Planning to finance an open source project through crowdfunding? Then don't quit your day job.
At first glance, open source and crowdfunding seem an inevitable match. After all, what could be more natural than software that nobody owns being funded by popularity? In theory, crowdfunding should allow developers to concentrate on what interests them, freeing them from the need to make a living or answer to an employer.
And, sometimes, the match works out. Linux Voice, a proposed new magazine, not only reached its goal of 90,000 pounds with several days to go, but exceeded it by over 33%. Similarly, A Raspberry Pi Build Cluster for Ubuntu reached its goal of 2,500 pounds several days before the end of its campaign, while Freedom Box, "a free software system built to keep your communications free" reached 134% of its requested $60,000.
However, these success stories are like those for self-publishing: they're the exception, not the rule. For every success story, there are at least a dozen failures, such as Yorba's effort to fund development of its Geary mail reader, or Canonical's efforts to raise $32 million for Ubuntu Edge, a proposed boutique phone. If anything, most crowdfunding efforts are even less successful than these examples.
If you want to evaluate your chances, the best place to go is Indiegogo's search page. Kickstarter, of course, is also a popular crowdfunding site, but its search engine is less useful for finding information about campaigns. Moreover, Kickstarter insists on original projects, which means that many hardware and software campaigns are ineligible for it, especially if early versions or prototypes already exist.
The first statistic is the success rate of projects that use the keyword "open source." If the first 12 pages of results is representative of the whole, then only 7.5% obtain their requested funding. In fact, only 3.5% receive over 50-75% of their goals, leaving 89% that raise under 50%.
The odds are slightly better for verified non-profit groups -- but still not good. 33% of campaigns by non-profits that mention open source obtain their goals, and less than 7% reach 50-75%.
Searching for the keyword "free software" offers similar results -- unsurprisingly, since anyone with any sense will use both "open source" and "free software" as near-synonyms in order to attract the widest possible range of possible donors.
By contrast, the less well-known synonym "FOSS" (free and open source software) returns 5 out of seven successful campaigns, or 72%, while the more obscure "FLOSS" (free/libre and open source software) returns 0 of 2. Possibly, these results suggest that a less popular term may be more successful than a popular one -- so long as it is not too obscure. The reason might be that while fewer potential donors are likely to browse a less popular term, those who are familiar with the term are less likely to donate.
No matter what keywords are chosen, both the funding type and the target amounts seem irrelevant to success.
Apparently, also irrelevant are perks -- the thank-you gifts given to donors. Whether campaigns succeed or fail, most offer finished samples of the software or hardware, acknowledgments on a web page, and schwag such as t-shirts and coffee mugs.
Browsing all these results shows a few successful software campaigns, such as Youtube Anywhere for Moodle and BlenderFDS 3. However, software campaigns number less than 8.5% of hardware campaigns. This discrepancy may reflect that software releases often have shorter life cycles than hardware, making them less interesting to donors.
What these results show is that crowdfunding is not a magical solution to funding. Having a unique project seems no guarantee of success, although working with popular products like Raspberry Pi or intense marketing efforts may help to tilt the odds in your favor.
Very likely, too, campaigning through another agency, rather than well-known sites like Indiegogo and Kickstarter, may make a difference. Last year, when MediaGoblin conducted a campaign managed through the Free Software Foundation, it raised only 77% of its $60,000 goal. However, those who heard about the campaign might have been more likely to donate than the general browsers of Indiegogo and Kickstarter. Had MediaGoblin used the better known sites, it might very well have been lost in the crowd and been less successful.
These uncertainties are only likely to become worse. If crowdfunding has not reached saturation point, it must be nearing it -- in other words, anybody who would donate is already donating. That means that more campaigns will be competing for the same number of donors.
Under these circumstances, crowdfunding is still a possibility. Increasingly, however, it is a possibility that needs to be approached carefully and with lowered expectations.