Starting a new business is always challenging, and one of the main factors is keeping up with your finances and with the government. We have the most used app for small businesses in Brazil, and there we used to provide a troubling experience for them to pay their taxes, causing a lot of confusion and friction.
As a product designer, I was part of a project with the goal of simplifying the payment of taxes for small businesses by redesigning the tax flow of Neon B2B, an app with more than 1 million active users who pay their taxes every month, where we increased the product’s retention by more than 10%.
@2020
I have omitted confidential information in this case study to comply with my non-disclosure agreement. I did a free translation of the texts on the screens from Portuguese to English.
The Tax Payment flow was the main reason for users to use our app because they needed to pay their taxes monthly, and because of that the Tax Payment flow was the most important product in terms of retention.
We noticed a decrease in the number of users accessing the Tax Payment flow and an increase in complaints. The UX of the Tax Payment flow had several usability problems, and with research and data analysis we found out users were feeling anxious because we were presenting outdated payment statuses, and that 50% of our users had past-due taxes in previous years, probably due to the lack of knowledge about their tax obligations and because our UX was failing to communicate this.
To tackle the issue, we went through a redesign that included automatic updates of payment statuses, a history of past-due taxes, and other AI improvements. To ensure that the redesign was successful, we conducted several rounds of design critiques with other designers and usability testing with active clients.
The results of the usability testing were positive, with users finding the tax list to be more organized and the past-due taxes more visible.
The redesign of the Tax Payment flow led to a 10% increase in retention, successfully achieving our goal.
The Tax Payment flow was the main reason for users to use our app because they needed to pay their taxes monthly, and because of that the Tax Payment flow was the most important product in terms of retention.
But something odd was happening:
Throughout 2020, we saw that the number of people accessing the Tax Payment flow decreased, in addition to a significant increase in complaints about the product, ranging from duplicated payments to not knowing that they had past-due taxes to pay.
We went from having 83% of all active users paying their taxes in January 2020 to only 71% in September 2020, so it's a huge decrease.
We thought about two main reasons why this has happened:
1. The 2020 economic instability due to the pandemic led a lot of small businesses to lose their source of income for a long period, so they weren't able to pay their taxes;
2. The UX of the Tax Payment flow had lots of already known usability problems.
Let’s deep dive into the UX of the Tax Payment flow.
On the main screen of the app, you could see the last 2 taxes and their statuses, but the thing is: the statuses you’re seeing may be out-of-date.
This happens because the app had no direct communication with the government to present the updated statuses automatically. To update them, you needed to access the Tax Payment list and tap on the “Update payments” button. This way you would send a backend request to search for new payments.
Every update request had a cost to the business, so updating the statuses wasn’t free. This means that updating all users in our user base wasn’t a scalable solution, and this was a technical constraint that we needed to work within.
In conclusion, if the user is checking their payment status only through the main screen, they’re probably seeing outdated payment statuses.
Showing out-of-date statuses wasn’t the only problem. Some other problems were:
With all the constraints and usability problems in mind, we defined our main goal:
We want to increase the number of active users of the Tax Payment flow by providing a seamless experience for them to pay their taxes.
To guide the next steps of our research, we gathered the team (PM, engineers, QA), stakeholders, and customer support to use a framework called CSD Matrix. This framework helps to outline and visualize what everyone involved in a project knows (Certainties), hypothesizes (Suppositions), and doesn't know yet (Doubts), which are the 3 columns of the matrix that we filled together.
The CSD matrix encompasses various types of knowledge, including user needs and behaviors, as well as organizational factors such as goals, processes, technology, and constraints.
With the matrix ready, I created a new board and put everything that was written on post-its to analyze what would be the best way to explore the doubts or validate the suppositions, separating between Qualitative Research, Quantitative Research, Analytics, Usability Testing, and others.
I’ve found out through desk research and talking to our accountants that paying taxes regularly has many benefits to the entrepreneur such as retirement pension, paid maternity leave, and so on. But, there are also downsides to not paying the tax, such as the interest rate for past-due taxes, and government debt which can even lead to revoking their business license.
With the help of our Data Analyst, we’ve discovered that:
They pay their taxes online
In 2020 the option to pay the tax through mobile banking has surpassed the option to pay it in person.
