The 2015 Crowdfunding Industry Report is live! Use discount code FIRST50 for 50% off.
I’m a huge proponent of research on the crowdfunding industry. There have been a lot of surveys conducted on KickstarterForum and CrowdfundingForum regarding backer psychology and why people support crowdfunding campaigns (example). I’ve also been approached by a handful of researches to help them collect data.
Whether it’s crowdsourcing.org’s industry research, reports put out by the crowdfunding centre (inaugural report), or these 10 crowdfunding research papers, data collection and analysis is always a great way to bring a spotlight to the industry and uncover the best practices.
If you’re conducting a research study on crowdfunding, I invite you to leave a comment below letting us know about it!
Since launching this blog, a lot of people have asked me how I actually got into crowdfunding and why I’m so passionate about Kickstarter. Aside from my own belief that crowdfunding will one day take its place besides venture capital and angel investment as a viable financial instrument, I was initially drawn to the industry when doing a research paper on Kickstarter.
While I don’t think my findings or research was significant or monumental (which is why I didn’t make it into a full paper), several readers have been asking to see it. I think it would be fun to put it up online, so I’ve included the PDF and my conclusions below.
Why I Wrote This Report (copied from report)
Professor Mollick, of the Wharton School of the University of Pennsylvania conducted a scholarly analysis of Kickstarter in his study, “The Dynamics of Crowdfunding: Determinants of Success and Failure.” In this report, Mollick examined data from Kickstarter to better determine what makes a crowdfunded project successful. He took into account the following factors for Kickstarter projects: geography, number of facebook friends of founders, number of tiers of rewards advertised, number of comments, number of updates given by founders, duration of project fundraising campaign, fundraising goal, percentage of goal funded, the number of backers, percentage of first time backers, and finally pledge per backer.
Mollick’s study was focused on the variables that affect the success of Kickstarter projects overall. He found that 47.90% of proposed projects were successful and that the average project goal was $9,755. In addition, he found that the average pledge per backer was $64.04 and the average number of backers was 57.90. Kickstarter.com itself reports the number of successful projects at 46%, the average goal as $5,000, the average pledge as $71 and the average number of backers as 85.
Mollick also determined that the “mean amount funded of failed projects is 10.3% of the goal. Only ten percent of projects that fail raise 30% of their goal, and only three percent raise 50% of their goal.” This data shows that projects generally will fail to reach their funding goal by a large margin. In addition, he found that “Twenty five percent of projects that are funded are 3% or less over their goal, and only fifty percent are about 10% over their goal. Only about 1 project in 9 receives 200% of its goal.” This shows that projects generally will succeed by a small margin.
Finally, overall, he determined that increasing fundraising goal level is negatively associated with success. Being featured on the front page of the website is strongly associated with success. “An unfeatured project has a 30% chance of success, while a featured project has an 89% chance.” Longer durations decrease the chances of success. “A 30 day duration project has a 35% chance of success, while a 60 day project has a 29% chance.”
It has been widely reported that gaming, music and film are the more successful categories on Kickstarter in terms of dollars raised, where as categories like photography and comics are less successful in aggregate funding. However, categories that have low amounts of total pledged dollars like dance can have high success rates (70 %). The weight given to the variables affecting success could vary from category to category.
Hypothesis and Test (copied from report)
Mollick’s analysis is centered on the variables that affect success for an average Kickstarter project. I will further his study by focusing more on how the weight given to the variables that affect project success vary from category to category. This could yield important information for Kickstarter project owners when both deciding which category to place their project and on which elements to focus. Once I establish the regression I will use the odds estimate ratios to test my predictions.
Statistical Hypothesis: Null = ? = 0, Alt = ? =? 0
y= ? + ?1 Goal + ?2 Pledged + ?3 Fper + ?4 Backers + ?5 Levels + ?6 Updates + ?7 Comments + ?8 Duration
Data (copied from report)
I will be using a data set comprised of 45,815 projects from 5/3/2009 until 8/1/2012 that have been gathered using a PHP scraper script (public data set). These projects have the following attributes: category, status (live/not live), outcome (success/failure meeting goal), fundraising goal, dollars pledged, funded percentage, number of backers, number of reward levels, number of comments, number of updates, duration of project fundraising campaign, and date funded. I scrubbed this data set by separating projects by category, removing outliers, and deleting projects that are “live” and therefore are still gathering pledges.
