The main ways of presenting data in a business presentation come down to four charts: a bar chart to compare things, a line chart to show change over time, a pie chart to show parts of a whole, and a scatter plot to show how two things relate. The skill is not knowing twenty chart formats. It is matching one of these four to the argument you are making. I have watched executives agonize over chart galleries when the real problem was simpler: they had not decided what the chart was supposed to prove.
A chart is an argument, not a decoration
Every chart type makes a specific kind of claim. When you choose the wrong one for your argument, you confuse the room even when your data is completely solid. The audience does not separate "good data, bad chart." They just feel confused, and confused rooms do not approve budgets.
Before you pick anything, write the one-sentence point of the slide. Then pick the chart that makes that sentence visible. The numbers in the sample charts below are made up for illustration.
When the point is "this one is bigger"
Revenue by region. Performance by team. Satisfaction scores side by side. A bar chart makes the case instantly: the eye sees which bar wins before you say a word. Bar charts are built for comparison, so save them for comparisons.
Revenue by region
When the point is "this changed"
Trends, growth curves, declines, seasonal patterns. The eye naturally follows a line and understands direction. A simple rule settles most chart debates: if time runs along your horizontal axis, you almost always want a line chart. The shape of the line, the dip, the acceleration, the plateau, is the story.
Signups by month
When the point is "this is the share"
A pie chart makes exactly one claim: these parts make up this whole. It stays readable with about five slices or fewer. Past that, the slivers stop arguing and start decorating. If you have nine categories, you do not have a pie chart. You have a bar chart waiting to happen.
Revenue by customer segment
When the point is "these two move together"
As this goes up, that goes up, or down. Price versus demand. Training hours versus performance. A scatter plot is the one chart built to show whether two variables move together. Say "related," never "caused."
Rehearsal hours vs. audience rating
Take the title off and look for three seconds
Here is the test I use on every chart. Cover the title. Could someone in the room still understand the point in three seconds? If the chart only makes sense with a long verbal explanation attached, it is not doing its job. A great chart makes one clear visual argument, the same way a good slide carries one idea.
If the honest answer is confusion, redesign it before you present, not after the meeting where it flopped. Cut categories, split the chart in two, or switch formats.
Sometimes the number needs a picture of its own
Not every figure needs a chart. When I review talks that are packed with numbers, the problem is rarely the chart type. It is the data dump: all the numbers with no context. My own attention starts to shut off at a certain point, even when I care about the topic.
So when one number is the point, say it in a plain sentence, pause, and put it in perspective. Compare a huge figure to something people already grasp, like the economic output of a whole country. A comparison people can picture beats a fourth chart.
What to do next
Open your next deck and run the three-second test on every chart in it. Then check that each chart type matches the claim it is making. If you are building the story around the numbers, start with what data storytelling really is and the three-part structure that makes data unforgettable. And if a high-stakes data presentation is on your calendar, get a quick quote and we will pressure-test the whole thing on camera.