Bachelor thesis
Criteria for choosing between the different forms
of crowdfunding
Name : Saskia Lodders
Student number : 10191224
Specialization : Economics and Finance
Field : Finance
Supervisor : Marijn Kool
2 Abstract
Since the financial crisis, banks have sharpened their conditions of giving loans and also funding through equity capital has become more difficult for organizations. Crowdfunding is an alternative to these more traditional lending sources. Crowdfunding can take the form of pre-ordering of the product, equity purchase, donations or a loan. Organizations that choose crowdfunding as a way of finance need to choose the right form and therefore the following research question is examined empirically: what are the criteria for organizations to choose between these different forms of crowdfunding? The success of a company’s crowdfunding campaign is partly determined by the form of crowdfunding chosen. Therefore, in order to find the criteria on which a company should base their decision for a certain form of crowdfunding, only successfully funded companies were looked at in this thesis. Since the dataset was too small according the form donations, research is only done for the other three forms. Research suggests that if a prototype is more complete, the probability of choosing pre-ordering of the product decreases. This conclusion contradicts the literature. Nevertheless, this study proved that if an organization has a product to offer, the probability of choosing the form pre-ordering of the product increases with 45,6%. When the initial capital requirement becomes larger, the probability of choosing the form pre-ordering of the product decreases and it is more likely an organization will choose loan-based crowdfunding or equity-based crowdfunding. If an organization first tried to obtain a loan from a bank, it is statistically significant that the probability of choosing loan-based
crowdfunding increases with 52,8%. 57,1% of the sample indicated that the choice for a particular form of crowdfunding is also determined by the particular form of crowdfunding the platform offers.
3 Table of contents
1. Introduction 4
2. Literature review 6
2.1 Definition of crowdfunding 6
2.2 Different forms of crowdfunding 6
2.3 Choosing between pre-ordering of the product and profit-sharing 7
2.4 Profit versus non-for-profit organizations 8
2.5 Determinants of success and failure on crowdfunding 8
3. Methodology 9 3.1 Data 9 3.2 Method 10 4. Results 12 4.1 Descriptive statistics 12 4.2 Margins 13
4.2.1 Margins of the indicator variables 13
4.2.2 Margins of the continuous variables 15
4.2.2.1 Margins of amount of funds 15
4.2.2.2 Margins of prototype 17
5. Conclusion 19
6. Discussion 20
7. References 21
8. Appendix 22
Section 1: Dutch survey 22
Section 2: English survey 24
Section 3: Multinomial logistic regression 26
Section 4: Overall margins of amount of funds 26
Section 5: Overall margins of prototype 27
4 1. Introduction
Since the financial crisis, it has become more difficult for start-ups to obtain financing through bank loans and equity capital (Belleflame, Lambert, & Schwienbacher, 2013, p. 3). To avoid the problem that firms remain unfunded, firms are forced to find alternative ways of funding.
Crowdfunding is an alternative to the more traditional lending sources and there are four main forms of crowdfunding: pre-ordering of the product, a loan, equity purchase or a donation.
The main goal of crowdfunding is to obtain funds from the crowd. Often, organizations post their business plans at crowdfunding platforms on the Internet. Platforms are the intermediaries between the organizations and the funders. Several platforms are Kickstarter, Symbid, Geldvoorelkaar and Oneplanetcrowd. At these platforms, a large group of investors have the option to invest small amounts of money in different ideas. There is a great diversity of ideas that people post at platforms, from creative ideas like musicians who seek funds to launch a cd to non-profit organizations seeking funds for charity.
However, funding does not have to be the only goal of organizations. Through crowdfunding, organizations can easily see the demand for their product. Also, crowdfunding is used for marketing purposes, creating interest in new projects in the early stages of development. Thus, financing through crowdfunding provides resources that go beyond capital which may be beneficial to organizations (Mollick, 2013, p. 3).
Crowdfunding is getting substantial attention from governments and newspapers. An
important event for equity-based crowdfunding was that Obama signed the JOB act in 2012. Besides donations and reward basis, the JOB act states that platforms in the United States may operate on an equity basis. Before the JOB act was signed, operating on an equity basis was not possible (Mollick, 2013).
However, since crowdfunding is a relative new source of financing, theoretical and empirical research is limited. The resources needed to understand the role of crowdfunding platforms are still limited, as well as the drivers for investors to choose a particular project and the reasons for organizations to choose a particular form of crowdfunding instead of one of the other forms.
Until now, not any empirical research has been done on the criteria for choosing between the different forms of crowdfunding. That is why this thesis examines the following research question empirically: what are the criteria for organizations to choose between these different forms of
crowdfunding? The success of a company’s crowdfunding campaign is partly determined by the form of crowdfunding chosen. Therefore, in order to find the criteria on which a company should base their decision for a certain form of crowdfunding, only successfully funded companies were considered.
To answer this question empirically, I will do a survey among Dutch organizations. The U.S., U.K. and the Netherlands are leading in crowdfunding (Vandendriessche & Walterus, 2013). There are many different platforms in the Netherlands that offer all the different forms of crowdfunding.
