So I want to talk a little bit more broadly about crowdfunding and how it works. And specifically talk about the issues of fraud and failure in crowdfunding projects. And this is based on a lot of research I've done and worked with other people on, as well. And I think it's a useful context for both understanding crowdfunding overall and also understanding your role as a project creator. So the first thing is, at least on Kickstarter, failure is remarkably low. So about 10% of projects, somewhere between 9% and 13%, in terms of how you're calculating, actually fail. And you can see the failure rate by pledge. And you can see it's sort of a u-shaped curve. So very large projects have a slightly higher chance of failure and smaller projects have a higher chance of failure. And that's probably because the smaller projects are less likely to have slack, less likely to have built out their system. Some of the things we discussed in a prior lecture on avoiding issues in your crowdfunding project. And larger projects are just more complicated, more subject to delay, more likely to fail. But even so, given these ranges, these are relatively small ranges, and the failure rates are pretty low. So the question is, why? And I actually have done a lot of research on this and want to sort of talk about some of those results, because there's an interesting set of contrasts. There's a famous book that came out about ten years ago called, The Wisdom of Crowds, that argues that crowds are really smart. That markets generally know what they're doing, that large groups of people can be smarter than experts. But if you go back 100 years before, there's actually another book called, The Madness of Crowds. Which explains that when you get crowds together you get things like witch hunts, and tulip buying manias and Ponzi schemes. So which is it? Are crowds wise or mad? And I've given you some of the answer already by telling you that failure rates at Kickstarter are low, but I wanted to kind of go more into detail. So I have this interesting paper with Ramana Nanda, who's currently at Harvard Business School. And Ramana and I decided to try and figure out whether crowds or experts were better at judging things and when they worked together. And we actually chose an area that would be particularly challenging for the crowd to match what experts do, and that was theater. So we actually looked at theater projects on Kickstarter. And this is especially interesting, because as of 2012 more money goes to the arts through Kickstarter than the National Endowment for the Arts, according to some counts. So it's really important to understand whether Kickstarter is actually doing a good job at giving money to the right kinds of projects. And the arts especially is challenging because there's a big difference between sort of high culture and low culture, right? So on one hand, you have cheesy community theater, which I'm a fan of, to be fair. But that's very different than a very high-end play put on in a avant-garde, sort of critically loved theater. So trying to understand where the crowd can work with experts is really important in this space. So what Robin and I did was we actually looked at experts, we brought in experts who judge contests like the NEA, the National Endowment for the Arts contest. And asked them to answer a bunch of questions derived from the criteria that NEA uses to figure out whether projects are good or bad. So, about the audience, the artistic merits of a project, whether it was feasible or not. And we had them judge Kickstarter projects. And what they didn't know, whether those Kickstarter projects had succeeded or failed. So we picked a bunch of projects that had succeeded by a large amount, some that succeeded by a smaller amount, and we tried to figure out what happened later in those projects as well. What we found was pretty interesting. We found that generally the crowds and experts agreed with each other. About 60% of the time the crowd and experts agreed. And the more successful something was with the crowd, the more likely experts would agree as well. So that suggested that on the most successful projects the crowd and experts agreed that these things were valuable. What was really interesting though is where the disagreement happened. It was by and large because the crowd was willing to fund things the experts were not. So that meant the crowd was willing to put money into a project that experts were not willing to do. So it's great to know that the crowd and experts generally agree with each other, and that the crowd is more willing to fund things than experts are. But was the things that were being funded any good? So we actually did a follow up and looked at the actual theater productions that were funded by Kickstarter in this case. And we classified them according to a few different criteria. A project was successful if it completed its theatrical run. So if it was supposed to be a three-day weekend run and it ran for three days, that was a successful project. We classified failures as any project that didn't run for its full length. And we classified commercial hits and artistic hits based on those rare projects that either turned into long-term off-Broadway style theater productions that made a lot of money. Or ones that won major prestigious theater awards. And we tracked the projects first that experts and crowd both agreed were good, right? So these are the ones that overlapped. And in those cases, we found the vast majority were successes, they ran their successful length. Very few of them were big stand outs. There's one