I will conduct a multiple logistic regression for each Kickstarter category to determine the effects of the variables gathered on the dependent variable of whether or not the project was successful, meaning that the fundraising goal was met. I will then compare these categories to determine whether certain independent variables have more effect on the dependent variable in one category than in others. The regression failure will categorically be represented by 0 and the regression success will be categorically represented by 1. The results with a lower AIC and SC show a better fit to the model. This would lead to less error statistically in those categories. See appendix for logistical regression results.
Download PDF (briggman_paper)
Download Data Set (click here – large file)
Download Logistic Comparison (click here).
Conclusion (copied from report)
From the results, it can be concluded that the weight given to the variables impacting fundraising success vary from category to category. However, not all variables had the same level of variation. Number of backers, number of levels, number of updates, number of comments, and project duration had the most variation.
Overall, for a one-unit increase in the number of backers in the categories of art and dance, there are stronger relative positive effects on the probability of fundraising success than for the rest of the categories and for a one-unit increase in the category of photography, there is a stronger relative negative effect.
For a one-unit increase in the number of reward levels in the categories of fashion, photography, and theater, there are stronger relative positive effects on the probability of success than the rest of categories and for a one-unit increase in the category of dance, there is a stronger relative negative effect.
For a one-unit increase in the number of updates by the project creator in the categories of dance and design, there are stronger relative positive effects on the probability of success than the rest of categories.
For a one-unit increase in the number of comments on the project in the categories of music and photography, there are stronger relative positive effects on the probability of success than the rest of categories and for a one-unit increase in the category of comics, there is a stronger relative negative effect.
For a one-unit increase in the duration of the project in the category of food, there is a stronger relative positive effect on the probability of success than the rest of categories.
Strangely, the category of dance had results that seem to vary from the norm for almost every variable except number of comments and the duration of the project. Even for those two variables, dance displayed moderate deviation from the norm. Should Kickstarter project owners decide to start a dance fundraising campaign, they should take these results into consideration, as the general guidelines established by Professor Mollick will not apply for this category. Otherwise, project owners will be less likely to conduct a successful fundraising campaign. Project creators should also take this study’s results into account when they create a campaign in any of the above listed categories, as the weight given to variables that affect success will deviate from Mollick’s findings.
Lastly, although several of these independent variables such as the number of backers and comments are out of the hands of project creators, many, like the duration of the project and number of updates or reward levels can be altered in order to improve the odds of fundraising success.
Since the variables in this study have different impacts depending on the project category, this information can help project owners when they are deciding which category to list their project on Kickstarter. This information holds enormous financial value. A great example is the famous Pebble: E-Paper Watch, which raised $10 M on Kickstarter. The owners decided to place this project in the design category, instead of the technology category, due to anticipation of better reception. Now, by utilizing the results of this study, project owners can help improve the odds of fundraising success and get a better idea of which project elements to focus on and to what degree, which varies from category to category.
Further study is required regarding the emerging crowdfunding industry. Recently, niche platforms have sprung into existence covering music, publishing, and many of the categories that exist on Kickstarter. Is it more beneficial for project owners to post their proposed project on Kickstarter or one of these niche sites? How does the weight given to variables affecting fundraising success change from platform to platform?
In addition, when equity crowdfunding is legalized towards the end of 2013 (good joke) after the SEC has finished crafting regulations, studies will need to be conducted to investigate these equity crowdfunding portals and how their activity differs from project crowdfunding platforms. The weight given to the variables affecting success may be completely different. Clearly, there is ample opportunity for further scholarly study of this new financial instrument.
Other Research Papers
If you took the time to read or glance through the study, I hope it was helpful in some way. Again, my motivation in sharing it is purely for nostalgia and I think it would be interesting. Check out these other research papers:
The Dynamics of Crowdfunding: An Exploratory Study
By Ethan R. Mollick. University of Pennsylvania – Wharton School.
