Problem Statement
Problem: The release dashboard use by product teams at Kessel Run to push releases and deployments to production environments is cumbersome, difficult to navigate, and requires a lot of assistance from the dashboard developers to be properly utilized.
Background: With detachment-wide development velocity increasing the use of the release dashboard has skyrocketed. User feedback has centered around cognitive overload, usability, readability, and a lack of functionality. I was called upon to clean up the application UI, establish more consistent design patterns, decrease a clutter, and highlight features that meet user needs.
A less cluttered dashboard view was implemented that bubbled a user’s assigned product teams to the top, while also giving leadership insight into DevOps Research and Assessment (DORA) metrics.
User Personas were created with the team to help develop empathy and an understanding of user pain points.
Methods:
Developed usability surveys, feature request logs, and other mechanisms for anonymous, direct feedback from users.
Facilitated persona creation exercises to help the team empathize with users of the dashboard.
Implemented a more consistent color palette that utilized colors more sparingly and increased focus on functions that meet user needs.
Added skeletons and loading indicators to acknowledge data bottlenecks that caused confusion and frustration during page loads.
Utilized tooltips, helper text, negative space, and repeatable patterns to lower the application’s learning curve and decrease cognitive load.
Results: Usability scores improved across the application, as well as user confidence. An anonymous survey of end users found that:
Users strongly agreed when asked if the product could be learned very quickly.
Users strongly disagreed when asked if the product felt cumbersome.
Users strongly disagreed when asked if the product felt unnecessarily complex.
Users strongly agreed when asked if they felt confident using the product.
A streamlined deployments page allows teams to sort through their production deployments a variety of ways without overloading them with too much information.