As I noted in this post from a few weeks ago, crowdfunding campaigns raised $2.7 billion last year, up more than 80 percent from 2011.

Now, research on crowdfunding by Wharton professor Ethan Mollick reveals some keys to the success of crowdfunded projects. Mollick looked at more than 48,000 projects on Kickstarter; of these, nearly half – 23,719 projects – raised the amount of money their founders sought.

In his paper, “The Dynamics of Crowdfunding: Determinants of Success and Failure,” Mollick reveals that the success of a project tends to be correlated to both its quality and the founders’ personal networks. Perhaps not surprisingly, the location of the project founders tends to influence the types of projects they choose. Finally, most of those receiving funding do their best to deliver their project, although delays aren’t unheard of.

Some of the findings:

1) Investors look for indications of the project quality, such as the quality of the founders’ preparation. While it might seem difficult to measure this in an online environment, Mollick turned to recommendations made by Kickstarter itself: that the founders include a video on the project and that they provide an update on the project status within three days of its launch. He also looked at the frequency of spelling errors in project pitches. The presence of videos and frequent updates were, indeed, more closely correlated with project success, while spelling errors reduce the likelihood of success.

The breadth and depth of the founders’ social networks also seem to be viewed as indicators of project quality. For instance, Mollick found that of the projects in the “film” category, a founder with 10 friends had a nine percent chance of succeeding, while someone with 100 friends had a 40 percent chance.

2) Failures occur by large margins; successful projects by smaller ones. For instance, just 10 percent of failed projects had raised even 30 percent of their goal, while only three percent had raised half their goal amount. On the flip side, one-quarter of successful projects were over their goal by three percent or less; just one in nine projects received more than twice its goal amount.

3) Geography influences the projects chosen. A few examples: music–based projects tend to dominate in Nashville, while film is big in Los Angeles and tech projects in San Francisco.

4) Delays aren’t uncommon. Just one-quarter of projects delivered on time. The delays tend to be longer on larger projects. In addition, projects that were over-funded by greater amounts also were more likely to be delayed.

Based on his research, Mollick offers several lessons for entrepreneurs who turn to crowdfunding to support their projects: the quality of the project matters, and founders should find ways to demonstrate that they’ve got a viable, quality project. Entrepreneurs should aim for reasonable funding goals, as it’s unlikely they’ll get more than what they ask for. They also need to have in place a plan for executing rapidly if their crowdfunding venture succeeds.