How to create charts in Excel

The short answer

Select the cells you want to plot, including the headers, and press Alt+F1. Excel builds a chart on the spot. Then use the Chart Design tab to change the type, and the plus icon at the chart's top right to switch elements like the title, labels and legend on and off.

That gets a chart on screen in about five seconds. The rest of this guide is about the harder half: making one that people can actually read.

Step by step, the long way

If you want the menu route rather than the shortcut:

1. Lay the data out simply. One row per thing, one column per measure, headers in the top row, no blank rows in the middle, no merged cells. Excel guesses your chart from the shape of the selection, so a tidy block gets a better guess.
2. Select the range, headers included.
3. Go to Insert, then pick a chart from the Charts group. If you are not sure, click Recommended Charts and look at Excel's suggestions.
4. Click OK. The chart lands on the sheet as a floating object you can drag and resize.
5. Use Chart Design > Change Chart Type if the first pick was wrong. Nothing is locked in.

To move a finished chart onto its own sheet, right-click it and choose Move Chart.

Pick the chart that fits the question

Most bad charts are the wrong type rather than a styling problem. Match the chart to what you are asking:

• Column: comparing a value across categories. Sales by region, cost by supplier.
• Line: change over time. Months, weeks, quarters, always in order along the bottom.
• Bar: the same job as a column chart, but horizontal, which is what you want when the category names are long.
• Scatter: the relationship between two numbers, like spend against revenue.
• Pie: parts of a single whole, with five slices at most. If you are squinting to tell two slices apart, use a bar chart instead.

If you need a legend to decode a pie chart with eleven slices, the chart is doing less work than the table it replaced.

Excel offers around fifteen chart families and you will use four of them. That is normal.

The three fixes that make it readable

A default chart is a starting point, not a finished thing. Three edits do most of the work.

Give it a real title

Click the title and type the finding, not the field name. "Support tickets fell 40% after the March release" tells the reader what to think. "Tickets by month" makes them work it out.

Fix the axis

Right-click the vertical axis and choose Format Axis. Set the number format (currency, percentage, thousands) so the labels stop reading as raw digits. Only start the axis somewhere other than zero if you have a reason and you say so, because a truncated axis makes a small change look enormous.

Cut the decoration

Delete the gridlines you are not using. Delete the legend when there is only one series, because the title already says what it is. Turn on data labels when there are few enough points to fit. The Microsoft 365 Excel page has the current feature list if you are unsure what your version supports, and Microsoft's Excel help documents every chart element by name.

Small things that save time later

• Turn your data into a table first with Ctrl+T. New rows then flow into the chart on their own.
• Sort before you plot. A column chart sorted largest to smallest is instantly readable. An alphabetical one is not.
• Copy the chart into PowerPoint with Paste Special > Picture when the numbers are final, so nobody opens your workbook by double-clicking the slide.
• Save a favorite. Right-click a chart you like and choose Save as Template, then use it from the All Charts tab next time.

When a chart is not the answer

Not every number needs a picture. Three figures in a sentence beat a three-bar chart every time, and a table is better when the reader needs the exact values rather than the shape. Chart when the shape is the point.

If your real problem is that the numbers take an hour to assemble before you can chart them, the fix is upstream: pivot tables and lookups, not prettier axes. That is the ground the Excel certificate covers, and Excel data analysis picks up where it stops. Both sit in the Business & Productivity school under one subscription.

Build the chart, then delete everything on it that is not carrying meaning. What survives is the chart.