has been an increase in the number of people and organizations using crowdfunding platforms, but Portugal is still one of the least developed countries in Europe. The Portuguese platform PPL obtained a success percentage of 45% of all launched It stands out that it was the use of social networks and projects . Crowdfunding can also be a financing alternative for startups, micro-enterprises and small and medium-sized enterprises (PMEs). The economic impact of this method of financing in Portugal has been acknowledged, and is regulated since August 24th, when Law No. 102/2015 was published in order to rule the practice of Portuguese platforms. This defines the duties of the platforms and what they are not allowed to do as well as establishing some limits on investment. Like other forms of funding, crowdfunding is based on the principle of trust and can be seen as funding that occurs in the initial state of a project . The main objective of this study is to understand the factors that better explain the success of a campaign and, thus, to contribute to the deepening of the factors that influence the success of a campaign. It is organized into four chapters besides the introduction: literature review, methodology, data analysis and conclusions. II. LITERATURE REVIEW Nowadays, companies have ceased to create value unilaterally as consumers became less passive. The consumer evolved to become a prosumer as first defined by Alvin Toffler in 1980 . The prosumer is actively involved in the design, production, communication and distribution of products and services. With the new technologies that have spurred forms of collective collaboration such as Wikipedia or "open source", consumers have been recognized as creative and productive forces. Fundraising is not new, especially with a social, solidarity or arts-related nature. But with the use of the internet, there are many possibilities that have made crowdfunding more meaningful and effective. Therefore, the spread of internet access and social networking platforms were essential to the increase of the crowdfunding phenomenon where entrepreneurs (the founders) and investors (the funders) can be more effectively combined, especially by the low costs of online research . A. Funders or Investors reward models Crowdfunding innovates by exploring the potential of social networks and other Web 2.0 functionalities to mobilize a large number of users around a community in a small period of time. One of the crucial aspects of crowdfunding is the issue of compensation, recognition or reward from investors, which can range from getting access to the product with low costs and before it is sold, to the simple recognition for having helped  . Based on the various types of returns that are offered to investors, there are different crowdfunding business models : . Donation which is tangentially associated with social causes and nonprofit projects (investors assume the role of philanthropists and expect no return on their investment); . Reward allows investors to receive a non-financial benefit in return for their investment  and to consider these funders as early customers giving them access to products in exchange for a lower price or other benefit ; . Loan allows the investor to provide financing, where the reward is the interest plus repayment of capital, that may take place, at the end of the loan period ; . Equity Crowdfunding provides equity investment by investors in exchange for company stock, profit sharing or voting rights . Consumers are increasingly creative agents who participate in creating value for a company by contrast with mere consumers or users of the value created by companies. On the other hand, companies turn to the crowd mainly for reasons of cost reduction . The third key player in crowdfunding is the platform, which acts as intermediary between entrepreneurs and the funders. While some platforms only provide a physical website, where a crowdfunding campaigns can be exposed, some manage the transaction and others offer also advice and other services helping to a higher level of success . B. Crowdfunding from the founders point of view One of the first definitions of crowdfunding from those who seek investment (the entrepreneurs), emphasize the idea that crowdfunding is a valuable alternative source of funding that is processed through the internet with the objective of obtaining financial resources to support a specific purpose initiative. Thus, building a supportive community is a critical ingredient for crowdfunding to be more profitable than traditional methods . While in traditional financing, the entrepreneurs recognize an