Cubic provides transportation solutions for large cities. Their Gridsmart software product is designed for traffic engineers to optimize traffic flow and safety at intersections.
Problem
Gridsmart's product leaders do not know what factors are causing customer trust to decline, how to improve it, and are uncertain about how to enhance the user experience.UX Design lead working with Product, Development, Marketing, and Computer Vision teams.
Outcome
The director of product was able to make 2 key business decisions based on my research and design work laying the ground work for an additional revenue stream from subscriptions. I provided actionable insights from 18 hours of interviews with key stakeholders and users from 10+ transportation agencies.
Fix existing system bugs before introducing new features to rebuild trust and improve people's experience using our product. Customers have lost trust in our product because software updates often introduce bugs. People hesitated to upgrade their software, fearing that updates would break their system. This mindset was preventing us from selling new features and creating additional value. Focusing on stabilizing the system before rolling out new features allowed us to rebuild customer trust.
Keep the current contract based pricing model until customer trust is regained. Our customers needed to see a massive increase in value before they would be willing to consider a subscription based model for our software. Our roadmaps and timelines were adjusted to ensure we had time to understand and deliver the value our customers expected in order to move them to a subscription.
Simplifying the interface
Monitoring intersections used to involve switching between three different applications. I created a simplified single interface to be implemented after the current bugs with the system were addressed. Adding a right details pane along with search and filter options on the left allowed traffic engineers to see all the information about the intersection they managing on one screen saving them an estimated 1.5 hours per day.
Search and Filter: Traffic engineers can quickly find intersections based on how the traffic lights are operating, alerts, or location.
Monitoring devices at intersections: Each row can be expanded to show traffic engineers which devices are at each different intersection and how they are operating.
Device info details: Each device has a detail pane showing it’s status, power level and other technical information about it’s health allowing the traffic engineer to understand how the hardware at an intersection is performing
Discovery
In order to better understand users, their needs, behaviors, and the context in which they interact with the product I did discovery research.
Stats
15 Moderated Interviews
4 Key Cross-Functional Stakeholders involved
In this section:
Research Goals
User Interviews, Analysis, and Insights
Goal Findings and Recommendations
Research Goals
Uncover challenges related to Gridsmart software that users encounter while monitoring their intersections.
Uncover challenges related to Gridsmart hardware that users encounter while monitoring their intersections.
User Interviews, Analysis, and Insights
To meet the project’s goals I organized, scheduled, and conducted interviews with 15 of our customers to gather insights about their needs, behaviors, motivations, and pain points. I collaborate with the product manager and director of product to determine the most important things we wanted to learn during the interviews. Then I crafted interview questions. I invited the product manager to join the interviews so they could hear from our users first hand. They joined when their schedule allowed and were delighted to hear the insights our customer’s were sharing. After the interviews I gathered and analyzed the data to identify common themes and actionable insights to inform design and development.
Analysis of user feedback grouped in themes
Sharing research insights Stakeholders: After all the interviews were complete I created an overview presentation deck for our director of product. He successfully used it communicate our user’s needs to his superiors for project prioritization and resources allocation. Product and Dev Teams: In order to help our team build empathy invited a key member of our development team as optional to interviews so they could hear from our users first hand. I summarized all interviews to 1 page summaries highlighting insights and opportunities that could be read in less than 5 minutes and made them available in to everyone. I continually documented and shared highlights from interviews with our team in confluence and Microsoft teams.
Goal 1 Findings
The first goal of this research was to “Uncover challenges related to Gridsmart software that users encounter while monitoring their intersections.” I uncovered 4 main issues and made recommendations on how to fix them to stakeholders to solve the issues and aid their decision making.
The system is unstable
The system does not perform well as more devices are added
Updating processors and logging in is time consuming and frustrating
Users cannot complete their work with the product without switching between multiple different interfaces
1. The system is unstable Users spend a lot of time diagnosing and troubleshooting. Cameras go offline frequently, setting up the network and keeping it running is challenging, updates can be unstable and may introduce new issues.
