Behavioral Science: Laying the Groundwork for More than a Nudge

Perry Wright, Senior Behavioral Researcher, Common Cents Lab at Duke University

Laying the Groundwork for More than a Nudge

Behavioral science is a popular topic. We want to know: can tuning our designs lead to better user outcomes? In this edition we bring you the experts from Common Cents Lab sharing what they are observing about personal finance. We think this is news you can use.

Here at Common Cents Lab we have the privilege of working with research and industry partners committed to improving the personal financial outcomes of their low- to moderate-income users, members, or clients using behavioral science. It’s challenging, rewarding work, and deeply affirming to partner with so many organizations dedicated to these material outcomes.

When a partner wants to improve financial health outcomes for their users, we collaborate with them to design an intervention. An intervention is a product or user contact revision informed by behavioral science that we structure as a science experiment. We do this so that we can test what works best. A lot of times, partners arrive with a handful of competing ideas already in mind for improvements, and the results of an experiment give everyone confidence before a partner rolls out changes across their entire platform.

We most often begin designing an intervention by analyzing the current situation. What decisions are users currently making within this product or service? How do these decisions reflect (or not) the intended outcomes of the product? Identifying the ways in which users are performing actions counter to your program assumptions, what we sometimes call misbehaviors, is one of the first steps in understanding how you can intervene to serve them more effectively.

Intention-action Gap

As you conduct this initial review, you may uncover what behavioral science calls an intention-action gap: users enter a decision-making environment or process with specific intentions and those outcomes are not fully realized due to an inability to follow through on those intentions. Large spans of time between steps or unclear expectations that require deep user attention can promote these intention-action gaps.

Simplify. One way to reduce the intention-action gap is to make the decision-making process as simple and straightforward as possible. For example, as new users interact with a retirement portal, consider presenting essential options (like increasing contributions) clearly and disclosing deeper customizations only with user action to avoid distracting their critical progress.

Give me a map. Another way to reduce the intention-action gap is to provide feedback along the way. For example, if you're asking users to complete a task that will take a few steps, let them know how many steps are left and what they need to do at each step. This way, they can track their progress and feel motivated to keep going.

Incomplete. You may also discover that you do not fully understand your users’ preferences, and so they are behaving in ways you don’t expect in order to accomplish a different, preferred outcome. Even if you have done thoughtful surveying and user testing, you may discover a mismatch between stated and revealed preferences.

Stated and Revealed Preferences

Most people think that their preferences are rational and thoughtful. They believe that they have a good understanding of what they want and why they want it. However, many times people’s stated preferences don’t match up with their revealed preferences. In other words, what people say they want doesn't necessarily correspond to what they actually choose when given the opportunity.

I thought I knew … There are a number of reasons why this mismatch can occur. One possibility is that people simply don’t have all the information when they state their preferences. For example, someone might say that they would be willing to save a certain percentage of each paycheck toward retirement, but when that percentage is converted to a dollar amount on their next pay stub they may go back and adjust their election, revealing an alternative preference. The missing information at the time of the decision is the equivalence of a percentage to dollars; when they got the new info, they revealed their preference.

Baby I’ve changed. Another reason why stated and revealed preferences can differ is that people’s preferences can change over time. What someone might want today could be different from what they want tomorrow. For example, someone might say that they want to take home more cash in one phase of life and later want to save a larger percentage of their pay for retirement.

Putting it to Work

Can I help you? At this point, you hopefully have some clarity about the current state of user behavior within your program. You’ve taken the time to understand where users are failing to realize their own intended outcomes and where they are revealing their preferences. Here is where you’ll often hear a call to present your users with a “nudge,” a small tweak to their experience with the existing process to guide them toward the correct action. But I want to encourage you to take a moment before proceeding.

Hold on. By making the effort to understand your users’ intention-action gaps and preferences, you’re in a much better position than when you started. You have transparency, and transparency gives you the information to responsibly design more structural interventions. Understanding user intentions empowers you to broaden the horizon of your possibilities to serve them. It might be that your users reveal preferences that are sub-optimal, or even harmful to them, and you have the ability to influence the actions that could help them adjust those preferences over time.

Design matters. Because here’s the thing. What you likely discovered in your review of user gaps and preferences is that your users aren’t totally at fault for these mismatches. It’s dispiriting to see, for example, a third of all users drop off at the exact same point in a signup flow, but it’s critical for understanding what needs adjusting (and it’s often not user behavior).

Behavioral science sometimes gets characterized as trying to trick people into performing some particular action—something closer to a hypnotist than a scientist. But, in truth, what we attempt to do is identify places where these gaps between a person’s intention and the action they perform exist, and then intervene to both influence the user and the system in which that user is acting, closing the gap from both sides.

… And so does empathy. I’m also sober about the fact that we can’t change our complex programs in an instant. What’s so powerful about first understanding specific misbehaviors is that we can strategically deploy system change—carefully adjusting one portal screen at a critical decision point can do the work of ten protracted communications campaigns. But you have to understand both sides of the gap to make that surgical change.

After Design Changes, What Next?

We haven’t even begun to map out the constellation of cognitive and behavioral biases that influence user decision-making for retirement, but this lays the foundation to put additional insights from behavioral science to work.

Over the next few installments, we’ll cover a few of the most important of these biases, starting with the power of defaults and the influence of anchoring.

Stay tuned for more! / Perry

Perry Wright is a Senior Behavioral Researcher at Duke University's Common Cents Lab, using behavioral science to create solutions that aim to increase the financial well-being for low- to moderate-income people living in the United States and abroad. The Common Cents Lab is funded by the MetLife Foundation and supported by BlackRock as part of BlackRock's Emergency Savings Initiative. For more information and to connect directly, you can reach Perry by email here.

This piece was featured in the July 21, 2022, edition of Retirement Security Matters. For more fresh thinking on retirement savings innovation, check out the newsletter here.