Part II
Visual Evidence: From Charts to Dashboards
Charts that answer, not decorate
This part is about visual argument: turning a clean dataset into a chart that answers a named business question without quietly overclaiming. Chapter 3 supplies the grammar, matching each visual family to the decision it serves and treating a dashboard as a structured memo from executive question to recommended action. Chapter 4 sharpens the craft — a chart's first choice is its baseline, confidence intervals describe variation rather than cause, and the honest endpoint of any dashboard is the next test to run. Together they move the reader from picking a chart type to defending a one-page board memo with its limits stated out loud.
2 chapters · 9 articles
What you’ll learn
- Map a business question to the right visual family — distribution, comparison, time, relationship, geography, or uncertainty
- Structure a dashboard as a memo: KPI, trend, breakdown, drilldown, and a recommended action
- Choose the baseline and index a series before choosing color, so comparisons read honestly
- Read confidence intervals and small multiples as descriptions of variation, not claims of causation
- Assemble grain, joins, metrics, and reshaping into a one-page board memo that names its own causal limits
Chapters in this part
Pick the chart from the question, not the question from the chart — and the definition underneath the metric matters more than the metric.
- 3.1Exploratory Visualization and Dashboards
- 3.2Chart Atlas
- 3.3Case Study: Market Concentration Metrics
Every chart is a comparison — name the baseline, show the spread, and end on a decision, not a pattern.