1
Crowdfunding: Quality Signals and the way to Success
Bachelor Thesis
Abstract
This thesis analyses whether having a video in the project description, being staffpicked, and frequent updates are positively related to the funding level of Kickstarter crowdfundig projects. For this research a panel dataset is used with data on 173 Kickstarter projects with a duration of 30 days, measured on three different time points every 10 days (collected on 10 May 2015). An OLS regression model and a random effects regression model are used to test the hypotheses. The estimation results show that having a video and posting updates are positively related to the funding level, moreover the results suggest that updates in the first 10 days of the project have the strongest positive association with the funding level, followed by the last 10 days. No significant relation has been found between being staffpicked and the funding level.
Johan Han 10269533
29-06-2015 2014/2015
Supervisor: J. Sol
2 This document is written by Student Johan Han who declares to take full responsibility for the contents of this document. I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
3
Table of contents
Abstract
1
Statement of originality
2
1. Introduction
4
2. Literature review
2.1 Kickstarter
5
2.2 Quality signals
6
2.2.1 Video and updates
8
2.2.2 Staffpicked
9
3. Methodology and data
3.1 Data
10
3.2 Empirical method
14
4. Empirical results
14
5. Discussion
16
6. Conclusion
17
7. References
18
4 1 Introduction
Obtaining funds is one of the most critical factors for prospective entrepreneurs (Gorman & Sahlman, 1989). In the early phases of a company’s life cycle, financing is typically provided by the founder, his friends and family and by business angels. When these funds are
insufficient, the business faces a funding gap (Collings & Pierrakis, 2012). Crowdfunding is seen as a way to reduce this funding gap in the early stages of new ventures. In recent years crowdfunding has emerged as a valuable source of funding for a variety of ventures seeking for external financing. Indeed on April 5, 2012 President Barack Obama signed the Jumpstart our business (JOBS) act into law to legalize equity crowdfunding, which drastically changed the landscape of crowdfunding platforms. The JOBS act made it easier for businesses and start-ups to gather funds by easing the regulations. This new crowdfunding method allows entrepreneurs to raise capital through an open call on the social web (Belleflamme, Lambert, & Schwienbacher, 2014). However partly because of a lack of sufficient value that can be pledged to financial investors and partly because of unsuccessful attempts to convince investors many entrepreneurial ventures remain unfunded (Chen et al., 2009). This raises the question about how to navigate through the concept that is crowdfunding and the underlying dynamics of successful crowdfunding. Although in the last few years there has been an emergence of online crowdfunding activity and practice and policy rapidly advances, this important and growing area of entrepreneurial activity is understudied. Even the basic academic knowledge on the dynamics of successful crowdfunding is still lacking (Mollick, 2014). Therefore I make an attempt to improve the analytical understanding of crowdfunding success, by studying how quality signals influence the funding level for projects on the online crowdfunding platform Kickstarter.
The remainder of this thesis proceeds as follows. First, I will elaborate on what
crowdfunding actually is, followed by a closer look at the crowdfunding platform Kickstarter. Second, I will identify quality signals through the use of existing literature. Third, I will discuss the method and results. And I will end with a brief conclusion and discussion points.
5 2 Literature review
2.1 Kickstarter
Belleflamme, Lambert and Schwienbacher (2010, p. 8) define crowdfunding as follows: “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”. They tend to draw funding from relatively small contributions from a relatively large number of individuals
(Belleflamme et al., 2014), with the goal to initiate a one-time project or as a viable source for seed capital to raise sufficient funds to start their venture (Kuppuswamy & Bayus, 2014; Schwienbacher & Larralde, 2010; Mollick, 2014). Other than these goals, crowdfunding has also been used as a marketing tool to create interest in the project or product in the early stages, receive feedback and improve their competitive advantage (Mollick, 2014; Gerber et al, 2012; Belleflamme et al, 2013). Also, founders use crowdfunding to demonstrate demand for their product, which may lead to positive press attention and potential funding from more traditional sources (Mollick, 2014). Crowdfunding can take the form of equity purchase, loan, donation or pre-ordering of the product/reward-based (Ahlers et al., 2015; Kuppuswamy & Bayus, 2013; Mollick, 2014). Reward-based crowdfunding has the largest number of online platforms and is the fastest growing form of crowdfunding. Therefore it is of academic importance to research this particular part of crowdfunding. Reward-based crowdfunding is based on principles that funders invest in projects, and therefore are expecting a successful outcome. Thus, funders have to evaluate an entrepreneur’s ability to produce and actually deliver the rewards by looking at possible quality signals provided by entrepreneurs (Mollick, 2014). In other forms of crowdfunding, like donation based crowdfunding factors other than potential rewards or monetary returns, like emotions, a political statement, personally supporting project founders, or entertainment, may be important for investors (Ahlers et al., 2015; Kuppuswamy & Bayus, 2013; Mollick, 2014). Ahlers, Cumming, Gunther and Schweizer (2015) presented an initial empirical examination of which start-up signals will induce small investors to commit financial resources in an equity crowdfunding setting. The literature mainly has focused on the signaling of start-ups toward capital investors or business angels. However there is little research on this signaling towards small investors in reward-based crowdfunding. Therefore I will focus on reward-reward-based crowdfunding in my study.