Hypothesis: With the pandemic in 2020, users who were used to paying their taxes in person started to pay them online through digital banking apps.
They’re not paying their taxes in advance
Less than 1% of our users paid their taxes before the month of the due date.
Hypothesis: This means that paying taxes in advance is not a reason why the number of active users has decreased.
They have past-due taxes from previous years
Half of our active users had at least one past-due tax in previous years.
Hypothesis: The current UX of the Tax Payment List was failing to show past-due taxes in previous years, and this made them not aware of being in debt.
Users are closing their businesses with pending past-due taxes
Almost half of our users close their businesses without paying past-due taxes.
Hypothesis: Besides failing to report past-due taxes in previous years in our UX, we also fail to communicate the consequences of not paying them, such as high-interest rates.
Based on the insights we got from the discovery so far, we defined the user interview goals:
Learn how users pay their taxes today, if they’re aware of benefits and losses related to paying taxes, and how they check their payment status.
I created the interview script and together with our UX Researcher we decided on the user profiles to recruit:
Small business owners, separated between 1. New businesses (under a year) and 2. Older businesses (over a year), so we can extract the different perceptions they can have based on how long they had their company.
The pandemic made them learn how to pay taxes online
Complementing what we saw in the data, they said they used to pay the taxes at the bank branch, but with the pandemic restrictions, they needed to learn how to pay them online with mobile banking — which they successfully learned.
Presenting outdated statuses made them feel anxious
Users mentioned that after paying their taxes, they feel anxious waiting for the app to show the updated payment statuses because they’re afraid of being in debt and what consequences this could lead to, such as revoking their business license. We learned that we can’t keep relying on manual updates because we have a great emotional responsibility to them.
Outdated statuses could lead to higher churn risk
They said they would lose confidence in our product if we keep showing outdated statuses, and consequently, it would lead them to uninstall the app.
Little knowledge about the benefits of paying the tax
They knew they could have access to retirement pensions if they paid their taxes, however, they didn’t know how the process to obtain the pension works.
They didn’t know their tax obligations when they started their businesses
Some users stated that they didn’t know that they needed to pay taxes when opening their business, and this made them indebted over time — their past-due taxes now have high-interest rates.
I looked at the biggest competitors that offer tax payment and discovered that so far, we had almost no advantage when compared to them, and all services were very similar, even in their problems. They didn’t provide real-time updates of payments or an easy way to view all past-due taxes.
After the discovery we defined two macro-problems that we have in our current flow and prioritized them in order of relevance, according to our research and data:
We are showing incorrect payment statuses
We learned from the research that users are feeling anxious about the incorrect payment statuses, and that presenting incorrect statuses could lead them to churn — we cannot keep doing this if we want to retain our users.
A large portion of the active users had past-due taxes in previous years
With data, we learned that 50% of our active users were in debt, which means that they’re accessing their Tax flow every month to pay new taxes but somehow are not aware of these forgotten debts, which are probably from the time they started their business since they didn’t know their tax obligations at that time. We need to change the UX to highlight these debts to avoid high-interest rates and even leading to revoke their business license.
At Neon, we had a Design System based on Material Design, and I used it to create our new tax flow.I created several versions of the screens and did 3 rounds of design critiques with other designers to get improvement insights and vote on which version was the best. Each round generated a lot of comments and together we co-created ideas.
After the third review, I iterated the screens and showed them to the development team to understand the technical feasibility and gather more feedback. Finally, together with the PM, we decided which would be the final interface to be taken to usability testing.
With the usability tests finished, I did the last iterations and decided on the final screens to go to MVP.I used Neon's Design System to create the screens, which is based on Google’s Material Design.
We’re now updating payments automatically. It was technically feasible and financially viable for the business to prioritize updating users who accessed the Tax List screen, this way we wouldn’t need to update the whole user base, only the ones that tapped to go to this screen.
Together with our UX Researcher, I defined the task success and the usability testing goals, and finally, I recruited active clients to participate in the test.
After validating the prototype with users, I designed the entire screen flow with business rules and edge cases, where over 15 new edge cases were added.
I found error messages that were being communicated as generic errors such as “Oops, something went wrong” instead of providing honest feedback to the user about what happened and what they can do about it.
The flow became bigger and better mapped than before.
As of this writing, the redesign of the tax flow has had a positive impact on our users and business: