<<< Back to all work ~/work/amazon-mcf.md
Product Innovation

Amazon MCF: transforming fulfillment

Amazon Multi-Channel Fulfillment, rebuilt for the businesses that run on it: simpler order management, sharper logistics, and orders that fulfill seamlessly across every sales channel.

>>> Role
Product Design Lead
>>> Skills
Product Design User Testing & Research Hackathon & Workshop Facilitation
>>> Company
Amazon · Contract
>>> Platform
Web & mobile
Amazon Multi-Channel Fulfillment platform on laptop and mobile
/// Amazon Multi-Channel Fulfillment
87%
Increase in seller adoption of the redesigned platform
4.2/5
Seller CSAT after launch, closing on the 4.3 target
24%
Reduction in support tickets

I led the design of Amazon's Multi-Channel Fulfillment platform (MCF): a solution built to simplify order management, improve logistics efficiency, and help businesses fulfill orders seamlessly across every sales channel.

01 What sellers were after

Drag-and-drop catalogs, real-time inventory, automated fulfillment. A growing seller is ready to take their product to the next level: selling more, and not just on one platform, but on every one of them.

But finding a fulfillment partner that can keep up is a real hurdle to that kind of growth. This is where Amazon's scale changes the math.

Amazon's advantage, by the numbers:

300+
Fulfillment centers worldwide
1.27M
Associates across the network
12.5K+
Trucks, vans, and sedans in the fleet

02 Where it broke down

MCF was powerful, but sellers, and their customers, were deeply frustrated. Even in the best moments, friction crept in: confusing language, no clear way to track an order, and tooling that assumed every seller looked the same. Enterprise sellers felt it most.

  • Opaque tracking: customers were left guessing where their order was.
  • One-size-fits-all flows: built for a single seller type, not enterprise volume.
  • Unnecessary friction: simple tasks took more steps than they should have.
  • Lost trust: the platform did not match how large sellers actually work.
MCF Seller feedback Churn risk
This platform was not made with enterprises like us in mind.
Seller
Brian
Role
Director of eCommerce
Company
General Mills

Brian wasn't an outlier. I went through the support data and talked to more sellers on the same tier, and the numbers said what he said.

15%
Decline in enterprise adoption
Increase in support tickets

03 What we set out to do

We set out to create a user-friendly, feature-rich MCF experience: one that gives sellers real-time visibility, flexible order creation, and tooling that scales from a single order to enterprise volume.

>>> Metrics
  • Decrease the number of support tickets.
  • Increase the live resolution rate.
  • Achieve a CSAT score of at least 4.3/5.

The metrics gave us a target, but not a starting point. For that we went back to the research, and one finding kept resurfacing: sellers were not sitting at desks. They were on warehouse floors, at markets, in transit between the two. That single observation set the direction for everything that followed.

04 Real-time tracking dashboard

Every shipment, in one place

All the critical shipment information, tracking and status included, brought together in one place. What follows isn't a set of screenshots, it's the shipment detail page itself, rebuilt in working code: click through live status on a route map, item-level contents, and problems surfaced before they become surprises.

~/mcf-dashboard.html ● live demo expand full size ↗
/// Click through it, this one is real. Live route map with carrier scan activity, item-level contents, and problems with a paid resolution path. Switch tabs inside the demo; the banner shortcut, refresh, and copy actions all work.

05 The orders hub

Customer orders, rebuilt as a triage tool

Before a shipment can be tracked, it has to be found. Every session starts on the customer orders page, and the original treated that job like a database dump: no way to triage, counts in three formats, orders with no identity beyond a number, language drifting between the nav and the page, and a dense seven-column table with no answer for the sellers who run their business from a phone.

The rebuild changes the page’s job: quiet when everything is on track, loud where a seller needs to act. And because the platform’s strength is mobile, the mobile-first commitment carries through, so triaging orders works just as well from a warehouse floor as from a desk. All five fixes are pinned on the live rebuild below. Hit walkthrough in the demo bar to step through them in place.