6 amount of funding; the founders then indicate what the money is needed for, and what is offered in exchange for the investment along with an end date for the project. Potential funders can browse the offers through crowdfunding platforms and decide to invest towards the target amount (Ahlers et al., 2015). Crowdfunding generally involves relatively small contributions of many consumer-investors over a fixed period of time (Kuppuswamy &
Bayus, 2013).
Start-ups have been able to raise substantial funds through crowdfunding platforms like Kickstarter, Indiegogo and Gofundme. Kickstarter is one of the largest and oldest
crowdfunding platforms on the internet. Since its launch in April 2009, Kickstarter has made it possible for companies active in categories like art, music, film and video, games, design, technology and more to raise hundreds of millions of dollars by several million community members (Ricker, 2011). According to kickstarter.com over 1,5 billion US dollars has been pledged by more than eight million people, funding over 80,000 creative projects since its start. ‘Backers’ of projects on Kickstarter do not receive financial incentives or equity in the project but instead receive rewards offered by the founders. According to Kickstarter the four most common rewards are: copies of things, creative collaborations of various kinds, creative experiences and creative mementos. Kickstarter operates on an all-or-nothing basis, meaning that no one will be charged for a pledge towards a project unless it reaches its
funding goal. If the goal is met before the end of the funding period, it can continue to receive additional funding until its deadline. If the project is successfully funded, Kickstarter will take 5% of total funds raised; on top there are payment processing fees between 3% and 5%. 2.2 Quality Signals
Researchers have identified several key quality signals that lead to investment in traditional investment settings. According to Michael (1974) these potential signals of quality are particularly important in the selection process. The presumption is that these signals reveal the underlying quality of the projects and therefore are more likely to receive funding (Agrawal et al., 2013). But online crowdfunding platforms are a whole different setting for investors and investors on online platforms are usually not professional investors with the same level of knowledge. They usually do not have the capability to extensively study and asses potential investments. Therefore founders need to find ways to clearly signal their value and quality to small investors on crowdfunding platforms to successfully raise funds (Ahlers et al., 2012). Mollick (2014) shows that quality signals, like having a big social network
7 (facebook friends), adding a video, being featured, and posting timely updates do appear to be linked to success in crowdfunding and that funders act like traditional venture capitalist in some ways by evaluating the quality of projects and the likelihood of success. Additionally Mollick (2014) has found that spelling errors in the project description or having longer project durations reduce the chance of success.
Kuppuswamy and Bayus (2013) state that there exists information asymmetry
between founders and funders. In the extreme case investors may not be able to determine the true quality of ventures which could lead to some potentially high-quality and good
performing ventures to unsuccessfully raise funds, because their quality is not easily recognizable (Ahlers et al., 2015). Hence, Ahlers et al. (2015) stress the importance of information going from the entrepreneur to the crowd. Companies should therefore try to reduce the information asymmetry by providing quality signals.
