For those trying to crowdfund a business, one of the earliest opportunities to go double-or-nothing comes before you've even raised any money for your company. Should you run a campaign that lets you keep whatever money you raise, or one in which you forfeit the proceeds if you don’t hit your fundraising goal?

A new study, from Toronto's York University and Universite Lille Nord de France in Lille, France, compares the results of those two different types of crowdfunding campaigns. The results suggest that if you're really serious about getting your project or product funded, you should probably grit your teeth and do an all-or-nothing fundraise.

"Overall, [all-or-nothing] fundraising campaigns involved substantially larger capital goals, and were much more likely to be successful at achieving their goals," write the researchers.

The appeal of keeping whatever you raise (the keep-it-all strategy) is obvious: Sure, you think you need $30,000 to get your company started, but if you were to get only $20,000, you figure you could work it out. You’ll get creative. Do some consulting on the side. Ask your friends to work for free. Whatever it takes.

Those who go the all-or-nothing route don’t have such comforts. If they miss their fundraising goals, they have to start over.

But investors, it turns out, get some comfort from the all-or-nothing strategy. They don't have to worry that the entrepreneur they've chosen will launch a weaker product than anticipated because he or she couldn’t make their funding goal. The researchers suggest that makes them more likely to fund an all-or-nothing attempt.

The report's researchers looked at 22,875 crowdfunding campaigns, each of which was trying to raise between $5,000 and $200,000. They pulled the information from Indiegogo, since that site is both large and well established, and lets entrepreneurs choose either a keep-it-all or an all-or-nothing campaign. Those who used the keep-it-all model were trying, on average, to raise $19,677. Those who went all-or-nothing, on average, were trying to raise more: $31,355.

Yet the all-or-nothing folks were more likely to hit their fundraising goals: 34 percent of their campaigns were successfully completed, compared to 17 percent of the keep-it-all campaigns. All-or-nothing campaigns attracted an average of 188 backers, while on average the keep-it-all variety attracted 73.

Maybe that's because the all-or-nothing entrepreneurs are trying harder. In 19 percent of the all-or-nothing campaigns, backers can get some kind of reward for ponying up as little as a dollar. That's only true for 12 percent of the keep-it-all campaigns.

They also tend to provide more information to potential backers: longer project descriptions, more pictures, more video pitches, and more updates and comments. In other words, these folks are going all in. In every way.