5 Naturally, there are differences between platforms in different countries. In the US for example, the biggest platform is Kickstarter, which is reward-based. In the Netherlands, the biggest platform is Geldvoorelkaar, which is based on loans. Also, the amount of funds that is pledged differs. In the US, the biggest amount pledged at Kickstarter is around $10.000.000 for Pebble and in the Netherlands the biggest amount raised is around €560.000 at Oneplanetcrowd for Snappcar. Since in the Netherlands there are sufficient platforms with all the different forms of crowdfunding, it is possible to do the survey among Dutch organizations.
Based on previous theoretical research of Belleflame et al. (2013), it is hypothesized that the research that will be performed provides the following results: Organizations prefer pre-ordering if the initial capital requirement is relatively small and when an organization has a complete prototype. When funding needs are high, it is expected that organizations will rather choose an equity purchase. If organizations first failed to obtain a loan from the bank, they will rather go for a loan through crowdfunding. Non-for-profit organizations will prefer donation-based crowdfunding.
This thesis is structured as follows. The next chapter gives an overview of the existing relevant literature on crowdfunding. The third chapter explains which method was used to obtain the data for answering the research question and which model is used for analyzing the data. Next, chapter four discusses the results of the data analysis. Chapter five concludes and provides suggestions for further research. Finally, this research is discussed in the chapter six.
6 2. Literature review
Crowdfunding is of growing importance. The amount of funds that was collected in the Netherlands in 2013 has more than doubled compared to 2012. The total amount of funds collected was €32.000.000 in 2013 compared to €14.000.000 in 2012 (Douw & Koren, 2014). Massolution (2013) examined the total amount of funds raised at crowdfunding platforms around the world. In 2012 the total amount of money raised through crowdfunding was $2.7 billion compared to $1.5 billion in 2011. Therefore, the growth rate was 81% in 2012 compared to 64% in 2011. Among other things, these facts indicate that crowdfunding has developed into an accepted alternative to public and private finance. In 2013, there was a clear transition from using crowdfunding as an innovative instrument, to using crowdfunding as an accepted form of financing (Douw & Koren, 2014).
This chapter gives an overview of the existing relevant literature on crowdfunding. Since crowdfunding is a relative new source of financing, theoretical and empirical research is limited. Much of the literature is focused on the investors, but not on the organizations that seek funds (Mollick, 2013). First, a formal definition of crowdfunding is given. Secondly, the different forms of crowdfunding are explained. Next, the literature on choosing between the different forms of
crowdfunding is discussed. Finally, the determinants for successful crowdfunding campaigns are explored.
2.1 Definition of crowdfunding
Crowdfunding originated from the broader concept of crowdsourcing. With crowdsourcing, profit oriented firms outsource specific tasks to the general public, generally through an open call via the internet (Kleemann Voß, & Rieder, 2008). Firms use the “crowd’’ to obtain ideas, services, solutions and feedback to develop corporate activities (Belleflame et al., 2013, p. 3).
As mentioned previously, the main goal of crowdfunding is to obtain funds from the “crowd” to obtain funds. Belleflame et al. (2013, p. 4) states the following definition for crowdfunding: “Crowdfunding involves an open call, mostly through the Internet, for the provision of financial resources either in the form of donation or in exchange for the future product or some form of reward to support initiatives for specific purposes.’’ The reward can be monetary (e.g. a loan) or
non-monetary (e.g. equity).
2.2 Different forms of crowdfunding
There are four different main forms of crowdfunding: donations, pre-ordering of the product, equity purchase or a loan. These four main forms are based on what investors receive for their contributions.
7 With donation-based crowdfunding, funders donate because they want to support the project. They do not need anything in return (Ahlers, Cumming, Gunther, & Schweizer, 2012, p. 8).
In contrast, funders who pre-order the product do expect something in return, the product. Pre-ordering of the product is part of reward-based crowdfunding (Mollick, 2013). Reward-based
crowdfunding offers a non-financial return in exchange for the investment. With pre-ordering of the product, the investor is treated as an early consumer. The growth rate and total amount raised for donation-based and reward-based crowdfunding are displayed in table 1.
Crowdfunding through equity is a third form of crowdfunding. With equity crowdfunding, small investors provide start-ups with funding in exchange for shares in the organization. Since Obama signed the JOB act in 2012, previously legalized in other countries, equity crowdfunding is legalized in the U.S. (Mollick, 2013). The growth rate in table 1 of 30% is mostly driven by growth in the number of European platforms (Massolution, 2012). Still, equity crowdfunding is relatively rare worldwide. It makes up less than 5% of all crowdfunding investment (Mollick, 2013). This is partly due to the high levels of regulation using equity crowdfunding.
The last model is the lending model. If founders use the lending model, funds are offered as a loan (Mollick, 2013). Funders will receive a fixed periodic income and expect the principal payment at the end. As displayed in table 1, loan-based crowdfunding reached the highest growth rate for the period 2011-2012 (Massolution, 2013).
