Sideline App V3 Amazon
About me Photo gallery Playground · Coming soon
Sideline App V3 · 2025

Streamlining workflows to address $30M+ in inefficiencies and cut unresolved defects by 50%

Lead UX Designer · Feb–Sept 2025 · Problem Solve associates worldwide

V2 vs V3 side-by-side comparison
Overview

A $30M problem hiding in a tool nobody talks about

V2 forced associates across Sideline, FCResearch, and other tools for every defect. V3 consolidates the workflow. This case study covers the two features with the most direct impact: Scan Item and Verify.

My role

Lead designer, end to end

Research across 13 fulfillment centers · Design sprints · Vertical slice testing · Conditional prototyping · Scope and strategy
The product

Where every process path sends its problems

Every step of a fulfillment center can produce a problem: an item scanned into the wrong bin, a count that doesn't match, a product the system doesn't recognize. When that happens, it gets flagged and routed to Problem Solve. Sideline App is the tool Problem Solve associates use to work through those flags: figuring out what went wrong, how to fix it, and how to get the item moving again. V3 was focused on the post-receive paths, decant, stow, and pick, where the volume is highest.

Diagram: Receive, Decant, Stow, and Pick each route exceptions into Problem Solve, handled in Sideline App V3
Every upstream path, decant, stow, pick, surfaces its own exceptions, all funneling into Problem Solve and Sideline App V3.
The problem

The real problem wasn't tool-switching. It was trust.

Associates brought a mental model built over years of training and tribal knowledge: understand why a defect happened before acting on it. They didn't trust that Sideline knew the problem or the fix, so for nearly every defect they left to investigate in FCResearch first, then came back to resolve. The tool-switching everyone could see was the symptom. The trust gap underneath was the cause.

Two moments drove most of that distrust: Scan Item surfaced no container contents, and Verify was missing the attributes associates needed to act with confidence.

$30M+
in annual inefficiencies from tool-switching
50%
of defects routed to Sideline went unresolved
3+
tools associates juggled per item defect
Diagram: the V2 workflow started in FCResearch before moving into Sideline App to resolve each defect
The most common V2 path. Behavior varied by associate, but the constant was the same: every defect required leaving Sideline for FCResearch before the work could start.
Constraints

V2's framework, worldwide reach, and habits that don't vanish overnight

01
Technical framework
V2's layout couldn't change drastically and backend API contracts were already set. Migrating to a new design system wasn't an option: the engineering lift had no backend justification. We had to build on top of what existed and ship in 7 months.
02
Audience range
Tenured associates versus someone on their day one, across different site types, operational cultures, and for many, English as a second language. The design had to work for all of them.
03
Tribal knowledge
These tools were built over ten years as one-off fixes, with training passed down informally rather than documented. That chain shapes how associates expect workflows to behave: if it's a damaged item, a label reprint, an expiration date, there's already a mental path for it. We were designing against patterns that run deeper than habit.
Going to the floor

Real problems don't live in Figma or at a desk

I led a one-day design sprint to align on V3 strategy, then went looking for what the sprint couldn't tell me. I shadowed associates across 13 fulfillment centers. Three patterns kept surfacing across all of them:

01
Item blindness at Scan Item
Every associate switched to FCResearch within seconds of scanning, just to see item details.
02
Attribute gaps at Verify
PSs cobbled together item details from multiple tools because V2 only surfaced weight, dimensions, barcode, and FNSKU.
03
Amazon.com as a verification crutch
For high-value items, associates left the tool entirely and checked the customer-facing listing.
Field research photo of associate at scanning station
Strategy

Field research as the foundation for V3

The challenge

Cognitive load. Problem Solve already carries more context-switching than any other workflow in the building. Adding FCResearch fields into Sideline wasn't automatically the right move.

The approach

Understand what information associates were actually looking at, why they needed it at that moment, and how it helped them make the right call. The goal wasn't more data. It was the right data, at the right step.

The space
Problem Solve workstation on the fulfillment center floor

The physical context associates are working in when they use Sideline App.

The screen
Sideline App running on the workstation laptop

What Sideline App shows them at that exact step in the workflow.

The solution

Fixing the two moments that broke the workflow

These are 2 of 70+ UX improvements and counting, the two with the highest UX impact. The moments that most shaped the experience, and the patterns they set ripple out to the rest of the screens down the road.

Feature 01 · Scan Item Page

Surfacing container contents before associates have to ask

Every associate switched to FCResearch within seconds of scanning. V3 brings the data they were looking for into Sideline at the moment it's needed.

Before · V2
V2 Scan Item

Scanning a container surfaced almost nothing: a barcode, a bin location, a count. To understand what was inside, associates switched to FCResearch and looked items up one at a time.

After · V3
V3 Scan Item

V3 surfaces every item in the container at the moment of scan: catalog images, product titles, quantities, and status. Associates can triage, prioritize, and handle easy fixes without switching tools.

Feature 02 · Verify Page

Restructuring around how PSs actually look for information

The Verify page runs as the same component across decant, stow, and pick. Getting it right here meant getting it right across the whole workflow.

Before · V2
V2 Verify

V2: weight, dimensions, barcode, FNSKU. Enough to confirm an item, not enough to resolve why it was a problem. Associates routinely pulled Amazon.com and FCResearch to fill the gaps.

After · V3
V3 Verify

V3 reorganizes around how PSs work: barcode hierarchy, bin fit, packaging requirements, expiration status, and the Amazon.com listing linked to the product name for visual confirmation without leaving the app.

Validation

The prototype that mimicked a warehouse

Problem Solve testing had to feel like the real thing. I built a conditional logic prototype in Figma that let associates pick up any item, scan it, and work through the full workflow as they would on the floor. Static screens couldn't validate something this branched.

Feedback didn't wait for formal sessions. A live operational channel kept associates, operations, and stakeholders in one place, so signal came in continuously and unfiltered. I brought builds onto the floor to test in context across sites, regions, and warehouse generations, from older Gen 1 facilities to the newest builds, confirming the workflow held up everywhere associates actually work, not just in ideal conditions.

01
Design sprint
Nov 2024
02
Design iterations
Mar–Apr 2025
03
Field testing
May–Jun 2025
04
Beta release
Jul–Aug 2025
05
Worldwide launch
Sept 2025
Impact

The numbers that moved

V3 launched worldwide in September 2025. Across 10,000+ associates, the consolidated workflow delivered measurable gains in efficiency, accuracy, and trust.

24%
less secondary tool usage
~3s
saved per resolution across 10,000+ associates
>10%
improvement in defect resolution
Adoption · Sep–Dec 2025

Within three months of launch, Problem Solve associates had migrated ~80% of processing volume to V3.

Feedback

What associates said

Feedback from Problem Solve associates following the worldwide launch in September of 2025.

Reflection

What this project taught me

01
Agile works here, but differently
UX worked ahead of engineering, using prototypes as the shared source of truth across design, eng, ops, and PM. Designs were finalized before engineering built, which meant bugs were caught before the pilot, not after rollout.
02
The floor is irreplaceable
Being onsite gave signal no research session can replicate. It kept iterations grounded in what associates actually needed, not just what requirements described.
03
Removing a tool isn't just a design problem
After launch, associates still opened FCResearch by habit, then noticed their screen looked different, then slowly started trusting the new information. That process takes time. It's also not fully solvable by design alone: edge cases in older generation sites and gaps in site-specific data mean there are problems the directed workflow can't fully capture.