
Project
Carvana logistics operational tool

Background
The current inefficient management approach results in higher costs and longer work hours. Field leadership, including logistics supervisors, managers, and regional logistics directors, lacks the resources, data, and demand visibility necessary to optimally staff and manage their teams. To address these challenges, a strategic management tool is needed to streamline expenses, enhance efficiency, and ensure that the warehouse operates at the lowest possible cost.
Role
User Research
End to end visual design
Interaction design
Prototyping & testing
Device
Desktop
Background
The current inefficient management approach results in higher costs and longer work hours. Field leadership, including logistics supervisors, managers, and regional logistics directors, lacks the resources, data, and demand visibility necessary to optimally staff and manage their teams. To address these challenges, a strategic management tool is needed to streamline expenses, enhance efficiency, and ensure that the warehouse operates at the lowest possible cost.
Role
User Research
End to end visual design
interaction design
Prototyping & testing
Device
Desktop
Challenge
How can we leverage existing data and managerial experience to effectively manage labor costs and reduce long work hours, ensuring that staffing levels and headcount align with productivity expectations?
The result
$6.56 CPU
Staffing planner
Reduce labor costs by accurately predicting staffing needs based on existing data, enabling leaders to effectively manage site productivity.
$1.64 CPU
Manual entries
Enhances the management of untracked work by manually inputting unrecorded activities, effectively bridging the gap in oversight.
$8.21 CPU
Activity tracker
By tracking the productivity of logistics associates in real-time, we significantly enhance efficiency and lower labor costs.
Recognize the obstacle
I conducted on-site appointments with managers and in-depth interviews to gather valuable insights not accessible from the office. By documenting essential information through notes, photos, and videos, I aimed to identify the root causes of high labor costs and inefficiencies.




Inaccurate staffing
Field leaders rely on past knowledge and workload sheets/assignment reports to determine each individual's daily staffing and job assignments. This approach makes it difficult to measure the accuracy of the plan and staffing levels.
Limited real-time visibility
80% of logistics activities are captured using apps, while 20% remain untraceable. We need to account for the untraceable activities and effectively use the trackable data to optimize headcount.
Tracking difficulties
Leader lacks the resources, data, and demand visibility needed to effectively staff and manage teams, hindering the warehouse's goal of low-cost operation. Addressing these gaps is essential for improving efficiency and performance.




Partner with leadership to optimize workload
After working alongside the general manager and transportation manager for several days, I gained valuable insights into their daily activities, tasks, and responsibilities. We can categorize all the tasks they need to complete and measure the time each task takes using existing data. By understanding shift times and the number of personnel available during each shift, we can determine how many people are needed for each hour and effectively distribute the workload across shifts.
Joshua
Associate General Manager
2.5 years experience
Daniel
Transportation Manager
5 years experience
Establish an optimized process for personnel allocation and track productivity.
Step 1: Preset total staffing headcount
~80% known Trackable logistics activities
+
productivity rate
Staffing headcount for Trackable activities
Total staffing headcount for ALL activities
~20% logistics Untrackable activities require manual time calculation for the remaining work
Staffing headcount for Untrackable activities
Step 2: Compare actual performance to preset benchmarks to assess staff productivity.
>
Beyond Exception
😀
🙂
Meet Exception
<
Below Exception
😐
Comparative analysis
Goal
To evaluate the current tools we use, identify their inefficiencies, and explore other products that could enhance our offering. By learning from these alternatives, we aim to refine our own product, integrating the most effective features to ensure a better fit for our needs.




Result
-
Quintiq UI is pretty bad, but it has all the information that users need.
-
MOMOS has built its own shift table which could be utilize to our performance table.
-
Deputy is pretty solid product, there are a lot of elements can be used in our product.
-
Monday and Insightful are the apps for shifts only, UI looks clean and has different graphics to show and represent data.
Define and structure the essential features
Through analysis, I have developed a method to calculate the required number of personnel for all tasks and assess whether actual work efficiency aligns with expected efficiency. This evaluation will help identify the necessary features and essential functionalities to be designed.
Show productivity rates for all tasks and allow shift time adjustments to ensure accurate data for users
Getting the right total headcount on each shift to complete the work for the day
Ability to allow leaders to count the work on the Inspection Center without using digital tools
Ability to see activity tracking placed over hours spent in each workflow by user.
Design explorations
Staffing planner explorations
After consulting with the data team, I prioritized displaying essential information to improve user experience, as including all details wasn’t feasible for the MVP.
We conducted several rounds of user validation to refine the design and ensure clarity and user understanding of the graphic elements. Here is the final version.


Activity tracker explorations
I explored several design approaches to differentiate between team and individual performance, ensuring users can easily interpret and act on the data.
Send out a Google Form with the prototype link and testing instructions. Users are already accustomed to this method.




Final key page design
Configurations
-
Integration of productivity rate and shift data: Feeds into Staffing Planner, Manual Entry, and Activity Tracker to provide comprehensive staffing insights.
-
Supports leaders: Enhances effective staffing management by aligning resources with actual needs.
-
Foundation for accuracy: Ensures precise headcount planning and helps meet performance expectations for better operational control.


Staffing planner
-
Data team model: Determines required staff based on task time, optimizing workforce planning.
-
Efficient scheduling: Allows better shift management to reduce costs.
-
Comprehensive staff overview: Provides visibility into daily and time-specific staff needs to enhance resource allocation.
Manual entry
-
Input: This table enables managers to input work that cannot be tracked through the app.
-
Time comparison: By comparing the allocated time for tasks with the time staff spends working, we can assess the appropriateness of our work assignments.
-
Data insights: entering this data provides valuable insights into the fieldwork conducted at each site.
-
Task Management: This process helps improve task management and enhances overall operational efficiency.



-
Integration of productivity rate and shift data: Feeds into Staffing Planner, Manual Entry, and Activity Tracker to provide comprehensive staffing insights.
-
Supports leaders: Enhances effective staffing management by aligning resources with actual needs.
-
Foundation for accuracy: Ensures precise headcount planning and helps meet performance expectations for better operational control.
Activity tracker
Week view
Day view

Overview

Details
Quality assurance
Our internal products often underestimate quality while favoring functionalities. To address this, I have prioritized product quality by encouraging engineers to resolve identified UI and UX issues. When encountering problems, I create a detailed Jira ticket outlining the issue for the engineers to address. They then work on the fix, which I review and approve before final implementation. This process elevates product quality and fosters a culture of continuous improvement.



Learning: Ideal vs Reality
As a designer, I strive for perfection, aiming to provide users with all the information they need—not just from an aesthetic standpoint. However, MVP cannot fully achieve this vision, and I must adapt the design to meet practical requirements.
During the development of the staff planner, I collaborated closely with our data science team to understand their model. Through our discussions, I learned that the data I believed would benefit users couldn't be provided because the model wasn't quite ready. As a result, I adjusted my design accordingly.
While things don't always go as planned, I consistently endeavor to deliver the highest quality in design and enhance customer satisfaction.