Step-by-Step

How to Make a Pie Chart from Survey Results

Last reviewed on 2026-05-22.

From raw response counts to a chart you can drop into a report

Before you start: pick the right question

Pie charts work for single-select questions with 2–6 answer choices: yes/no, multiple choice, Likert scales. They don't work for "check all that apply" questions because the responses don't sum to 100% — use a bar chart for those instead.

Step 1 — Export your raw counts

SurveyMonkey

Open the question summary view. Note the response count or percentage next to each answer option. You can also export the full results as XLSX and read the counts from the summary sheet.

Google Forms

Open the Responses tab. Each question shows a built-in chart and a percentage per option. Use those numbers directly.

Typeform

Go to Results → Insights. Each question has a breakdown table with counts and percentages. Copy the values into the maker.

Step 2 — Decide: counts or percentages?

Use percentages in the slice labels ("42%") and put the sample size in the chart title ("Customer Satisfaction (n=312)"). Percentages make the chart comparable across questions with different sample sizes; the n= keeps you honest about the base.

Step 3 — Enter the data

Open the maker (or one of the survey templates).
For each answer option, add a row — the option text as the label, the percentage (or count) as the value.
Set the chart title to something like "How satisfied are you with our support? (n=312)".
For Likert-scale questions, keep the options in natural order (Strongly Disagree → Strongly Agree). Set Sort to "Input order" so the maker doesn't reorder them.

Step 4 — Color with intention

For ordered scales (agreement, satisfaction, frequency) use a red→green or red→blue gradient — colors carry meaning. For unordered options (which feature, which brand) use a qualitative palette with high contrast between adjacent slices.

Step 5 — Export and caption

PNG for slides, SVG for print or web. Add a one-line caption below the chart with the sample size and the date range ("Q1 2026 customer satisfaction survey, n=312"). The caption is what makes a chart citable rather than just decorative.

Heads up: If your sample is under 30 responses, flag that clearly in the caption. A single respondent moves the chart by 3+ percentage points at that size, and your audience needs to know.

Frequently Asked Questions

Should I show percentages or raw counts on the slices?

Percentages on slices, raw count (n=) in the title. That way the chart is comparable across questions and honest about the underlying base.

How do I visualize Likert scale results in a pie?

A 5-slice pie chart for one question, with options in natural order (Strongly Disagree → Strongly Agree). For comparing many questions, switch to a stacked horizontal bar chart.

What about 'select all that apply' questions?

Don't use a pie chart — the percentages will sum to more than 100%, which is misleading. Use a horizontal bar chart, one bar per option, showing the percentage of respondents who picked it.

Can I make multiple pie charts for one survey?

Yes, but be careful. A grid of small pies is hard to compare. For multi-question summaries, a single stacked bar chart usually communicates faster.