In recent decades, innovation challenges, hackathons and other open innovation competitions have become quite popular, particularly in technology. But the broader use of contests as a powerful, cost-effective way to reward effort spans all industries. US government agencies run prize competitions that encourage the public to solve problems. NGOs regularly hold fundraising contests that encourage donors to compete for prizes or recognition. Managers hold contests—although they may not explicitly describe them as such—to encourage employees to compete for a coveted promotion. And in the distant world of cryptocurrencies, Bitcoin miners compete with each other to win Bitcoin by being the first to verify transactions on the blockchain.
“There are many different contexts where we see contests being used in practice,” says George Georgiadis, associate professor of strategy at Kellogg.
However, despite their ubiquity, conducting an optimal competition—that is, a competition that wins the most money in terms of incentivizing competitors to achieve a desired goal—is surprisingly difficult. After all, the contest designer has a number of important decisions to make. When exactly should the contest end? How should the prize be distributed? And what is the best strategy for providing feedback to contestants about their performance and the performance of their competitors?
New research by Georgiadis and his colleagues Luis Rayo, professor of strategy at Kellogg. Jeffrey Ely, professor of economics at Northwestern (and courtesy professor of managerial economics and decision science at Kellogg); and Sina Khorasani, a postdoctoral fellow at the University of California, San Diego, offers answers.
The researchers conclude that, in many common scenarios, the deadline for completing the competition should be temporary, meaning that if no one reaches the final goal, the deadline is extended. The researchers also find that in such scenarios, the optimal competition divides the prize equally among all competitors who succeed before the deadline and provides competitors with limited, but strategic, feedback.
“Our goal is to try to find the best possible competition under certain conditions,” says Rayo. “Perhaps our results help us understand some features of real competitions, but they also suggest improvements.”
Reviewing a Contest
Organizations regularly use incentives to motivate individuals and groups to work hard. Often, this takes the form of a contract that dictates how an employee is compensated: they may receive a base salary, for example, but also have the ability to earn a bonus if they reach a performance milestone. Or maybe their pay increases linearly as their performance increases.
Researchers have long studied how to design the optimal contract in different situations. But there are times when it might make sense for organizations to use a contest.
“You can imagine a boss who has some prize money and wants to use that money to motivate employees,” says Rayo. One option is to offer a contract that will incentivize one employee at a time, perhaps offering each a chance to earn a bonus. But another option is to offer a contest where multiple employees compete against each other for prize money.
“Which is better? It depends,” Rayo says. “One of the things we show in this paper is a situation where it’s actually better for workers to compete with each other.”
Specifically, the researchers demonstrate that when employers are able to control the flow of information and strategically withhold information about how their competitors are doing, organizations can potentially make more money by offering a competition than by drawing up individual incentive contracts. .
“When you can control that flow of information, that’s a big reason why you might want to run a tournament instead of rewarding every player regardless of how everyone else is doing,” Rayo says.
The Best Competition
To better understand when contests are most useful—and how to design the optimal one—the researchers created a mathematical model that describes how workers will work under different compensation systems. In their model, employers want employees to work hard for as long as possible. In the process, they might achieve a specific goal, such as creating an algorithm that is 10 percent more efficient, hitting a certain sales goal, or becoming worthy of a promotion. Once the goal is achieved by one of the contestants, the employer is notified.
The researchers then modified various aspects of the competition to determine which design would motivate contestants to work as hard as possible for the same prize money.
“We’re trying to expand that prize money as much as possible,” says Rayo.
One thing that differed was how the employer shared information with the contestants. What if the employer notified all contestants once only one contestant met the goal? Alternatively, the employer could regularly inform competitors that no one had yet succeeded. Or the employer could decide to share information randomly, so that sometimes competitors are informed of each other’s performance and sometimes not.
The researchers also considered how the competition should end. And here there were several possibilities. The contest could end once one contestant reaches the desired goal, or all contestants could be given a hard deadline, after which the contest would end, even if no one had finished yet. Or contestants could be given a temporary deadline, but the contest could be extended.
Finally, the researchers varied how the prize was distributed: Should it be winner-take-all, randomly assigned, or shared equally among all successful contestants?
They found that the optimal tender used a temporary deadline. Contestants did not hear anything about others’ performance until that deadline. However, if the temporary deadline passed and a new one was set, contestants could conclude that no one had yet succeeded. This would continue until one or more contestants were successful, at which point the contest would end at the next provisional deadline and the prize would be shared equally among each successful contestant.
Why is this optimal contest design?
For as long as possible, even after a competitor is terminated, the employer wants to stay mum so as not to discourage those still employed. So continuing the contest (and keeping quiet) is their best bet.
But after a certain period of time, competitors are likely to become disillusioned and lose the will to compete. Thus, being able to extend a temporary deadline allows the employer to offer some much needed incentive by stating that “it is safe to continue. No one has succeeded yet,” says Rayo.
And dividing the prize equally among all those who succeed is ideal, because if it went solely to the contestant who finishes first, the contestants might lose their motivation over time, assuming that even if they finished before deadline, they are unlikely to terminate first. “It’s important not to discourage people from working towards the end of the cycle,” says Rayo.
Georgiadis points out that all the various design elements complement each other to create an optimal competition. If, for example, the interim deadlines are set very close to each other, then the rule of equal distribution of the prize money is very encouraging, because the competitors will assume that many of them will not reach the target before the deadline, so there will be larger shares than the prize money paid out. But the competition will end prematurely, leaving the effort on the table. On the other hand, if the interim deadlines are too far apart (or fall off altogether), then the equal sharing rule is not very encouraging because contestants will assume that the prize will be shared among too many contestants.
“The feedback policy goes hand in hand with the allocation rule,” says Georgiadis.
In the real world
The research highlights ways in which existing competitions can potentially be improved. Take Bitcoin, for example, which uses a competition to incentivize miners to verify updates to the blockchain. “What we show is that if you replace the current system with the one we proposed, you can do the same effort while having to issue less Bitcoin,” says Georgiadis.
More generally, the researchers were surprised by how cost-effective this type of competition can be. “As a company, what you want to do is create a lot of value and capture as much of it as possible,” says Georgiadis. “What this competition does is it creates the maximum value it can create given the prize money, and manages to capture it all.”
This is what makes contests superior to individual incentive contracts — at least in cases where the employer retains control over the flow of information.
“If you were going to contract employees one-on-one, that means you’re probably going to waste the prize,” says Georgiadis. But the competition format allows essentially the same prize money that would otherwise have been wasted on Contestant 1 to motivate Contestant 2. Contestants’ uncertainty about their competitors’ success—and therefore whether it’s really worth it to keep working— it is what allows the organization to extract more miles from the prize.
Another way of thinking about it, however, is that if employers capture all the value, employees capture none. “Employees are not thrilled about it,” says Rayo.
He points out that many organizations have other goals—such as building morale or loyalty. In this case, deliberately designing a less efficient contest could be the ticket.
“If you wanted employees to keep some value for themselves, you’d make those deadlines closer to each other, instead of going silent for, say, three months and then revealing that nobody made it,” Rayo says. . This would provide competitors with much better information about their chances of winning a large share of the prize, helping them make informed decisions about whether to continue working.
“You keep them in the dark less,” says Rayo.