Table 1: volumes and growth rates for different forms of crowdfunding
Crowdfunding volume
2012 Growth rate (2011-2012)
Donation-based and reward-based crowdfunding $1.4 billion 85%
Equity-based crowdfunding $116 million 30%
Loan-based crowdfunding $1.2 billion 111%
Source: Massolution, 2012 & 2013
2.3 Choosing between pre-ordering of the product and profit-sharing
When an organization decides to obtain funds from the crowd, they need to choose a particular form of crowdfunding. Until now, no empirical research has been conducted on the criteria for
choosing between the different forms of crowdfunding.
Belleflame et al. (2013) compare two forms of crowdfunding theoretically: pre-ordering of the product or investing a fixed amount of money in exchange for a share of future profits or equity securities. This decision partly depends on the amount of funds that has to be raised. If the initial capital requirement is small, an organization has the possibility to choose the form pre-ordering of the product. However, when the amount of funds that has to be raised is too large, either the distortion in price discrimination between crowdfunders and regular consumers becomes excessive, or the amount
8 that has to be raised requires a lot of investors when an organization does not want the prices for products to get high. At this stage, it becomes more profitable for the organization to choose
crowdfunding based on profit-sharing or equity issuance. Massolution (2012) confirms this and shows in practice that crowdfunding for financial return (equity-based and lending-based crowdfunding) raises the largest sums of money per campaign. Donation-based and reward-based crowdfunding drive less funding volume.
Belleflame et al. (2013) also suggest that in the case of information asymmetry, profit-sharing schemes are more preferred compared to pre-ordering. Information asymmetry is more pronounced when the project is in an early stage of development. Having a complete prototype reduces the asymmetric information between the organization and the crowdfunder, because it is clear for the investor how the product will work and what it will look like. Therefore pre-ordering of the product typically requires the existence of a prototype at time of funding.
2.4 Profit versus non-for-profit organizations
It is also possible to distinguish the different forms of crowdfunding according to passive investments and active investments. Lambert & Schwienbacher (2010) empirically found that companies tend to involve more often in active investments and fewer passive investments. Active investments are investments which enable the funders to provide input or to vote on the project. Non-for-profit organizations tend to involve more in passive investments. Passive investments are
investments with a promise of compensation. The investor however, as opposed to the active investor, has no involvement in the project. Organizations that seek passive investments therefore are only interested in obtaining funds.
2.5 Determinants of success and failure on crowdfunding
In order to choose the best form of crowdfunding for an organization, it is useful to know what determinants are for a successful crowdfunding campaign. Mollick (2013) empirically examines the determinants that predict the success of a crowdfunding campaign. He suggests that, based on data of the platform Kickstarter, success is linked to quality of the product or the service that is offered. Preparation is a first signal of quality. Mollick uses two determinants of preparation: a video and updates of the projects soon after launching them. Using a video the founders can show their
product/service quality more easily. Furthermore, if the founder provides updates quickly, this should indicate a prepared founder. An additional measure for quality is spelling errors. Mollick states that signals of quality predict success. Videos and frequent updates are associated with greater success and spelling errors reduce the chance of success.
9 Lambert & Schwienbacher (2010) find that non-for-profit organizations are significantly more likely to achieve their target level in comparison with other organizational forms. When an investor is unable to evaluate quality of the products, there is more information asymmetry at profit-driven organizations compared to non-for-profit organizations, because investors may think that profit-driven organizations focus too much on profits at the expense of quality and therefore do not know what the product quality will be. There is no such incentive for non-profit organizations.
Another factor that determines success is the network size of the founders (Mollick, 2013). Mollick used the number of Facebook friends of the founders. These findings indicate that if the founders have a large network, it is associated with successful fundraising. A large network refers to the role of family and friends as funders for early stage ventures. In the early stages of fundraising, there will be less asymmetric information about the product or service for family and friends, therefore family and friends are more likely to invest. This funding creates a signal for later funders through accumulated capital (Agrawal, Catalini, & Goldfarb, 2011, pp. 2-5).
3. Methodology
3.1 Data
To gather data for answering the research question, I performed a survey among Dutch organizations. As previously mentioned, the survey was held among successfully funded
organizations. When taking a closer look at factors that determine the particular form of crowdfunding for these organizations, it is clear that these organizations succeeded in obtaining the total amount of funds with their particular form of crowdfunding.
The organizations in my dataset have used different platforms. The following platforms are used: Symbid, Geldvoorelkaar, Oneplanetcrowd, Kickstarter, Kapitaal Op Maat, 1%Club and CrowdAboutNow. Using these different platforms, data is gathered from all different forms of crowdfunding. For example, Symbid is an equity-based platform, Geldvoorelkaar is a lending-based platform and 1%club is categorized as a donation-based platform.
In total, 47 organizations have been contacted and 33 complete surveys have been received. The response rate of this survey is therefore 70%. Section 1 in the appendix contains the survey. Since the survey was held among Dutch organizations, the survey was in Dutch. Section 2 in the appendix contains an English version of the survey.