Bikhchandani, Hirshleifer and Welch (1992) introduce the concept of information cascades in their study as follows: “An informational cascade occurs when it is optimal for an individual, having observed the actions of those ahead of him, to follow the behavior of the preceding individual without regard to his own information”. This can lead to peer-effect that is defined as a social process where group behavior influences individual outcome (Ward, 2010). According to Ward this is due to information overload and low knowledge about goods. This results in difficulties in assessing the quality of projects as it is harder for possible investors to evaluate projects and because of the high search costs. Potential
investors therefore may base their actions on behaviors of others. Herding behavior is another form of peer effects. Burtch (2011) finds in his study about a crowdfunding project in funding and selling T-shirt designs that there indeed exists evidence for herding behavior in
crowdfunding context. This herding phenomenon will increase because of a growing number of observable decision makers within the crowdfunding market that lowers the average level of private knowledge in the market. Not only Burtch (2011) but also Bikhchandani et al. (1992) point out that herding behavior and information cascades can lead to negative consequences, as potential investors disregard their own information that could have led to better decision but instead follow the herd to their disadvantage.
Not all start-ups receive funding, thus we can derive from this that small investors on Kickstarter can interpret some of the information provided by start-ups as signals of quality. Stuart, Hoang, and Hybels (1999) state that because the quality of start-ups often can’t be observed directly potential small investors must appraise the start-up based on observable
8 attributes that are thought to co-vary with its underlying but unknown quality. The perceived value of a reward-based venture is based on whether the investors believe the project creator and are compelled to the proposed project (Kuppuswamy & Bayus, 2013). This corresponds with Connelly et al. (2011), who say that potential investors try to evaluate the unobservable characteristics of start-ups by interpreting the signals sent by entrepreneurs and Mollick (2013), who says that crowdfunding success appears to be linked to project quality, in that projects that signal a higher quality level are more likely to be funded. Thus I can conclude that investors on Kickstarter appraise ventures based on observable attributes e.g. quality signals of the company.
2.2.1 Preparedness (video and updates)
According to Chen et al. (2009) preparedness of the entrepreneur is a quality signal to potential investors that positively impacts decisions to fund ventures. Preparedness can be seen in the quality of business plans, how thoroughly business plans are prepared to ensure that these business pitches conform to the standards of success. This is in accordance with Galbraith et al. (2013), who also confirm that perceived presentation preparedness and attractiveness influence the assessment of the proposal. Kickstarter mentions as a standard of success the importance of having a video in the first page of their Creator Handbook, as this shows the minimum preparedness of entrepreneurs. These entrepreneurs who took time to create and upload a video, probably also put more effort in other parts of their project and may therefore be an indicator of a higher-quality project. Crowdfunding experts also emphasize the importance of having a video in the project description (Mollick, 2014; De Witt, 2012; Steinberg, 2012). Therefore I propose that having a video in the project description is positively related to the funding level of the project.
Hypothesis 1: Including a video in the project description is positively related with the funding level.
Kickstarter also recommends the use of updates in their Creator Handbook as a standard for success. Creators can communicate with their backers by posting public updates that
everyone can see. Many industry experts also stress the need for project creators to develop and execute an effective campaign that communicates with the media, bloggers, and the potential backers (Steinberg, 2012; Dushnitsky & Marom, 2013). This is important as Crowdfunding platforms depend on an online social community (Gerber et al., 2012).
9 Kuppuswamy and Bayus (2013) have found that successful projects have significantly more project updates than unsuccessful projects and add more backers each day. Many experts also highly recommend updates as a means to generate visibility and excitement around the crowdfunding project (Budman, 2012). According to Lambert and Schwienbacher (2010) trust building is essential for successful crowdfunding; by updating frequently with information about the development of the project founders can better build such a trust relationship. Therefore I propose that having frequent updates is positively related to the funding level. I will also look into the effect of the timing of updates on success.
Hypothesis 2: Having (frequent) updates is positively related to the funding level. 2.2.2 Staff picks
The editorial team at Kickstarter Headquarters spends a big part of their time reviewing projects. When somethings sticks out as particularly compelling, like a fun video, creative and well-priced reward, a great story or an exciting idea they make the project a Staff Pick. They do this throughout the day and throughout the life of a project (kickstarter.com). Qin et al. (2013) found that being featured on Kickstarter front page is associated with greatest positive effect in pledges compared to other forms of advertising. Likewise, Ward and Ramachdran (2010) have found that being listed on a popularity list is positively correlated with successful achievement of the funding target. Crowdfunding platforms like Kickstarter may reinforce the herding behavior that is influential in online communities by introducing popularity data, shortlisting of projects and staff picks which results in the narrowing the choices (Frydrich et al., 2014). Being featured by Kickstarter also carries a `seal of approval', and there might be prestige effects that make the project more desirable (Ackerberg, 2001). Being staff picked will probably carry the same ‘seal of approval’ and therefore I propose that being staff picked is positively related to the funding level.