opportunity and persuade a limited group of potential investors, in crowdfunding, the public represents a large dimension. Through crowdfunding, a larger and heterogeneous number of people has to be persuaded and recognize the opportunity that the entrepreneur presents . C. Crowdfunding from the funders point of view One of the first definitions of crowdfunding from those that can be viewed as financers was proposed in an article published in the New York Times, announcing a new business model in which consumers has a new role, the role of investors. One could read that "The consumers invest a small amount, sometimes as small as $1 - in a product - often an album of a new musician, clothing, or jewelry from an aspiring designer" . Others emphasize the efforts of individuals or teams (cultural, social, either no-profit and for-profit) to finance their projects, attracting small contributions from a large number of individuals, using the internet, without the financial intermediaries . And, others define crowdfunding as the process that allows the acquisition of a certain monetary amount of various individuals through an online platform . These definitions are not contradictory, as they convey the same concept. D. The management of the network and the campaign Previous research emphasizes two keys success factors for the projects to be financed: the network management and the campaign management . Applying their research to the film producers’ crowdfunding, Hobs et al.  highlighted the importance of the management of both the network and the campaign, rather than merely the size of the network or the quality of the project, which were the findings of previous studies, such as . III. METHODOLOGY Crowdfunding seems to be an easy and less demanding way of funding for new projects, but there are many factors to consider for the campaign to reach its funding goal. It is not irrelevant to note that about 65% of campaigns in Kickstarter and 55% of campaigns in the Portuguese platform PPL do not reach their proposed objectives, although it is still far from the performance of other markets. What did not go well? Was the promotion of the project poorly developed? Target misidentified? Or simply the result of a bad idea? This work intends to find about the do's and don'ts of a campaign in the specific area of arts and sports. A. Hypotheses While the amount of money per supporter is expected to be positively correlated with the success of a certain project, the duration of the project is expected to be negatively related to the success of a campaign . Thus, the first two hypotheses of this research are The search for success of a project justifies the abovementioned hypotheses. B. Modelling crowdfunding The quantitative independent variables include the number of images per project ( ), the number of days that the project is online ( ), the number of views per project ( ), the number of comments per project ( ), the number of funders per project ( ), and the raised amount per funder ( ). The independent variables that are qualitative are treated as dummies. One dummy variable is the level of requested amount 1 and the reference level is < 3000 . Other set of dummies are the bimonthly periods to start a project ( ( )/ ( ) ), where the period January/February is the reference one. Finally, the use of the network Facebook ( ) is another dummy where no is the reference category. The model is specified as follows : = H1: The number of days the project was online (the runtime period) negatively explains its success. p 0 H2: The amount raised per funder explains positively the success of a project. The greater the communication about the project, the better the chances of obtaining the necessary funding . A good communication among supporters through a social network facility and a good explanation about the project, either descriptively or video graphically, is positively related to its success . Thus, being connected to the social network Facebook, the number of images per project as a proxy of good explanation of the project, the number of views, and the number of comments as proxies of good communication of the project help to get the necessary funding. At last, the number of funders per project also explains the success of a project. H3: The Facebook connection increases the success of a project. Considering the importance of having a large network , it is expected that: H9: High levels of requested amount negatively explain the success of a project (the higher the value requested, the lower the success). 