Instructional Scheduling Assistant

The Project

UC San Diego has over 70 academic departments who plan and submit a quarterly class schedule to the central Registrar’s office for processing over 30,000 classes per year. Over the course of 10 years, 5 different departments developed homegrown solutions to address the class scheduling process for staff and instructors in 70 different departments. Some departments were using those 5 systems. Others were still using good old fashioned paper and pencil (literally), whiteboards, and systems of surveys, spreadsheets, and email. My group was hand selected to join a new initiative called Distributed Application Development, whereby we would collaborate with 2 other departmental IT groups to collectively come up with a unified solution for all campus.

The Problem

Technical:

The front end team decided to use Angular to make an SPA whereas the backend developer, who was in a different department, had independently started laying the groundwork based on an entirely different model, with incomplete knowledge of the requirements. There were no official specification or requirements documents despite months of committee meetings.

Organizational:

Bringing together 5 front and back end developers, 4 of whom who had never worked with UX design concepts before, was a large challenge. We didn’t have a project manager or a standardized development methodology to use between groups.

Change Management:

Unifying both the administrative staff and the faculty of 70 different academic departments is no small task. There were many pockets of resistance to the change, for myriad technical, social, and political reasons.

My Role

For the first 6 months or so, I started out as the UX Lead. Over the course of the project, I teamed up with a temporary project manager to help shape our agile methodology and adopt Jira as our issue tracking and organizational tool. After another 6-8 months, that person rotated off and I assumed full project management duties on top of UX lead. My role ultimately ended up as full Product Manager and included training development, ticket management, and frequent presentations to all levels of staff, faculty, and leadership.

My Approach

I have a user-first approach and always start by trying to deeply understand the user’s problems and priorities. I intentionally used simple tools like index cards for cardsorting so that the ideas could be easily shuffled around without technology getting in the way of the discussion. I also created very low fidelity sketches and wireframes when first designing so that the users wouldn’t get stuck on the look and feel or precise placement of elements.

Research

I started by getting familiar with the 30 person committee at their monthly meeting. It was clear that a committee meeting was not the best way to get requirements, so I distilled that group into 7 power users who met with me over 8 sessions for user story mapping and card sorting. This group was then reformed into a 10 person weekly workgroup who I lead through process discovery, policy decision-making, feature prioritization, usability testing, user acceptance testing, and change champion activities.

This was the original User Story Map which was initially created on index cards pinned to the conference room wall.

This was the original User Story Map which was initially created on index cards pinned to the conference room wall.

I transcribed the index card version into a tool called StoriesOnBoard, which became our roadmap for the entire project.

I transcribed the index card version into a tool called StoriesOnBoard, which became our roadmap for the entire project.

Methods and Tools

  • Personas

  • User Story Mapping (StoriesOnBoard)

  • Card Sorting (Trello)

  • Static and Dynamic Prototyping (Balsamiq)

  • Process Mapping (Lucidchart)

  • Usability testing (Balsamiq and in QA environment)

  • User Acceptance Testing

  • Ticket Management (first Kayako, transitioned midway to ServiceNow ITSM)

  • User Training (ServiceNow KB)

We used Trello to organize our Personas.

We used Trello to organize our Personas.

Detail view of the “Scheduler” persona.

Detail view of the “Scheduler” persona.

Very early wireframe I created in Balsamiq.

Very early wireframe I created in Balsamiq.

First working version in our QA environment, based on my wireframe.

First working version in our QA environment, based on my wireframe.

A detail modal.

A detail modal.

Outcome

Six major releases were adopted by all 70 departments over the course of 3 years. Each of the 6 releases were on time and on budget. Our designs dramatically reduced staff processing times while increasing data accuracy and reducing the amount of manual communication between staff and faculty, resulting in over $615,000 cost savings annually.

Success lead to creation of 3 new spinoff data governance workgroups and funding approval for an expanded next phase of the project.

We won the 2019 California Higher Education Collaborative Focus on Efficiency Award, based on “Improved Performance, Service, and Outcomes”.