The Unfulfilled Promise of Behavioral Science in Fintech

Let’s take a look at the poster child for beh sci in fintech: Lemonade, an insurance company that was one of the first to incorporate behavioral science principles into its operations. This integration was largely under the guidance of Dan Ariely, a prominent but now controversial figure in behavioral science due to allegations of data fabrication in his research.

Lemonade used behavioral insights to simplify the insurance process, allegedly encouraging integrity in claims through tactics such as the Honesty Pledge — or making the user append their signature before answering qualifying questions (for example, how many claims they’ve had). Even Lemonade’s go-to-market press releases revolved around the now-debunked Honesty Pledge, which, on the surface, didn’t exactly bode well for them. Just take a quick look at their stock price, which has lost over 76% of its value in the last 5 years. With this fiasco in mind, it’s hard not to think that all the talk about beh sci’s potential impact on the sector was just that — all talk with little substance. 

Of course, this is only one company, with its own unique set of challenges, earnings, ardent supporters, as well as short sellers. But let’s dig a bit deeper: Lemonade’s headcount has increased close to 20% since its share price peak in 2021, and according to InvestorObserver, “its customer count jumped to more than 2 million while its gross earned premium rose by 27%.” Not too shabby. Just as the much-publicized honesty pledge is just one feature among many others that make up their product and service offering, we know that there are many sides to this story. 

We see it often — a new product is not hitting the benchmarks leadership wanted in terms of new sign-ups. Or maybe a new service that has recently been rolled out is not being upsold as much as expected. At that point, the organization brings in a behavioral science company to run a diagnostic and propose new ways in which they can solve the problem. The crux of it is that until now, a vast majority of the situations in which behavioral science has had the chance to make its mark revolve around these or similar problems. This is in collaboration with the multiple “state of fintech” reports coming out of organizations that overwhelmingly focus on "initial user adoption" or "uptake of premium services." 

Predictably, the go-to solution for practitioners for such problems usually revolves around behavioral interventions related to improving trust or perceived value. This goes along with strategies like framing financial information in a more engaging manner, leveraging social proof to encourage sign-ups, or creating urgency around limited-time offers.

We want to be clear that these problems — encouraging users to engage with the product, get through an onboarding flow, or check out other relevant services that are part of a package — are, in most cases, the most critical ones! We know this since we’ve helped build products from the ground up. If these issues are left unaddressed, companies or entire product lines can die quite quickly. That being said, what, in our opinion, is a disservice to the field and what explains the somewhat lackluster results or the half-fulfilled promises is that practitioners who are brought in often try to attack these issues directly. 

More often than not, the research efforts practitioners use for assignments — including studies, interviews, and literature reviews — delve deep into the behavioral science itself. However, the problem is that more often than not, this approach doesn’t fully grasp the rich, contextual tapestry of the product or service in question from the business or operational side. Without this interdisciplinary understanding, interventions that are recommended by behavioral science consultants or experts often either do not see the light of day in the company (as they are not fully anchored in reality) or produce results that are underwhelming. While the traditional research efforts mentioned above yield valuable insights, they often lack a holistic understanding, merely scraping the surface of the product or service's historical evolution and its complex interplay within the broader business reality. 

This oversight, however, is not entirely the fault of the practitioners. The onus also falls on the organizations and their leadership, who often make two (what we consider to be) mistakes. 

First off, when companies look to try a “behavioral science” driven solution, they look for consultants that complement their own team’s expertise. They do not see the use in bringing in an outside team composed of diverse expertise such as a behavioral scientist with a PhD, a product manager, and an analyst — they want as much “science expertise” per head as possible! Just make them work with your own product manager and your own analysts. They’re already getting a salary, right?! 

