UPS · Global E-Commerce · Partner Compensation
AI-Enhanced Partner Rewards System
Designed a global partner compensation system rewarding frequent shippers with cash back — then added an AI layer that flags at-risk partners before they churn.
The Challenge
UPS needed an application where frequent and volume shippers would be rewarded with cash back on their shipping spend. The system served three distinct user levels — admins who managed tiers and approvals, T&C users who configured compensation terms, and data entry users who processed partner information — across a truly global, diverse user base.
The design had to be appealing and the forms easy to fill out for everyone, with maximum usability and accessibility. My initial concepts were built around the admin user, then broadened usability outward to all levels.
My Design Approach
User research revealed an everyman-type user profile with extensive global usage. I designed an interactive image landing page showing small shops, people, trucks, and planes in different countries — connected by UPS's global network — to create an emotional connection that static status banners couldn't achieve.
- Three-tier access model: Admin, T&C, and Data Entry users each saw role-appropriate views of the same underlying system
- Revenue and Volume Base Tiers: Configurable compensation tiers based on either shipping spend or package quantity
- Accessible forms: Every input, dropdown, and date picker designed for maximum usability across skill levels and languages
- Global partner filtering: Search and filter across partner name, client ID, vendor, SSC code, and creation date
What Happened
The design was well received by the UPS client. However, it was ultimately rejected by the agency's marketing team — not on usability or client feedback grounds, but because it didn't replicate the status-banner style marketing had used to win the account. A junior designer conformed to marketing's direction but dropped the accessibility and usability commitments that had been promised to the client.
The lesson: A design that serves users and a design that serves an internal pitch deck are often two different things. This experience reinforced my commitment to fighting for the user — even when the pressure to conform comes from inside the house.
AI Enhancement: Partner Risk Intelligence
The original system showed partner status as a binary: Active or Inactive. But by the time a partner goes inactive, the relationship is already lost. An AI layer can analyze shipping patterns, payout utilization, tier proximity, and engagement signals to flag at-risk partners before they churn — giving account managers time to intervene.
Click any orange-flagged partner below to see the AI's reasoning and recommended intervention.
Interactive prototype — click any orange-flagged partner to see AI churn analysis and recommended intervention