Based on the survey and the literature review, there are six variables collected which may have influence on the particular chosen form of crowdfunding, where these variables are all properties of the organization. The variables are:
10 1. Profit/Non-Profit
2. Amount of funds gathered through crowdfunding 3. Tried to obtain a loan from the bank first: yes/no 4. Firm offers a product or service
5. In what state was the prototype, indicated in percentage of the finished product, when the crowdfunding campaign started
6. Platform important for choosing a particular form: yes/no
3.2 Method
The multinomial logistic regression model will be used to test how different variables affect the likelihood of observing a particular form of crowdfunding. The dependent variable in this model has a finite number of outcomes: the four different forms of crowdfunding. Since the variables (the independent variables) are properties of the organizations, a multinomial logistic model is required (De Boer, et al., 2004, pp. 463-473).
However, when doing the multinomial logistic regression, it became clear that the dataset was too small regarding to the form donations. In the Netherlands, the form donations is particularly used by foundations or non-profit organizations (Douw&Koren, personal communication, June 4, 2014). It is difficult for organizations that are established as profit organizations to ask for donations if they make a profit and the funder is not rewarded. Douw&Koren (personal communication, June 4, 2014) states that if we compare the Netherlands to the United States based on how people are investing in companies through crowdfunding, the people of the United States are more willing to give donations. Goodwill in the United States is regarded as more common compared to the Netherlands. Since the dataset is too small regarding the form donations and it is relatively obvious that foundations/non-for-profit organizations use donations in the Netherlands, the donation-based form is left out of the model.
In the multinomial logit model, using three dependent variables now, a set of coefficients is estimated: β(pre-ordering), β(equity purchase) and β(loan). In an unidentified multinomial logit model, a set of coefficients is estimated as following:
Pr (y=i) =
This model is unidentified, because it is possible to fill in different betas and obtaining the same probabilities for y=1, y=2 and y=3.
Therefore, the model must be identified by setting one beta equal to zero. It does not matter which form will be the base. The form that will be used in this model is equity purchase. Using equity purchase as the base, the remaining coefficients will measure the change relative to the form equity purchase and the equations of the probabilities become:
11 Pr (y=pre-ordering) = Pr (y=equity purchase) = Pr (y=loan) =
The results of the multinomial logistic model are difficult to interpret. Since the coefficients only compare the probability of a given outcome with the base outcome, the sign of parameter βi cannot
always be interpreted as the sign of the effect of the explanatory variable on the probability to choose that particular alternative (De Boer et al., 2004, pp. 463-473).
To explore the meaning of the regression results, it is useful to analyze the marginal effects after running the regression. The marginal effect measures the effect of changing the values on the probability of observing an outcome (De Boer et al., 2004, pp. 463-473).
In the model, four independent variables are indicator variables. The indicator variables are: profit/non-profit (1), tried to obtain a loan (3), product or service (4) and platform important (6). Amount of funds (2) and prototype in percentages (5) are continuous variables. With continuous variables, there are an infinite number of values at which the margin can be computed. Therefore, when interpreting margins of continuous variables, there should be a value indicated at which it is worth knowing the margin. With continuous variables, the margin or probability is the difference between the coefficients of two values.
In practice, Douw&Koren (personal communication, June 4, 2014), leading in
crowdfunding consultancy, suggests that organizations that wish to obtain an amount greater than €50.000, reward-based crowdfunding may be difficult. Douw&Koren advises a combination of reward-based crowdfunding and another form of crowdfunding above this amount. Based on this information, it is worth knowing the margin for the amount of funds that is valued at €50.000, because we may see that around this amount of funds organizations will rather choose equity-based
crowdfunding or loan-based crowdfunding than pre-ordering of the product. The probabilities will be considered for an extra amount of funds of €10.000 (€60.000), or for an amount of €10.000 less (€40.000). A deviation of €10.000 in amount of funds is chosen, because with the high amounts of funds it is not useful to look at an increase of just €1.000, since these margins will be very small.
The independent variable prototype contains values of 0 to 1. The margins of 0, 0,5, and 1 will be used to come to a conclusion on the influence of a prototype on the form of crowdfunding.
12 4. Results
In this section the results of the data analysis will be discussed. First, descriptive statistics will be discussed. Next, the margins for the independent indicator variables will be considered. Finally, the margins of the continuous independent variables will be discussed.
4.1 Descriptive statistics
First, the dependent variable and the independent indicator variables are considered. Table 2 gives an overview of the descriptive statistics of the dependent variable and the independent indicator variables. The frequency is shown for each indicator variable. Pre-ordering appears most frequently in the dataset. Most organizations in the dataset offer a product and it is clear that most of the
organizations are established as profit organizations. Furthermore, most organizations did not try to obtain a loan first. 20 out of the 35 (57,1%) organizations indicated that the choice for choosing a particular form of crowdfunding also depends on the opportunities particular crowdfunding platforms offered. For example, the company Doodle3D suggested that their choice for a particular form of crowdfunding was also based on the opportunities of the platform. Doodle3D indicated that they chose the platform Kickstarter, because this platform has the biggest audience who are passionate about innovative technologies. However, the company Snappcar indicated that it is logical to choose a platform after the organization knows which form of crowdfunding it will use.
