Crowdfunding mobilises $1.3 billion for innovation funding annually and that amount is projected to more than double in the next five years. Platforms like Kickstarter and Indiegogo have not only broadened access to funding for companies that might struggle in the capital markets, but have also transformed the way companies connect with consumers during product development, replacing focus groups with real customers who have a stake in the final product.
Despite crowdfunding’s many benefits, however, numerous campaigns still fail. On Kickstarter alone, 59% of campaigns do not reach their initial funding goal, let alone progress further. To understand why, we embarked on an empirical analysis of 18,173 campaigns for physical products in the technology and design categories on Kickstarter.
Our data set contained detailed information about campaign characteristics (such as the amount of customer involvement) along with product descriptions from the beginning and end of each campaign. We fed these descriptions into an unsupervised natural language processing (NLP) model called latent Dirichlet allocation (LDA) to measure the number of product features. By comparing product features at the beginning and end of a campaign, we can track how much a product has been improved and how much of that is due to customer feedback.
What we found was that many companies often present initial products that are so fully developed that customers don’t feel that their input will materially change the product.
While the precise reasons for a campaign’s failure are difficult to pinpoint, Mike Cecot-Scherer’s 2014 campaign for his innovative camping tents offers an instructive example. The campaign description highlighted numerous features, such as a large net roof for stargazing and warm weather, and pre-installed emergency zipper repair sliders. Mike explained to us in an e-mail that he saw an extensive feature list as a way to differentiate his product from its competitors: “Since my tents were neither inexpensive nor ultra-lightweight, I had to introduce my potential customers to the wide world of other design improvements that might be made.”
While penetrating a market with more developed products is usually a sound strategy, it may backfire in a crowdfunding setting. Anyone who is not an expert on tent technologies can easily feel overwhelmed by the large range of features presented and find it difficult to propose improvements. As one customer commented: “Trouble is, you’ve given detailed information on the Home page, so I’m not sure what’s left to add.” Ultimately, the campaign raised only about one-fifth of its original goal, leading to its failure.
It’s also clear that Kickstarter funding participants place a high value on being able to contribute to the development of the products they invest in. It doesn’t have to be their own ideas either: Our data showed that the addition of a single customer-suggested feature nearly doubles the likelihood of a successful Kickstarter campaign from 27% (if no features added) to an impressive 62% (if one feature added). In effect, companies that use Kickstarter are being paid by customers to receive their insights instead of having to pay for expensive market research. What’s not to like?
Reid Hoffman, the co-founder of LinkedIn, once famously said “If you are not embarrassed by the first version of your product, you’ve launched too late.” Our study puts this notion into test in crowdfunding, and a key takeaway for entrepreneurs from our analysis is to avoid overdeveloping. To be sure, an “embarrassing” first version can be too raw, risking sending your customers down all sorts of rabbit holes. But on the whole you’ll engage your customers productively if you offer them a stake in a work in progress.
The article first appeared in Harvard Business Review.