Nubank
Desigining challenging lending products for a digital bank with over 50 million customers.
(Aug 2021– Current)
About
Apart from the leverage to work wit a lover brand, what really got me into Nubank was its culture of data-informed decisions. One of the perks of working on a lending product is that it requires a lot of funnel analysis and constant A/B testing. Knowing that it was expected that I'd constantly test and iterate my designs was like music to my ears.
Not that I didn't have any challenges, though. I have to balance providing a sort of e-commerce funnel vision with an experience that doesn't hide fees or try to outsmart users. Being transparent is precisely what sets Nubank apart.
So hitting the sweet spot of user convenience and business metrics is the quintessential expectation for my role as a product designer.
My role First and foremost, I'd like to give a big shout out to the amazing team I worked with on this project. Because we were a growth squad, we naturally did a lot of cross-team collaboration. I'm proud I worked with such a talented team of Product Managers, Product Marketing, Business Analysts and Engineers.
As of my role in specific, was in charge of designing the interface and experience of the lending sales funnel. It was expected that I made smart decisions based on business data, which included analysis done by our BAs as well as results from usability tests.
Speaking of which, I contributed on a lot of user research and moderated user testing. Again, we were a growth-driven team, so findings new business opportunities was key.
The product Nubank's lending product offers a few options for users, such as personal loans, secured loans and other new features that I currently am in charge of doing the design discovery.
On my onboarding, I began working on finding low hanging fruits on the the unsecured personal loan which then escalated to other kinds of loans.
Overdraft project
Following funnel optimization, my second feature at Nubank was to design an overdraft feature. That is something that every major bank in Brazil offers, but has a very bad reputation among clients because of abusive interest rates and lack of clarity on limit among other problems.
Prioritization on backlog took place after doing an opportunity sizing with Business Analysts. I was responsible for planning a survey and some guerrilla interviews to find potential use-cases and the BA team took it from there. It became obvious there was room to build a fair experience while resolving a urgent business need.
Apart from being common ground in competitors, it was a featured that was often requested by clients on surveys and interviews. So, in order to expand our product portfolio, my squad decided to tackle this bet.
Data gathering
We mapped from the beginning that a lot of user research and interview was required, so right on my first month the team's UX researcher and I led a workshop session to gather questions to be addressed from PMs and BAs.
Following the workshop session, we asked for a specifically segmented user base that would cover most of the profiles we needed.
Apart from qualitative analysis, the BA chapter helped us analyze a lot of data on a deep dive and things got quite business-driven, so I can't really disclose numbers here. Discovery
Problem 1) Clients took out light loans and found it hard to manage
After updating our credit policies and allowing clients to have more than one loan active at the same time, we observed a considerable chunk of our clients hired many loans within a short period of time. Because it's easy to hire and have money right away, around 15% of the active base had at least 3 current loans simoutaniously. With different due dates and many other financial responsibilities, we saw an big spike on risk giving clients were not keeping up with their loans as much.
Problem 2) Clients credit scores were getting lower and lower and many didn't even know
Having more than one loan active at a time definitely lowers down credit scores. Problem is many clients were not aware of that we reported them automatically to credit agencies. They felt Nubank wasn't fair to them, especially those who paid every installment on time. By the end of 2021 we saw our product NPS get lower.
Problem 3) Light loans NPV was very low and eventually made clients ineligible
Making clients ineligible not only hurts our NPS but it also made it hard for us to achieve for our recurrence goals. On top of that, NPV (which is basically measures how "worthy" it is to borrow someone money was low compared to larger ones.
Main findings from guerilla interviews
Users see overdraft as a last resort when their financial life needs an uplift. They don't want to count on it for life nor does Nubank. Because of risk, we wanted this to be more of a simple loan rather than an addicting practice that people would use monthly.
Previous research had shown borrowing money from a bank was a reason for many users to feel shame. Financial issues are a delicate matter and it deals directly with people's emotions. So, cutting to the chase, we clustered interview results in three categories, as below. Then, I mapped out opportunities based on the connection they felt with product.
Heuristic evaluation
After gathering data from the design discovery, I set out a couple of workshop sessions to evaluate the legacy flow to understand how we could promote overdraft in the interface.
So in collaboration with other designers to do critiques and then evaluate the experience using Nielsen's 10 Usability Heuristics. Below is the main screen of the old flow.
Tweaks on the sales funnel that led to experiments
New loan available
We figured this area took way too much space in the viewport (more than 1/3) but still did not carry a lot of useful information.
No way to personalize limit whatsoever.
Limit available
You don't have any active loan
There's no need in giving this sort of negative message which can come across as error.
Users mentioned they were afraid a loan was gonna be hired as soon as they clicked the button, which wasn't the case, since they'd have to go through the funnel to hire. So it was misleading.
