Column Usage Behavior Analysis
Impact of Customizable Columns on User Behavior
Project Duration
Q4 2024 - Q1 2025
My Role
As Lead UX Researcher, I spearheaded the quantitative analysis of column usage, applying statistical methods (e.g., t-tests, correlation analysis) to uncover behavioral shifts and deliver data-driven recommendations for UI optimization and feature enhancement.
Team
Project Overview
This quantitative UX research analyzed financial advisor interactions with customizable data columns (approx. 1000 user sessions) in a trading platform. By statistically comparing pre and post-feature delivery data, the study identified significant shifts in column usage, saved view creation (788% increase), and information hierarchy, providing actionable insights for optimizing the platform's information architecture and improving user workflow efficiency.
Methods
Tools
The Challenge: Taming Information Overload
Financial advisors were drowning in data. The trading platform presented a vast array of columns, but this often led to information overload, hampering quick decision-making and efficient workflow management. The existing system lacked the flexibility to easily tailor data views to specific analytical needs.
Data Overwhelm
Too many columns, too little clarity?
Data Overwhelm
Advisors struggled to manage and interpret over 50+ potential data columns, making it difficult to isolate critical information quickly.
Inefficient Workflows
Was column setup a time sink?
Inefficient Workflows
Default column layouts were often suboptimal, forcing users into repetitive manual adjustments and slowing down analysis (avg. 24+ mins/session).
Lack of Personalization
Could users truly make the view 'theirs'?
Lack of Personalization
Limited and cumbersome customization meant users couldn't easily save or switch between tailored views for different tasks, with only 22% using saved configurations.
Impact of Inflexibility
- •Time wasted on manual column setup.
- •Increased risk of missing critical data points.
- •Frustration due to rigid, non-intuitive interface.
- •Inability to efficiently switch between analytical views.
Our Research Aimed To:
- •Quantify time spent on column configuration.
- •Identify most valuable columns for key advisor tasks.
- •Understand user needs for personalized data views.
- •Lay groundwork for an intelligent column management system.
The Quest for Clarity and Control
The challenge was clear: empower advisors by transforming data display from a rigid structure into a flexible, personalized tool. This research aimed to understand their core needs to build a system that surfaces the right data at the right time, effortlessly.
Key Research Questions
We aimed to understand how financial advisors actually use columns in their daily workflows to optimize the information architecture and improve the overall user experience with data presentation.
Understanding User Adaptation & Behavior
To analyze how the introduction of customizable columns shifted user interaction patterns and column usage.
- Q1:
How did the introduction of customizable columns affect overall user behavior and workflow efficiency?
- Q2:
Which specific columns saw the most significant changes in usage frequency (increase/decrease) post-customization?
- Q3:
What were the key differences in column interaction patterns (e.g., sorting, visibility) before and after the feature update?
Investigating Customization & Configuration Patterns
To identify common patterns in how users created and utilized saved column templates and configured their views.
- Q1:
What patterns emerged in the creation and usage of saved column templates (e.g., common column groupings, naming conventions)?
- Q2:
How did the visibility and prioritization of specific columns change within user-customized views?
- Q3:
What correlations were observed between different types of columns being used together in custom configurations?
Informing Design & Optimization Strategies
To translate behavioral insights into actionable recommendations for improving default settings and future enhancements.
- Q1:
How can insights into frequent column combinations and saved templates inform better default column arrangements?
- Q2:
What are the actionable design implications for improving the discoverability and usability of less-utilized but potentially valuable columns?
- Q3:
Based on observed behaviors, what future enhancements or new features could further optimize the column customization experience?
Guiding Principles: From User Behavior to Optimized Design
These research questions were strategically formulated to bridge the gap between observing raw user behavior with customizable columns and deriving actionable design principles. By dissecting how users adapted to, configured, and valued different data points, we aimed to establish an empirical basis for refining the platform's information architecture. The answers to these questions directly informed the quantitative usage analysis, comparative behavioral studies, and pattern mining methods employed, ensuring that our subsequent research efforts were laser-focused on creating a more intuitive and efficient user experience.
Research Methods
To understand advisor interaction with data columns and drive UI optimization, we employed a multi-faceted quantitative approach.
Quantitative Usage Analysis
Tracking Column Interactions at Scale
Analysis of approximately 1000 user sessions, focusing on column visibility, interaction frequency (e.g., sorting, filtering), and configuration patterns within the trading platform's 'Account and Groups' tab.
What this revealed: This large-scale analysis provided a clear picture of which columns advisors used most, ignored, or struggled with, forming the basis for data-driven UI improvements.