Table 2: descriptive statistics of the dependent variable and the independent indicator variables
Variable N Frequency Percent
Form of crowdfunding 35 pre-ordering 14 40% equity purchase 8 23% loan 13 37% Profit / non-profit 35 non-profit 2 6% profit 33 94%
First a loan: yes/no 35
no 29 83%
yes 6 17%
Product/ service 35
product 22 63%
service 13 37%
Platform important: yes/no 35
no 15 43%
13 For continuous variables, it is useful to look at other statistics. Graph 1 and 2 show the
frequencies of amount of funds and prototype. There is great dispersion in the amount of funds with a minimum of €3.000 and a maximum of €558.625. Most of the organizations have collected an amount of funds between €3.000 en €100.000. Prototype is measured in decimals from 0 to 1 and it can be seen that on average the prototype was about 50% finished before organizations started to collect funds through crowdfunding. Most organizations in the dataset indicated that they had a complete prototype or not a prototype at all when they started to collect funds through crowdfunding.
Graph 1: amount of funds Graph 2: prototype
4.2 Margins
Before computing the margins, the multinomial logistic regression is conducted. However, as already explained in the research method, the coefficients of this regression are difficult to interpret and because of that the margins will be discussed. The results of the regression with equity purchase as the base are displayed in section 3 of the appendix.
After running the regression, margins are computed. The margins of each form will be considered in this part and the different forms of crowdfunding are compared according to the margins.
4.2.1 Margins of the indicator variables
14
Table 3: average marginal effects for the different forms of crowdfunding
Margins
Variable Pre-ordering Equity purchase Loan
Profit (1)/non-profit (0) -0,456 0,182 0,274
(6,254) (199,37) (194,773)
First a loan: yes(1)/no(0) -0,227*** -0,301*** 0,528***
(0,077) (0,072) (0,098)
Product (0)/ service (1) -0,361*** 0,226 0,134
(0,056) (0,280) (0,281)
Platform important: yes(1)/no(0) 0,105 0,051 -0,156
(0,092) (0,132) (0,141)
*significant at the 0.10 level, **significant at the 0.05 level, *** significant at the 0.01 level
The average margins for when an organization first tried to obtain a loan from a bank are all significant at the 0.01 level. If an organization first tried to obtain a loan from a bank , the probability of choosing a loan-based crowdfunding increases with 52,8%. This makes sense, because
organizations that are rejected when trying to obtain a loan from a bank first, will probably try to get a loan through crowdfunding. When an organization is first rejected when trying to obtain a loan from a bank, the average probability of choosing the form pre-ordering of the product or equity purchase decreases. The average probability of choosing pre-ordering of the product decreases with 22,7% compared to a 30,1% decrease in the probability of choosing the form equity purchase.
The margin for having a product or service as an organization is significant at the 0.01 level when choosing the form pre-ordering of the product. If an organization wants to offer a product, the probability of choosing the form pre-ordering of the product increases with 36,1%. It increases, because when an organization has a product to offer, it has the possibility to offer this product to its investors using this form of crowdfunding. This makes it more attractive for people who are interested in the product to make an investment through this form of crowdfunding. The margins for having a product or service are not significant for equity-based crowdfunding and loan-based crowdfunding. Therefore, nothing can be stated about having a service or product when it comes to the probability of choosing equity-based or loan-based crowdfunding.
The margins for importance of the platform in choosing between the different forms of crowdfunding are all insignificant. Since the margins for a particular form of crowdfunding are all insignificant, it is concluded that 57,1% organizations of the sample indicated that their choice for a certain form of crowdfunding is also determined by the form the platform offers. However, there is no possible conclusion on suggesting that this result is significant for a certain form of crowdfunding.
Almost all organizations in the dataset are organizations that are established as a profit organization, because most of the non-for-profit organizations chose donation-based crowdfunding and these organizations are omitted from the dataset. Only two organizations in the dataset are established as a non-for-profit organization and they both chose the form pre-ordering of the product.
15 Unfortunately, the margins for variable profit/non-profit are not significant. Non-for-profit
organizations can not choose the crowdfunding form equity-purchase, because they cannot distribute profits to owners. Therefore, it makes sense that if the organization is established as a non-profit organization the probability of choosing equity-based crowdfunding decreases, as can be seen for the margin of equity-based crowdfunding.
4.2.2 Margins of the continuous variables
4.2.2.1 Margins of amount of funds
When discussing the margins of the continuous variable amount of funds, it is useful to consider the overall margin. The overall margin holds nothing constant and the average value of the predicted probabilities is reported. This margin is not useful when drawing conclusions. However, it is useful to see if the predictions for this variable are significant. Section 4 in the appendix contains the results of the overall margin of the three different forms of crowdfunding. The variable amount of funds is significant at a level of 0.05 for pre-ordering the product and loan-based crowdfunding. The overall margin of the variable amount of funds of equity-based crowdfunding is significant at a 0.10 level. The p-values of the overall margins are used to indicate the significance of the margins, because the p-values at particular values of the variable only indicate whether the value is significant if the variable changes with 1 unit.
