Melissa Carmichael - Fractional Design Services

Company: Whole Foods & Amazon Fresh
Role: Product Design Lead
Timeline: 6 months (2 months Discovery, 4 months Design)
Released: 2024
Note: Due to the sensitive nature of this project, wireframes are only shown. Additional designs can be produced upon request.

Summary

A re-platformed tool to enable Distribution Center Buyers to forecast, buy, and monitor the supply and demand flow of $3B worth of perishable products, unlocking ~$10M in annual entitlement for Whole Foods & Amazon Fresh Grocery.

Challenges

This project had a false start with the Amazon Engineers at the helm; they had created a draft of the requirements document that did not meet the internal business stakeholders acceptance criteria. I was tapped to step in and quickly get up to speed, led and conducted interviews with users and business stakeholders to understand the problem space, paired with a PM to define a new set of requirements, and prototyped and designed a new user interface that would set the foundation of a multi-year, multi-phased migration plan. I was able to quickly show evidence that we were keeping the internal user goals in mind, build trust and alignment with the internal teams, while innovating on the new solution.

Discovery

My first step with any project is to orient and understand the existing solution. In this case, it was a macros-heavy Excel spreadsheet, saved and uploaded locally through each user's desktop. The goal was to build out as a web application, integrating AI/ML to optimize the forecasting and mangement of ordering at a Distribution Center level, which averaged thousand of pallets of perishible items.

I led user research Discovery interviews across 20 Regional Buyers to understand their workflows, challenges, and opportunity areas to establish project scope and areas for improvement. Above is an image capture of my notes to document the functionality, as well as call out pain points and opportunity areas for improvements.

There were 12 distinct data tables across 4 tabs and pop-ups in the existing experience, and users leveraged a multitude of external data sources to enable and de-risk their decision-making process.

Project Goals

The product team's goals covered two broad categories for the project:

• Create unified data view with new roles, permissions, and data storage mapping
• Ensure adoption & engagement with new tool through a seamless migration of experience.

We not only wanted to keep at feature and functionality parity, but unlock opportunity areas to build trust with the ML forecasting recommendations and make this tool their single source of truth when managing and monitoring orders.

Design Optimizations

NOTE: Due to the sensitive nature of this project, wireframes are only shown in this case study. Additional designs can be produced upon request.

• Optimized View: Show all relevant data in an accessible view to enable informed decision-making
• Smarter Automation: Reassess where automations in the ML recommendations could be easily turned off or overridden manually
• Streamline Workflows: Reduce time-on-task by consolidating repetative actions and converging multiple data sources
• Customizations: Identify areas that could be configurable based off of user preference
• Design Patterns: Align with common paradigms in existence with the Excel workbook experience to lower adoption friction

Outcomes

Several rounds of prototyping and testing were conducted prior to handing off to engineering for build. The Whole Foods leadership team were kept up to date and informed on the project's process and had full buy in. The pilot test was with a few Distribution Centers in Q1 of 2024.

The internal business stakeholders had confidence that the new tool would have easy adoption, and by conducting the up front discovery interviews and prototype testing, the Distribution Center buyers felt they had a direct line for feedback and have their voices heard, also helping to smooth the tool transition, increase adoption, and advocate it's use across the team.

This project was the foundation step of a multi-year, multi-phased migration plan to unlocking an estimated $10M annually for Whole Foods & Amazon Fresh Grocery through leveraging AI/ML to reduce waste and mistakes in large scale ordering processes.