6 / ≥3000 4 4/ 4 4/ 4, 12 5, / 5 5 / 5 5 / 8 1, 7 11 2 2 2, . (1) 9 13 6, Because a multiplicative form of the model is used the quantitative variables were ln-transformed by taking the natural logarithm of the original variables. The specification of the theoretical model for the financed projects is = 0 + 12 H8: The bimonthly period to launch a project at the platform influences its success. 3 / 1 is the difference between the raised amount and the where requested amount in euros of the financed project i; and, is the random variable that counts for the influence of omitted variables in the model. H5: The number of views explains positively the success of a project. It is also expected that to launch a project in certain time period can contribute for its success. Thus, / 3 10 3, 4 H7: The number of funders per project explains the success of a project. 3 3 6 H4: The number of images per project explains the success of a project. H6: The number of comments explains the success of a campaign. 1 8 1 1 4/ + + 1, 5, + , ≥3000, 5 9 13 5 + 2 2 / , + 3 3 + 6 6 / , + 7 + 2, 3, + 11 10 + 6, /, / , 4, + + +. (2) IV. DATA ANALYSIS The data were gathered from the PPL Crowdfunding Portugal, the platform with the most notoriety in the national market from August 15, 2011 to March 29, 2016. Some data characteristics can be seen in Table I. The majority of the projects in both areas were financed (57.9% for the area of Music and 81% for the area of Sports); the preferred months to launch a project that was financed in the area of Music are February and December (14 and 18, respectively) and in the area of Sports are July and August (6 on both months); the majority of the financed projects (63.6%) in the area of Music requested an amount smaller than 3000 euros in the area of Music and the opposite is registered in the area of Sports where the dominant financed projects (78.6%) required an amount equal to or larger than 3000 euros. TABLE I. SOME DATA CHARACTERISTICS CONCERNING THE CROWDFUNDING AREA OF MUSIC Financed project? Financed projects:month to start the project? Level of requested amount in euros No Yes Total January February March April May June July August Sepetember October November December < 3000 € ≥ 3000 € Area of Music n % 72 42.1 99 57.9 171 100.0 4 4.0 14 14.1 9 9.1 9 9.1 11 11.1 4 4.0 4 4.0 6 6.1 4 4.0 8 8.1 8 8.1 18 18.2 63 63.6 36 36.4 Area of Sports n % 8 19.0 34 81.0 42 100.0 5 14.7% 1 2.9% 3 8.8% 5 14.7% 4 11.8% 1 2.9% 6 17.6% 6 17.6% 2 5.9% 1 2.9% 0 0.0% 0 0.0% 9 21.4 33 78.6 estimations in both crowdfunding areas once they are significantly correlated with other independent variables (Tables II and III). Therefore, hypotheses six and seven are not going to be studied. A. Validation of the hypothesesof the linear regression model Before estimating both models, the hypotheses of the regression models should be validated: . Normality of the random error - Table IV shows the validation of this hypothesis once the value of the p-value in both models for each area is greater than the level of significance. TABLE IV. Pearson’s coefficientsa ( ) ( ) ( ) ( ) ( ) (n=98) (n=99) (n=47) (n=98) (n=98) 1 98 0.082 98 -0.239** 98 -0.097 46 -0.052 98 0.028 98 0.082 98 1 99 0.081 99 0.168 47 -0.038 99 0.140 99 -0.239** 98 0.081 99 1 99 0.431*** 47 0.542*** 99 -0.218** 99 -0.097 46 0.168 47 0.431*** 47 1 47 0.069 47 -0.055 47 -0.052 98 -0.038 99 0.542*** 99 0.069 47 1 99 -0.410*** 99 0.028 98 0.140 99 -0.218** 99 -0.055 47 -0.410*** 99 1 99 a. **. Correlation is significant at the 0.05 level; ***. Correlation is significant at the 0.01 level. TABLE III. ( ) ( ) ( ) ( ) ( ) p-value 0.200* Statistic 0.965 df 32 p-value 0.373 a. Lilliefors Significance Correction; df stands for degrees of freedom; *.- This is a lower bound of the true significance. . Variance of the random error is constant - it is validated in both areas once Figures 1 and 2 show that the dot circles are randomly distributed around a line parallel to XX’ with a slope equal to zero. Dependent variable: ln Gap ESTIMATED CORRELATIONS BETWEEN THE INDEPENDENT VARIABLES , IN THE AREA OF SPORTS Regression Standardized Predicted Value Figure 1. Homogeneity of the residuals in the Music area. Pearson’s coefficientsa ( ) df 92 (n=98) (n=98) (n=99) (n=47) (n=98) (n=98) 1 34 0.081 34 0.413* 34 0.321 16 0.070 34 0.030 34 0.081 44 1 34 0.006 34 -0.173 16 0.263 34 -0.076 34 0.413 * 34 0.006 34 1 34 0.439 16 0.558** 34 -0.485** 34 0.321 16 -0.173 16 0.439 16 1 16 0.441 16 -0.371 16 0.070 34 0.263 34 0.558** 34 0.441 16 1 34 -0.663** 34 .030 34 0.076 34 -0.485** 34 -0.371 16 -0.663** 34 1 34 Dependent variable: ln Gap Regression Standardized Residual ( ) (n=98) Area of Sports Shapiro-Wilk test Regression Standardized Residual