User interview
Recommendations to solve the issue Before creating new features, focus additional development resources on fixing existing bugs. Find a way to test the product’s stability in a variety of different network environments and firewalls configurations. This could mean testing software with a couple of partner transportation agencies with different network configurations.
2. The system does not perform well as more devices are added Users may experience lag when trying to stay connected to 15+ (sites) intersections. Instability increases with scale as more sites are added.
Recommendations to solve the issue Allow users to to view and monitor their intersections in groups and improve the systems performance so that it can accommodate monitoring and management of hundreds of intersections.
3. Updating processors and logging in is time consuming Users do not have an easy way to update all systems. They are often required to re-enter their login credentials multiple times during a working session.
User Quote: “And there's sometimes, I can come in the morning and log in and I'll have to verify (type in username and password) like 15 different intersections and I don't know why.” - Michael R
Recommendations to solve the issue Once the user logs in persist their session across all their intersections until they end their session.
4. Users cannot complete their work with the product without switching between multiple different interfaces. Users must switch between multiple applications and interfaces to complete there work. This is not efficient.
Recommendations to solve the issue Create a single unified interface where users can execute all tasks
Goal 2 Findings
The second goal was to “Uncover challenges related to Gridsmart hardware that users encounter while monitoring their intersections.” I uncovered issues with the camera’s mounting hardware.
The camera mounting hardware is not secure, causing shifts in the view Users frequently redraw vehicle detection zones in the software application because of shifts in the camera’s position. This can be caused by weather, i.e. wind, storms, heat (which makes the pole the camera is mounted to sag) etc.
Recommendations Explore options to reduce the amount of work a user must do to have vehicle detection zones working properly.
Option 1. Automatically redraw zones:Collaborate with the computer vision team to adapt and expand computer vision’s capabilities and automatically redraw detection zones to keep them aligned to the position they were first set to. This could be achieved by taking a snapshot of the camera’s view upon initial setup and then making continual adjustments when needed to keep the view as close to the reference image as possible. More conversations are needed with the computer vision team to determine the impact on the processors resources and to define the best approach.
Option 2 Hardware upgradeCollaborate with the hardware team to discuss upgrading the mounting hardware such that the camera readjusts itself or does not shift and move due to weather events like storms or high heat. This could mean mounting the camera on a gyroscopic device and or producing or purchasing higher quality brackets that clamp tighter to the mounting pole.
Solution - UI Concepts
After I finished research and shared the results with the stakeholders, dev team, and product team I begin drafting a new interface.
Early drafts of new interface
Early low fidelity drafts of the new interface
Updated home dashboard
Original Layout The original layout did not have all the data users needed to make decisions about the sites they were monitoring. It also did not have an easy way to drill into a specific site for more info.
Updated Layout The updated modernized layout allows users to
Search and filter across all their sites to find specific information
Get an overview of all sites quickly and expand a sites details for more information
Decide what column headers they want so they can view data by the categories most relevant to their current task
Get a quick overview of the status and health of all their sites
Results
2 Key business decisions made Based on my research our VP of product was able to confidently decide the following
Prioritize fixing existing system bugs before introducing new features. I discovered that customers were reluctant to purchase new offerings due to a lack of trust in our ability to deliver reliable features. By addressing these existing problems first, we were able to focus resources on rebuilding customer trust, which ultimately will drive revenue more effectively than developing an upgraded system.
Our customers need to see a massive increase in value before they would be willing to consider a subscription based model for our software. Our roadmaps and timelines were adjusted to ensure we had time to understand and deliver the value our customers expected in order to move them to a subscription
Conclusion
This work demonstrates the transformative power of UX research in revitalizing a legacy enterprise product. Through just 15 moderated interviews over 2.5 months and collaboration with key stakeholders, I identified critical user needs and pain points that had been damaging the product’s reputation for 5+ years. By integrating UX research into the product development process, we began to regain customer trust and aligned the product with user expectations. This approach not only addressed immediate issues of uncovering user challenges with the system but also established a foundation for new features and functionality in the future.
