Recently I attended the Nudge.It North Conference, a virtual behavioral science conference that featured a number of compelling speakers from a variety of different backgrounds, including best-selling authors, research professors from top universities, and a World Series of Poker Champion. I’m always learning from Tim Houlihan and Kurt Nelson who did a fabulous job organizing the event. The conference’s stated goal was to bring “world-class experts to share knowledge and experience in applying behavioral science to business, design, and strategy development.” Despite being just a one-day event, it nonetheless offered some important insights that can be directly applied to the world of incentives and loyalty program design.
Of the presentations I attended, there were a few that really stood out to me. The first was a talk given by Annie Duke, an author and former World Series of Poker Champion. During her presentation she touched on the importance of making good decisions by working through the ways in which those decisions might go bad. In behavioral psychology, this is referred to as a “pre-mortem,” where the decision-maker asks themselves two key questions before making an important decision. These are:
What do we know at the time of the decision?
What can we anticipate might go wrong with the predicted outcomes?
In high-stakes poker tournaments, the ability to make sound judgments and influence the behavior of others is of the utmost importance, so Duke was particularly well-positioned to share her experiences on this topic, after winning 3 World Series of Poker bracelets. This concept of the pre-mortem is one of those mental exercises that seems really simple to understand but is actually quite hard to do in practice. In order to understand “what we know,” we have to consider a few necessary components:
Ultimately, predicting the outcomes of our decisions seems to be a good business habit in general, but specifically in incentives and loyalty program design it can also be particularly useful as we attempt to predict how our programs will affect the behavior of the participants. If we can anticipate how the various levers of our program will create positive, negative, or neutral behavioral outcomes, we can adjust these programs accordingly.
Another insightful presentation came from Reuben Kline, a professor of behavioral game theory at Stony Brook University. In his talk, Meritocracy or Malfeasance?, Kline shared some insights into people’s rationale behind lying and cheating. He discussed a recent study conducted in the U.S. and Italy that attempted to answer the questions of whether or not violations of meritocracy can lead to malfeasant behavior—in other words, why people may rationalize cheating.
The cross-national study revealed that in “un-meritocratic” situations where merit was de-emphasized—meaning low-performing individuals were rewarded more than high-performing individuals for the same task—the high-performers were more likely to lie about their behavior. I thought this was a pretty intriguing data point.
In recent years, loyalty programs have been seeing an increase in fraudulent behavior, and I think that a key to solving this problem has to go beyond just simply stricter policing within the program. This study would seem to suggest that in fact the rules structure of a program also makes a difference. By making sure the rules a program design truly reward those who deserve it and closing loopholes that would enable underperformers to take advantage, we can perhaps do a better job of discouraging fraudulent behavior.
Finally, I also attended an interesting talk by Charlotte Blank, currently the Chief Behavioral Officer at Maritz. She mentioned how behavioral science should really be thought of as more of a verb, or an action, rather than a noun or a thing to implement. It’s about experimenting and figuring out what’s really going to drive particular results. Blank brought up two common types of automotive dealership incentives: “cash on the hood” incentives that focused on the consumer, and “cash in the trunk” incentives that focused on the dealership as a business entity. Blank shared some research that focused on the dealership performance.
The study focused on a field experiment centered around the concept of loss aversion, which suggests that we value things we lose more than the things we already have. In the experiment, automotive dealerships were split into two groups: one that was given a bonus at the end of the month if a sales target was reached, and another that was given a bonus at the beginning of the month, but which had the bonus “clawed back” if the sales target was not reached. The question was, does loss aversion work better as an incentive strategy than a more traditional strategy of rewarding upon goal achievement?
The results were compelling. The test group—the “loss aversion” dealership—ended up performing 4% worse than the control group with the straightforward end-of-month reward. Somewhat ironically, this led to an estimated “loss” of $45 million in revenue. In part, this was due to the fact that the test group was so focused on avoiding losses (“loss-framing”) and ended up spending significantly more effort trying to make easier, but lower-margin, sales. In the long run, this ended up inadvertently hurting results.
The key takeaways from this session were, first, that program rules focusing on loss-framing are not always successful, which is somewhat counterintuitive based on what we know (or what we think we know) about loss aversion. Secondly, the study reinforced the importance of A/B testing and “leading like a scientist” when it comes to incentive program design.
Our team at HMI Performance Incentives prescribes that field tests can not only inform our best practices in terms of program design moving forward, but they can also potentially improve the return of the program and get you closer to the bullseye, not just on the target.
I always enjoy attending conferences like Nudge.It North because I usually walk away with at least a few key actionable insights. Especially during these unprecedented times, the ability to attend these conferences virtually has been a great way to stay on top of and continue improving on industry best practices. Whether it’s through the study of behavioral economics, game theory, human psychology—or yes, even high-stakes poker—I’ve found that there’s lots that can be gleaned for incorporating a cross-disciplinary approach to incentives.
Please feel free to share your thoughts with me here.