ESTIMATED CORRELATIONS BETWEEN THE INDEPENDENT VARIABLES, IN THE AREA OF MUSIC Area of Music Kolmogorov-Smirnov a test Statistic 0.075 The data were treated in order to avoid influent outliers through the analysis of the residuals. TABLE II. THE NORMALITY HYPOTHESIS OF THE STANDARD RESIDUALS a. **. Correlation is significant at the 0.05 level; ***. Correlation is significant at the 0.01 level. The variables (number of comments per project) and (number of funders per project) are eliminated from the Regression Standardized Predicted Value Figure 2. Homogeneity of the residuals in the Sports area. . Random errors are not correlated – if the value of the Durbin-Watson (DW) statistic is close to two, the rule of thumb tell us that the model does not suffer from the autocorrelation problem, which is the case in both models as Table V shows. Thus, the hypothesis that the error terms are not correlated is not rejected. . Independence of the independent variables – if the values of the coefficient Variance Inflation Factor (VIF) are less than 5 , the model does not suffer from multicollinearity problem , as it is shown in Table V. B. Models’ estimations Table V also allows us to say that the quality of the adjustment in the Music area is lower than it is in the Sports area = 0,141; = 0,330 . Just 14.1% or 33% of the variations on the dependent variable are explained in both areas by the estimated models, after correcting the determination coefficient from the number of independent variables specified in the model. TABLE V. Areas Music Relevant Independent variables Level of requested Amount per funder ≥ 3000 euros ln N Views Facebook ln N Images Sports ln Runtime of the project 2 4 a. ** / / MODELS’ ESTIMATIONS Estimations a Estimated coefficients VIF DW -0.709** (0.016) 1.145 0.141 2.042 0.457 ** (0.032) 1.356* (0.071) 0.722 ** (0.063) 2.063 *** (0.003) −1.162 ** (0.053) −2.082 ** (0.069) 1.074 1.0324 1.074 0.330 2.036 1.041 1.101 - Correlation is significant at the 0.05 level; ***. Correlation is significant at the 0.01 level); b. values between parentheses are the p-values; Dependent variable: ln Gap;. In the Music area, projects with requested amounts ≥ 3000 euros are more prone to failure; that is, for a unit increase in requested amount equal to or greater than 3000 euros, it is expected that the dependent variable decreases in 70%, if the effects of the other independent variables are hold constant. However, holding the effects of other independent variables constant, the number of views per project has a positive impact on the success of the project because; that is, when the number of views per project varies 1%, the estimated effect on the dependent variable increases 45.7%, which reveals an inelastic relationship. In the area of Sports, the variables runtime of the project in terms of days that a project is at the platform, the number of images per project and to be connected by Facebook explain positively the success of a project in this area; by contrast, the periods March/April and September/October to launch a project online explains the failure of a project in this area. Thus, hypotheses five and nine are validated for the area of Music. In the area of Sports, hypothesis one is partially validated as a positive relationship was not expected and the hypotheses three, four and eight are validated (Table VI). TABLE VI. SUMMARY OF THE VALIDATION OF THE HYPOTHESES H1: Number of days the project was online (the runtime period) negatively explains its success. H2: Raised amount per funders explains positively the success of a project. H3: Facebook connection increases the success of a project. H4: Number of images per project explains the success of a project. H5: Number of views explains positively the success of a project. H8: The bimonthly period to launch a project at the platform influences its success. H9: High levels of requested amount per funder lower the success of a project. Area of Music Area of Sports Not validated Partially validated Not validated Not validated Not validated Validated Not validated Validated Validated Not validated Not validated Validated Validated Not validated V. CONCLUSIONS Crowdfunding brought a new dimension to the financing of entrepreneurship, offering those who want to develop a new project, a methodology based on digital tools. In addition to funding, crowdfunding brings other benefits to entrepreneurs, such as testing the product on the market, getting feedback and improving communication. Whether for cultural or other reasons, such as different market needs or different perceptions of crowdfunding, different markets may react differently. The