New loan call to action
Delivery
After critique sessions and ideation phase, I got to design the new product. There was a huge expectation on launching overdraft given it was a feature that would not only boost metrics such as conversion rate, amount lent and unit conversion but also users really wanted to use that feature at Nubank.
Lending homepage revamp
We tested this and found users feel comfortable knowing they can turn it on and off at a glance.
Loan progress transparency
One of our main tenants as a brand is to provide a transparent experience, especially when it comes to finances. Making users overspend is bad for them and for business since it increases risk.
Overdraft settings
We made it easy for users who were eligible for overdraft to deactivate the product and change limit amounts whenever they felt like. All they have to do is press a switch button.
Editable overdraft limit + Switch to give a sense of control
Overdraft details
We asked customer service to gather the most requested information and showed them here to handle user's objections
Objection handling
Results: usability was okay but perception was pretty disastrous
We ran a couple of surveys before jumping into guerrillas and usability tests. They had promising results as users mentioned they were interested in having quick access to light loans.
Following interviews and usability tests, however, show the perception around the product was pretty bad. Some key insights:
1) Users consider harmful having an at-a-glance overdraft button
We knew from the beginning that overdraft is a practice done by traditional banks with abusive interest rates and lackluster transparency. Although we avoided all that, clients still understood the product as something potentially harmful to their financial life. 2) Clients asked us how to make button disappear not to fall into temptation
Allowing clients to quit the product altogether was definitely a blocker for us. We discussed deeply this user pain with stakeholders and one of the major concerns was drop in NPS and lack of trust in Nubank – especially considering it is such a popular brand in social media and relies on word of mouth for acquisition.
3) "Special check" as it is translated to Portuguese is neither special nor check
Although we came up with several value propositions, none of them seemed to cover the use-case we needed. In fact, many clients were upset that there was no benefits to them, as they'd rather to have small loans and quit them altogether as soon as they had access to money. In sum, overdraft did not solve their problem.
We took a step back to focus on the core problem: management
We quickly got back into our problem statement and figured we should focus on addressing the needs by providing an easier and more robust management experience rather than necessarily offer a new product.
Sometimes all a client wants is to avoid complexity
Overdraft adds a new level of complexity to a subject that's already sensitive: financial decisions. In the macroeconomic scenario we are in 2021, financial responsibilities are one the main drivers for anxiety, and overdraft seemed to worsen it rather than help.
It was surprisingly painless for me as a designer to convince stakeholders that the optimal decision was to focus on management of a loan.
Pros of pivoting to Loan Management rather than create a new product:
1) Decrease risk
By designing a post-hiring experience in which it is easy to pay for a loan, with reminders and visual metaphors;
2) Faster learning
More agile approach with quick iterations over crucial screens such as dashboard;
3) Creation of roadblocks to prevent debt
We came up with a series of blockers to the policies that mitigated clients getting in debt.
Loan management research
We ran a series of interviews to find out what would happen clients keep on being on-time. We began internally, running workshops with stakeholders and fellow designers to come up with questions to be asked to our clients.
As we were running out of time and risk was increasing worldwide, talking to users about Loan Management became our top 1 priority.
Hypothesis
Users are looking for payment flexibility to get their financial lives on track.
So let's keep it lean yet effective.
New products add more complexity while payment features is what makes difference.
On phase 1 of discovery we found out it is common for users to hire a loan to pay for another loan. This gets them into an infinite snowball of debt, so giving options for them to refinance their loan gives them more flexibility while gives business more predictability.
Editable loan cover image
We found on user testing this causes an emotional connection to the product, which leads to clients keeping up with their payments
Highlight open balance as a compelling event
Create a sense of urgency
We tested a bunch of fake door actionable items for very small samples to have proxies on how adoption would be if we offered such features.
The importance of quick testing with fake doors
As soon as we realized these concepts fitted users better and would still generate more revenue to Nubank, we were eager to actually test features in production. Creating buttons with fake doors to test adoption was crucial for us to decide which option to focus on.
1) Borrow more money generated huge interest
More than 1/2 of clients who viewed our dashboard in production got interested in clicking in "borrow more" feature, which led us to design a funnel dedicated to this.
2) Satisfaction was much better
Clients understood the features we were proposing as a way to sort of "create a relationship" with the bank. One in which they'd show responsibility with their money and the bank on the other hand would provide fair interest rates and conditions.
3) Clients score increased over time
With the roadblocks we created, we saw a significant impact on the score the system automatically reported to credit agencies. Although this is something that happens in the underwriting process – that is – clients don't really get to "see" this benefit, it allowed us to leverage new and more responsible loans in the future.
