We created a transparent consumer banking application that utilizes learning and prediction to provide better service. We spent eight months researching, designing, prototyping, and user testing our app in the Portuguese market. Our client was Exictos, a banking software company. Echo lets you master your finances.
For sure this will be a piece of work we are going to introduce into our roadmap of products. It was nice to see the way you understand financial service on channel solutions. We are going to merge this with the research we are doing in-house with machine learning and artificial intelligence. [It] was a pleasure.
– Pedro Camacho, Head of Development Banking Channel Solutions at Exictos
Exictos, a bank software leader in the Lusophone (Portuguese-speaking) market, approached our team, Elementary, to create a contextually-aware consumer banking experience. Preferably, something that can be highly adaptable to the various users that use Exictos products.
After four months of user and market research, we found that consumer banking applications are opaque, meaning that users don’t really know what is happening. When people look at their finances, they are asking three questions:
Instead of helping their users, banks (often) only provide a list of transactions from which users are supposed to intuit their financial health.
In response to the opaque market and our user research, we created a transparent consumer banking application for the Portuguese market that utilizes learning and prediction to help people understand their finances.
Our work was split over two four month periods: one of research; one of designs, prototypes, and user testing.
Project Manager: I created the structures and systems that allowed us to work efficiently and effectively. These included thought structures for thinking about and explaining our project, Google Ventures’ Sprint process, and how the team functions (everyone leads one aspect and contributes to the other parts). I facilitated our critiques, leading the team to effective outcomes.
Prototyping Lead: I utilized my previous front-end and prototyping experience to build our website, hi-fi Framer prototypes and experiments, and wrote the design suggestions for our machine learning algorithms.
Elementary is a group of four master’s students in the Carnegie Mellon University and University of Madeira Master of Human-Computer interaction dual degree program. We are a multi-disciplinary team with backgrounds in the humanities, psychology, and computer science. When working, we each take the lead on one aspect of the project, then collaborate to completion.
Exictos (our client), formerly known as Promosoft, was founded in 1989. They specialize in the production and implementation of core banking software, serving more than 60 banks in Portuguese-speaking countries. Exictos expanded their offerings to include both mobile and online banking software as a layer over their core platform. In 2015, Exictos was acquired by the Polish Asseco group to strengthen its presence in Africa and to open up new opportunities for expansion in South American markets, especially Brazil.
Our research spanned our first four months of work. To guide our investigation into Portuguese consumer banking, we created a series of core questions. Each question lead us to a research method.
To interpret our data, we created models including a flow model, sequence diagrams, a cultural model, and a competitive analysis grid. We also printed out the photos and captions users created in their photo diaries. Everything went up on our walls so we could be immersed in our research during the later design phase.
Concluding our research phase, we compiled a short list of critical insights made from our research data. These are the most relevant insights.
Users strongly distrust banks and their intentions, often taking steps to obfuscate their purchasing habits.
Users make conscious decisions for how they will pay for things, taking control of their money use. users often have to or choose to go through multiple steps in order to achieve their goals.
Banks are not providing services in a way that pleases their users. When customers have problems, they are unlikely to contact the bank about it. Instead, users get frustrated and begin looking for new banks. This is especially true when banks charge “maintenance fees”. Without adequate communication, users wonder what these fees are for and suspect that the banks just want to take their money.
When users want to know where their money is going, they track it. But, the tools provided to them by banks are insufficient. our data shows that most users do some form of expense tracking or budgeting. Excel and mental-tracking are the most popular forms of estimating budgets.
Often aware of savings options, few users have savings accounts or other investments. In interviews they remark that they “should have an account, but don’t”.
In order to create a transparent consumer banking application, we created a five part vision that guided our design work.
By continually providing value to users, banks are more likely to retain customers who are less likely to get frustrated. We are providing opportunities for customers to truly understand their finances. We are doing so in ways that no popular bank has, setting our client, Exictos, and their clients apart from their competition.
Based our research, we understood the user’s core questions: How am I doing? How will I be doing? If something changes, how does that affect me? In our last four months of work, we answered these questions by building Echo.
We explored each part of our five part vision in four week-long design sprints (our fifth part was explored throughout). With the user feedback, we were able to condense what made sense, get rid of what didn’t, and update what needed improvements.
In the time after our sprints, we continuously critiqued our work, iterated on it, and tested with users weekly. At the end, we presented two hi-fi prototypes: one showcasing visual design and another for interaction design.
Transparency Users have questions about their transactions, groups of transactions, and patterns. We made it easy to answer these questions.
Foresight We give users just enough insight into their future so they can be prepared. These insights are based on the user’s previous actions.
Relation of user to others When you’re doing something for the first time, you don’t know what to expect. We allow users to anonymously learn from each other by showing them how “other users like you” have acted.
As the project manager, I implemented the Sprint process as outlined by Jake Knapp, John Zeratsky, and Braden Kowitz. We worked through four week-long sprints which each include complementary research, sketches, a prototype, and user testing. It is a process that uses divergent and convergent thinking, harnessing the strengths of each member of the team. In this phase, each prototype is independent from the others.
After our sprints, we had near-daily group critiques and continued our weekly user testing. Most of this was on our home island, Madeira, but we also spent a week testing in Lisbon.
Our prototypes were built off of our research, constructed in individual brainstorms, synthesized in a critique, and informed weekly from user feedback sessions.
We built two hi-fi prototypes: one for visuals and one for interactions.
Framer experiments In addition to our primary work, I also made small experimental interactions in Framer.
Our final product, Echo, allows users to master their finances. The information architecture is centered on three pieces: transactions and categories, goals, and a financial simulator. Additionally, we let our users search the system with a fuzzy search (not shown here).
I want to feel in control of my finances.
– User interview
Wow. It’s good it shows [your available balance and expected spending], ‘specially if you are struggling to make ends meet.
– User feedback
When a user opens the application, they're greeted with their historical and upcoming transactions, as well as their balance, expected spending, and critical insights.
I want to save for my wedding dress. I’ve always dreamed about it.
– User in our photo diary study
When is this application out for download? Sell it to [my bank,] Millenium!
– User feedback
Goals act as savings buckets, meaning that users can save towards different objectives such as a new vehicle, a trip, or paying off a loan. Goals are flexible and can be paused, money can be transferred in and out, and goals can be continuous. Perhaps most importantly goals create achievable outcomes based on existing mental models.
Money virtually and automatically transfers between the user’s main account and goal account to keep a distinction between allocated and unallocated funds. Their money is automatically balanced between a current/checking account and a high interest savings account based on the user’s habits.
I want the bank to help me understand my habits and their implications.
– User interview
Great, I like this part! This is good for me because of my house loans, I’m still paying for them.
– User feedback
When we think about what might happen to us, we wonder how it will affect our finances. The simulator section allows users to explore these possible futures. Similar to goals, we use minimally structured questions to help users get a meaningful answer without bogging them down with input. We also give them contextual information to inform their decisions. Users may star their entry to save it for future use.
Some simulators can be found on the web and others can only be accessed when sitting down with a banker. Echo democratizes financial simulation by empowering users.
We delivered two final prototypes, one to show interactions and a second to show visuals.
This material was adapted from work my counterparts and I completed for Exictos and the Madeira Interactive Technologies Institute.