10 3. Methodology
This chapter describes the data and methodology used in this thesis. It will elaborate on the data description of the data collected from Kickstarter projects and the empirical method used. 3.1 Data
Our dataset consist of naturally occurring data scraped from the Kickstarter website through the web scraper import.io. The 201 most recently started projects (not more than 3 days ago) with a duration of 30 days have been selected on 10 May 2015. The decision to only use projects with a duration of 30 days has been made due to time restrictions for this thesis. By selecting the most recent projects of the whole Kickstarter website, I try to achieve a valid and generalizable dataset of the whole site without a specific selection of the different categories; Technology, film, music, publishing, games, food, art, fashion, crafts, theater, design, comics, photography, journalism, and dance. The chosen projects will be scraped every ten days (three time periods) to map the progression. Out of the 201 projects 28 have been omitted because they have been prematurely canceled (25 canceled, 1 suspended), leaving us with 173 projects in our sample.
The following key variables have been gathered through the use of the webscraper. This data is summarized in table 1 and 2.
%funded: The percentage of the project’s funding goal that has been raised. Projects that raise at least 100% are successfully funded, and only then do the founders receive the raised capital, as Kickstarter follows an all or nothing model. However, projects can also receive
more funding than their funding goal.
Category: Kickstarter categorizes projects into 15 different categories. Technology, Film, Music, Publishing, Games, Food, Art, Fashion, Crafts, Theater, Design, Comics,
Photography, Journalism and Dance.
Goal: The amount of funding that founders want to obtain. According to Mollick (2014), higher goals make projects less likely to succeed, due to the possibility that this conveys less confidence in the project by the founder and thus a lower expected quality by funders. Video: This dummy variable measures whether the project contains a video or not, and takes the value 1 for when the project includes a video. Kickstarter and other crowdfunding experts stress the importance of having a video included in the project’s description as a standard of
11 0,0 5,0 10,0 15,0 20,0 25,0 Art Co m ics Cr af ts De sig n Fas hi on Film Food G ame s Jo ur na lis m M us ic Ph ot og rap hy Publ ishi ng Te ch no lo gy The at er Da nce Dataset Kickstarter success.
Staffpicked: This dummy variable measures whether the project has been selected by
Kickstarter as a staff pick, and takes the value 1 for when the project is chosen as a staff pick. Researchers have found that being featured on the Kickstarter homepage, or a popularity list is positively related to reaching the target goal and make the project more desirable (Qin et al,
2013; Ward and Ramachdran, 2010; Ackerberg, 2001).
Number of Backers: the number of backers investing in the projects. Amount pledged: The amount of funding that has been pledged by funders. Updates: The number of updates that has been placed by the founder. Updates are a way for
founders to reach out to their current and potential funders, to inform them about successes, developments of the project, and more. These updates can be placed during and after the deadline of the project. For this study only updates before the deadline have been recorded, because updates after the deadline do not influence the percentage of funding anymore. Kickstarter also mentions updates as a standard for success and Kuppuswamy and Bayus (2013) have found that successful project have significantly more updates than unsuccessful ones. Updates are also important because crowdfunding projects depend on an online social community (Gerber, et al., 2012).
The data represents 173 projects with a combined amount of $564651 in pledges, of which 47 (27.2%) have been successful in reaching their funding goal, raising a total of $401511. The percentage of successful projects is lower than Kickstarter’s overall success rate of 37.49% . Figure 1 shows the distribution of categories among the projects in the dataset compared to the Kickstarter website. Notable are the categories Film, which is underrepresented in the dataset, and the category Technology, which is overrepresented in the dataset. Nonetheless, the sample seems to adequately represent the category distribution of Kickstarter.
