Stacked Charts

What Are Stacked Charts?

Stacked charts show how multiple variables contribute to a total over time, with bars composed of colored segments stacked on top of each other.

Purpose: Visualize part-to-whole relationships, show how components contribute to total, track mix changes over time.

When to Use

Best For:

  • Media mix analysis (how budget is allocated)

  • Showing composition over time

  • Part-to-whole relationships

  • Budget allocation tracking

  • Proportion changes

Examples:

  • Marketing spend by channel over time

  • Revenue by product category

  • Budget distribution across teams

  • Contribution by marketing group

How to Create

Requirements:

  • Select 2 or more variables

  • Click "Stacked Chart" button (📚)

  • Click "Generate Chart"

Chart Structure:

  • X-axis: Time periods

  • Y-axis: Total value

  • Each bar: One time period

  • Segments: Individual variables (stacked)

  • Bar height: Sum of all variables

Reading Stacked Charts

Total Bar Height:

  • Shows combined total of all variables

  • Changes in overall sum visible

Segment Heights:

  • Each color = one variable

  • Taller segment = larger contribution

  • Changes in individual contributions

Proportions:

  • Visual sense of relative sizes

  • Mix shifts visible over time

  • Dominant contributors clear

What Variables to Visualize

Marketing Budget:

  • TV_Spend, Digital_Spend, Radio_Spend, Print_Spend

  • Shows: How budget is split across channels

  • Insight: Mix evolution over time

Channel Groups:

  • Traditional_Media, Digital_Media, Promotions

  • Shows: High-level budget allocation

  • Insight: Strategic shifts

Product Mix:

  • Product_A_Revenue, Product_B_Revenue, Product_C_Revenue

  • Shows: Revenue composition

  • Insight: Which products growing/declining

Team Budgets:

  • Team_A_Budget, Team_B_Budget, Team_C_Budget

  • Shows: Resource allocation

  • Insight: Investment priorities

Common Patterns

Stable Mix:

  • Proportions remain constant over time

  • Consistent strategy

  • No major reallocation

Shifting Mix:

  • Proportions change over time

  • Example: Digital segment growing, TV shrinking

  • Insight: Budget reallocation visible

Growing Total:

  • Bar heights increasing

  • All segments may grow

  • Or mix shift while total grows

Dominant Player:

  • One segment much larger than others

  • Example: TV dominates spend

  • Insight: Concentration in one channel

Use Cases

Budget Evolution:

Variables: All marketing channel spends
View: How has mix changed over 12 months?
Pattern: Digital growing from 20% to 40% of total
Insight: Shift from traditional to digital

Proportional Analysis:

Variables: Channel contributions (from decomposition)
View: Which channels drive most value?
Pattern: Media 60%, Price 30%, Promo 10%
Insight: Media is primary driver

Planning Scenario:

Variables: Current allocation vs. proposed
View: How would new plan change mix?
Pattern: Rebalances TV down, Digital up
Insight: Visualize proposed changes

Stacked vs. Grouped Bar Charts

Stacked Shows:

  • Total sum (bar height)

  • Proportions within total

  • Composition over time

Grouped Shows:

  • Individual values side-by-side

  • Direct variable comparison

  • Not part-to-whole

MixModeler uses stacked by default for composition analysis.

Interactive Features

Tooltip:

  • Hover over segment

  • Shows variable name, value, percentage

  • See contribution at that point

Legend Click:

  • Hide/show variables

  • See remaining variables restack

  • Analyze subsets of mix

Zoom:

  • Focus on specific time periods

  • Detailed view of mix changes

Tips

Variable Count:

  • 2-5 variables: Easy to read

  • 6-8 variables: Still manageable

  • 8 variables: Consider grouping or separate chart

Color Choice:

  • Use distinct colors

  • Assign consistent colors to same variables

  • Consider colorblind-friendly palettes

Interpretation:

  • Focus on both totals AND proportions

  • Look for mix shifts

  • Identify growing/declining components

Order Matters:

  • Largest at bottom usually

  • Or logical grouping (all media together)

  • Consistent order across time

Comparison with Decomposition

Stacked Charts:

  • Can visualize any variables

  • Pre-model exploration

  • Raw spend or revenue data

Decomposition Charts:

  • Show model contributions

  • Post-model analysis

  • Predicted impact, not just spend

Both use stacked format but different data sources.

Summary

Stacked Charts Show:

  • Part-to-whole relationships

  • Total sum and composition

  • Mix evolution over time

  • Proportional changes

Best for understanding how a total breaks down into components and how that breakdown changes over time.

PreviousBar ChartsNextCorrelation Heatmaps

Last updated