Coming next: The Second Factory →

Back to Library
Dashboards & Visualization✓ Principle

Sometimes Just Use the Number (When a Chart Makes Things Worse)

The Problem

Dashboard showing "Customer Satisfaction Score." It's a gauge chart. With gradients. And a needle pointing to 8.2. Looks fancy. But here's the problem: the gauge took up 20% of the dashboard space to show ONE number. And it's harder to read than just showing "8.2" in big text. Worse: comparing this month's gauge to last month's gauge requires mental math—you can't see the change easily. You built visualization theater. It looks like you're doing analysis, but actually you just made a number harder to read. Sometimes the simplest solution—just showing the actual number—is the best solution.

The Principle

Not everything needs a chart. If you're showing 1-3 numbers, especially a key metric, just show the number. Big. Bold. Add context with small text ("+0.4 vs last month"). That's it. Gauge charts, donut charts with numbers in the center, progress bars for single metrics—usually worse than just the number. The magic happens when you realize that 67% of dashboard space can often be simplified to large numbers with small context. Think manufacturing: you don't build a complex machine to measure one thing when a simple gauge works better. Same with viz—use the simplest tool that communicates clearly.

🎯

3 Steps to Take Action

  1. 1

    **Audit Single-Metric Visualizations** Find every chart showing ONE number: - Gauge charts - Donut charts with number in center - Progress bars (just showing % complete) - Single-bar charts - Pie charts with 2 values Ask: Would a large number with context be clearer? Test: Show both versions to 3 users. Time how fast they read the value. Numbers are usually 3-5x faster.

  2. 2

    **Replace with Big Numbers + Context** For single metrics: - Show the number large (40-60pt font) - Add small context below: * Change vs last period: "+2.3 vs last month" * Target: "Target: 8.0" * Trend indicator: ↑ ↓ → arrows * Sparkline (tiny line chart) if showing recent trend Example: ``` 8.2 ↑ +0.4 vs Nov | Target: 8.0 ``` Simple. Fast. Clear.

  3. 3

    **Reserve Charts for Complexity** Use charts when they add value: - Trends over time (line charts) - Comparisons (bar charts) - Distributions (histograms) - Relationships (scatter plots) - Geographic patterns (maps) DON'T use charts for: - Single numbers (just show the number) - 2-3 categories (table or numbers work fine) - Decoration (to "make it visual") Rule: Chart must answer a question numbers can't. If a table or number is clearer, use that. Bonus: Dashboards become scannable. Execs read KPIs in seconds instead of minutes.

💡 Inspired By

Stephen Few's guidance on dashboard design. Edward Tufte's principle of maximizing data density. Visual perception research showing numbers are fastest for single values.

This is what Analytics Pitfalls & Principles is about: people helping people.

In Practice

A healthcare analytics team had an executive dashboard with 12 KPIs. Each was a gauge chart, donut chart, or progress bar. Looked impressive. Execs complained: "Takes too long to read." Team redesigned: Each KPI became a large number with small trend indicator. "8.2 ↑ +0.4" instead of a gauge chart. Dashboard went from 12 complex visualizations to 12 simple numbers. Comprehension test: executives found and interpreted KPIs 4x faster with numbers vs charts. Space freed up was used for 2 actual trend charts showing the WHY behind changes. One exec said: "Now I spend 10 seconds reading KPIs and 2 minutes understanding trends, instead of 2 minutes decoding gauges." Simple numbers aren't boring—they're efficient. Save charts for when they actually add value (trends, comparisons, distributions). For single-point metrics? Just use the number.

More on Dashboards & Visualization

From the book

The book opens on exactly this: forty-seven dashboards, three of them heavily used, and nothing anyone could decide from.

This card is one of 57 drawn from The Auction Block by Brian Cariveau — the story of an analytics leader with 90 days to fix what the cards describe.

Each leadership pattern falls into a few pitfalls most often. Find your pattern.The Analytics Leadership Mirror: free, 5 minutes, no account