This case study demonstrates how user-centered design principles can transform complex data analysis tools into intuitive experiences. By implementing the ability to filter on aggregated values directly within the platform, we eliminated the need for users to export data to external tools, drastically improving their workflow efficiency and analysis capabilities.
This case study demonstrates how targeted usability testing led to practical solutions for power users, how to balance technical constraints with usability needs, and how thoughtful information architecture can dramatically enhance the data analysis workflow. Whether you're interested in data visualization, reporting tools, or redesigning complex systems, this study offers actionable insights into our approach to solving challenging filtering problems.
Problem: Discovery Boards, A powerful in-house BI tool, while, powerful for data visualization, lacked the ability to filter based on aggregated values, forcing users to export data to Excel for further analysis, causing workflow disruptions and inefficiencies.
Solution: I redesigned the filtering experience to include "Filter on Aggregations" capability, allowing users to apply measure filters directly within Discovery Boards while working within existing technical constraints.
Impact: The new design eliminated the need for data exports, reduced analysis time, and significantly improved user satisfaction by enabling complete data analysis workflows within a single platform.
Filtering on aggregated values shouldn't require leaving your analysis tool.
When users needed to filter data based on aggregate metrics in Discovery Boards (like showing only departments with average salaries above $80,000), they faced a frustrating roadblock. Despite having powerful visualization capabilities, Discovery Boards couldn't filter on these aggregated values—a critical gap in functionality that disrupted analytical workflows.
Users were forced to export their data to Excel or other tools, apply the needed filters there, and then potentially return to Discovery Boards with new insights. This wasn't just inconvenient—it was breaking the analytical flow and causing significant inefficiencies in data-driven decision making.
The challenge was clear: How could we enable filtering on aggregated values within Discovery Boards while working within technical constraints and maintaining a user experience that aligned with existing mental models?
