GovFlow
A travel booking and expense module within the Travel.mil application. After a decade without updates, the system was long overdue for a redesign.
Role
User research, wireframing, and UI design. Develop comprehensive design system. Establish brand identify and art direction.
Result
Decrease in Help Desk tickets. Reduction in user training costs. Built trust and developed strong relationships with the customer.
Project Overview
GovFlow is part of Travel.mil, a decades-old travel management system for the U.S. Department of Defense (DoD). The application was notorious for its poor user experience. With confusing navigation, error-prone workflows, and minimal user feedback, it prioritized function over usability, earning its reputation as one of the worst travel systems ever created. In 2015, changes in DoD security policy required significant back-end updates, prompting the Program Management Office (PMO) to address long-standing user concerns. I was brought in to redesign the itinerary screen, a key component of the travel authorization workflow.
Initial Challenges
The customer never had a UX designer on staff so they didn’t see the need to provide me access to users. Instead, I received design direction solely from the PMO. In spite of this lack of insight, I analyzed their documentation and business rules to develop a deep understanding of how the existing system worked.
The legacy itinerary, modeled after a paper form, was inefficient and overwhelming. User metrics revealed over half the form fields could be hidden or pre-populated with the user’s profile data. Applying standard UX best practices such as linear form layout and progressive disclosure resulted in faster completion time and reduced input errors dramatically. The redesign was fully responsive and optimized for any device.
Legacy Itinerary Design
Legacy Itinerary Form
Redesigned Itinerary
Results
Initial testing was extremely positive, with time on task to complete the Itinerary form dropping from an average of 2 minutes to approximately 10 seconds. The integration of UX best practices, like system feedback and smart defaults, increased the task success rate from around 70% to over 95%, as fewer users abandoned the form in confusion and required Help Desk support. These enhancements led to significant time savings for the user community, translating to real cost efficiencies for the government. Recognizing the value of UX, the PMO was ready to begin working on a complete redesign of the entire application.
The Next Challenge
Following the successful redesign of the Itinerary workflow, the PMO shifted focus to the expense system, a longstanding pain point for both users and management alike. The system was inefficient and cumbersome, contributing to improper expense payments reportedly costing the government over $1 million per day! The PMO sought a solution and believed they had identified the fix.
The Importance of User-centered Design…
A successful UX process begins with understanding user needs; without this, it’s impossible to effectively address their problems. The PMO was unfamiliar with UX and therefore did not see the need for thorough user research and testing. They envisioned an expense system based on their own assumptions about user interactions and instructed me to design accordingly. They incorrectly assumed users would conform to their mandated process and believed that confusion could be mitigated through training. In reality, the system needed to accommodate over a dozen different user personas, requiring flexibility and intuitiveness rather than rigid adherence to a single workflow.
…and the Pitfalls of Designing by Committee
The PMO provided low-fidelity concepts that described a process whereby users would first upload all expense documentation to a "Trip Workbook” which served as a repository for receipts and travel documents. Then, in a subsequent workflow, users would create new expenses and drag and drop receipts from the "Trip Workbook" onto the expense to link them together. Compounding the complexity, the PMO assumed travelers preferred collecting paper receipts, taping them to a physical sheet of paper, and uploading an image to the system. They envisioned the need to crop individual receipts from the uploaded image for expense assignment. Despite voicing concerns that this approach was convoluted and impractical for mobile use, my feedback was overlooked, and the PMO instructed me to proceed with their design.
Their Initial Design Concept
Create Expense
Expense List
Drag-and-drop Design
Expense Workflow
On paper, the problems with this design were obvious:
The user workflow was too fragmented.
It wasn't designed with responsiveness in mind.
It was extremely rigid and inflexible.
Despite my critique of their design, the PMO proceeded with development approval. When the expense update went live, it yielded mixed results. While the customer appreciated the enforcement of policy, improved documentation traceability, and enhanced administrative processes, users struggled with the new workflow and found the interface difficult to navigate. Consequently, negative feedback began to flood the Help Desk almost immediately.
“There are too many steps to creating a new expense.”
“Not many people copy groups of receipts at a time so cropping is not a huge benefit.”
“Get rid of the Trip Workbook section entirely!”
“The confusion over this new redesign is outweighing the functionality.”
User Research and Personas
Following the negative reception of the initial Expense release, the PMO sought my assistance. I emphasized the importance of in-depth research and analysis, beginning with a better understanding of GovFlow users. I identified several roles that make up our user community and created user personas to review with the customer. These tangible UX artifacts clarified the problem space and guided the PMO on the types of users I needed to engage for answers.
Rethink and Simplify
While waiting for the PMO to gather users, I revisited the system metrics utilized for the Itinerary design. The program had generated significant usage data since the expense release went live a month earlier, allowing me to analyze the frequency with which users accessed the crop tool and the number of cropped receipts uploaded to the system.
Cropping Tool Usage in a 30-day period:
Total receipts uploaded: 1,051,945
Substantiating documents uploaded: 613,138
Total number of documents cropped: 469
Out of 1.6 million files uploaded, less than 0.0004% were cropped.
Data-driven Design
With the usage data in hand, I presented my findings to the PMO. It was clear that users were confused by the cropping tool and the “Trip Workbook” concept. The PMO’s design demanded that users conform to its structure rather than tailoring a solution to their needs. Interviews with actual users revealed the majority of travelers only used GovFlow occasionally, underscoring the need for a simple, intuitive expense process. Consequently, I eliminated the “Trip Workbook” entirely and introduced a multi-step wizard, allowing users to create an expense and attach a receipt in just a few clicks, with the option to add documentation later. This design embraced a mobile-first approach, ensuring functionality across all devices.
Expense List
Expense Details Card
Expense Creation Process
Step 1: Select an expense type.
Step 2: Enter the expense details.
Step 3: Attach a receipt.
Results
The redesign led to an 85% reduction in Help Desk tickets for the PMO, with the remaining inquiries primarily related to policy issues. The rejection rate for travel reimbursements due to errors dropped significantly, from over 50% to around 10%. The PMO was pleased with the new design, and initial user testing resulted in a high satisfaction rating. The simplicity and intuitiveness of the new design allowed the PMO to save around $25k in the development of user training modules which they then redirected towards developing new feature in the application. The PMO recognized the value of my work and began to trust the UX process. They opted to proceed with the next major set of enhancements for Travel.mil.