Smart Staffing Board

Stage 1: User Research

Contextual Inquiry Interview Questionnaire

Fulfillment centers (FCs) are complex facilities where many different kinds of management and collaboration occur. During an on-site visit at fulfillment center BFI1, an operations manager mentioned that they would like to upgrade their physical staffing board to a digital one because it’s time-consuming for people to use it.

To understand how people use the existing staffing board, identify main problems and underlying needs, I spent tons of time with real targeted users. By using the methods listed above, I collected valuable first-hand data and did observations of associates in their actual work context.

Stage 2: Identify and Prioritize Problems

Affinity Diagram

After went through all the data that gathered from the previous phase, similar responses begin to emerge naturally. This helped me identify patterns of how people interacted with the staffing boards. And it also helped me discover problems, find design opportunities and understand design constraints.

To deliver something in a limited internship time, I chose the following problems as high priority problems to solve:

1. Associates have difficulties finding their tiny name magnets on a big and chaotic board in the crowds.

2. Associates are supposed to check the staffing board four times a shift, but more than 80% (9 out of 11) of them only check it two times.

3. Managers have to go to find associates before they can change their assignments.

Stage 3: Divergent Design

Persona Storyboard Information Architecture Brainstorm

Based on what I found in the previous steps, I created personas and storyboards to envision how associates interact with the Smart Staffing Board. Besides that, a chart with all the information involved in the system was helpful to understand what information was available and what their relationships were.

Multiple directions and solutions were brainstormed and compared. The first direction was showing people's work information along with co-workers who work next to them on that specific day because I thought it would be easier for them to find their work stations if they see one of them. The second direction was only showing each individual's information but with more details, such as their skillsets, shifts, ambassadors, etc.

After showing different directions and solutions to associates, managers and got some feedback from them, I finally decided to push forward with a new direction that combined different advantages of each idea.

Stage 4: Iterative Design

Design with Real Data

The visual design was iterated with real users' data because real data can help me find visual design problem as early as possible.

One of the design decisions I made here was to put the check-in status bar on top of personal information. Initially, I thought it might help users to identify themselves quickly if they saw their personal information first. However, since one of the design goals was to urge associates to check the staffing board four times a shift, it makes more sense to let people know their check-in status every time after they log in the system.

Stage 5: Hardware + Software Prototyping

High-fidelity Prototype Code Refactoring Object-oriented Programming

The iteration process not only appeared in digital design phases, but also could be found during hardware and software prototyping stages. Insist on the highest standards is always one of my tenets. Although there was a version of the fully functional code, I did another round of code refactoring by using Object-oriented Programming, which reduced 23.2% of the code to finish the same job.

Stage 6: User Testing

A/B Test Questionnaire

After finished a fully functional prototype for associates, I was ready to test it. The primary goal of the test was to verify the hypotheses that a digital staffing board can accelerate the checking process for associates and improve their user experience.

I applied an A/B testing to compare associates’ performance of using the Smart Staffing Board against the existing one. The simulation also answered the question of how useful and efficient the design based on short qualitative surveys.

To get more accurate test results, I generated the UI based on associates’ real assignments of the test day. Associates could check out their personal assignments by scanning their Amazon badges.

Test Results

Data Analysis

The average time spent by 11 associate participants under a controlled condition (check the board one by one) was reduced from 5.15s (on the current board) to 2.74s (on the Smart Staffing Board), which decreased by 46.80%.

11 out of 11 associates reported that the work information on the Smart Staffing Board was “clear and easily findable.”

9 out of 11 (~82%) associates gave positive feedback on the user experience. These associates said they felt the Smart Staffing Board was faster; simpler; easier to read; more efficient, organized and private that the existing board.