specification model used in this study is credible as it identifies determinant factors to explain success or failures of a project in different crowdfunding areas such as Music and Sports. Therefore, attention should be paid to the number of views left on the project’s page which is associated to the number of potential funders as well as to the number of comments that potential funders can make on the project’s page because they induce positive impacts on the project’s success. Therefore, the entrepreneurs should describe their projects properly, using images, to be seen as credible projects in terms of providing navigation link(s) to be used by the potential funders. Also, an option should be provided to share the web address of the project via e-mail or social network sites such as Twitter, Facebook, or/and LinkedIn in order to extend the contact list of the founder(s). From these channels, the potential number of funders can increase and so can the number of views. But, in order to get the funders’ attention and donation, the requested amount of a project per funder should not be high, knowing that projects with small amounts of requested amounts can deliver success to the projects in the area of Music. While the runtime of a project can be extended in the area of Sports in order to explain success, in the Music area this action is not relevant. In conclusion, there are specificities for different areas and, because of that, each area should be treated separately. However, this study faced some limitations concerning the lack of information about the funders’ profile that could not be done. Also, the presence of significant correlations between independent variables led to the elimination of some potential independent variables (number of comments and number of funders per project) from the model’s estimations. Some theoretical questions are raised and remain to be answered in future studies, such as: • What is the market's adherence to the product concept? • Can a greater number of donors be associated with a greater interest for the project? • Are links and references in the social media a determinant to the crowdfunding campaign success? ACKNOWLEDGMENT The authors acknowledge the support of PPL Platform for the access to data used in this study and the BUSINESS RESEARCH UNIT (BRU-IUL), LISBOA, PORTUGAL. REFERENCES   Lukkarinen, A, Teich, J. E., Wallenius, H. and Wallenius, J, (2016). "Success drivers of online equity crowdfunding campaigns". Decision Support Systems 87, 26–38. Gerber, E. M., Julie S. Hui, and Pei-Yi Kuo (2012). "Crowdfunding: Why people are motivated to post and fund projects on crowdfunding platforms". Northwestern University, Creative Action Lab, Sheridan Drive, Evanston.  Kozinets, R. V., de Valck, K., Wojnicki, A. C. and Wilner, S. J. S. (2010). "Network Narratives: Understanding word-of-mouth marketing in online communities". Journal of Marketing, 74, 71-89.  http://ppl.com.pt/en/infographic  Brown, T., Boon, E., Pitt, L. (2016). Seeking funding in order to sell: Crowdfunding as a marketing tool, Business Horizons.  Mollick, E. (2013). "The dynamics of crowfunding: An exploratory study". Journal of Business Venturing, vol(29), Issue 1, 1-16.  Kotler, Philip. (1986). "The Prosumer Movement. A New Challenge for Marketers". Advances in Consumer Research, 13, 510-513.  Agrawal, A. K., Catalini, C., and Goldfarb, A. (2011). "The geography of crowdfunding", (No. w16820). National Bureau of Economic Research.  Hemer, J. (2011). "A snapshot on Crowdfunding". (No. R2/2011). Working papers firms and region.  Collins , L. and Pierrakis, Y.(2012). "The venTure Crowd: Crowdfunding equiTY inveSTMent inTO buSineSS". Nesta.org.uk, 8-33.  Belleflamme, P., Lambert, T., Schwienbacher, A. (2013). "Crowdfunding Tapping the right crowd". Journal of Business Venturing, 29, 5, 610-611.  Lehner, O. (2013). "Crowdfunding Social Ventures: A Model and Research Agenda". Venture Capital, 15(4), 289-311.  Ordanini, A, Miceli, L., Pizzetti, M., & Parasuraman, A. (2011). "Crowd Funding: transforming customers into investors through innovative service platforms". Journal of Service Management, 22(4): 443-470.  Hobs, J., Grigore, G. and Molesworth, M. (2016) "Success in the management of crowdfunding projects in the creative industries, Internet Research, Vol. 26 No. 1, pp. 146-166.  Evers, M. (2012). Main drivers of crowdfunding success: a conceptual framework and empirical analysis. Masters Thesis in Business Administration. Rotterdam School of Management, Erasmus University. Kutner, M., C. Nachsheim, & J. Neter, (2004). Applied Linear Statistical Models. 4th. McGraw-Hill; Irwin.