data-visualization

name	Data Visualization
description	Select the right chart type by story intent, apply colorblind-safe palettes, annotate insights, and remove clutter for clear data communication.

Purpose

Turn data into visuals that communicate. A good chart answers one question clearly. A bad chart confuses many questions at once. This skill covers chart selection, color theory, annotation, and the principle that the chart title should BE the insight.

Steps

1. Start with the question: what does the audience need to understand?
2. Select the chart type based on the relationship in the data
3. Apply the colorblind-safe palette
4. Write the title as the insight, not the metric name
5. Annotate the key data point
6. Remove all clutter: gridlines, borders, unnecessary legends, 3D effects
7. Add source attribution

Chart Selection by Story Intent

Intent	Best Chart Type	Example
Compare categories	Horizontal bar	"Sales by region"
Show trend over time	Line	"Monthly revenue, 2023-2024"
Part of whole	Stacked bar or pie (max 5)	"Revenue split by product line"
Correlation	Scatter	"Marketing spend vs. revenue"
Distribution	Histogram or box plot	"Customer age distribution"
Ranking	Sorted horizontal bar	"Top 10 customers by revenue"
Composition over time	Stacked area	"Channel mix evolution"
Flow or conversion	Funnel or Sankey	"Sales pipeline stages"
Geographic	Choropleth map	"Revenue by country"

Colorblind-Safe Palette (Mandatory)

Use the Tableau 10 palette for all data series:

Order	Color	Hex Code
1	Blue	#4e79a7
2	Orange	#f28e2b
3	Red	#e15759
4	Teal	#76b7b2
5	Green	#59a14f
6	Yellow	#edc948
7	Purple	#b07aa1
8	Pink	#ff9da7
9	Brown	#9c755f
10	Gray	#bab0ac

Never use arbitrary colors. Never rely on red/green distinction alone.

Title = Insight Principle

Bad Title (Topic)	Good Title (Insight)
"Q3 Revenue"	"Q3 Revenue Grew 18% Driven by Enterprise Segment"
"Customer Satisfaction"	"Satisfaction Dropped Below Target for First Time in 2024"
"Headcount by Department"	"Engineering Grew 40% While Sales Stayed Flat"

Annotation Rules

• Highlight the single most important data point with a contrasting color
• Add a callout or label to that point
• Gray out non-essential data series
• Use direct labels instead of legends when there are 3 or fewer series
• Place the source in small text at the bottom left

Clutter Removal Checklist

• Remove chart border and background fill
• Remove or lighten gridlines (light gray if needed)
• Remove 3D effects
• Remove unnecessary legend (use direct labels)
• Remove data labels on non-key points
• Ensure axis labels are readable (no rotated text)
• Verify the y-axis starts at 0 for bar charts

Output Format

• Chart specification: type, data mapping, colors, title (as insight)
• Annotation guide: which data point to highlight and why
• Source attribution text

Guidelines

• One chart, one message. If a chart tells two stories, make two charts.
• The audience should understand the point in under 5 seconds
• Start bar chart axes at 0 to avoid exaggerating differences
• Pie charts: maximum 5 slices. If more, use a horizontal bar chart.
• Time always flows left to right on the x-axis
• If the chart needs a paragraph of explanation, the chart has failed