12 The crowdfunding projects in our sample tend to succeed by small amounts; almost 45
percent of the successful projects only raise 10 percent or less in surplus of their target goal. Furthermore, projects that fail tend to do so by large margins. The mean amount funded of failed projects is only 7.94 percent, with 75 percent of the projects pledging only up to 10 percent of their target goal. In fact, 64 out of the 126 failed projects (50.8%) managed to crowdsource no funding at all. See fig. 2 for the histograms of the pledge level of successful and failed projects. 116 projects (67.1%) out of the whole sample contained a video, while 42 (89.4%) out of the 47 succeeding projects contained one. Additionally 74 (58.7%) out of the 126 failed projects contained a video. Also, 9 projects (5.2%) had been chosen as a staff pick, while this number was 5 (10.6%) for the succeeding projects.
Fig. 2a Pledge level of successful projects
13
Table 1
Summary Statistics.
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) Variables All Funded Art Comics Crafts Design Fashion Film Food Games Journalism Music Photography Publishing Technology Theater Success 0.272 (0.446) - 0.500 (0.522) 0.500 (0.577) 0.125 (0.354) - 0.455 (0.522) 0.182 (0.395) (0.426) 0.214 0.067 (0.258) 0.500 (0.707) 0.320 (0.476) - 0.389 (0.502) 0.214 (0.418) 0.500 (0.548) %funded 44.006 (79.549) 140.681 (100.021) 82.000 (80.509) 67.500 (75.743) 32.500 (51.542) 10.800 (19.058) 61.182 (72.738) 24.182 (41.254) 25.857 (45.630) 31.333 (80.472) 55.000 (77.782) 42.640 (46.664) - 42.500 (52.312) 60.786 (147.395) 60.833 (62.506) Goal 18803 (36892) 7305 (8782) 1591 (1153) 12392 (12197) 3595 (2078) 45200 (60015) 11442 (10389) 16600 (22561) 28812 (30975) 31549 (75605) 52500 (67175) 12107 (29565) 11750 (12467) 6326 (7183) 34480 (47335) 6615 (2987) Updates 1.584 (3.281) 3.617 (4.803) 0.833 (1.642) 1.250 (1.893) 0.875 (1.126) 2.200 (3.834) 1.818 (3.188) 0.455 (1.336) 1.071 (1.900) 1.200 (1.934) 0.500 (0.707) 1.760 (2.818) - 4.056 (7.511) 1.357 (1.830) 3.667 (3.724) Comments 2.624 (10.037) 6.000 (12.706) 0.250 (0.622) 2.750 (3.202) 1.625 (4.597) 3.600 (4.506) 1.545 (2.841) (10.426) 3.229 1.571 (3.368) 3.267 (7.035) - 0.640 (1.114) - 4.389 (14.097) 7.107 (20.691) 1.000 (1.550) Video 0.671 (0.471) 0.894 (0,312) 0.417 (0.515) 1.00 0.500 (0.535) 0.800 (0.447) 0.727 (0.467) 0.636 (0.492) (0.514) 0.571 0.867 (0.352) 0.500 (0.707) 0.680 (0.476) 0.333 (0.577) 0.611 (0.502) 0.821 (0.390) 0.500 (0.548) Staffpicked 0.052 (0.223) 0.106 (0.312) 0.083 (0.289) - - - 0.091 (0.302) 0.045 (0.213) 0.214 (0.426) 0.067 (0.258) - 0.040 (0.200) - 0.056 (0.236) - - Amount pledged 3264 (8599) 8543 (10407) 1268 (1429) 3384 (3123) 681 (1038) 1918 (3740) 4200 (6754) 1893 (3598) 3428 (8750) 517 (858) 2763 (3871) 3420 (6585) 11 (18) 3553 (8194) 7204 (17033) 3428 (4385) Number of backers 38.11 (92.05) 110.96 (143.09) 21.42 (26.58) 44.75 (47.70) 10.75 (17.65) 16.00 (21.60) 29.36 (37.60) 24.68 (73.15) 40.50 (92.86) 15.20 (27.15) 46.00 (23.00) 37.76 (51.02) 0.67 (1.15) 66.94 (152.28) 67.11 (155.37) 42.33 (49.75) Observations 173 47 12 4 8 5 11 22 14 15 2 25 3 18 28 6 Table 2 Correlations.
