Joshua Litchfield

My role

Senior Product Designer

Team

1 Product Manager, 6 engineers, subject matter experts

Duration

2 quarters

Process overview

Product thinking, sketching, primary and secondary research, visual design, prototypes for feature validation, SME feedback sessions

Personas

Branch Manager, Care Manager, RN Care Manager, Care Coach, Care Coordinator

Outcome statements

• Professionals can manage their caseloads efficiently at scale from a single view.
• Professionals can view their panel based on personal preferences, workflows, or program goals.
• Branch Managers need to view a Professional's panel for caseload balancing, assessing coverage needs, and reviewing care team activity.

Technical constraints

• Performance requirements: data might come from Vela Pro, a system of record, or other services. We need to ensure the Panel will be performant. Working closely with Engineering around this topic was essential.
• Data failures: we want to ensure that if one service fails, it doesn't kill the entire Panel service. If a failure does occur, how can we allow the Pro to recover from that?
• No machine learning or AI: Vela doesn't use ML or AI, nor did we have folks with those skills. Also, using these technologies for several, often experimental, use cases might be too costly.

Development statements

• List view: Pros need a new list view with columns for activity and other care room information.
• Filtering: Pros need to filter their care team to see activity and other data.
• Sorting: Pros need to prioritize sorting to get to what matters most to them. For example, I want to sort by my primary care teams, then secondary.
• Data: as more data becomes available in the Panel, Pros must hide those irrelevant data, not germane to my workflow or program objectives.
• Efficient navigation: Pros need to efficiently go in and out of care rooms, which the current experience could do better. Red routes and modality will allow Pro to complete tasks with speed and efficiency.

Secondary research

• I needed access to users, but I didn't have it. However, user feedback from the Product and Technical Support teams informed the problem statement and user needs.
• Next, I conducted secondary research to evaluate best practices for filtering, sorting, data density, and complex tables while considering our users' skill levels.

Design principles

• Performance: if the panel is performant, it will retain its utility.
• Efficiency: easy access to data, red routes, and modality significantly decreases interaction costs, allowing the Professional to spend less time in the application and more time with care teams.
• Extensibility: filtering and sorting must be extensible enough to allow different data types we have yet to think of to support furure use cases.
• Straightforward filtering and sorting: considering our users' technical proficiency, users need to be able to quickly determine how the panel is filtered and sorted with a low cognitive load.
• Just enough information: prior knowledge and my data density research helped validate our assumptions about data overload.
• Information hierarchy: for those pertinent data, Pros want to move panel columns to show the most critical information first.

Iteration one

The first iteration of the panel utilized existing components to give the panel a consistent experience—for example, the use of cards as rows and applied filters.

Quick filters exploration

I played around with quick filters with a single click instead of a dropdown menu with higher interaction costs.

Exploring filters and long lists

Filters with short lists are displayed in dropdown menus in Vela Pro. Given many filters have long lists, I experimented with the right drawer.

Iteration two

Sometimes, I build out the experience when working on a project, considering the long-term vision. It helps with story mapping and planning. During the second iteration, I began considering the requirements and design principles the team and I agreed on. The utility menu here accounts for views, filter and sort, and other menu items.

Rule builder filters

My desk research informed me that sentence-style wording, instead of parametric-style, makes it easier for Professionals to comprehend, even if longer.

Advanced sorting

In the previous implementation, we allowed the Professional to choose a primary sort and then make some assumptions about the secondary sort. In the future, the Pro will determine the order of her care teams.

View Options

Data overload would eventually be a genuine concern. Also, some data would be necessary for some Professionals rather than others. They need a way to hide and order panel columns.

Expanded columns

Performance is always a consideration, so I worked closely with my Engineering colleagues. Using the principle of progressive disclosure, we can show just enough data about the care team. With another call, the Professional can access care team data without leaving the panel.

Panel views

Having complex filtering and sorting and the ability to configure fields a Professional cares about is a game-changer. Still, it can be taxing to configure them regularly, even if the application has a memory. Being able to save the settings into a view will be crucial. Of course, the professional needs to be able to rename, duplicate, or delete views.