16
Graph 5: loan
Graph 3, 4 and 5 show the margins of variable amount of funds in a graph. When the initial capital requirement of organizations becomes larger, the probability of choosing the form pre-ordering of the product decreases. This result matches the literature, because when funding needs are high, it is more difficult to obtain funds through pre-ordering of the product.
First, when the initial capital requirement becomes larger, the graphs show that it is more likely that organizations will choose loan-based crowdfunding. However, when looking at the amount of funds around €100.000, the probability of choosing loan-based crowdfunding goes from increase to decrease, where the probability of choosing equity-based crowdfunding is still increasing.
As noted before, Douw&Koren (personal communication, June 4, 2014) suggests that if the initial capital requirement is higher than €50.000, there is a higher probability of choosing equity-based crowdfunding or loan-equity-based crowdfunding compared to pre-ordering of the product. Table 4 contains the margins of each form of crowdfunding at different amount of funds. These margins give the amount of change in probability with a change of one unit in amount of funds.
Table 4: margins of variable amount of funds
Margins of variable amount of funds Amount of funds Pre-ordering Equity purchase Loan
€ 40.000 0,247*** 0,273*** 0,479*** (0,056) (0,079) (0,085) € 50.000 0,188*** 0,298*** 0,515*** (0,042) (0,084) (0,086) € 60.000 0,151*** 0,315*** 0,534*** (0,027) (0,087) (0,087)
*significant at the 0.10 level, **significant at the 0.05 level, *** significant at the 0.01 level
When it comes to continuous variables, the marginal effect is the difference between two margins. It is shown that if the amount of funds increases with €10.000 from €40.000 to €50.000, the probability of choosing pre-ordering of the product already decreases. If the amount of funds increases
17 with €10.000 from €50.000 to €60.000, the probability of choosing pre-ordering of the product
decreases with 3,7%. However, the probability of choosing equity-based crowdfunding increases with 1,7%, where the probability of choosing equity-based crowdfunding increases with 1,9% when the amount of funds increases with €10.000 from €50.000 to €60.000. These increases and decrease in probabilities are very small. Nevertheless, these increases indicate an increase in probability of choosing the form equity purchase or a loan and a decrease in probability of choosing the crowdfunding form pre-ordering of the product.
4.2.2.2 Margins of prototype
The last independent variable that will be considered is prototype. First, the overall margins are displayed in section 5 of the appendix. It is shown that only the overall margin for pre-ordering of the product is significant.
Graph 6, 7 and 8 show the margins of the independent variable prototype in a graph. In section 6 of the appendix, details of graph 8 are shown enlarged.
Graph 6 : pre-ordering Graph 7: equity purchase
18 The first result is remarkable, because theory suggests that if an organization has a complete prototype, there is a high probability the organization will choose for the crowdfunding form: pre-ordering of the product. However, graph 6 indicates that if a prototype is more complete, the probability of choosing pre-ordering of the product decreases. This result is also shown in table 5. Again, the marginal effect for a particular increase or decrease (e.i. 0,5) is the difference between two margins. If there is a complete prototype the probability of choosing pre-ordering of the product decreases with 44,4% compared to not having a prototype (prototype=0) and as already mentioned, the overall margin of pre-ordering of the product is significant at the 0.10 level.
Concerning loan-based crowdfunding and equity-based crowdfunding, the average margins are not significant. Therefore, there is no conclusion possible regarding the crowdfunding forms equity purchase or a loan.
Table 5: margins of variable prototype
Margins of variable prototype Prototype Pre-ordering Equity purchase Loan
0 0,622*** 0,068 0,310*** (0,046) (0,046) (0,063) 0,5 0,392*** 0,224*** 0,385*** (0,037) (0,074) (0,080) 1 0,178 0,47 0,352 (3,958) (3,383) (0,412)
19 5. Conclusion
Crowdfunding can take the form of pre-ordering of the product, equity purchase, donations or a loan. Organizations that choose crowdfunding as a way of finance need to choose the right form and therefore this thesis examined the following research question empirically: what are the criteria for organizations to choose between these different forms of crowdfunding? Only successfully funded Dutch organizations have been used in this research and there are several variables collected that may have had influence on choosing between the different forms of crowdfunding. Unfortunately, due to a small dataset, this research was conducted for all forms except the form donations. For donation-based crowdfunding, it can be stated that in the Netherlands this form is particularly used by non-for-profit organizations and foundations (Douw&Koren, personal communication, June 4, 2014).
Firstly, if an organization has a product to offer, it is statistically significant that the average probability of choosing the form pre-ordering of the product increases with 36,1%. This makes sense, because an organization can show the investors exactly what the product will look like, which will make people that are interested in the product more willing to invest.
