As an operations professor at Duke’s Fuqua School of Business, Robert Swinney was naturally drawn to the online crowdfunding industry.

“The nuts and bolts of a campaign – how you design it to optimize revenue, how success targets should be set, what kind of rewards should be offered and how they should be priced – these are operational questions,” he said.

Crowdfunding – the act of raising small amounts from hordes of investors – raised an estimated $34 billion worldwide in 2014. Swinney studied Kickstarter-style campaigns, in which campaigners are seeking to get a product or service off the ground. Contributors pledge funding – often associated with a reward in addition to the product – but no one is charged until and unless a pre-set goal is reached.

Working with Ph.D student Soudipta Chakraborty, Swinney created mathematical models to solve two problems faced by every crowdfunding campaign. First, they show that a high campaign goal is the best way to show potential contributors that the product or service is high-quality; second, having rewards tied to early contributions is the best method to nudge consumers to get involved rather than waiting on the sidelines, a tendency that can doom a campaign.

Swinney discusses their work in this Fuqua Q&A.

Because contributors can’t touch a brand new product or do any research on it, how can you communicate the quality of what’s on offer?

Many researchers have studied how to manipulate price as way of signaling to consumers, to tell them something they don’t know about the thing we want them to buy. This is the first time that’s been looked at in a crowdfunding context. What’s different about crowdfunding is that we have a success target – that doesn’t exist in regular retail.

We find the success target is a better way to signal quality than price. You can signal you have a high quality product by setting a higher campaign target. That sends the signal that you’re confident you have a really great campaign, because a lower-quality product won’t copy that.

This turns out to be better than signaling quality with price, because manipulating the price can hurt your profit even if the product turns out to be a huge hit. In contrast, setting a higher campaign target only hurts your profit if the product was going to be dud, or a modest success.

What are the risks to that approach?

Though it gives you gains in other ways, signaling is costly, not least because there’s a greater risk you won’t reach your goal. A side-effect we found is that the costliness of signaling can deter people from creating a high quality product. Instead it gives campaign creators the incentive to create lower-quality products.

So crowdfunding platforms should do everything they can to minimize this information gap by making campaigns as transparent as possible. Many do versions of this already, with rules prohibiting renderings of products and requiring physical prototypes and videos of the product in operation. Ultimately it’s a problem inherent to crowdfunding that it encourages lower-quality products on these platforms.

Even if you successfully communicate the quality of the product, you still need people to contribute. Why do so many contributors take a wait-and-see approach, and what can campaign creators encourage potential contributors to get involved earlier?

In retail settings, consumers like to wait. For example, we know swimsuits are expensive at the start of summer and cheaper when fall begins, so some people wait until fall to save money on a swimsuit for next year. In crowdfunding you’re not waiting for a discount, you’re waiting to learn whether the campaign is going to succeed or not. It’s the same kind of strategic intentional delay of a purchase, but in a different setting.

There are two different contribution patterns that emerge over time. Some start with a really high level of contributions and then fizzle out. But some pop back up at the end, with contributors rushing in at the end of the campaign. These patterns reduce the profit of the creator. When people are waiting for others to go first, you need more contributions to succeed. When contributors intentionally delay, they could forget, or get distracted and not come back.

Our model shows that when contributors delay, they are waiting to see a critical number of contributions before they start to participate. So if the campaign doesn’t hit that critical number then the campaign peters out. But if it does, contributions ramp up as you get closer to the goal.

This critical point differs for different campaigns, but our model can calculate it for any campaign based on whatever the parameters are.

So how can you nudge people to contribute rather than staying on the sidelines and waiting for this critical point?

The key is in the rewards. Our model finds the optimal way to design a campaign is to tie the price a contributor pays to how early they get involved. So the later you contribute, you more you pay. Each contributor sees a greater likelihood of success based on the amount other people have already contributed, so ideally every contributor would pay a little bit more than the previous one.

It’s not feasible to design an actual campaign this way, because you would need an infinite number of prices. But what we find works well is having a limited number of rewards at a low price, and then raising the price after success has been attained – like an early bird discount. So you could sell 500 units at $20 and an unlimited number at $25.

If you design the campaign so that you reach your goal when the discounted items sell out, that’s going to perform close to this optimal model.  The way to implement this is to post them both at the same time. People can see they would pay a lower price until those items are exhausted. That works because you make it costly for people to wait. It shifts the contribution time earlier in the window, it makes success more likely and it increases the project’s revenue.

Of course it also helps the contributors, because increasing the campaign’s chance of success makes it more likely they’ll get the product they want, since if the campaign doesn’t reach its goal then the product won’t get made.