Fidelity - Moving From The Mainframe

Fidelity Investments was making some big changes, and I was along for the ride.

Starting off with a disclaimer: My work in this area is protected by nondisclosure agreements, and as a result, I cannot show my work in graphical format or talk about it in detail. That being said, I will do my best to outline my procedure and research in this project.

In 2008, I got a call to come interview for a very interesting project. Fidelity Investments was making some big changes, and needed someone with expertise in web development and specifically user interaction. At this time, the specialization of fields into UI/UX was really just getting started in a formal way, so this was a great opportunity to expand into this newly developing field.

The Challenge

This project is a bit hard to describe in generalities, but we'll go back to the beginning. If you are familiar with the development of computers, and especially ones which were built to serve large datasets and very large corporations, you'll know that a lot of these types of institutions had most of their information stored on mainframes.The early years of the new millienium saw companies more willing to experiment with web-based environments, and Fidelity was no exception. For this project, the challenge was to take a thin-client appplication which was in use company-wide and translate it into a web based software. All the while respecting the users who were very attached to using this client-side application, and felt any change would only slow them down! Understanding the user experience was key to succeeding.

The Process

Obviously, user research was the first place to start! This was a bit difficult, since the company's branches were so numerous and there was, at the time, not really any central repositories of usable data about their behaviour like there would be for a website, for instance. With a website, you can collect a lot of data in a very short time about how people are using your site. Visits, clicks, time spent, and the like are all very easy to find and begin to analyze. Sadly, this type of data was not present and collectable for this software. So there were many hurdles to start with in terms of building objective datasets. Turning to some older methods, I settled on user surveys, ride-alongs and interviews to begin to understand how people used the program.

The User Experience

All of the data collected pointed to the fact that their users were attached to this software because they were very familiar with it's workings. There were many shortcuts and keystrokes that helped with efficiency, and because it was simple, it was easy to understand and use. However, there were limitations and shortcomings that would actually improve once the data was brought to a web-based environment. More powerful searches were be a huge bonus. The ability to perform their own analytics on the data collected was another, and finally, the presentation of data in a visual dashboard that would allow them to react to and the act on the data in real-time instead of waiting for reports to be processed was the biggest selling factor.

The Lab

One great asset while I was involved with this product was that I was allowed use of the formal User Experience lab that the company used for their web-based research. Since we were bringing Windows software workflows and processes into a web environment, I was able to design screens and perform tests with actual users of the product. This experience was a great introduction to a formal user research environment, and I definitely learned a lot in the time that I spent there. Ultimately, I was able to achieve high satistfaction marks even from the most skeptical or critical of users. My time at Fidelity was a true adventure and it still ranks among my top personal career achievements.