However, if the capital requirement increases, the probability of choosing the form pre-ordering of the product decreases and the probabilities of choosing loan-based crowdfunding or equity-based crowdfunding increases. This decrease in probability of the form pre-ordering of the product can be explained using the literature, because when the capital requirement increases, an organization either needs more investors or it needs to ask a higher price per investor. Around the amount of funds of €100.000, the probability of choosing loan-based crowdfunding decreases, where the probability of choosing equity-based crowdfunding is still increasing.
Thirdly, if an organization tried to obtain a loan from a bank, it is statistically significant that if the requested loan was denied and an organization seeks finance through crowdfunding afterwards, the probability of choosing the form loan-based crowdfunding increases with 52,8%.
Results regarding the determinant prototype contradict the theory. This research suggests that if a prototype is more complete, the probability of choosing pre-ordering of the product decreases. However, theory suggests that pre-ordering of the product requires the existence of a prototype at time of funding (Belleflame et al., 2013). This contradiction could be caused by a small dataset.
Finally, 57,1% of the organizations in the sample indicated that their choice for a particular form of crowdfunding is also determined by the particular form a platform offers. Nevertheless, there is no conclusion possible on suggesting if this result is significant for a particular form of
crowdfunding.
To summarize, there are several variables that influence the decision of organizations to choose between the different forms of crowdfunding. However, to improve this research, future research could be done using a larger dataset and it could incorporate more variables that influence the organizations decision between the different forms of crowdfunding.
20 6. Discussion
There are some limitations to this research. Firstly, when trying to examine this question empirically, it became clear that when looking at the form donations, the dataset was too small. In the Netherlands, organizations who seek funds through donation-based crowdfunding are particularly foundations or non-profit organizations (Douw&Koren, personal communication, June 4, 2014). However, goodwill is, in some other countries like the United States, much higher when comparing to the Netherlands. In these countries it is more common for organizations that are established as profit organizations to seek funds through donation-based crowdfunding. When using a sample including other countries, results might be more relevant according to the form donations.
As noted before, having a complete prototype reduces the probability of choosing the crowdfunding form pre-ordering of the product is a remarkable result. This result contradicts with theory, that suggests that if an organization has a complete prototype, it is more likely the organization will choose pre-ordering of the product. Adding more organizations that choose the form pre-ordering of the product to the dataset would increase the accuracy of these findings and the conclusion
regarding the variable prototype might be different.
Some of the results according to the independent variables are not significant and to make this research more reliable, sample size should be larger. More organizations that are successfully funded could be contacted to get more respondents.
Finally, it can be argued that there are more variables that might have influence on choosing between the different forms of crowdfunding. The variables that are used in this research are derived from the questionnaires that are held among the organizations. However, there may be other variables that influence the decision of choosing between the different forms of crowdfunding. For example, the sector in which the organization is active may have influence on which form of crowdfunding is chosen.
21 7. References
Agrawal, A., Catalini, C., & Goldfarb, A. (2011). The geography of crowdfunding (NBER working paper No. 16820).
Ahlers, G.K.C., Cumming, D., Gunther, C., & Schweizer, D. (2012). Signaling in equity crowdfunding (SSRN working paper).
Belleflame, P., Lambert, T., & Schwienbacher, A. (2013). Crowdfunding: Tapping the right crowd.
Journal of business venturing.
De Boer, P., Franses, P.H., Heij, C., Kloek, T., & Van Dijk, H.K. (2004). Econometric Methods with
Applications in Business and Economics. New York, United States: Oxford University Press.
Douw, S., & Koren, G. (2014). Crowdfunding in Nederland 2013. Douw&Koren.
Fiddelaar, S., Geluk, G., & Zuijderhoudt, B. (2014). Crowdfunding in Gelderland. Direction Subsidies & Groeifinanciering.
Kleemann, F., Voß, G.G., & Rieder, K. (2008). Un(der)paid innovators: the commercial utilization of consumer work through crowdsourcing. Science, Technology & Innovation studies, Vol. 4, No. 1, 5–26.
Lambert, T., & Schwienbacher, A. (2010). An empirical analysis of crowdfunding. Social Science
Research Network (SSRN).
Massolution. (2012). The crowdfunding industry rapport. Massolution. Massolution. (2013). The crowdfunding industry rapport. Massolution.
Mollick, E. (2013). The dynamics of crowdfunding: An exploratory study. Journal of business
venturing.
Vandendriessche, F., & Walterus, B. (2013). Vlaamse Overheid – Agentschap Ondernemen. KPMG & Stibbe.
22 8. Appendix
Section 1: Dutch survey
Ik schrijf mijn bachelorscriptie over crowdfunding. De voornaamste vormen van crowdfunding zijn: vooraf bestellen van het product, doneren, aandelen of een lening. Voor mijn onderzoek heb ik de volgende onderzoeksvraag opgesteld:
What are the criteria for organizations to choose between these different forms of crowdfunding?
Voor deze onderzoeksvraag heb ik enkele vragen voor u opgesteld. U kunt bij meerkeuze vragen het antwoord dat voor uw organisatie van toepassing is dikgedrukt maken.
