A “Wicked Problem”: UX Research and the continuous search for framing the problem at Nubank

Complexity and importance of understanding the problem, a task often more important than finding a solution.

people working at nubank

Reviewed by Amanda Legge, Tabata Milanez and Maria Fernanda Lopes

When I joined Nubank in mid-2021, I was assigned as a UX Researcher to the credit card limit policies/analysis team in order to help solve a specific problem faced by part of our customer base: not receiving the limit they desire on their credit card.

In this article, I’ll explore the complexity and importance of understanding the problem, a task often more important than finding a solution.

The nature of “wicked problems”

Rittel & Webber (1973) introduced the concept of “wicked problems” in their seminal work, “Dilemmas in a general theory of planning”. These problems are inherently complex, often born out of other issues and exhibiting a web of interdependencies. The path to a solution is rarely straightforward, making the precise definition of the problem a significant challenge.

As UX researchers, a relevant part of our role is focused on framing problems. This means that we investigate issues that require more depth using different research techniques and, through them, we seek to define with greater precision what is the problem we need to solve and which are the paths that have the greatest potential to solve the problem. Then the product team as a whole will develop solutions, test them, and measure the impact.

If the solution has a good impact, it will be used. If not, depending on the level of confidence we have about the problem framing, the process can go back to either reinvestigating the problem or testing other paths of solutions. It is not a 100% linear process, since you can go back to previous stages, but the process does not usually escape the traditional stages of: discovering the problem, defining the problem, developing and testing solutions, and finally implementing the solution.

The scale of problems varies, ranging from minor to major, and even wicked. While minor problems usually follow a linear process, wicked problems demand a relentless pursuit for better problem framing.

Wicked problem discovery

As part of our commitment to enhancing customer experience, we keep a close eye on feedback from various channels, including social media and support tickets. One recurring problem we noticed was customers wanting to receive a higher limit on their credit card, which ended up being one of these “wicked problems”. Let me explain why.

A wicked problem is a nexus of interconnected issues. For simplicity, I’ll highlight two problems related to our main issue (the lack of credit card limit) that we need to analyze to better frame our primary problem and guide us towards the most effective solutions.

Before we proceed, I’d like you to do a quick exercise: based on your own knowledge, try to identify the root cause of the problem we just mentioned. Ask yourself, “Why aren’t customers receiving the credit limit they desire?”. As you read on, I’ll guide you through a brief discovery and, by the end, you’ll be able to compare your initial thoughts with our findings. So, let’s go!

Problem #1 – Financial literacy deficit in Brazil

The PISA (Programme for International Student Assessment) 2018 survey by the OECD (Organization for Economic Cooperation and Development) evaluated the financial literacy of 15-year-old students from 20 countries to understand how prepared they are for facing common money issues of adult life in their future. Brazil ranked towards the bottom in financial literacy scores, 17th out of the 20 countries included in the study.

General financial literacy results by country on PISA 2018.

Analyzing the results of these young Brazilians based on the 5 levels of financial literacy classified by PISA (“below level 1” being the lowest and “level 5” being the highest), Brazil presented the respective distributions (with the explanation of its meaning):

2% of respondents at level 5 (the highest score)

They can apply their understanding of a wide range of financial terms and concepts to contexts that may only become relevant to their lives in the long term.

10% of respondents at level 4

From this level upwards, they can or have the necessary knowledge to apply in financial contexts that will be relevant for their adult life, such as bank account management, and are able to make decisions taking into account long-term consequences, such as knowing how to identify the overall costs applied to a long-term loan.

20% of respondents at level 3

From this level upwards, they can consider the consequences of financial decisions and can make simple financial plans in their family contexts.

25% of respondents at level 2

Level 2 can be considered a baseline level, below which students may need support in order to answer questions in financial literacy. At this proficiency level, they can present skills that are essential for full participation in society as independent and responsible citizens.

29% of respondents at level 1

They display basic financial literacy skills. They can recognize the difference between needs and wants, make simple decisions on everyday spending, and apply basic single numerical operations (addition, subtraction or multiplication) in financial contexts of their lives.

14% of respondents below level 1 (the lowest score)

They can’t even present basic financial literacy skills.

Percentage of students at each level of proficiency in financial literacy divided by country on PISA 2018.

This information shows a harsh reality: most young Brazilians grow up without having a minimum level of financial education necessary for their lives, which corroborates for them to become adults with a disorganized financial life and inadequate planning for their future. For instance, through the research we do at Nubank, we have already seen people who use their credit card not only as a method of payment (as it should be), but also as a recurrent extension of their monthly income, spending money that they don’t have.

Data from the CNDL (Confederação Nacional de Dirigentes Lojistas, roughly translated as the National Confederation of Store Representatives) and the SPC Brasil (Serviço de Proteção ao Crédito, roughly translated as the Credit Protection Service) show that this is true for 20% of credit card users.

Complementing this vision, research carried out by the same cited agencies showed that 67% of Brazilians are unable to save money throughout the month. Although this fact is impacted by the financial situation of part of the population in lower socioeconomic classes, the research also indicates that, even for classes A and B (the top two socioeconomic classes in Brazil), the proportion of those who do not save anything in the month is 54%. By definition, this means that many Brazilians are living a life standard at the limit of their income.

So, what does this have to do with the main problem of customers not receiving the limit they expect to receive?

Well, as we learned through this investigation, it has everything to do with it…Since by living a standard of living at the limit of their income, these people expose themselves to a high chance of not being able to afford their expenses when unexpected situations happen. And believe me, sooner or later they happen (as “Black swan theory” defends).

In the end, this greatly increases the risk of providing more credit to these people: on one hand, risk for the bank that will suffer a financial loss; on the other hand, a high risk for the customer, who will be in default, living with an open debt that grows over time due to interests, and may even impact other spheres of that person’s life and have an even more devastating effect.

