Research-Driven Discovery Process
With limited understanding of user goals, needs, expectations, and frustrations at the project outset, we needed to answer fundamental questions about how people actually used the system. We initiated a thorough primary and secondary research process to build an evidence-based foundation for design decisions.
Competitive Analysis
Evaluated 2-3 leading business analytics tools to understand industry standards, user expectations, and opportunities for differentiation. We consolidated findings into benchmarks that informed our unique approach.
User Interviews
We conducted more than 8 user interviews to gain deep insights into user workflows, expectations, and pain points.
To ensure effective interviews, we followed a structured process:
- Defining interview objectives
- Preparing open-ended questions
- Conducting pilot testing with internal stakeholders
- Performing detailed post-interview synthesis
Interview Objectives
We focused on uncovering:
- Crucial KPIs users prioritize
- Information hierarchy users require
- Common use-cases for the tool
- Features users expect (reporting, sharing, comparisons, drill-downs)
- Current challenges using the existing Analytics Hub
Interview Questions
Questions were prepared thoughtfully and categorized into themes:
KPIs, workflows, pain points, reporting needs, roles, and expectations.
KPIs, workflows, pain points, reporting needs, roles, and expectations.
Pilot Testing
A pilot interview was conducted with a stakeholder to refine the questions and ensure a smooth conversational flow.
Domain Training
Immersed ourselves in understanding the Analytics Hub structure, system terminology, user challenges in managing data, current upload flows, and the existing information hierarchy to speak the language of our users.
Information Hierarchy (Existing) : The existing hierarchy required manual uploads from data uploaders at every level, resulting in: High workload, Delays, Error-prone processes
Ideal Information Hierarchy: Data should ideally flow directly from the source to the dashboard, eliminating the need for intermediaries.
Key Research Insights
Interview data was carefully analyzed and synthesized, revealing four critical problem areas that would guide our design strategy. These insights represented the voice of our users and became the foundation for prioritizing features and improvements.
Data Reporting Challenges
- Difficult to analyze data and pinpoint specific issues
- Need for seamless visibility from tower-level to portfolio-level
- Desire for one combined tool for both reporting and analytics
- Request for a single unified dashboard to view all KPIs at once
Data Extraction Pain Points
- Multiple disparate systems required to extract needed data
- Collection process demanded several different tools
- Strong preference for direct integration from source systems
- Manual intervention causing delays and errors
Navigation Frustrations
- Absence of advanced search capabilities
- Unclear paths for drilling down into detailed data
- Poor structure for moving across different datasets
- Difficulty finding relevant information quickly
User Access Issues
- Need for role-based access controls and permissions
- Content visibility should match user responsibilities
- Different views required based on organizational level
- Security concerns with open access to sensitive data
Define & Ideate
We moved into defining structure, strategy, and experience using:
- User Personas
- Affinity Mapping
- Impact–Effort Matrix
- User Flows
- Information Architecture
User Personas
We categorized personas into four user groups:
Data Consumers
(Company Head, Country Head, Tower Heads)
Monitor KPIs, identify risks, track business performance.
Monitor KPIs, identify risks, track business performance.
Data Uploaders
(Portfolio Managers, POCs)
Collect and upload data; need simplified and automated workflows.
Collect and upload data; need simplified and automated workflows.
Data Downloaders
(Tower Heads, Directors, Managers)
Download KPIs and reports for reviews and presentations.
Download KPIs and reports for reviews and presentations.
Admins (Tool Managers)
Manage access, permissions, data configuration.
Affinity Mapping
By clustering similar findings, we were able to reveal clear patterns and insights that guided more informed and user-centric design decisions.
Impact and Efforts Metrics
Insights were prioritized by user value and complexity, helping us shape an MVP that delivered maximum impact early.
Information Architecture
Design Iteration
The design iteration process was broken into smaller steps to manage the extensive data representation. The process included:
- Low-fidelity wireframes
- Visual exploration (cards, graphs, colors)
- High-fidelity wireframes
- Branding & identity alignment
1. Low-Fidelity Wireframes
Explorations focused on: Navigation, Data placement, Overall layout, Dashboard structure, Hierarchy and quick scan patterns
2. Visual Representation — Cards, Graphs & Tables
Once card structures were finalized, we tested them across multiple datasets to ensure:
- Scalability
- Visual consistency
- Semantic use of color
- Clarity in hierarchy
- Quick interpretation
Card Elements
During card finalization, we validated the visualizations with different data sets to make sure they scaled well and remained clear.
Graph Elements
3. High-Fidelity Wireframes
After multiple cycles, we arrived at a final high-fidelity design aligned with the company’s visual system.
Login & Security
Before: Open access to all users with no restrictions on sensitive data, creating significant security risks and making it impossible to control who saw what information.
After: Implemented role-based access with clear permission structures, dramatically improving security while ensuring users see exactly what's relevant to their responsibilities.
Tile Data Display
Before: High cognitive overload from inconsistent design, no minimalist principles applied, and slow, tedious processes for switching between datasets. Users felt overwhelmed by visual clutter.
After: Clean, easy country-level access with intuitive navigation, download options readily available, clear insight tags marking "Highlights" and "Lowlights," and a consistent, standardized visual design language throughout.
Graph Visualizations
Before: Cluttered, inconsistent visuals with no aesthetic balance or adherence to minimalist design principles. Users struggled to extract meaning from poorly designed charts.
After: Clear visual hierarchy using semantic color usage that conveys meaning, clean and aesthetic graphs following minimalist principles, and thoughtful data-to-ink ratios that highlight insights rather than decoration.
Table Data Display
Before: Overwhelming cognitive load from dense, unstructured tables that made deriving insights extremely difficult. Users spent too much time scanning for relevant information.
After: Clean data flow with strategic whitespace, relevant contextual insights displayed alongside tables, and simplified scanning patterns that enable quick comparison and decision-making.