1. Wat is de naam en website van uw organisatie?
2. In welk jaar heeft uw organisatie geld verkregen via crowdfunding?
3. Hoeveel mensen heeft uw organisatie op dit moment in dienst?
4. Is uw organisatie voor profit of non-profit doeleinden opgericht?
5. Indien uw bedrijf voor profit doeleinden is opgericht, wat was de omzet en de winst in 2013 (indien u deze gegevens aan mij wil/kan verstrekken)?
Omzet: Winst:
6. Via welke vorm van crowdfunding heeft uw organisatie geld verkregen? A) de investeerders konden het product vooraf bestellen
B) aandelen uitgeven C) via leningen D) donaties
23 7. Hoeveel geld heeft uw organisatie verkregen via crowdfunding?
A)€0 - €15.000 B) €15.000-€30.000 C) €30.000-€50.000 D) €50.000-€100.000 E) €100.000-€200.000 F) >€200.000
8. Heeft uw organisatie eerst geprobeerd geld te lenen via de bank en is dit vervolgens afgewezen? A)Ja
B)Nee
9. Levert u een product, dienst of beide?
A) Product, namelijk B)Dienst, namelijk
10. Indien het een product is, hoever was u al met uw prototype in procenten toen u fondsen ging werven via crowdfunding?
A) 10% F)60%
B)20% G)70%
C)30% H)80%
D)40% I)90%
E)50% J) 100%
11. Welk platform heeft u gebruikt?
12. Welke andere platformen heeft u nog bekeken?
13. Is de keus voor een bepaalde vorm van financieren ook gebaseerd op de mogelijkheden die het platform aanbood?
A) Ja, B)Nee,
24 14. Wat waren voor uw organisatie de redenen om voor deze specifieke vorm van crowdfunding (aandelen/ een lening/ vooraf bestellen van het product/donatie) te kiezen?
Section 2: English survey
I am writing my thesis about crowdfunding. Crowdfunding can take the form of pre-ordering of the product, equity purchase, donations or a loan. I have proposed the following research question:
What are the criteria for organizations to choose between these different forms of crowdfunding?
I’ve prepared fourteen questions to be able to come up with an answer for this research question. When you arrive at a multiple choice question, you can make the answer that applies to your company bold.
1. What is the name of your organization and the website of the organization?
2. In what year did your organization gain money through crowdfunding?
3. How many employees does your organization have at the moment?
4. Has your organization been established for a profit or a nonprofit purpose?
5. In case your organization has been established for profit reasons, what was the turnover and the profit in 2013 (if you would like to like to share this data with me)?
Turnover: Profit:
25 6.Through what form of crowdfunding has your organization gained its money?
A) pre-ordering of the product B) equity purchase
C) a loan D) donations
7. How much money has your organization gained through crowdfunding? A)€0 - €15.000 B) €15.000-€30.000 C) €30.000-€50.000 D) €50.000-€100.000 E) €100.000-€200.000 F) >€200.000
8. Has your organization first tried to lent money from a bank and has this proposal then been rejected?
A) Yes B) No
9) Does your company deliver a product, a service or both?
A) Product, …. B) Service, ….
10. In case yours is a product, how far were you producing your prototype when you started collecting funds through crowdfunding?
A) 10% F)60%
B)20% G)70%
C)30% H)80%
D)40% I)90%
E)50% J) 100%
11. What platform did you use?
26 13. Has your choice for this particular form of financing also been based on the opportunities that the platform offered?
A) Yes, B)No,
14. What were the reasons to choose this particular form of crowdfunding (equity-purchase/loan/pre-ordering of the product/donations)?
Section 3: Multinomial logistic regression
base: equity purchase
Coefficients
Variable Pre-ordering Loan
Profit (1)/non-profit (0) -20,724 0,775
(7772,127) (9691,754)
Amount of funds -0,0001* -0,0000024
(0.000) (2,27*10^-6)
First a loan: yes(1)/no(0) 13,921 18,728
(3836,026) (3836,025)
Product (0)/ service (1) -10,984 -1,017
(6,775) (3,361)
Prototype (decimals) -13,764 -2,511
(8,480) (3,815)
Platform important: yes(1)/no(0) 0,995 -0,808
(1,854) (1,120)
Constant 35,916 1,719
(7772,133) (9691,755)
*significant at the 0.10 level, **significant at the 0.05 level, *** significant at the 0.01 level
Section 4: Overall margins of amount of funds
Amount of funds
Form of crowdfunding Margin Standard error P-value
Pre-ordering -0,0000115** 4,63*10^-6 0,013
Equity purchase 4,16*10^-6* 2,24*10^-6 0,063
Loan 7,31*10^-6** 3,36*10^-6 0,029
27
Section 5: Overall margins of prototype
Prototype
Form of crowdfunding Margin Standard error P-value
Pre-ordering -0,944* 0,492 0,055
Equity purchase 0,576 0,391 0,141
Loan 0,368 0,436 0,399
*significant at the 0.10 level, **significant at the 0.05 level, *** significant at the 0.01 level
Section 6: Margins of prototype for loan-based crowdfunding
