Dispute Management System
Helping Fraud analysts identify, understand, and prevent bad travel experiences on EG's platforms.
CASE STUDY
Overview
Fraud analysts at Expedia exist across several levels of access, experience, and domains. They use two tools to inspect, process, and decision potentially fraudulent transactions - the Risk Case Management tool, which is the legacy tool, and the Dispute Management System, which is its in-process replacement. As the company has grown and acquired new brands and products, new capabilities are added to the roadmap by the PMs overseeing various arenas of fraud (e.g. Debit Memos, Supplier, etc). Some functionality has been added adhoc and needs to be retrofitted with better UX in mind, and some will need to be designed from scratch.
MY ROLE
UX Designer
User research
Interviews
Information architecture
Wireframing
Prototyping
TEAM
Platform XD
Team lead: Radhika Mehra
PMs: Mark Watson, Jaimon Barot, Ilya Padel
TOOLKIT
Sketch Figma, JIRA, Confluence
TIMELINE
July 2019 - January 2022
Problem Statement
RCM's UI and workflows are out of date, and lack functionality to cover new areas of fraud like Account Takeover (ATO) and new brands and products. Management would like to migrate RCM's functionality to RCM, but many analysts are reluctant to stop using RCM out of habit / familiarity. They would also like to add an "entity level" view to the tools, which the analysts could use to drill down on a single piece of metadata (a credit card, an address, an IP address, etc.) and investigate any/all linked transactions. However, there is no clear consensus on how this will integrate into the existing site map and workflows.
Using extensive research and analyst feedback, DMS can be re-designed with enhanced functionality, clearer information architecture and more streamlined, flexible workflows.
User Research
I began my research by creating user personas to get a better holistic view of how EG's fraud prevention tools are used and supported internally, and the role they play in ensuring a seamless experience for partners, suppliers, and travelers by preventing trust damage and limiting financial loss.
User personas for various stakeholders, both direct and peripheral, of the Dispute Management System
I conducted several one-hour user research sessions with EG fraud analysts at all levels of access and experience, using both RCM and DMS tools. These sessions involved both in-depth interviews using a standardized set of questions and a field study of current workflows, probing for pain points and opportunities for easy win improvements.
Following each session, I compiled my notes and observations into reports to present to stakeholders, with an emphasis on UX issues, accessibility issues, and operational dead weight.
Compiled notes from several Analyst shadowing and research sessions
What these sessions showed was that both tools as they were currently being used were out of date and did not cover all fraud scenarios, and that analysts had to rely on workarounds and clunky workflows to complete their tasks. I highlighted several opportunities for low-lift/high-upside improvements alongside notes on established project goals.
Design
While iterating on design improvements to the existing tools and conceptualizing how the added capabilities could be integrated, I learned via the company's internal UX slack channel that a new Design System was being developed and deployed, to be used on all externally facing EG products and services going forward.
This development created a host of issues for this project - I had already begun designing using what would soon become outdated components, and several piecemeal solutions were already being built by engineers. Thinking more long-term, and knowing that there had been internal discussions around externalizing fraud monitoring tools as a SaaS product for partners, I felt it would be best to begin implementing the new Design System - even though there was no clear mandate for internal tools, and even though it would mean significantly overhauling the designs already in progress.
I met with Fraud & Risk group leadership to explain the implications of a new design system and the dilemma we faced with the DMS re-design already in progress. The team agreed that, to avoid any issues with externalization and to bring our products into better design alignment with the rest of the company, that we would accept the sunk cost of the work already built and re-start the design phase using new components and a re-designed layout.