Data provided by Serasa (a Brazilian institution focused on improving financial health) about defaults in Brazil made available in June 2023 show us that this is already a reality. Practically half of Brazilians are in default (44% of the population), and this trend goes upward month after month. It is also not surprising that between basic utilities (like water, gas and electricity), credit card, and retail, credit card is the segment with the highest number of Brazilians in default, accounting for 29% of the population.

Percentage of Brazilians who are in debt, broken down by the 27 states in Brazil.
Monthly evolution of the amount of Brazilians in debt, in Millions.
Percentage of Brazilians with debts in each segment.

Concluding the illustration of the first problem intertwined with our main problem, we now know that, for customers who do not receive the credit limit they desire, part of the problem is rooted in the lack of financial education added to the life standard they end up living.

However, an important question that appears along with this finding is: if it is people’s financial life that is preventing them from receiving more credit, do they also have problems receiving more credit limit from other banks, or is this a Nubank-specific problem?

This question leads us to a second problem intertwined with our wicked problem, which is essential for us to understand more deeply in order to continue improving our problem-framing.

Problem #2 – Some banks offer excessive credit limits

If the origin of the problem were only in people’s financial lives, they would have the same difficulty in getting more credit at any bank. However, each bank has its own criteria of risk analysis to understand whether or not to provide more credit to each customer. Naturally, some banks are more conservative, while others are more aggressive. Each bank’s culture also influences how much weight they place on their customers’ financial health compared to the company’s profitability.

While customers may initially be pleased with a high credit limit, many may not realize the risk this available credit represents. If they don’t manage it carefully – and as we’ve seen, there is a lack of financial education to do so – serious consequences can happen. These range from being blacklisted by credit institutions to accumulating unmanageable debt.

Perhaps some of you might be thinking that the responsibility for the proper use of the credit limit should lie solely with the customers. However, when we consider the data shown above about the population’s lack of financial literacy, it’s quite obvious that the majority of people will not be able to manage their  spending due to lack of knowledge. This, consequently, means that the bank is also responsible, at least partially, for offering too much credit to these people.

By analyzing the chart below, which shows the percentage of customers at different banks in Brazil who were between 15 and 360 days late paying their credit card bill (the chart uses 15 days as a baseline to account for people who simply forgot to pay their bill on time), it is possible to see how the variation from one bank to another is high, reflecting in part – even if only superficially – their previously mentioned loose evaluation criteria to give customers more credit. In addition, it also reminds us that, although the decision to give a high line of credit to all customers is an easy path to increase customers’ satisfaction, it is also a path that has serious consequences.

Percentage of 15+ days late bills on each bank in the Brazilian market. It was analyzed 15+ days as a baseline to account for people who only forgot to pay their bill on time. Source:https://www3.bcb.gov.br/ifdata/#

Sometimes, percentages can feel a bit impersonal, so let’s break it down with real numbers. Taking Nubank as an example, we see that in March 2023, the company had 34 million active credit card customers. This means that every 1% of people who can’t pay back their credit card debt equal 340,000 Brazilians. That’s a huge number! In my opinion, banks have a big role and responsibility here. Their easy-going credit rules can lead to Brazilian families, who don’t know much about managing money, getting into debt because they’re offered high credit limits.

Even though some customers want a higher credit limit on their Nubank credit card, the bank’s strict checks are actually there to keep these customers safe. They also stop the bank from taking on too much risk of people not paying back their debt. It’s a situation where both the bank and the customer win if things go well. However, the lack of knowledge about money and the common practice of offering high credit limits can stop some customers from understanding what the right credit limit for them is, so they don’t end up in trouble.

To wrap up this second problem, we can say that, on top of the issues we talked about in problem 1, customers also have unrealistic expectations for high credit limits. This is made worse by some banks in Brazil giving their customers too much credit, and by customers not knowing how much credit is safe for them to have.

The process of framing is continuous

Now that we’ve delved into two of the problems related to our wicked problem, do you remember what problem frame you defined before reading the rest of this article? Does your hypothesis still hold? Did the data show a path you hadn’t imagined? Through these examples, we hope I’ve demonstrated how the construction and refinement of the problem frame is essential to guide the solution towards the problem we need to solve.

If we only took into account the first problem, we might think that the root cause of our main problem is exclusively the lack of financial education, and could start trying to solve it. However, when we delved into the second problem, we saw that, in addition to the lack of financial education, there is also a misalignment of customers’ expectations about the amount of credit they should receive. And knowing this is fundamental to creating solutions that can really solve the problem.

If you find yourself grappling with a wicked problem and the process isn’t fluid, don’t get frustrated. Although I’ve used linearity to explain the problems investigated, the process of framing a wicked problem is rarely simple or linear. Similarly, the process of my team framing this problem was also not linear.

We started the process with limited knowledge, conducted various research studies, and only after learning more were we able to test solutions. We learned from those attempts and continued to refine the problem over time, feeding it with our learnings to reach a conclusion.

To assist in this complex but necessary process, we recommend using the “Opportunity Solution Tree” framework, created by Teresa Torres and explained in her book “Continuous Discovery Habits.”

Opportunity Solution Tree structure, taken from Teresa Torres article.

Using this framework, we managed to break down a wicked problem down into smaller problems, explore solutions for each of them, and delve into small experiments to validate whether these paths can solve our problem. With the “divide and conquer” logic, the framework narrows the focus area to help solve several small problems over time that, together, will have a significant impact on the desired outcome. This is how we continue to provide better service to our customers. Unfortunately, we can’t share much about the actual solutions because they’re still in progress, but perhaps that’s a topic for a future article.