Analyzing

Definition

Analyzing is an investigative mental mode where users are examining data, information, or content to extract insights, identify patterns, and derive meaningful conclusions.

Analyzing belongs to the ‘Cognitive’ Mode Family: focused on deeper thinking, reasoning, and reflection. Synonyms include: Investigating, Examining, Interpreting.

Contextual Relevance by Role

• Analysts: Run in-depth comparisons across KPIs or segments.
• Managers: Examine performance trends, people metrics, or team health indicators.
• Data Scientists: Drill into complex queries and apply models to evaluate trends.
• Researchers: Investigate behavioral, survey, or product usage data.
• Business Leaders: Synthesize insights across domains to inform strategic decisions.

Mental Model

• Pattern recognition across datasets
• Hypothesis formation and testing
• Causal relationship identification
• Synthesis of insights from multiple sources

Emotional Context

• Intellectually curious and engaged
• Satisfaction from discovering insights
• Frustration with incomplete or inconsistent data
• Excitement when identifying meaningful patterns

Behaviors

• Manipulating data views and visualizations
• Drilling down into details
• Comparing metrics across dimensions
• Exporting or sharing findings

Journey Stage

When in the user journey this intent typically occurs:

• Mid-to-late journey
• Often triggered post-monitoring or reporting
• Used when an event, anomaly, or hypothesis prompts further investigation

Measuring Analytical Clarity

How effectively users can extract meaningful insights and patterns from data and information.

Quantitative Metrics

• Insight discovery rate
• Time to insight (minutes to reach conclusion)
• Data comprehension accuracy
• Collaboration rate (shares, comments, exports)

Qualitative Indicators

• Confidence in conclusions
• Satisfaction with depth and flexibility
• Perceived trust in available data

Design Implications

1. Provide Flexible Data Manipulation Tools

Users in analyzing mode are forming and testing hypotheses. They need to manipulate dimensions, groupings, and metrics in real time to explore multiple angles. → Support drag-and-drop pivoting, ad hoc calculations, dynamic filters, and toggling between views.

2. Support Multiple Visualization Types

Different questions demand different visual representations (e.g. time series, heat maps, scatter plots). → Let users choose or switch visualizations without friction, and offer smart defaults aligned with the data shape.

3. Enable Filtering and Segmentation

Analyzing often involves zooming in on specific cohorts, outliers, or dimensions. Users must isolate slices of data that reveal hidden patterns. → Include intuitive controls for filtering, drilldowns, and segment comparisons, with visibility into what filters are active.

4. Allow Saving and Sharing of Analyzing States

Analysis is iterative and collaborative. Users want to preserve a moment-in-time view (with filters and charts configured) for later reference or sharing. → Support bookmarking, named views, and shareable links or exports that preserve configuration.

5. Provide Context and Benchmarks for Interpretation

Analysts need help determining what’s good, bad, normal, or surprising. → Include peer benchmarks, historical comparisons, reference lines, or annotations to help users interpret data confidently.

UX Domains

• Data Analysis
• Business Intelligence
• Reporting Tools

UX Context Examples

• Analytics dashboards
• Data visualization tools
• Report builders
• Business intelligence platforms

Components and Patterns

• Analytics Dashboard
• Data Table
• Chart Builder
• Filter + Segmentation Controls

Do’s and Don’ts

Mistaking Analyzing for Monitoring

• Analyzing requires active investigation; Monitoring involves passive tracking.
• Don’t confuse dashboards that update over time with tools that enable true data exploration (e.g., comparing cohorts).

Over-Reliance on Static Data

• Users need to pivot, slice, and manipulate data.
• Static reports limit hypothesis testing and iterative exploration.

Data Without Context

• Data must be accompanied by benchmarks, comparisons, and labels.
• Context drives clarity and helps users trust their interpretations.