June 26, 2013
“Crowdfunding allows founders of for-profit, artistic, and cultural ventures to fund their efforts by drawing on relatively small contributions from a relatively large number of individuals using the internet, without standard financial intermediaries.
Drawing on a dataset of over 48,500 projects with combined funding over $237M, this paper offers a description of the underlying dynamics of success and failure among crowdfunded ventures. It suggests that personal networks and underlying project quality are associated with the success of crowdfunding efforts, and that geography is related to both the type of projects proposed and successful fundraising.
Finally, I find that the vast majority of founders seem to fulfill their obligations to funders, but that over 75% deliver products later than expected, with the degree of delay predicted by the level and amount of funding a project receives. These results offer insight into the emerging phenomenon of crowdfunding, and also shed light more generally on the ways that the actions of founders may affect their ability to receive entrepreneurial financing.”
Leaning In or Leaning On? Gender, Homophily, and Activism in Crowdfunding
By Jason Greenberg – NYU. Ethan R. Mollick – University of Pennsylvania – Wharton School.
July 3, 2014
“Female founders seek and receive less startup capital than male entrepreneurs. One reason for this disparity is a lack of female representation among funders of startups, and a potential solution is to increase the proportion of women in decision-making roles. Both the problem and the solution implicitly rely on homophily – that women will support other women given a chance.
However, a lack of clarity over when and how homophily influences individual choices makes it uncertain when better representation is actually advantageous. Using data from crowdfunding, we empirically examine whether higher proportions of female funders lead to higher success rates in capital-raising for women. We find that women outperform men, and are more likely to succeed at a crowdfunding campaign, all other things being equal.
Surprisingly, this effect primarily holds for female founders proposing technological projects, a category that is largely dominated by male founders and funders. This finding stands in stark contrast to expectations concerning homophily. A laboratory experiment helps explain how this pattern might emerge and allows us to theorize about the types of choice homophily driving results.
We find that a small proportion of female backers disproportionately support women-led projects in areas where women are historically underrepresented. This suggests an activist variant of choice homophily, and implies that mere representation of female funders without activism may not always be enough to overcome the barriers faced by female founders.”
Crowdfunding: Tapping the Right Crowd
By Paul Belleflamme CORE and Louvain School of Management, UCL (Université Catholique de Louvain) ; CESifo (Center for Economic Studies and Ifo Institute) Thomas Lambert UCLouvain ; Université Lille Nord de France – SKEMA Armin Schwienbacher Univ. Lille Nord de France – SKEMA Business School
July 9, 2013
“With crowdfunding, an entrepreneur raises external financing from a large audience (the “crowd”), in which each individual provides a very small amount, instead of soliciting a small group of sophisticated investors. This article compares two forms of crowdfunding: entrepreneurs solicit individuals either to pre-order the product or to advance a fixed amount of money in exchange for a share of future profits (or equity).
In either case, we assume that “crowdfunders” enjoy “community benefits” that increase their utility. Using a unified model, we show that the entrepreneur prefers pre-ordering if the initial capital requirement is relatively small compared with market size and prefers profit sharing otherwise.
Our conclusions have implications for managerial decisions in the early development stage of firms, when the entrepreneur needs to build a community of individuals with whom he or she must interact. We also offer extensions on the impact of quality uncertainty and information asymmetry.”
Crowdfunding: Disintermediated Investment Banking
By Brian J. Rubinton McGill University
April 11, 2011
“This paper introduces crowdfunding as a concept and model for the evolution of investment banking. Crowdfunding, an application of crowdsourcing, is defined as one party’s attempt to finance a project by offering three types of investment opportunities to potential investors. The investment opportunities are donations, passive investments, and active investments.
From this foundation I develop a model in which interdependent agents operate in a dynamic, discrete setting. Potential investors decide whether or not to invest in one of three opportunities each period while the entrepreneur sets the parameters of the game to maximize the probability of successful financing. I then simulate the model to analyze the effects changes in key parameters have on the results of the game.”
