Maersk
Enhancing Cargo Cut Visibility at Maersk
Maersk is one of the world’s largest integrated container logistics companies, specializing in connecting and simplifying trade through its extensive global network of shipping and supply chain solutions.
The Problem
Teams responsible for the optimization of intercontinental cargo ships lacked visibility to streamline their actions effectively. This gap led to inefficiencies in maximizing ship capacity, handling cargo drop-offs, and rolling over shipments to subsequent voyages.
The Solution
A digital product designed to:
Visualize key operational data.
Integrate machine learning (ML) analytics to forecast cargo drop-offs, enabling proactive decision-making
Discovery Phase: User Interviews and Workshops
Participants: 15 professionals, including managers and coordinators, from three critical teams:
Overbooking Team: Manages capacity exceeding and scheduling.
Cargo Cutting Team: Decides on drop-offs before the ship departs.
Rollover Team: Manages cargo shifted to future voyages.
Objectives:
Understand user pain points, motivations, key performance indicators (KPIs), workflows, and system expectations.
Sample Interview Questions
"Can you show me your current system?"
"What do you like and dislike about the existing tools?"
"What does your typical workday look like?"
Wireframe Development
First Iteration
Initial wireframes were created based on insights from interviews and internal discussions. Feedback from users highlighted areas for improvement:
Lack of clarity in terminology.
Unclear representation of forecast data.
Overemphasis on absolute numbers, where percentages were more relevant.
Second Iteration
Revised wireframes incorporated user feedback:
Enhanced visibility for key metrics.
Graphical elements to draw attention to critical data.
Clear focus on the central role of forecast analytics.
Balanced use of percentages and absolute values based on context.
Improved nomenclature for operational alignment.
New filters for refined data analysis.
Operational Terminology
Voyage: Refers to the specific ship journey, identified by direction (North/South), route name, and operational week.
POL (Port of Loading): Origin port where cargo is loaded onto the ship.
POD (Port of Discharge): Destination port where cargo is unloaded.
Booking Types:
Contract: Reserved slots for clients with ongoing agreements.
Bucket: Additional spaces contracted by the same clients.
Slot: Direct bookings made through Maersk's website.
Usability Testing
Moderated Testing
Participants: 4 users.
Goals:
Measure the time taken to identify primary actions (e.g., noticing the forecast).
Determine the success rate for identifying relevant voyage parameters.
Unmoderated Testing
Participants: 20 users.
Metrics Captured:
Time to complete specific actions (e.g., expanding a column).
Comprehension of actions.
Perceived value of the interface.
Results
Task: Expand analysis for a specific voyage.
Success Rate: 100%
Drop-off Rate: 0%
Usability Score: 86%
Direct Success: 100%
User Feedback:
Interface Intuitiveness:
"Very intuitive": 71%
"Intuitive but improvable": 29%
Overall Impression:
"Excellent": 75%
"Good": 25%
Meets Expectations:
"Partially": 57%
"Completely": 43%
Final Adjustments
Added new filters and refined nomenclature based on additional user suggestions.
Enhanced data visualization for better operational insights.
Introduced more actionable features for improved decision-making.
Handoff and Implementation
The backend was developed in-house, requiring close collaboration between design and engineering teams. I worked closely with machine learning engineers to integrate the forecast model and with frontend developers to ensure operational feasibility. I also facilitated knowledge sharing to align the entire team on data limitations and operational nuances.
Usability Testing