~/mcf-orders.html ● live demo

Customer orders

Fulfilled by Amazon across all your connected channels
Order
Created
Ship speed
Status
Delivery
· 12 min agoShopify · Maya Alvarez
12 min ago
Standard
Processing
Est. Oct 24On track
· 3 hr agoWalmart · Priya Raman
3 hr ago
Standard
In transit
Est. Oct 25Delayed · was Oct 22
· 5 hr agoEtsy · Jonah Reyes
5 hr ago
Standard
Processing
On holdMissing Amazon barcode
· 8 hr agoShopify · Amara Diallo
8 hr ago
Priority
Out for delivery
Arrives todayOn track
· YesterdayBigCommerce · Sam Porter
Yesterday
Expedited
Delivered
Delivered Oct 18
· Oct 16eBay · Chris Bailey
Oct 16
Standard
Canceled
Canceled by seller
· Oct 15Walmart · Lena Fischer
Oct 15
Standard
Returned
Refund issued Oct 18
1–7 of 5,149
/// Live rebuild in HTML/CSS/JS. Tabs, search, and bulk select work. Hit walkthrough above to see the five design notes pinned on the UI. Sample data, not connected to Amazon.
/// The orders hub, rebuilt. Quiet when everything is on track, loud where a seller needs to act: the delayed order and the missing barcode surface instantly. On a phone, the rows become cards.

06 Flexible order creation

From a single order to enterprise volume

Order creation needed to flex from one order to bulk enterprise volume, without ever making the simple case feel complicated. Like the shipment tracker, this isn’t a screenshot — it’s the create-shipment flow rebuilt as a working LTL quote engine. The old screen was a form that asked for numbers and then showed a price it couldn’t justify; this one earns the price. Set the lane and dock accessorials, add commodities by handling unit, and every figure is computed live and traceable: weight and volume set the density, density sets the freight class (never typed), and class plus distance plus accessorials produce the rate — shown across three service levels so cost trades honestly against transit time.

~/mcf-create.html ● live demo expand full size ↗
/// The quote engine, rebuilt. A labelled four-step flow — Lane & Facilities, Freight Details, Rates, Review & Book — with a live-quote rail that always reconciles. Freight class is derived from density, accessorials move the price, and the three rate options trade cost against transit time. Hit Continue to walk the whole flow, or expand it full size.

07 Mobile first

Designing for where sellers actually work

So we committed to it: mobile is not a smaller screen, it is where many sellers run their business. Plenty manage everything from a phone, so I designed for mobile first and let the desktop follow. Every experience in the sections above was shaped by that decision.

No screenshots here either: the two demos above are running below at phone width, exactly as a seller sees them. Scroll inside each device — same code, no separate mobile build.

~/ · 390 px viewport ● live demo

01  Built for speed. The tracker — status, route map, and live scan feed — stacked for a glance from the warehouse floor.

02  Manage anywhere. Booking LTL freight from a phone — stepper, accessorials, and the live-quote rail all reflow.

/// One responsive codebase. These are the same pages as the demos above, at a 390 px viewport — the layouts you get on a phone, not mockups of them. Scroll inside either device.

08 What I'd take forward

  • Look beyond the assumptions. Designing for a true Amazon FBA branch, rather than a simple restock, taught me to question the brief and design for the problem underneath it.
  • Mobile first, going forward. Prioritizing smartphone users can dramatically increase adoption, especially for sellers in fast-growing and developing markets.
  • Continuous improvement. Shipping is the start, not the finish. Continuous feedback and iteration are what drive lasting success and higher satisfaction.
/// End of case study. Next time I'd start on the warehouse floor.
>>> Thanks for reading this far

Say hi

/// No newsletter, no funnel. /// Just email me if this resonated. /// I always write back.
Contact me →
/// Listen · summary 0:00 / 0:36