Manual refresh controlled awareness
Employees refreshed to find new orders, then printed the oldest orders first. Priority had to be visible before work began.
Featured product case · Pick Pack Ship
PVH: Ship from Store
A three-month product effort that connected order data, store inventory, employee movement, paper documents, packing, and shipping so closed stores could operate as local fulfillment points.
PVH · Tommy Hilfiger2021Embedded UX leadThree-month delivery
Featured case at a glance
Challenge
Turn closed stores and their inventory into local fulfillment points.
My role
Embedded UX lead across research, workflow design, prototyping, testing, and delivery support.
Delivery
A connected pick, pack, and ship workflow developed with a product owner and engineers in three months.
The business problem
Turn stores into local fulfillment points.
During COVID store closures, online demand and in-store inventory were separated. PVH needed a clear way for store teams to receive orders, locate products, pack items, handle exceptions, prepare documents, and complete shipments within existing systems.
3 months
Discovery to delivery
3 stages
Pick, pack, and ship
2 contexts
Mobile and desktop
3 rounds
Employee feedback
My role
Employee research, workflow design, prototyping, interaction detail, and delivery support.
Team
Product owner, engineers, store managers, and store associates.
Constraint
Fit new work into existing inventory, order-management, devices, documents, and store spaces.
Delivery
A connected workflow, training guidance, and a prioritized enhancement backlog.
01 · Research in context
The workflow started in a stockroom, not a screen.
A store walkthrough and employee interview revealed a process spread across a crowded back room, printed pick lists, nearby inventory, a shared workstation, another computer upstairs, and the till.
One store interview
59
orders on the first sale day
80–100
orders per day at the Christmas peak
3–4
employees typically involved
10
orders in a printed pick batch
This is an operational snapshot from one employee interview, not fleet-wide performance data.
Paper was a portable interface
Printed documents moved with employees when the shared workstation could not. The product had to support that bridge, not erase it.
Store layout changed the job
Stock location, device access, staffing, and handoffs determined whether employees worked order by order or in batches.
02 · Model the service
Map the entire job before polishing the interface.
The wireframe board connected order intake, prioritization, picking, packing, printing, final checks, and shipment. It also made the physical handoffs between desktop, mobile, paper, stockroom, and till visible to the whole team.
Prototype A · Batch plan
Select several orders, then pick.
Employees could make a plan, create a printable batch, move through inventory, and return to packing by order.
Prototype B · Persistent overview
Keep priorities and order state visible.
Employees moved from a stable order queue into item-level picking and packing while status remained available for handoffs.
03 · Compare and refine
Prototype comparison made the tradeoffs discussable.
This was comparative prototype testing, not a live production A/B test. Employees and stakeholders could walk through both models and react to speed, clarity, physical fit, priority visibility, and recovery when inventory did not match the order.
What we compared
Batch selection versus a persistent order overview, plus variations in status, terminology, document printing, and exception handling.
What we listened for
Where employees hesitated, lost context, needed paper, asked another person for help, or could not tell what would happen next.
How the product evolved
Later prototypes combined a visible queue with printable batches, explicit item states, recovery paths, final checks, and order history.
Production A/B measurement plan
Hypothesis
Keeping urgency and the next action on each order card will help employees begin the right order faster without increasing fulfillment errors.
Variants
- ●A: Time-left indicator and primary action shown directly on each order card.
- ●B: Priority sorting with urgency summarized separately from the order card.
Primary metric
Median time from opening the queue to the first pick action.
Guardrails
Wrong-order starts, unavailable-item errors, and reassignment rate.
04 · Connected prototype
Walk through the desktop fulfillment flow.
The prototype begins at the picking queue and continues through packing, unavailable items, final checks, document printing, completion, and order history.
Start with an order in the PICK queue. The connected path covers both the happy path and operational exceptions.
Open full prototype ↗05 · Design beyond the happy path
The product had to remain clear when reality changed.
Order fulfillment depends on inventory accuracy, staff availability, printers, packaging, and handoffs. The interface needed to make those changing conditions visible and recoverable.
Keep priority visible
New-order notices, time-left indicators, and clear primary actions helped employees decide what needed attention first.
Bridge screen and paper
The workflow supported mobile movement, desktop review, and printed pick lists because the real job crossed all three.
Show item state clearly
Picked, packed, unavailable, and completed states stayed visible so the next employee could understand what had happened.
Design recovery paths
Out-of-stock, missing, and damaged items led to explicit recovery and reassignment decisions instead of a dead end.
Close the operational loop
Final checks, document printing, completion history, and tracking details supported the work after picking ended.
06 · Stay through delivery
Approved screens were not the finish line.
I stayed close to the product owner and engineers as the application moved into development. The work included interaction logic, system constraints, edge cases, employee guidance, and future improvements.
✓
Flow and state logic
Detailed behavior across order priority, picking, packing, exceptions, reassignment, printing, completion, and history.
✓
Engineering partnership
Ongoing clarification of interaction rules, system constraints, and edge cases as the engineering team built the product.
✓
Employee readiness
Training guidance and a prioritized enhancement backlog supported rollout and future improvement.
Relevance to bol
Product design inside complex physical and digital operations.
Relevance to bol
✦Learning from frontline employees in the real service environment.
✦Designing employee tools for time-sensitive operational work.
✦Connecting mobile, desktop, paper, and physical handoffs in one service.
✦Comparing prototypes before committing to a workflow model.
✦Designing clear exception handling and recovery paths.
✦Working directly with a product owner and engineers through delivery.