Banking App Redesign: Financial Modeling

Financial Modeling User Flows
Financial Modeling User Flows

This project was interesting because, unlike standard banking functionality, it required me to build something with which many people may not be familiar. The participants all had either never used a financial modeling tool, or had used one of which I was not familiar and which formed the basis of their mental models for controls and navigation.

It was a challenge to try to anticipate how users would perceive functionality, where things should live, why they are trying to accomplish the task I set up, and if they would ever use the features I was building in a natural, unprompted setting. 

1. Results: Many testers were not familiar with financial modeling or budgeting apps and tended to have manual hacks that were laborious, such as reviewing all transactions for a month by pulling up banks statements for 8 different accounts (Laura). I believe that good design should not push a feature onto a user. It should rather make obvious a simpler way to accomplish his or her goals. 

Learnings: In order to do this, there should be a useful tool within the accounts that leads to the financial modeling functionality, but only through a tool the user might find useful when reviewing their individual accounts. Discovery can happen after a positive iteration with a useful tool. 

2. Results: Participants were confused about the safe to spend and, occasionally about the savings goal because they had not set those up to begin with. The participants who had little or no experience with financial modeling functionality relied on the wording of the goal and clicking around rather than prior experience. 

Learnings: Because of this, some of my testing felt too contrived. While testing a new product may require a user to perform a task they would not usually try, there should be careful thought in what the task will be and how it is worded, so that it feels natural for the user to explore the product and dicker the functionality – not force themselves through it. For instance, instead of asking users to see if their Spotify promotion ran out by comparing this months bill to previous months, I could have asked them to see if they are spending a lot more on gas recently. 

3. Results: By contrast there were different styles of exploration by testers, which affected the participants’ experiences and the quality of feedback I received. Some were slow and asked me for reassurance and I felt confident in asking them what they thought. However, some participants wanted to click around and explore everything just out of curiosity and to get a sense fo the whole environment. The result was a rapid tour of the app that ended up frustrating one participant in particular after he found the screen that accomplished mission #1, but clicked out of it without taking it in. He spent ten full minutes searching for that screen because he though eh had already tried that avenue.

Learnings: I am still practicing setting up and facilitating a user test. I improved in setting up expectations for how a desktop test would function, explaining pacing, and talking out loud. All of the participants seemed prepared and unfazed by the limited functionality and were great about sharing their thoughts with regular or minimal prompting. 

However, I was not prepared for participants to just want to explore the environment and for those who did that quickly. I wanted to give each of them space to familiarize themselves with the app, but failed to slow people down sometimes. 

Part of the issue was that participants kept wanting to go somewhere familiar to accomplish the mission, such as their accounts. Since I had not anticipated this, I had neither connected the accounts flows, nor built financial functionality into them. Tying back to my first learning, I believe linking the financial modeling functionality in the accounts could be more useful to users as they explore.

In my next round of testing, I intend to be more structured in how I regulate the speed of the test. I will set the user up to know that I will ask them to pause on most of the screens to tell me about them first and then I will keep that as a firm structure.