However, what really brings in game-changing results is the integration of behavioral science expertise with the rest of the disciplines. Think about it. The behavioral scientist understands human behavior and decision-making processes. The product manager contributes insights into the product's functionality, user experience, and market fit. The analyst brings in the data perspective and understanding of trends, user behaviors, and performance metrics. Together, they can holistically assess both the scientific and practical aspects of a product or service. But here’s the unfortunate reality: it’s extremely rare that this synergy can be achieved between the client's team (of product managers and analysts) and the external consultants.  

The second mistake that leaders make is bringing in experts with an extremely constrained, problem-specific focus. This approach tends to seek immediate, surface-level solutions rather than committing to a more integrated, thorough methodology that would consider the complexities and nuances inherent in the problem at hand. What we’ve seen is that such a narrow lens does a disservice to the potential of the interventions, relegating it to either minor adjustments or major overhauls unanchored in various realities (like technical, cultural, or cost-related ones). 

We’ve now seen that there’s more than just one party at fault, whether that be the behavioral scientists hired or the organizations hiring them. So what can be done to move towards unlocking the potential of behavioral science (or whatever name it will take in a few years) in fintech?

Start from the core mandates 

This one is easier said than done. When we talk to executives, whether they are at a top bank or a niche fintech product company, all they essentially want to do is solve the problem at hand with as little overlap as possible with other areas of the business. 

The reality is that unless the practitioners that come in are granted a broad enough latitude (not just to understand user behavior but to delve into a business and organizational context), whatever they come up with will have lower chances of succeeding or even seeing the light of day. We’ve seen some great ideas from competitors, such as 100 slide decks that are well-researched and backed by data and experiments, meant to be presented with great fanfare to teams. We would only come to find out three years later that the company hadn’t even started the implementation process, as the ideas, although great, were so uncoupled from reality that they had to be abandoned altogether.  

Leaders who bring in these teams need to do whatever they can to enable practitioners to align their insights and interventions effectively with the company's long-term strategic goals, market position, and most importantly, operational nuances and limitations. Practitioners must fully understand the business's core mechanics, how their teams work with each other, and the history of the market challenges and successes that a product or service has had. Only then can practitioners craft solutions that resonate deeply with both the company’s objectives and users’ needs, not to mention ones that are actually… implementable. 

Multi-disciplinary teams

I know… I know. This might sound a bit preachy, but the reality is that when people work across disciplines and closely together, they manage to achieve stunning results. Say what you want about Elon Musk, but one strategy that he successfully used across his companies, and the one that is credited with getting SpaceX off the ground, is forcing his engineers and designers to work very, very closely together.

Secondly, when enlisting the help of behavioral scientists and consultants, leaders must prioritize hiring professionals who possess dual expertise: deep knowledge of behavioral science and an equally profound understanding of the business domain in which they operate. These experts should be adept not just in the theoretical aspects of behavioral science, but also in applying these principles in a specific context, such as product design, service design, or finance. Oftentimes, this means that the most impactful teams are not going to have three (or sometimes even one) PhDs as part of them. This can be surprising, but we’ve been called numerous times to “unlock” a team of PhDs that were stuck in 6 month research projects with little to show, especially if they were brought in directly from academia, where rigor trumps many other things. 

Focus on long-term user needs

We’ve seen that practitioners themselves must adopt a holistic perspective and collaborate in multi-disciplinary teams. The synergies resulting from working across disciplines in the same team are hard to beat compared to when an outside scientist is paired with an internal company team member. 

However, while it's crucial for fintech companies to succeed commercially, they will ultimately fail if the users’ long-term interests and preferences do not align with the behaviors they are encouraged to adopt. As a social enterprise, this is something we hold dear above everything else. No matter what product, service, or intervention is proposed, it needs to involve a sincere commitment and a corresponding execution to genuinely enhance the financial well-being and decision-making capacity of the users — and in a way that is fully transparent to them.

We’ll continue to unpack this topic in the next article in this series! Stay tuned, and if you have any comments, suggestions, or things you have beef with in this article, do write to me at dan@thedecisionlab.com!