From complex workflows to user centered experiences, explore four projects where I solved business challenges through UX/UI and product design. I was the only UX designer on Global Directory, Supply Chain, and Next Generation. On My Choice for Business I worked with two other UX designers. Select a project below to open its case study.
Each project opens here as a complete case study. Select another project at any time to switch.
Case Study
Global Store Locator Redesign Project
The UPS Global Store Locator is how people and businesses find a UPS location for shipping, postal services, and drop offs, anywhere in the world. When I was at UPS, the locator had high bounce rates and poor feedback, and its design had fallen behind the rest of the site. I was the only UX designer on the team, in a full time role, and I redesigned the locator for desktop and mobile.
Only UX designer on the team3,500+ PDF downloads a month8 developers on the delivery teamFull time January 2019 to February 2021
Team
UPS development team of 8, 2 product owners, business and executive team
My Role
Only UX and UI designer
Tools
Adobe XD, Figma, Adobe Suite
Timeline
January 2019 to February 2021
The redesigned Global Directory. Search first, results beside the map, and the location details where the click lands.
The Problem
High bounce rates and poor feedback
A store locator has one job, and people were leaving before they finished it. The feedback pointed at navigation and service information that were hard to make sense of, and the dated design made the tool feel neglected next to the rest of UPS. The bounce rate was the symptom leadership watched.
The legacy locator. A cramped search panel, a small map, and a page that looked years out of date.
High bounce rate
Visitors arrived looking for a location and left without one. Bounce at that level points at usability, not at demand.
Poor feedback
Users reported challenges with navigation and with understanding service information, which hurt satisfaction and repeat use.
Dated design
The locator looked older than the rest of UPS, and that first impression quietly argued against trusting it with a shipment.
Research
Data first, then faces
I dug into behavior data and analytics to find the confusing steps and decide what to fix first, partnering with the product owners and the analytics team throughout. With departmental approval, I interviewed customers face to face, and categorized the feedback that was already coming in. Affinity mapping with the team turned all of it into the redesign priorities.
Heuristic walkthrough
Finding. The navigation and hierarchy made the search harder than the search itself, and error handling left people stranded in the middle of a task.
Decision. The walkthrough, reviewed with the product owners, set the floor for what the redesign had to fix.
Affinity mapping
Finding. Interview and feedback themes kept returning to the same few steps in the search, the steps where people gave up.
Decision. Those steps were fixed first, and the rest of the redesign followed their order.
Key Moves
Three decisions that shaped the redesign
Move One
Search and map, front and center
The redesign leads with the one thing every visitor came for. One search, the results and the map together on the screen, and the details of a location, its services, hours, and directions, right where the click lands. Nothing about the visit matters until the location is found.
Finding: people arrived with a single question, where is my location, and the legacy page made them work for the answer. Decision: the search is the page. Everything else supports it.
Move Two
One system for every screen
Desktop, tablet, and mobile had drifted into separate experiences, so I built a reusable design system in Adobe XD, Sketch, and Figma, applied WCAG 2.1 AA guidance to the shared patterns, and standardized the UI patterns so product and engineering teams could reuse them across UPS global products.
Finding: three screen sizes had become three products to design and maintain. Decision: shared patterns, designed once and applied everywhere, with accessibility built into the pattern itself.
Move Three
Test findings became fixes
Usability testing ran through the redesign, and its findings were turned into specific fixes across the shipping and locator experience rather than filed as reports. Working in two week sprints with the development team of 8, each round of testing ended with a fix list that shipped in the next cycle.
Finding: testing kept surfacing small frictions that added up to the bounce rate. Decision: every test ends in fixes, not a report, and the fixes go into the next sprint.
Challenges
Working under real constraints
Everyone is the user
A one time shipper and a business shipping desk land on the same locator. The design had to serve both without modes or forks.
Three screen sizes, one experience
Desktop, tablet, and mobile had to be designed together, or they would drift apart again within a year.
A global product
The patterns standardized here were meant for reuse across UPS products and markets, so every decision had to travel well.
A legacy site that could not pause
The locator sat inside a large, busy legacy site. The redesign had to fit what UPS could actually build and ship.
Outcome
Approved designs, and a system that kept working
Approveddesktop and mobile designs, ready for development handoff.
Design systemreusable patterns built for new features across UPS products.
