Modernizing a Corporate Banking Cash Ordering Platform

Our product enabled companies to order hard currency via an online platfrom. We made the decisive move to modernize the platform after years of users enduring our legacy experience. Paralyzed by what to include in the revamped product, I facilitated workshops to define an MVP that I tested with users.

We had years of anecdotal feedback from customers, and I had run interviews with customers when I joined the team. When the decision to modernize was firmly made, we were overwhelmed by how to start: we didn't know how to evaluate which user needs were most critical and weigh the complexity of supporting those needs.

My approach was to standardize all of these inputs to make them actionable

I rallied the team into an MVP workshop. The framework was:

• Standardize the user data
  Convert the user feedback into a standardized statements, e.g., As a user I need to import a file to place orders, and As a user I need reorder the same order multiple times.


• Define criteria
  Product managers, dev, and UX met as groups to define primary and secondary critera to evaluate the standardized statements. The teams' criteria were formed as questions, e.g., have we heard this from customers before? and Does this make ordering cash easier?


• Evaluate the user needs
  Each user need was evaluated against each teams' criteria. If the need met primary criteria = +3 points; Secondary criteria = +1 point. Example: Has the need to import files to place orders been heard from customers? + 3 points.


• Keep score and prioritize
  After summing up the scores, the needs were prioritized into 3 groups with the highest scoring needs defining the MVP.

From here the product team had clear user stories to write and there was transparency with the UX and dev teams about the work. A testable prototype was designed efficiently.

I tested the prototype in 2 stages:

1. I moderated an initial set of 5 moderated usability tests, gathered feedback and issues, and then shared with the team to address.


2. UX iterated on the prototype which I then tested with 5 more users in moderated testing. We confirmed previous testing issues were addressed and took note of other behaviors.

The testing was comprised of 2-3 tasks users had to complete: placing orders, reviewing orders, and canceling orders. We found most users could complete these tasks with minimal friction, but there were key findings about how users interacted with tables, used a currency counting widget, and even more broadly about their processes for ordering cash.

The impact of the internal collaboration and testing was not just certainty we designed to users' expectations. We also:

• Discovered different cash ordering processes
  We suspected nuances in how companies ordered cash. In research we defined 4 distinct processes.


• Questioned assumptions about value
  We were confident that reordering would be an important feature for users, but after hearing about the cash ordering process we decided to later test this assumption (and good thing we did!)


• Learned about behaviors needing more support
  Tasks like printing orders inspired thinking about to support record keeping.