A good chart makes the answer obvious in seconds; a bad one confuses or misleads. Visualisation is where analysis becomes communication — and it's a skill employers value highly.
Match the chart to the message
Choose based on what you're showing:
- Trend over time → line chart.
- Comparison between categories → bar/column chart.
- Part-to-whole → stacked bar or (sparingly) a pie with few slices.
- Relationship between two variables → scatter plot.
- Distribution → histogram.
- A single key number → a big, bold KPI card. If a table communicates better (precise values people will read off), use a table — not every insight needs a chart.
Design for clarity
- Remove chart junk — 3D effects, heavy gridlines, needless colour and decoration all reduce clarity (Tufte's "data-ink ratio": maximise ink that shows data).
- Label directly where possible, rather than forcing readers to decode a legend.
- Use colour with purpose — highlight the point; don't rainbow everything. Check colour-blind safety.
- Sort bars by value, not alphabetically, so the story is obvious.
- One message per chart — if it needs a paragraph to explain, split it.
Avoid distortion (ethical visualisation)
- Don't truncate the y-axis on bar charts — it exaggerates differences.
- Don't cherry-pick timeframes to imply a trend.
- Keep scales consistent when comparing charts.
- Show uncertainty where it matters. Misleading charts destroy trust — and your credibility.
Tell a story
Order your visuals to build an argument: context → what happened → why → what to do. Give each chart a takeaway title ("Revenue fell 12% driven by the North region") rather than a generic one ("Revenue by region"). The title should state the insight.
Put it to work
Take a chart you've made and improve it: pick the right type, strip the junk, add a takeaway title, and fix any axis distortion. Then build a full dashboard in the Executive Dashboard lab.