Crowdfunding platforms such as Kickstarter and Ulule have emerged over the last years as the place to go to get the funding necessary to make projects come true – be it a business start-up, a creative work or a social activity. Many individuals chip in and as a crowd they contribute money to help artists produce their records, graphic novels, video games; or to allow geeks to produce a tool to transform bananas in a piano, or a set of tools to make robots with drinking straws. More: a project founder might ask for, say, fifty thousand dollars to produce a wheeled plastic cooler, and receive support for thirteen million, 260 times as much.
What is behind crowdfunding’s success? Is simply asking for help enough to tap into the reservoir of altruism present on the Internet? Well, it is not so easy. Only about half of all crowdfunding campaigns actually get funded, and the extent of funding varies considerably across projects. Many dismally fail.
What makes the difference to reach the project’s funding goal and what actually happens in the process of the funding campaign? Are crowdfunders affected by the funding decisions of others and to what extent is project success path-dependent, determined by very early pledges? Do project launchers try to exploit the relative anonymity of the Internet to contribute to their own projects?
In a recently published article, we explore these questions using data provided by Startnext, the biggest German crowdfunding platform. Startnext gave us access to its whole database – more than 100,000 pledges to more than 2,000 projects. The answers indicate that it is never too late to get a campaign on a successful track – provided its creator manages to get a surge in pledges rolling.
“It’s never too late”
This finding is rather striking. The message of nearly all the early literature on crowdfunding is to “launch hard or go home”: success is predicted by and depends on what happens in the early stages of a project campaign. This is called path-dependence: attracting lots of support in the, say, first day or two, is a crucial indicator of quality, attracts more and more pledgers, and nearly always leads to success. This pattern was considered so strong that dedicated web services appeared (and were later shut down due to legal complaint) that predicted project success or failure after the first day of a Kickstarter campaign.
However, we find that at Startnext a large fraction of successful projects are late bloomers. Let us be more precise: your average crowdfunding campaign starts off with a healthy amount of pledges, levels off in the middle part of the funding phase, and then (eventually) gets a boost in the last few days. Picture it as an inverse S shape. This pattern is closely followed by a relative majority of projects – about 23% at Startnext.
We find that even projects that fall short of this average path to success have a high probability to make it in the end. Nearly 90% of the projects that are under track after two thirds of the campaign still become success stories. Even more strikingly, 40% of severely under-track projects still make it in the end. These are projects that are more than 70% off their targeted amount with a handful of days to the deadline. The key insight of our analysis is that it is never too late in crowdfunding.
But what’s behind the late surges of under-track but eventually succesful projects?
We conducted a survey among projects whose campaigns looked bleak for a long time to uncover what made the difference. The survey responses suggest that the eventually successful projects never give up and increase their communication activities. They interact with the crowd, via videos, blog or social media posts. In contrast, projects that fail to reach their funding goal do not step up their communication efforts along the course of the campaign. It seems important to keep the projects alive and update the community to harness the power of the crowd.
“Angels in the crowd”
But this is not the only reason behind the late blooms. Looking closely at pledges, we found that projects that were severely under track do not receive more pledges toward the end of the campaign, but larger ones – more than three times as large than all other project categories. Our survey responses indicate that these large pledges originate from “angels in the crowd”, external investors that look around for promising ventures to support.
Another potential reason explaining the late surges is self-help. Nothing bars project launchers to create a nickame on the platform and pledge relative large amounts of money to make it past the threshold in time. Uniquely among crowdfunding platforms, Startnext allows project creators to fund their own campaign – thus giving us the opportunity to assess the extent of self pledging, something that can’t be done with, say, Kickstarter data. We find that project launchers do indeed self pledge, moderately overall, but with a clear pattern and extreme cases. For some projects self pledges are substantial: 6% of all projects are self funded by a quarter or more of their funding target. Why self pledge? The distribution of self pledges clearly identifies three main motivations: to start a campaign that has seen little action in the first hours or days, to revive interest after a period of slack and to secure funding once the deadline approaches. Nevertheless, the late surges at (severely) under track projects are overwhelmingly driven by external funders – the “angels” – thus confirming that the “it’s never too late” result does not depend on self-help.
To summarize: in crowdfunding, a strong start leads almost universally to success. But the converse is not true: a slow start is no death sentence, as booms can happen at any time. What is the secret of late boomers? Our evidence suggests that even very late efforts at extended communication matter – be it for attracting a crowd or an angel.
 | Mailchimp)