Swept Away by the Crowd? Crowdfunding, Venture Capital, and the Selection of Entrepreneurs
By Ethan R. Mollick University of Pennsylvania – Wharton School
March 25, 2013
“Venture Capitalists (VCs) are experts in assessing the quality of entrepreneurial ventures. A long tradition of research has examined the signals of quality that VCs look for in new ventures, and the biases that result from the VC selection process.
Recently, an alternative form of new venture funding has arisen in the form of crowdfunding, which relies on the judgement of millions of amateurs about which entrepreneurial projects are worth funding. Little is known about the degree to which amateurs respond to the same signals of quality as VCs, and whether they are subject to the same biases.
To address this gap, I examine 2,101 crowdfunded projects that match characteristics of more traditional VC-backed seed ventures. Despite the radical differences in selection environments, I find that entrepreneurial quality is assessed in similar ways by both VCs and crowdfunders, but that crowdfunding alleviates some of geographic and gender biases associated with the way that VCs look for signals of quality.”
Crowdfunding Models: Keep-it-All vs. All-or-Nothing
By Douglas J. Cumming York University – Schulich School of Business Gaël Leboeuf Univ. Lille Nord de France – Skema Business School Armin Schwienbacher Univ. Lille Nord de France – SKEMA Business School
June 2, 2014
“Rewards-based crowdfunding campaigns are commonly offered in one of two models: “Keep-it-All” (KIA) where the entrepreneurial firm sets a fundraising goal and keeps the entire amount raised regardless of whether or not they meet their goal, and “All-or-Nothing” (AON) where the entrepreneurial firm sets a fundraising goal and keeps nothing unless the goal is achieved.
We provide large sample evidence consistent with the view that the usage of AON is a credible signal to the crowd that the entrepreneur commits not to undertake the project if not enough is raised. This signal reduces the risk to the crowd, thereby enabling the AON entrepreneurial firms to set higher goals, raise more money, and be more likely to reach their stated goals.
In contrast, KIA projects tend to be less successful, since the crowd bears the risk that an entrepreneurial firm undertakes a project that is underfunded and hence more likely to fail after the campaign. Entrepreneurs use the KIA model for scalable projects; that is, projects that are still feasible with partial funding.
Further, we provide evidence that the crowd is much more sensitive to information provided by AON projects. We show that these findings are robust to a number of robustness checks, including but not limited to use of instrumental variables and propensity score matching.”
Gender Dynamics in Crowdfunding: Evidence on Entrepreneurs, Investors, Deals and Taste Based Discrimination
By Dan Marom Hebrew University of Jerusalem – Jerusalem School of Business Administration Alicia Robb University of California, Berkeley – Coleman Fung Institute for Engineering Leadership ; Kauffman Foundation ; University of Colorado at Boulder ; Federal Reserve Banks – Federal Reserve Bank of Atlanta ; University of Deusto – Basque Institute of Competitiveness ; Marin Economic Consulting Orly Sade Hebrew University of Jerusalem – Department of Finance
May 29, 2014
“In this paper we investigate whether a new form of venture financing – crowdfunding – reduces the barriers of female entrepreneurs to raise capital. Specifically, we investigate gender dynamics and biases in the process of raising funding to new projects via the leading crowdfunding platform – Kickstarter.
We find women made up about 35% of the project leaders and 44% of the investors on the platform. On average, men seek significantly higher levels of capital than women for their projects, and also raise more funds than women. However, women enjoy higher rates of success in funding their projects, even after controlling for category and goal amount. Only about 23% of projects that men invested in had female project leads. Conversely, more than 40% of projects that women invested in had female project leads.
Multivariate analysis indicated significant positive correlation between the gender of the project leader and the percentage of the same gender investors. Questions arise around what factors explain the fact that female-led projects are predominantly financed by women.
In an attempt to disentangle taste-based discrimination from statistical discrimination we conducted a survey of investors from the Kickstarter platform and find evidence that some of lower investment in female-led projects by men can be attributed to taste-based discrimination.”
Find more studies here.
If you’re conducting a research study on the crowdfunding industry, leave a comment below!