2 yearsfull time, from research to handoff.
The redesigned locator was approved, with final desktop and mobile designs ready for development handoff. The design system outlived the project, giving product and engineering teams shared patterns that sped up work on new features. UPS did not hand me post launch bounce numbers for this page, so the claim here is the approval, the handoff, and the system other teams kept using.
Lessons learned
A locator is a utility, not a brochure. People give a tool like this seconds. Design for the short visit, and the short visit succeeds.
Analytics tells you where, people tell you why. The data found the steps that lost people. The interviews explained what those steps were doing wrong.
Systems work is portfolio work. The screens got the approval, but the design system is what changed how fast every later feature could move.
Supply Chain, where the fix turned out to be the words
UPS Supply Chain Solutions was launching a new site around the Forwarding Hub, a portal where medium and large companies quote, book, and track air and ocean freight in one place. The Global Directory sits inside this site. I was brought in for the branding and content updates across the site, and for a second problem nobody had managed to solve.
Only UX/UI designer on the projectApproved and put live on the siteThousands of complaints traced to two labels
Team
UPS Supply Chain stakeholders and the site team
My Role
Senior UX/UI Designer, the only designer
Scope
Branding, content updates, and entry point clarity
Timeline
Full time at UPS, 2019 to 2021
The Forwarding Hub. Quote, book, and track freight in one portal, built for companies, not parcels.
The Problem
A freight door that looked like a parcel door
The second ask was the hard one. Small package shippers kept landing on the Supply Chain site, entering the Forwarding Hub, and discovering they could not ship a parcel with a freight tool. They complained to customer service in the thousands. The product worked. The users arriving at it were simply the wrong ones, and every label on the way in told them they were in the right place.
Wrong users, right product
The Hub was built for freight. Nothing at the entry said so in words a parcel shipper would act on.
Complaints by the thousand
Customer service carried the cost of the confusion, one frustrated call at a time.
One door, two audiences
The same pages had to welcome freight customers and turn parcel shippers toward the right tool, politely and early.
Research
Read the clicks before touching the design
Stakeholders shared the analytics on how users landed on the Supply Chain site and what they clicked once they arrived. I paired that with a heuristic evaluation of the entry points: the menus, the main call to action, and the exit links, looking for the exact words that were inviting the wrong audience in.
Landing and click analytics
Finding. Small package users arrived through the main Supply Chain pages and took the biggest, most general button on the page, because its words promised shipping of any kind.
Decision. Treat the labels as the problem to fix, not the users as the problem to block.
Heuristic evaluation
Finding. The main call to action said Quote & Ship, which a parcel shipper reads as an invitation. The one link meant for them was labeled UPS.com, which reads as leaving, not as the parcel door.
Decision. Rename both, in the vocabulary of the two audiences: freight words for freight, parcel words for parcels.
Key Moves
Three decisions that solved it with words
Move One
See the problem as a reader
Before proposing anything, I read the entry pages the way a small package shipper reads them: fast, hopeful, and looking for the word ship. Every step of that reading worked, which was exactly the problem. The path was well designed for the wrong promise.
Finding: the flow was not broken, the invitation was. Decision: fix the words at the two decision points and leave the working product alone.
Move Two
Rename the main door
The main call to action, Quote & Ship, became Air & Ocean Quotes. The new label names the freight the Hub actually quotes, so a parcel shipper knows in one glance it is not for them, and a freight customer knows it is.
Finding: Quote & Ship promised every kind of shipping. Decision: let the button say air and ocean, the two things the Hub sells.
Move Three
Give parcel shippers their own sign
The exit link labeled UPS.com became UPS Small Package, in the header and the menu. Instead of a dead end, small package users got a labeled door to the tool built for them, placed before they could enter the freight flow.
Finding: UPS.com read as abandoning the task. Decision: name the destination, so leaving feels like arriving.
Challenges
Working under real constraints
A problem design work had not solved
The pages had been through design attention before. The miss was not visual, it was verbal, and verbal problems hide from visual reviews.
Two audiences, one front door
Turning one audience away without insulting them, while rolling out a welcome mat for the other, in the same few labels.
Freight words, plainly
The labels had to stay accurate for freight professionals while becoming unmistakable for everyone else.
Outcome
What changed
2labels carried the whole fix, the main call to action and the small package link.