big commercial hit out of this group that was liked by both crowds and experts. But otherwise, they were sort of conventional theater projects. Incidentally, we were also able to look at some of the projects that weren't funded by Kickstarter but that the experts liked and were later staged. All of them were successes, no fantastic hits out of that group either. Things get more interesting when we look at the things that the crowd liked but the experts were unwilling to fund. Again, most of them were successes, but here we have our only failure. In this case, it was a project that actually played in New York. And I still remember the New York Times review which said, in this case, that during the production, I was trying to find something good to say about this particular play. And around intermission, I realized that the fonts in the play bill were quite nice. So that was a pretty damming review, production shut early as a result. But interestingly, despite this one failure, there were actually more artistic hits and more commercial hits out of this group. So major prestigious theater awards were won by the projects that the crowd liked but the experts wouldn't have funded. And also, these projects were more likely to end up successfully touring on Broadway. So as a result, what we see is that the crowd may actually have wider tastes Iand be more tolerant of successes and failures than just the experts. So the experts are good at minimizing risk, but maybe less good at picking things that could be a tremendous success or tremendous failure. So the crowd and experts can really work together here to come up with the right kind of results. So I think there's value in experts definitely and especially in avoiding errors, failures and that sort of thing. But the crowd can add a dimension of picking high risk, high success projects that maybe the experts alone would be less likely to fund. What's also interesting is going beyond this idea of failure to looking at fraud. So I get interviewed a decent amount about crowdfunding and people always ask me, is there a lot of fraud on crowdfunding? And I can tell you that on Kickstarter, the site I've studied most, the fraud rates are very low. And the reason why fraud rates are low, I believe, is due to something called Linus's law. Linus's law is named after Linus Torvald, the inventor of Linux, the operating system. And Linus's law states that with enough eyes, all bugs are shallow. What that means is if you have enough people looking at a problem, to someone the answer's going to be trivial and they'll know how to solve it right away. The same thing happens in crowdfunding. If you have a successful project then you'll attract a lot of attention to that project from communities of people who care about that. So if you're putting up a project on virtual reality, you'll find lots of people who are interested in virtual reality will start to come to your project if it's successful. And having lots of potential experts from lots of different areas look at your project means somebody's much more likely to spot something that would indicate your project is fraudulent or not going to work. Than if you didn't have these sorts of eyeballs on a project, or just one person was evaluating it. So the crowd can actually do a pretty good job at spotting potential fraud. But you have to have a couple of things happen to make this work. One, you need to make sure that the platform is actually engaged and willing to shut down projects that people flag as potentially fraudulent. And that there's some sort of threshold that allows you to give time to evaluate the project before people get the money. So people can't run away with the cash necessarily. And you also need to make sure there's lots of people looking at the project. So one of the things that I worry about a little bit with small equity crowdfunding platforms is there's not that many people looking at the platform. So Linus's law is less likely to work, and that means fraud could be more likely to happen. But when you have the crowd looking at something and really a lot of eyeballs on a project, the chance of fraud really does drop. I estimated that less than a quarter percent of the money going to Kickstarter goes to projects that are intentionally not planning on delivering but trying to take away the money. And that's a high estimate, so it may be even lower than that. So, what does this tells us? It tells us that the crowd is surprisingly smart. The crowd tends to use rational criteria when backing projects. If you remember, in a prior lecture, I discussed the fact that a single spelling error decreases your chance of success by 13%, if you have a spelling error in your project pitch. That's because the crowd actually is looking for signals of quality and can't be easily fooled when those signals of quality aren't there. Second, the crowd allows high quality, high variance projects to succeed. Experts tend to be more conservative in many cases than the crowd. So the crowd gives more people a chance to succeed or fail, perhaps, than the experts alone might be willing to do. And finally, fraud is very low in crowdfunding, but that's because Linus's law at work. So you have to be careful for crowdfunding platforms that don't have enough eyeballs, that don't have enough diversity of viewpoints, and where platforms aren't willing to shut down projects that look potentially fraudulent. Those protections against fraud in the crowd probably won't work. But overall, the crowd is actually pretty wise and can make some pretty good decisions about what to fund.
