My role: Lead design, Interaction Design, Prototyping
Team: Product manager, 3 Engineers, Business analyst, Design system lead
Duration: February to June 2023
Backgound
Coupa was seeing signs of customer dissatisfaction, not just through support tickets or complaints, but through real business losses. In several cases, including large enterprise customers like Disney, DHL... we noticed that even when Coupa helped save millions through sourcing optimization, the user experience within procurement remained a bottleneck, slowing down invoice cycles and affecting supplier enablement.
One clear signal came from internal performance metrics: Pending Buyer Action was consistently one of the longest wait states in the procurement lifecycle. Requisitions would sit idle for days, sometimes weeks, because buyers couldn’t tell which requests were theirs — or worse, teams would duplicate efforts or miss approvals altogether.
Requests were sitting idle because there was no clear ownership
• Missed SLAs for approvals
• Manual tracking
Rather than jumping straight into design, I stepped back to define a high-level strategy to tackle the root problem. I broke it down into three milestones, which were presented during the October 2022 walkthrough and targeted for completion by February 2023.
In the early stages of the project, I mapped out the existing requisition handling process and identified major workflow breakdowns.
Users like Sara could create requests, but ownership was unclear—every buyer saw every request, leading to confusion and delays. Buyers often relied on Slack, Nexora, and Excel spreadsheets to manually coordinate assignments, causing duplicate work, idle requests, and even missed approvals.
Based on the pain points identified—especially the lack of ownership and the manual burden of assigning requests—we set out to design a more structured and scalable assignment workflow.
Our goal was to give buyers back control, eliminate duplicate efforts, and improve SLA performance by ensuring every request had a clear owner.
Auto Assignment Exploration
To reduce manual effort and speed up request handling, I explored ways to automate the assignment process based on predefined business rules.
After testing both options, we moved forward with the Single-Screen layout as our final design. It proved most effective for our power users, allowing them to view and complete all rule details in one place—minimizing clicks and making it easier to cross-check inputs. To enhance clarity, we introduced inline validations and collapsible sections, improving focus and flexibility without overwhelming the screen.
Milestone 2: Ensuring Accurate Ownership
While many requisitions are now auto-assigned through rules, some still fall through—either due to missing fields or unmatched criteria. These edge cases require manual review and assignment. I wanted to ensure we supported not just efficiency, but also accuracy—making it easier for users to quickly understand what needs attention and take the right action.
Design System
I created a card component that balances clarity and scalability. It standardizes spacing, grouping, and visual hierarchy—making it easier for buyers to scan, take action, and adapt to future system needs.
The result of our iterative design process, enriched by stakeholder feedback and cross-departmental collaboration, was a dashboard that went beyond mere analytics. It became an essential tool for Buyer Managers to orchestrate their team's workflow effectively. With the redesigned buyer datatable, the dashboard transformed into a strategic command center, enabling managers to make informed assignments that optimize team performance and expedite the procurement cycle.
The implementation of the revamped requisition datatable and manager dashboard has significantly elevated the user experience for our vast user base. The improved data visualization and streamlined functionality have been instrumental in enhancing the productivity of over 60,000 users, empowering them to navigate the procurement landscape with confidence and strategic insight.
Looking ahead, one exciting opportunity in this project could later have AI-powered smart suggestions. As users perform frequent, repetitive actions, such as reassigning similar requisitions the system could detect these patterns and proactively suggest automation rules or bulk actions. We feel this could really level up with our system by helping users save time, reduce manual effort, and create a more intelligent, responsive workflow especially for teams managing high volumes of assignments.
