DESIGN @ UW
Data Vizard : A data
visualisation dashboard
ROLE
Product Designer
TEAM
1 Designer, 2 Researcher
DURATION
12 weeks
TOOLs
Figma, Miro, Notion
CONTEXT
Analysts and data scientists deal with a lot of complex data every day, but presenting it in a way that’s clear and engaging can be a challenge. Tools like Tableau and Power BI are widely used, but they often feel too rigid or complicated when it comes to telling a story with data.
That’s where Data Vizard comes in. It’s designed to make data visualization easier and more flexible, helping users create better insights without the frustration. The goal is to provide an intuitive tool that helps people work more efficiently and collaborate more easily with others.
USER RESEARCH
To better understand what makes an ideal dashboard, we talked to a mix of professionals—2 early-career data scientists, 2 mid-career business analysts, and 1 VP of Data Science.
The goal was to dive deeper into their needs and challenges when it comes to designing dashboards and presenting insights to stakeholders.
RESEARCH FINDINGS
After talking to users, we found several common challenges they face when working with data visualization tools. These discussions uncovered pain points that impact ease of use, flexibility, and effective storytelling.
To make sense of the insights, we organized our findings using affinity mapping. This process helped us identify recurring patterns and group similar challenges together.
User Needs and Expectations
Users include data scientists, business analysts, project managers, and researchers, each with different needs.
Dashboards must be customizable to suit stakeholders and easy for the target audience to use.
Visual appeal is key—well-designed dashboards are more credible and persuasive.
Limitations and Flexibility Issues
Tools like Tableau and Power BI are useful but lack flexibility, especially for AI integration and custom visuals.
These tools are often complicated and not intuitive, requiring users to take lessons to fully understand them.
COMPETITIVE ANALYSIS
In this section, we compare two leading data visualization tools—Microsoft Power BI and Tableau—based on key user experience criteria. This analysis highlights the strengths and weaknesses of each platform, providing insight into how they cater to different user needs, preferences, and organizational requirements.
UX EVALUATION
After exploring the features and functionality of Microsoft Power BI and Tableau, I took a closer look at their visual design through a heuristic evaluation. This involved analyzing key interface elements like layout, readability, and visual hierarchy to see how well they support user needs.
DESIGN QUESTION
How might we design a data visualization dashboard that promotes creative data presentation while maintaining flexibility?
USER GOALS
This section outlines the key tasks and actions that users should ideally be able to accomplish within the data visualization dashboard.
Users should be able to import various types of raw data files, including CSV, Excel, and TXT, or manually enter data.
Users should have the ability to create and customize visualizations, with options to adjust charts, colors, and layouts to their preferences.
Users should be able to interact with and explore the data by applying filters, drilling down for more details, and engaging with the visualizations.
Users should be able to incorporate custom code or integrate with external tools like R , Python to create personalized visualizations.
Users should be able to share and export their data visualizations in multiple formats for presentation to stakeholders.
CARD SORTING
To better organize dashboard features, we conducted a card sorting exercise with users. They grouped elements like data sources, charts, and filters based on their understanding and workflow.
IDEATION TESTING
Keeping our design goals in mind, we tested our wireframes with users.
Import data
I explored ways for users to import data, including formats like CSV, XLS, third-party software integrations, and links from online databases. The goal was to create a solution that supports all these options seamlessly.
Actionable Tabs
I ideated variations to help users visualize and filter data without feeling overwhelmed. The goal was to provide easy access to key features while keeping the interface simple and intuitive.
FINAL DESIGNS
Keeping our design goals in mind and after a lot of iterations, we came up with the following designs.
Import data
Users can import various types of raw data files, including CSV, Excel, and TXT, or manually enter data.
Visualise and Customise data
Users can create and customize visualizations, with options to adjust charts, colors, and layouts to their preferences.
Filter data
Users can interact with and explore the data by applying filters, drilling down for more details, and engaging with the visualizations.
Share projects
Users can share and export their data visualizations in multiple formats for presentation to stakeholders.
FUTURE SCOPE
What more can we do?
How might we create a platforms that enables data scientists to integrate machine learning model explainability using SHAP game theory values?
How might we reduce friction and promote collaboration where users can work with each other real time?