How to Make a Pie Chart with Percentages

The most common pie chart request is also the most automatable one: “show the percentages.” You should never compute them yourself. Here is the workflow, plus the two details, rounding and label layout, that decide whether the finished chart looks professional.

Enter what you actually measured

Start with raw numbers. If 87 people answered “very likely”, enter 87, not the 40.7 percent you worked out on a calculator. Raw values have two advantages: they contain no rounding errors for the chart to inherit, and they let labels show both the count and the percentage, which is the most informative combination.

Paste works as well as typing. Copy a column of category names and a column of numbers from Excel or Google Sheets, paste into the data table, and the parser handles headers and formatting. The chart preview updates as the data lands.

One toggle for percentages

Open the label settings and enable Show percent. Every slice immediately displays its share of the total. Change any value in the table and all percentages recompute instantly, which is the point of letting the tool own the math: the chart can never drift out of sync with its data.

Two more toggles live next to it. Show name puts the category on the slice, and Show value adds the raw number. For a chart that stands alone, name plus percent is the classic combination. For financial charts, value plus percent answers “how much” and “what share” at once.

The rounding problem, solved

Here is the detail most tools get wrong. Round each percentage independently and the printed numbers frequently sum to 99 or 101. A third, a third, and a third becomes 33, 33, 33, and someone in your audience will add them up. They always add them up.

This tool rounds with the largest-remainder method: percentages are rounded down, then the leftover points go to the slices that lost the most in rounding, until the total is exactly 100. Displayed as whole numbers, one third three times shows as 34, 33, 33. Each label is within half a point of the truth and the sum is beyond criticism.

You can also show one decimal place. Do it when slices are close, say 24.6 versus 25.1, and the whole-number labels would print identically. Skip decimals otherwise: “17.0%” carries a false air of precision, especially over survey data.

Placing labels so they fit

Percentages on fat slices are easy: the label sits inside, centered. Thin slices are where charts fall apart, because a 3 percent slice cannot contain its own label.

The auto placement mode handles this: labels sit inside slices that have room and move outside for slices that do not, connected by leader lines so ownership stays unambiguous. If many slices are thin, force all labels outside for consistency, or better, group the small slices into one Other slice first, which the editor does in one click.

Also give the labels enough contrast to survive export. Dark text on light slices, or bump the font size for slides. Both controls are in the label settings.

When percentages need context

A percentage without a base invites misreading. “62% preferred option A” means something different from 8 respondents than from 800. Put the base in the title: “Preferred lunch option (n=214)”. For money, the total serves the same job: “Q2 spend by channel, $84k total”. The slices say what share; the title says share of what.

Check, then ship

Before exporting, read the chart the way a stranger would. Do the percentages visibly total 100? Is every label attached to an unmistakable slice? Does the title name the whole? Then export: PNG for documents and slides, SVG or PDF for print, or copy to clipboard and paste it directly where it is going. The percentages are baked into the image, correct by construction, and you never touched a calculator.