Livestakeholders approved the change and it was updated on the site.
0new pages or features built. The product never needed them.
The stakeholders approved the solution and the site was updated. I have no complaint volume numbers from after the change to quote, so the claim is the one the evidence supports: a problem the team had not solved was solved by reading the page the way its wrong users read it, and changing two labels.
Lessons learned
Read the sign before rebuilding the door. When users flood the wrong entrance, the fastest research is reading your own labels as a stranger.
A call to action is a promise. Name the thing it actually does, for the people it is actually for.
Seeing is a design skill. The fix was not more design. It was noticing what was already there.
Next Generation, modernizing the system UPS shipping runs on
E2K is the database application that UPS global shipping has run on for more than 25 years, and it still runs today. Next Generation was the program to rebuild it as a modern cloud application, section by section. I was the only UX/UI designer, and I started where every task in the system starts: search and results.
Only UX/UI designer$500k a month to run the legacy systemFirst section designed: search and results
Team
Next Generation program team and UPS stakeholders
My Role
Senior UX/UI Designer, the only designer
Tools
Sketch, Adobe XD, pen and paper
Timeline
Full time at UPS, 2019 to 2021
The redesigned E2K search and results, the first section of the modern application.
The Problem
A giant that could not be replaced all at once
E2K was built in the early 2000s and carries the details of global shipping in dense, aging screens. The objective from stakeholders was to copy one section at a time into the new application, modernizing the look and feel as it moved, until the legacy system, costing upwards of $500k a month to run and maintain, could retire to the cloud.
E2K as it still looks. Dense dialogs from the early 2000s, memorized by the experts who use them.
Too big for one leap
A system this central cannot stop. It had to be rebuilt while the business kept shipping on it.
Expert users, aging screens
The people who use E2K are experts. Their speed comes from memory, not from the interface helping them.
Internal only
No mobile version and no public audience. The design target was the working day of UPS operators.
Research
Learn the machine before drawing over it
I started with a heuristic evaluation of E2K and turned the findings into interview questions for the people who run global shipping on it every day. I researched applications similar to E2K, built user flows with pen and paper for each section, and kept two checklists side by side: every element on the old screens, and every element in the new designs, so nothing the business depended on could quietly go missing.
Heuristic walkthrough
Finding. The pain clustered where the work starts: setting up a search and reading its results across stacked dialogs.
Decision. Make search and results the first section redesigned, and the pattern the rest would follow.
Interviews with E2K users
Finding. Operators had built personal workarounds for screens that fought them, and trusted the data even when they distrusted the interface.
Decision. Keep the data, the terms, and the order they trust. Change the interaction, not the language of their work.
Element inventories
Finding. A screen that looks cluttered can still carry fields the business cannot lose, and nobody could say from memory which ones.
Decision. Account for every element in writing before any element moved.
Key Moves
Three decisions that started the rebuild
Move One
Map it on paper first
Each section began as pen and paper user flows, checked against the element checklists. On paper, a misplaced field costs an eraser. In a system like E2K, it costs a shipment.
Finding: the sections looked separate but shared the same search habits. Decision: design the shared pattern once, then apply it section by section.
Move Two
Design the new search and results
Working in Sketch and Adobe XD, I designed the initial search, the results, and their interactions as the first buildable section, following UPS brand standards exactly: the standard color scheme, the standard UI elements, and the standard fonts.
Finding: developers needed components they could build in the cloud environment, not pictures of screens. Decision: deliver the section as reusable UI components with the search pattern settled.
Move Three
Finish one section completely
The completed designs covered the search and results section from sign in to the filtered results view, so the development team could build the first cloud section and prove the approach before the harder sections followed.
Finding: a program this size needs an early, buildable win. Decision: finish one section completely rather than sketching five.
Challenges
Working under real constraints
25 years of business logic
Every screen carried rules nobody had written down. The research had to surface them before the design could move them.
Rebuild it while it runs
Global shipping could not pause for its own redesign. Sections had to move one at a time, in the right order.
A hard stop from outside
Budgets moved when COVID 19 hit. Large programs with long timelines were the first to be cut, this one included.
Outcome
Designed, ready, then shelved by budgets
25+years the legacy system had been running global shipping.
Firstsection, search and results, designed completely and ready to build.
