Understand what drives change and which segments mattered most

business analytics

slug: recipe-business-analytics-understand-what-drives-change-and-which-segments-mattered-most

Recipe: What drives change?

category: business analytics

Problem

Executives and analysts often see only a total change metric (e.g., revenue grew 10%), but don’t know why it changed or which segments drove it.

Examples:

• Revenue increased this quarter — which products, regions, or channels contributed most?
• Customer churn dropped — which segments improved retention?
• Marketing spend went up — which campaigns delivered the bulk of the return?
• Growth may be concentrated — was it broad-based or driven by a few standout segments?

Solution

Follow this pipeline to quantify changes and highlight the top contributors:

• Aggregate measures across dimensions with cube step
• Calculate period-to-period contributions with contribution step
• Identify the vital few segments driving the majority of change with pareto step

Step Sequence

cube step -> contribution step -> pareto step -> chart step

Input Datasets

• Aggregated transactional or performance data (e.g., sales, revenue, units)
• Must include dimensions for grouping (product, region, channel) and measure(s) to analyze

Output Dataset

• contribution_pareto — dataset with contribution metrics and Pareto classifications
• Key columns: dimensions, baseValue, compareValue, contribution, contributionPct, growthRate, rank, cumulativePct, abcCategory, paretoFlag

Step-By-Step Explanation

Step	Purpose	Notes
cube step	Aggregate measures across chosen dimensions	Example: total sales by product × region × quarter
contribution step	Compute changes between periods for each segment	Calculates absolute and percentage contribution
pareto step	Identify top contributors driving most of the change	Classify segments into ABC categories and mark the vital few
chart step	Visualize results	Optional bar chart, Pareto chart, or dashboard integration

Variations & Extensions

• Use filter step upstream to focus on specific time periods or segments
• Combine with index step to benchmark high-contributing segments against average performance
• Feed into dashboards or executive reports for trend monitoring
• Include multiple measures for multi-dimensional contribution analysis

Concepts Demonstrated

• Contribution analysis for understanding change drivers
• Pareto analysis for prioritizing the vital few segments
• Sequencing of multiple DAZL steps to produce actionable insights
• Integration of analytics into visualization or reporting

Related Recipes

• How to identify the vital few segments (Pareto analysis)
• How to evaluate segments relative to a benchmark (Index analysis)
• Cube-building for multi-dimensional aggregation

Notes & Best Practices

• Ensure upstream data is clean and aggregated correctly
• Document the comparison periods for transparency
• Interpret top contributors in context — high contribution doesn’t always equal high value
• Use visualizations to communicate insights clearly and effectively

Metadata

title: "How to understand what drove change and which segments mattered most"
category: "business analytics"
difficulty: "Intermediate"
tags: [cube, contribution, pareto, analytics, executive insights]
inputs: [aggregated transactional data]
outputs: [contribution_pareto]
steps: [step-cube, step-contribution, step-pareto, step-chart]
author: "Tom Argiro"
last_updated: "2025-10-25"
doc_type: "recipe"