Success Funded Goal Updates Comments Video Staffpicked Amount pledged Number of backers Success 1 Funded 0.7654* 1 Goal -0.1310* -0.1555* 1 Updates 0.2193* 0.2462* -0.0199 1 Comments 0.2197* 0.3483* 0.2613* 0.2499* 1 Video 0.1803* 0.2343* 0.0452 0.2067* 0.1441* 1 Staffpicked 0.0414 0.0851 -0.0245 0.0864* 0.0598 0.1544* 1 Amount pledged 0.3294* 0.2788* 0.2791* 0.2672* 0.8184* 0.2155* 0.1419* 1 Number of Backers 0.4549* 0.5314* 0.0614 0.3700* 0.7172* 0.2256* 0.2191* 0.7469* 1 *p<.05
14 3.2 Empirical method
To investigate how the percentage funded of Kickstarter projects is affected by the updates, we estimate a model using OLS regression. The percentage funded and updates variables have been adjusted to respectively change in percentage funded and updates compared to the previous period. The dependent variable is change in percentage funded, and the independent variables are the change in updates, goal, video and staffpicked. All the computed variables have been winsorized at the 1st and 99th percentile to limit extreme values and control for outliers. The regression is performed for the whole dataset, but also for each individual time period to investigate the differences between them.
∆Percentage Fundedit = β 0+ β1 ∆ Updatesit + β2Goalit+ β3Videoit+ β4Staffpickedit+ 𝜖𝜖t
Furthermore we estimate a second model with a panel data regression. Panel data is useful because our data set contains different projects with changing values over time. This method is particularly beneficial because it allows us to include fixed effects and control for
unobserved variables that differ across projects, but remain constant over the measured time. Our panel data deals with two-dimensional observations: the entities are the panel’s variable project’s id, and the time, which contains three different measuring points. To decide whether to use a random or fixed effects model a Hausman test is performed. The panel data is
strongly balanced, because there are no missing observations. 4. Results
The results of the OLS regression for the effect of the change in total amount of updates, goal, video and staffpicked on the change in total percentage funded are presented in table 3. All the coefficients are significant at the 1% level, except for staffpicked, which is insignificant throughout all time periods, and goal and video in time period 1, with a significance level of 5%. First, the coefficient for goal is negatively associated with the change in percentage funded of projects, meaning that setting a higher funding goal has a negative effect on percentage funded. Second, the change in updates has an overall positive effect on the percentage funded; this effect is strongest in period 1, followed by period 3, and weakest in period 2. The results are in support of the hypothesis, because it shows that an increase in updates increases the percentage funded of the project. Third, the coefficient for video is also positively related to change in percentage funded throughout the periods. Lastly, the
15 coefficients for staffpicked show that the positive effect of being staffpicked is greatly larger in the third period than the first two periods, this effect however is not significant. Therefore there is no support for the hypothesis.
Table 3 OLS regression Dependent variable: change in percentage funded
Full sample Time period 1 Time period 2 Time period 3 Regressor Goal -.0001*** (0.0000) [-.1580] -.0002** (0.0000) [-.1763] -.0001*** (0.0000) [-.1533] -0.0001*** (0.0000) [-.1616] Updates 8.0380*** (1.5364) [.3403] 8.8982*** (2.8059) [.3600] 4.9618*** (1.6540) [.2557] 7.9938*** (2.1037) [.3297] Video 9.1421*** (2.0246) [.1624] 10.0947** (4.5011) [.1478] 8.3866*** (2.1145) [.2219] 9.6195*** (3.4472) [.1705] Staffpicked 6.2593 (4.5411) [.0497] 4.9159 (7.2106) [.0322] 4.7044 (5.0436) [.0556] 10.6252 (9.9222) [.0842] constant 5.5093*** (1.3155) 8.7350*** (3.2613) 2.3016*** (0.7345) 5.6672*** (2.0545) Summary statistics observations 519 173 173 173 F 29.93 10.83 11.94 11.33 R2 0.1999 0.2051 0.1736 0.2123
