Just a moment...

Building a centralized data dashboard:

LF Logistics data project

As the sole design person in a team of data scientists, I used my summer internship as an opportunity to independently conduct an end-to-end UX project for helping transport planners get the most relevant information they need at a glance.

Role

Data/UX Intern

Time

May 2019 - August 2019

Deliverables

Data dashboard, research report

Skills used

UX/UI design, user research, data visualization

Process overview

1. The challenge
2. Mobile app
3. TV dashboard
4. Final design

Lawrence Zheng

1. The challenge
2. Mobile app
3. TV dashboard
4. Final design

How might I

design to eliminate a company bottleneck at a logistics firm?

LF Logistics had recently transitioned to a new database management system, posing a challenge for transport planners who lack technical skills to access the data. They often relied on IT or the Supply Chain Analytics (SCA) team to create Tableau visualizations on demand, causing bottlenecks and inefficiencies.

While working as part of the SCA team to address this issue, I observed that my team, primarily data scientists, were focused on technical aspects rather than considering how end-users would interact with the data visualizations.
‍
Recognizing the need for a user-centered design approach, I advocated for ongoing iteration and end-user involvement. I clarified that UX encompassed more than aesthetics and proposed extending the development timeline for proper user research. Independently, I established and implemented a UX design process for the data visualizations.

The opportunity:

I can consolidate the most commonly used data into a user friendly platform

Through active engagement with various stakeholders within the organization and conducting user interviews with transport planners, I identified 3 main pain points that provided opportunities for improvement:
Based on the pain points, I felt that there was a great opportunity to consolidate the most commonly used data into a user friendly platform, instead of having many scattered visualizations that people don't know where to find.

Initial idea:

Mobile app for accessing data

I brainstormed solutions for data consolidation and decided that a dedicated mobile app would be suitable as a design solution. I felt this would address each of the three pain points I identified most effectively.
Re: confusing to navigate
A dedicated mobile interface would be able to consolidate existing visualizations into one place and make it easier for planners to find specific information.
Re: time/bottlenecks
If a request is commonly made for a certain type of visualizations, it could just be added to the app once and permanently be made available, rather than having the same request be made repeatedly.
Re: lack of convenient access
All planners should have access to their mobile phones, and they would be able to see information even while on-the-go. Lock screen notifications could also be utilized to alert planners of abnormalities.
I created wireframes showing the concept of creating custom “quick access” groups where planners would be able to personalize their experience and add visualizations they need to see frequently.

By selecting a group, planners would be able to view basic information for each visualization they added. Planners could also “press into” each visualization to view more detailed information (e.g. breakdown by facility), and be able to toggle between viewing flat numbers and percentages. The ability to set customizable lock screen notifications (through a “manage alerts” screen) addresses the need for users to view urgent information quickly.

The problem:

Mismatch with user context and organizational context

After creating basic mobile wireframes and conducting user testing and stakeholder feedback sessions, I received two critical insights that led me to halt further development of the mobile design:
Mismatch with user context
Transport planners were concerned about appearing distracted or unproductive by frequently using their phones. Given the predominantly office-bound nature of their work (managing remote communication with drivers and workers), a mobile solution appeared less appealing.
“I feel like if my manager saw me constantly on my phone, they would think I am slacking off browsing memes or playing Candy Crush.”
- Interview quote from transport planner,
illustrating why they wouldn't want to use a mobile app for data access
Mismatch with organizational context
Existing efforts within the IT and data warehouse teams focused on deploying mobile visualizations to the Tableau Mobile app, supported by upper management. It was challenging to justify a separate mobile application when resources were already allocated to the existing solution.
The IT team was already migrating dashboards to Tableau Mobile

Design solution:

Using a TV screen for real-time data

While at the office space of the transport planners, I noticed that there was a large TV that was turned off. I used that as inspiration for a TV dashboard idea.

Due to the layout of the control center the planners worked in, the TV dashboard could be deployed in a location where all the planners could just look up from their desks and instantly see the required information.

The head of the transport planning department was enthusiastic about this idea when I talked about it with him, and the team in general received the idea very positively.
Diagram showing birds-eye view of the office space for transport planners

Early low fidelity screens, used for A/B testing

Design A
Design B
Through A/B testing, I found that planners preferred the design and layout of Design A and were able to more quickly find the information they needed using it. However, there was some concern that certain components, such as the volume accumulated statistic, could be made easier to find. Research also made it clear that certain components, such as having a map or comparisons to past data, would not be very helpful for the transport planners.

Final design

This project from many years back taught me to embrace ambiguity and not get too attached to an idea that may seem "intuitive". I was initially discouraged by having to switch gears from a mobile design; however, thinking of it from a different way, I managed to develop a better end product that was more suited to the context.

NEXT PROJECT

Combating implicit biases

Bookstore Simulator