Shelvedby the COVID 19 budget cuts, not by any verdict on the design.
The project was cancelled by senior leadership when the COVID 19 budgets were cut, and the work was shelved until further notice. What did not get shelved was the designer. UPS moved me to my next project, which says more about how the work was valued than any launch metric from a cancelled program could.
Lessons learned
Inventory before invention. In a legacy system, the bravest design move is accounting for everything that already matters.
Design the pattern, not the page. The first section set the rules every later section would inherit.
Cancelled is not the same as failed. Programs die of budgets. Work that was ready to build stays ready to be judged.
My Choice for Business, delivery status at a glance
My Choice for Business gives business customers one place to watch their deliveries. I was asked to develop UI ideas for displaying delivery status (delivered, on schedule, and date changed) and to rework the Delivery Timing section. This was a team effort: two other UX designers from a third party company, plus stakeholders, managers, business analysts, and copy writers.
One of three UX designersDelivery Timing section, my main buildBuilt for testing with current customers
Team
Two other UX designers, stakeholders, managers, business analysts, copy writers
My Role
Senior UX/UI Designer
Tools
Axure RP 8
Timeline
Full time at UPS, 2019 to 2021
My delivery status concepts. In the first design, selecting a segment of the chart highlights the matching status beside it.
The Problem
Status that took a second look
A business checking deliveries thinks in three states: what arrived, what is on schedule, and what changed. The ask was to explore displays that answer in one glance, and to evaluate the current and developing concepts with the other designers before anything went in front of customers.
My Choice for Business as customers knew it: a customized dashboard with branded tracking, returns, and claims.
Research
Evaluate together, then design
The team worked heuristically, evaluating the current product and each others concepts in the open. My concepts were built to be tested with current customers, so the evaluation question was always the same: can a busy business user read this in one glance?
The current display
Finding. The product promised a lot at once: dashboard, tracking, returns, claims. Status, the thing checked most, had to compete with all of it.
Decision. Give status the summary position and let the rest support it.
Design team evaluation
Finding. Static summaries informed nobody at a glance. The concepts that worked made status touchable, with the detail one click deeper.
Decision. Lead with an interactive summary and keep the detail one step down.
Key Moves
Three decisions that shaped the concepts
Move One
Rebuild Delivery Timing in Axure
My main build was the Delivery Timing section, designed in Axure RP 8 as working, high fidelity screens the team could click through: inbound and outbound views, date ranges, and the status cards for on schedule and date changed.
Finding: timing is the question behind every status check. Decision: give Delivery Timing its own clear section with the two answers side by side.
Move Two
Make status something you can touch
The lead concept is an interactive summary. The chart segments and the status buttons are the same control: click segment A, B, or C and the chart expands that segment while the matching status text highlights. Delivered, on schedule, and date changed stay connected instead of competing.
Finding: people trust a display they can question. Decision: one control, two views of the same truth.
Move Three
Variants beat arguments
Alongside the interactive concept I produced plainer variants of the data view, with and without the quick view panels, and presented all of them to the team for collaboration and modification. The team could react to real options instead of opinions about options.
Finding: three designers converge faster on screens than on principles. Decision: bring variants, and let the team improve them in the open.
Challenges
Working under real constraints
One slice of a big program
My Choice was ongoing and larger than any one designer. My work had to fit decisions made before I arrived and survive after I left.
Three designers, one product
Two other designers worked in parallel. Alignment came from evaluating together, early and often.
Designing for a test, not a launch
The concepts were headed for customer testing. They had to be clear enough to judge, not just to admire.
Outcome
Delivered, then on to the next project
3status states carried through every concept: delivered, on schedule, date changed.
1section owned in depth: Delivery Timing, from concepts to clickable Axure screens.
Movedto my next UPS project when my part was done.
My concepts were delivered to the team, and when my part of the program finished, UPS transitioned me to another project. I rotated off before the final results, so I will not claim numbers I did not see. What this project shows is the collaborative side of the UPS years: shared evaluation, variants on the table, and a handoff the next designer could pick up.
Lessons learned
Status is a promise kept in public. A delivery display earns trust in one glance or loses it in the same glance.
Bring variants, not verdicts. Options on the table move a team faster than one defended design.
Design for the next designer. In an ongoing program, your handoff is part of your design.