***p<0.01, **p<0.05, *p<0.1. Robust standard errors are given in parentheses. Betas are given in brackets.
Table 4
Panel data regression Dependent variable: change in percentage funded Regressor Goal -.0001*** (0.0000) Updates 9.8587*** (2.4104) Video 9.4301*** (2.5242) Staffpicked 7.0572 (7.0223) Time2*Updates -7.1158*** (1.9714) Time3*Updates -1.7271 (2.3780) constant 5.5762*** (1.5591) Summary statistics observations 519 rho 0.3685 R2 0.2124
16 Table 4 presents the outcomes of the random effects panel data regression. The Hausman test justifies the use of the random effects regression (P>chi2 = 0.6628). The significance levels of the estimates are similar to the OLS regression, as all the variables are significant at the 1% level except staffpicked and the interaction of updates and time period 3, which are not
significant. Furthermore we can see a similar pattern of the percentage funded by change in updates to the OLS regression, namely that the effect of updates on percentage funded is the strongest in period 1, followed by period 3, and weakest in period 2. However, in the panel data regression the effects in period 2 and 3 are relatively weaker compared to period 1 than in the OLS regression. The rho shows us that 36.85% of the variance is due to the differences across panels. The results of the panel data regression support our hypothesis that updates and having a video are positively related to the funding level. The hypothesis for staffpicked remains unsupported.
5. Discussion
Stimulated by the JOBS act, crowdfunding has experienced a rapid growth throughout the years. However, this new phenomenon is largely understudied. While this thesis provides some insights into the dynamics of successful crowdfunding, they still have a number of limitations. First, this thesis focuses only on reward based crowdfunding, rather than the array of other forms like equity crowdfunding or donation based crowdfunding. The crowdfunding phenomenon is eminently larger, and the dynamics of crowdfunding presumably differ between these forms. Second, due to time and tool restrictions the research has been limited to a limited amount of projects, only with a duration of 30 days, and measured over three time periods. Studying more projects with a more diverse variety of durations and having more measurement points will greatly contribute in obtaining deeper insights into the
crowdfunding dynamics. For instance, the effect of being staffpicked, which was insignificant in this thesis, could be further explored. Additionally, there are numerous other quality
signals that could influence crowdfunding success that have been left out of this thesis. To further increase our knowledge on crowdfunding, researchers should continuously study and search for these factors in the future.
Finally, there are some clear implications for entrepreneurs. First, although some researchers argue that visual elements in crowdfunding pitches have become such a standard,
17 that the controlling character to predict success is weakened (Frydrych et al. 2014), this thesis shows that there is still a significant positive effect of including a video in the description on the funding level. Entrepreneurs, who seek crowdfunding, should therefore try to always include a video into their project. Second, in addition to the confirmed positive effect of updates on the funding level, evidence has been found that the timing of these updates is of great importance. Founders of projects not only should focus on making updates, but take into consideration when they make these updates, as we have seen a strongest positive relationship of updates and the funding level in the first, followed by the last 10 days of the project. It is therefore recommended to prepare a video as well as some updates before the launch of the project.
6. Conclusion
In recent years crowdfunding has emerged as a novel method for start-ups to raise sufficient seed capital for their venture. Crowdfunding platforms like Kickstarter successfully facilitate the funding of thousands of crowdfunding projects. Nevertheless, knowledge on the
dynamics of successful crowdfunding is lacking and this field of study is still understudied. This thesis analyses empirically whether updates, having a video, and being staffpicked positively affects the funding level of Kickstarter projects. The dataset contained panel data on 173 projects with a duration of 30 days, collected from the Kickstarter website through the use of a webscraper. Data was collected every 10 days, for a total of three time periods.
It was expected that having more updates would positively affect the funding level. Furthermore including a video in the project description and being staff picked was expected to be positively associated with the funding level. To test whether these assumptions hold, an OLS regression and random effects regression have been performed.
Results suggest that an increase in the updates as well as having a video in the project description was positively related to the funding level. Additionally, the results showed that updates in the first 10 days of the project are strongest related to the increase in funding level, followed by the last 10 days. The results for being staffpicked were not significant and hence the hypothesis was not supported.
Although this study successfully contributes to the academic knowledge on the dynamics of successful crowdfunding there is still much to learn. This study will hopefully
18 inspire others to further research the dynamics of this rising and disruptive crowdfunding phenomenon.
7. References
Ahlers, G. K., Cumming, D., Günther, C., & Schweizer, D. (2015). Signaling in equity crowdfunding. Entrepreneurship Theory and Practice.
