Location Management — E2E user journey & discovery research.
A three-phase UX research programme uncovering inconsistent location UI patterns across IBM Cloud — from competitive analysis through 14-participant user journey research to remote generative discovery interviews and conceptual wireframes for a unified location model.
How it started
IBM Cloud exposes location-related controls across dozens of services — Virtual Server for VPC, Watson Studio, Power Virtual Server, Satellite, and more. Each product team independently designed their own location UI: some used drill-down dropdowns, others showed maps, others plain text links. There was no consistent mental model, no shared pattern, and no single access point for location management.
The initiative was ignited at the April 2024 Eureka Hackathon, which surfaced the problem: "How do we optimise and train data for the IBM AI Assistant to retrieve location-based resources accurately?" — exposing a deeper structural issue: the location concept itself was broken across IBM Cloud.
I was brought onto the project by JP Pollard alongside Andras Kuti, Derek Bressler, and Virginia Nicholson to lead the UX research and design workstream, in partnership with BJ Scheid and Vijay Agarwal.
Three phases — from audit to generative research
What the research revealed
- Inconsistent design elements — Drill-down menus, dropdowns, maps and text links used interchangeably with no shared pattern.
- Increasing cognitive load — Users had to re-learn location controls on every service. No persistent location context anywhere on the platform.
- Disconnected dropdowns — Country selections did not cascade to Region or Zone. Multiple redundant clicks required.
- No admin control point — Client administrators had no way to manage or enable locations across their IBM Cloud account.
- On-premises blind spot — IBM Worldwide, Co-Location, Business Partner, and 3rd Party locations had no representation in the existing location model.
"Users encounter difficulty when attempting to locate resources based on their location — inconsistent design elements increase cognitive load, leading to more clicks to achieve user goals."
Problem Statement — Location Management ResearchWhat was designed
Based on synthesis, I developed a unified location filter model using a cascading hierarchy: Location Type → Geography → Country → Metro → Region → Zone. The filter persists as a global access point, not per-service.
- Location Type — Off Premises / On Premises (IBM Worldwide, Co-Location, Business Partner, 3rd Party)
- Persistent global filter — Accessible from the landing page hamburger menu and catalog landing page
- Admin control point — Location Administration panel for client admins to manage company locations
- Conceptual prototype — Figma prototype demonstrating the full filter cascade with live data counts
