Sofvie

Sofvie Internal Dashboard

Designed to simplify and enhance client success analytics

Over a four-week sprint, I had the opportunity to work with a team and design an internal dashboard for key stakeholders at Sofvie, a company specializing in hazard and risk management tools.

I worked on this project with 3 other designers, (Nancy Liu, Kaitlin King, and Xue (Snow) Yu) and Sofvie's data science team to build out the application.

Role

UX Designer

Time

March 2023
(4 Weeks)

Tools

Figma, Figjam

Team

4 UX Designers
Sofvie's Data Science Team

Challenge

Sofvie needs a customer service operations dashboard

Sofvie is a software company that creates hazard and risk management tools for workforce safety in dangerous physical environments like mining and construction.

My team's role was to design a client/customer success operations dashboard to serve as an internal tool for key stakeholders at Sofvie.

Setting up for Success

Aligning Design Expectations & Technical Requirements

Sofvie's data science team, our main point of contact throughout the project, provided us with an in-depth list of each data point expected in this dashboard, compiled from discussions they've had with the stakeholders of the company.

This list outlined the data points that they felt were necessary for the Sofvie team.

📋 Key Takeaways from the kick-off call

Users

Users of this dashboard have different roles including: CFO, Client & Success, Quality & Services, and Sales.

Dashboard View

While this is the case, all users will access the same dashboard.

Development

Sofvie’s Data Science team will likely use Chart.js and Python libraries to develop the dashboard

Design Constraints

Hold off on customization for a future version, include all listed data points, and keep in mind library capabilities while designing.

Project Timeline

We had four weeks to complete this project. With this in mind, we planned accordingly to include feasible user research and testing as well as time to make adjustments to the final deliverables.

User Research

First, we interviewed Key Stakeholders with card sorting

With the time and this challenge in mind, we came to the idea of interviewing the key stakeholders of these different teams to learn more about what an ideal dashboard could look like that would aid them in making valuable business conclusions.

Ideation

Turns out, Key Stakeholders had conflicting understandings of this Dashboard

Too much data

During our last interview with the CFO we learned that he had envisioned a simpler version of the specs provided to us by the data analytics team and ruled out several data points due to confidentiality reasons.

Missing data

Additionally, during the interviews, stakeholders brought up new data points that they would find helpful as well viewing these points at a company level and site level and being able to compare the data to a different time period.

Discovering commonalities across interviews to make meaningful groupings

As we began to review the groupings from each key leader’s card sort, we were challenged with trying to find common ground in their results. Our approach was to utilize the home page as the area with the most important points among key leaders.

I proposed a method to my team to sort our data points by importance based on stakeholder feedback, identify prioritized points, and eliminate ones that didn't hold enough priority and be relocated elsewhere.

This approach ended up being effective and after considering all feedback from the team leads, we were able to provide suggestions on our end for when we next met with the data science team.

We created Site Map to visualize our final sort

After working through our card sort results/discussions, we grouped the data points into three pages: The Home Page, Engagement, and Customers.

We also proposed additions to allow for more nuanced and meaningful data.

From our interviews, we learned that the data science team’s specs may not align best with the CFO and the key leaders.

We also discovered opportunities to enhance the dashboard and make it more functional for our users needs.‍

Low to High Fidelity

Proposing our new, revised low-fi design

When we shared our low fidelity mockups with the data science team, we went through each chart and resolved all confusions and issues as they came up in one session.

Building our component system

When we shared our low fidelity mockups with the data science team, we went through each chart and resolved all confusions and issues as they came up in one session.

User Testing

Validating our Design through User Testing

After completing our high fidelity prototype on Figma, we conducted a user test on the key leaders to assess the usability of the dashboard and observe their interaction with it.

Refining after Feedback

The feedback we received across all key stakeholders ended up being quite positive! Aside from minor fixes and suggestions, everyone was satisfied with the functionality of the dashboard and the information it provided and they were grouped.

In our final handoff we provided the team with all of the examples from Chart.js and Python for chart visualizations, all major/minor changes we made to the initial project specs, and all remaining uncertainties we had that were out of our scope of involvement and for their internal team to decide.

Concluding Thoughts

Communication is Key

Early on, we learned that there was a disconnect between the groups at Sofvie on what they wanted for this dashboard. Initiating a conversation with all teams and stakeholders involved played a huge role in resolving this and checking in helped us keep everyone up to date and in agreement - it allowed us to improve this project for the better and resulted in a successful project that everyone was happy with!