Ackerberg, D. A. (2001). Empirically distinguishing informative and prestige effects of advertising. RAND Journal of Economics, 316-333.
Agrawal, A. K., Catalini, C., & Goldfarb, A. (2013). Some simple economics of crowdfunding (No. w19133). National Bureau of Economic Research.
Belleflamme, P., Lambert, T., & Schwienbacher, A. (2014). Crowdfunding: Tapping the right crowd. Journal of Business Venturing, 29(5), 585-609.
Bikhchandani, S., Hirshleifer, D., Welch, I., 1992. A theory of fads, fashion, custom and cultural change as informational cascades. J. Polit. Econ. 100 (5), 992–1026. Budman, Y. (2012), Quora Answer to Question “In a Crowdfunding Campaign, What is the
Donation Pattern”. http://www.quora.com/In-a-crowdfunding-campaign-what-is-the-donation-pattern-Peaks-and-valleys-vs-time-into-the-campaign
Burtch, G. (2011). Herding behavior as a network externality.
http://aisel.aisnet.org/icis2011/proceedings/humanbehavior/28
Chen, X. P., Yao, X., & Kotha, S. (2009). Entrepreneur passion and preparedness in business plan presentations: a persuasion analysis of venture capitalists' funding decisions. Academy of Management Journal, 52(1), 199-214.
Collins, L., & Pierrakis, Y. (2012). The venture crowd: Crowdfunding equity investment into business. Nesta.
Connelly, B.L., Certo, S.T., Ireland, R.D., and Reutzel, C.R. (2011): Signaling Theory: A Review and Assessment, Journal of Management 37, 39-67.
19 Dushnitsky, G., & Marom, D. (2013). Crowd monogamy. Business Strategy Review, 24(4),
24-26.
Frydrych, D., Bock, A. J., Kinder, T., & Koeck, B. (2014). Exploring entrepreneurial legitimacy in reward-based crowdfunding. Venture Capital,16(3), 247-269.
Galbraith, C. S., McKinney, B. C., DeNoble, A. F., & Ehrlich, S. B. (2013). The impact of presentation form, entrepreneurial passion, and perceived preparedness on obtaining grant funding. Journal of Business and Technical Communication,
1050651913513902.
Gerber, E. M., Hui, J. S., & Kuo, P. Y. (2012, February). Crowdfunding: Why people are motivated to post and fund projects on crowdfunding platforms. In Proceedings of the International Workshop on Design, Influence, and Social Technologies: Techniques, Impacts and Ethics.
Gorman, M., & Sahlman, W. A. (1989). What do venture capitalists do?.Journal of business venturing, 4(4), 231-248.
JOBS ACT. (2012). http://www.gpo.gov/fdsys/pkg/BILLS-112hr3606enr/pdf/BILLS-112hr3606enr.pdf
Kuppuswamy, V., & Bayus, B. L. (2014). Crowdfunding creative ideas: The dynamics of project backers in Kickstarter. UNC Kenan-Flagler Research Paper, (2013-15). Michael, S., 1974. Job market signaling. Quarterly Journal of Economics 87, 355. Mollick, E. (2014). The dynamics of crowdfunding: An exploratory study.Journal of
Business Venturing, 29(1), 1-16.
Ricker, T. (2011). Kickstarted: How One Compnay is Revolutionizing Product Development. The Verge, December 20, http://www.theverge.com/2011/12/20/2644358/kickstarter-success-product-development-revolution
Stuart, T. E., Hoang, H., & Hybels, R. C. (1999). Interorganizational endorsements and the performance of entrepreneurial ventures. Administrative science quarterly, 44(2), 315-349.
20 Schwienbacher, A., Larralde, B., 2010. Crowdfunding of small entrepreneurial ventures.
SSRN Electronic Journal.
Steinberg, D. (2012). The kickstarter handbook. Real-life Crowdfunding success stories. Ward, C., & Ramachandran, V. (2010, December). Crowdfunding the next hit: Microfunding
online experience goods. In Workshop on Computational Social Science and the Wisdom of Crowds at NIPS2010.
De Witt, N. (2012). A Kickstarter's Guide to Kickstarter: How to Successfully Fund Your Creative Project. BookBaby.
