story

Behavioural economics in the real world: Three key lessons

By Garrett Meccariello, Tobias Nasgarde and Perjan Duro
Posted on April 18, 2019

As you may have read in a recent interview, the Think Forward Initiative (TFI) offers financial support to a promising collaborative field research project led by Perjan Duro (MoneyCoach), Garrett Meccariello and Tobias Nasgarde (Prospect Labs). A major hurdle has been taken: the researchers now understand how to implement their research design into an actual app.

The project title – ‘Nudging healthier financial behaviour’ – sounds straightforward enough. The researchers aim to uncover what types of messages or alerts are the most effective in convincing the users of MoneyCoach. This personal finance app, that has been downloaded by over a million people, helps people to set short-term financial goals for themselves.

But untangling how and why people engage in certain financial behaviours, and identifying actual ways to jog them in the right direction, is easier said than done, the researchers discovered. ‘When we set off, our understanding of the literature that surrounds our theories was very solid,’ Garrett Meccariello says. ‘But we had to learn what to take into account when designing field experiments in partnership with other-than-research stakeholders. This is Research 2.0, a whole new league.

To help you appreciate the way Research 2.0 works, and perhaps even to boost the way other researchers might design their experiments, the team formulated three key lessons:

“​ KEY LESSON 1 – When collecting data, team up with the best.”

When attempting to control for all potential variables in the study design, we were not properly aware of how our requests built metaphorical barriers for the MoneyCoach product design team.

Collecting personal data is a hassle these days, with the all-seeing eyes of the European GDPR regulators everywhere. ‘Simple’ data points (age, gender) must be cautiously derived and handled. Product developers are not as fortunate as researchers, who have the luxury of compensating participants of their lab or survey-based experiments.

This challenge required a unique solution that would not have been an option without the MoneyCoach team. Without the team’s knowledge of user experience design and the extrapolation of demographic data, our intervention would have required users to provide their information explicitly. Not only does this approach feel unnatural, it would also have acted as a potential barrier to our original intentions.

“KEY LESSON 2 – Don’t overcomplicate; think Research 2.0!”

The process of conducting research is known to be time-intensive, and slow-moving at best. Compared to the speed at which product developers usually operate, the rather methodical pace of our initial approach (Research 1.0) hampered our ability to theorise, design, and test interventions in a scalable manner.

We spent so many hours discussing the appropriate framing of our behaviour changing messages. We talked and thought and thought some more. Even though we were being the most productive researchers ever, we were actually falling deeper and deeper into the rabbit hole known as ‘Research 1.0’. Endless thinking may make many researchers feel safe and content, but the truth is that any thinking work should be cut up in small doses, neatly distributed throughout the process.

We now believe that ‘Research 2.0’ should be the gold standard for changing behaviour at scale. Rather than spending dozens of hours conducting in-depth literature reviews, formulating pinpoint hypotheses, and zinging various data points, simple theories can (and should) be tested as frequently as possible in mini-evaluations. Rest assured that product managers, who want to influence healthy behaviour without overcomplicating the approach, will be thrilled.

“KEY LESSON 3 – Be a true team: enjoy each other’s strengths”

Anyone who has worked in a cross-functional team will say that patience and understanding are critical to work together successfully. When experts of various fields come together to solve a common problem, new viewpoints are uncovered, that may lead to alternative solutions to problems – even problems previously thought to be unsolvable.

As soon as we realized that our research team lacked an understanding of user experience design, we got in touch with the MoneyCoach team to create a seamless intervention best suited to change behaviour. And it worked both ways: the guidance provided by the research team complemented the developers’ skills.

These three lessons may have changed the way you think about research. They certainly helped the research team to translate the power of rich academic research into a more agile, effective way of changing behaviour. ‘As we continue to roll out our interventions to the MoneyCoach community, we are now identifying the most effective message to increase savings behaviour,’ Garrett Meccariello says. ‘The success of our tests will ultimately inform the iteration shown to all one million plus users.

Research 2.0 just isn’t about slow journal publications, but about immediate behaviour change. Whilst each experiment should still be conducted with the same standards and integrity as before, the 2.0 approach may yield the most actionable insights.