Usage Frequencies (~1k sessions)
Comparative Behavioral Analysis (Pre/Post Feature)
Pre/Post Customizable Column Feature
Statistical comparison (t-tests) of column usage metrics (M=79.3 to M=67.2) before and after the introduction of customizable columns, demonstrating a significant shift (p=.001) in user interaction patterns.
What this revealed: By comparing behavior, we precisely measured the impact of the new customization feature, confirming users were adapting workflows and identifying where adoption was strongest.
Column Usage Shift
Column Configuration Pattern Mining
Discovering Common User Setups
Identification of frequently co-occurring columns in user-saved views through correlation analysis, revealing common information grouping strategies and workflow-specific data needs.
What this revealed: This helped us understand *how* advisors organized their information, uncovering natural data relationships that could inform better default layouts and template suggestions.
Saved View Patterns
New Feature Adoption & Engagement Tracking
Measuring Impact of New Columns
Measuring the adoption rate of newly introduced columns (e.g., 'SWP Amount' at 29%) and the overall engagement with customization features, evidenced by a 788% increase in saved view creation.
What this revealed: Quantified the success of new features and the general appetite for customization, highlighting which additions were most impactful and where further user education might be needed.
Adoption & Engagement Increase
Foundation for Process: These quantitative methods (Usage Analysis, Pre/Post Comparison, Pattern Mining, Adoption Tracking) formed the backbone of our research. They allowed us to statistically measure behavioral shifts, identify user-preferred configurations, and quantify the impact of new features, directly informing UI optimizations for column management.
Research Process
Our data-driven approach to understanding column usage patterns and optimizing workflows
Column Usage Analysis
Data Extraction
Power BI Visualization
Pattern Recognition
PM Implementation
Column Usage Analysis
Analyzed which columns are used extensively by financial advisors and identified potential template combinations to streamline workflows.
Data Extraction
Located and extracted scattered data from multiple databases, untangling complex JSON file structures using Power Query.
Power BI Visualization
Imported processed data into Power BI for cleaning and visualization to understand column usage patterns by different user types.
Pattern Recognition
Applied machine learning and regression analysis to identify meaningful connections between column selections and user behaviors.
PM Implementation
Delivered comprehensive data and visualizations to product managers, enabling them to create more effective column management components.
Key Findings
Our analysis of approximately 1000 user sessions revealed critical insights into how financial advisors interact with AG Grid column configurations
Increased View Engagement
A 788% surge in weekly saved view creation (from 1.8 to 14.2) followed new capabilities, with 100s of new unique views created in three months, showing strong adoption.
Shift in Column Usage
Original column usage significantly dropped (Mean: 79.3 to 67.2, p=.001) as users adapted workflows to the new customizable options.
Selective New Column Adoption
Adoption of new columns was selective. 'SWP Amount' was most used (29% of views), while 'Est Tax Rate' saw minimal uptake (4%).
Terminology Impact on Usage
Consistent naming significantly influenced column co-usage. Related terms like 'PIP', 'SWP', 'G/L' were frequently grouped, aiding discoverability.
Key Research Impact
The analysis revealed a dramatic increase in saved view creation (788%) and identified opportunities to improve column discoverability through consistent terminology. New columns showed selective adoption patterns, with SWP Amount leading at 29% usage.
Impact & Outcomes
How our column management system redesign transformed data workflows and delivered measurable efficiency gains
Research Insights
Users spent 14.5 min/day managing columns
68% frequently recreated similar column sets
73% struggled finding relevant columns
Implementation
Created intelligent column grouping system
Built template system with one-click application
Implemented AI-driven column search & suggestions
User Impact
83% reduction in column configuration time
Templates shared across teams increased 217%
94% of users reported easier data access
Time Savings
Annual time savings across the organization from streamlined column management and template sharing.
Data Accuracy
Decrease in data interpretation errors due to consistent column templates and better organization.
Financial Impact
Productivity gains and improved decision-making translated into significant cost savings.
Workflow Transformation
Before Redesign
Manual column selection for each new analysis
No ability to save or share configurations
Difficult column discovery with 500+ options
After Redesign
One-click application of saved configurations
Enterprise template library with 350+ templates
Smart search with predictive suggestions
Visuals
Research Objectives Overview
Outlines the research framework for evaluating impact of platform deliveries on user workflows.
Defines both quantitative and qualitative research approaches for column usage analysis.
Establishes methodology for integrating insights to understand user adoption behavior.
