# Mean vs median: differences and a worked example

> Calculate the mean and median of five response times and decide what each number says when one case is much slower.


**The arithmetic mean adds all values and divides by the number of observations. The median is the middle value after sorting them.** Both summarize a dataset, but they can describe different experiences when some cases sit far from the rest.

If a report says your team responds “in 12 minutes on average,” establish which average it uses and what happened to the slowest replies.

## How to calculate the mean and median

Imagine five customer inquiries handled during one morning. Time runs from arrival to the first human reply. Every observation uses elapsed minutes and the same definition. These are hypothetical data.

| Inquiry | Minutes to first reply |
|---|---|
| A | 2 |
| B | 3 |
| C | 4 |
| D | 5 |
| E | 46 |

The mean is **(2 + 3 + 4 + 5 + 46) ÷ 5 = 60 ÷ 5 = 12 minutes**. Every case contributes to the sum, including the 46-minute wait.

For the median, sort the values: 2, 3, **4**, 5, 46. With five observations, the third occupies the middle. The **median is 4 minutes**: two observations lie below it and two above it.

For an even number of observations, average the two middle values. For 2, 3, 4 and 5, the median would be **(3 + 4) ÷ 2 = 3.5 minutes**. OpenStax explains these rules and the effect of extreme values in its [measures of center](https://openstax.org/books/introductory-business-statistics-2e/pages/2-3-measures-of-the-center-of-the-data) chapter.

**Mean and median calculator**

Use the five example times or enter other first-response times, in minutes.

Hypothetical initial example:

- Response times in minutes: 2; 3; 4; 5; 46

Mean = total time ÷ inquiry count. Median = middle sorted value; for an even count, average the two middle values.

- Mean: 12 minutes

- Median: 4 minutes

- Included inquiries: 5 inquiries

- Accumulated waiting: 60 minutes

After sorting the times, position 3 contains 4 minutes: that is the median.

One value per line, up to 1,000 values. Semicolons also work. Use a decimal point or decimal comma, without thousands separators or empty lines between values.

Use times from zero to 1,000,000,000 minutes per case, with up to six decimal places. Keep the measurement rule consistent. The sum is accumulated waiting across inquiries, not staff handling time.

The web version lets you edit values, calculate and reset the example. Results are rounded to up to four decimal places; very small values use scientific notation.

## What an extreme case changes

In the five-inquiry dataset, 46 minutes pulls the mean upward. If that case had taken 6 minutes, the total would be 20 and the mean would fall to 4. The median would still be 4. This shows why the median is less sensitive to the precise magnitude of an extreme value.

It does not justify deleting the slow case. If the record is correct, someone waited 46 minutes. Investigate whether it belongs to another reporting period, reflects a measurement error or involves a relevant condition, such as needing another team's information. Split datasets when the question or definition warrants it, and state what you include.

## Choose what the report needs to show

To describe the center of these times, present both measures with their base: “Five inquiries: median 4 minutes, mean 12 minutes and maximum wait 46 minutes.” The median describes the center while the maximum preserves the much slower experience.

To reconstruct the total waiting time in this same dataset, the mean and count are sufficient: **12 × 5 = 60 minutes** of accumulated waiting across inquiries. The median cannot reconstruct that sum. Those 60 minutes are not staff working time: this dataset measures waiting until a reply, rather than active handling time.

A target such as “every inquiry answered in under 10 minutes” requires counting cases that met the threshold. The mean or median alone cannot answer it. Here, four of five were below 10 minutes: **80%**. Five cases teach the calculation. To assess your operation, define a period and collect the full relevant set.

## Make the calculation reproducible

Keep one row per case, its unit and the measurement rule. An unknown duration should remain missing; entering zero would claim an immediate response. Check whether business minutes and elapsed minutes, or human replies and automatic acknowledgments, have been mixed.

[Download the five fictional times as CSV](/blogs/stats-five-20261104/example-response-times.csv) to repeat the calculations. When combining team averages with different numbers of inquiries, continue with the [weighted average](/en/guides/weighted-average/): averaging figures without their underlying counts can change the result.

One concrete use of the median on social media: spotting when another account's post takes off without one earlier hit moving the bar. We apply it in [Instagram competitor tracking](/en/solutions/instagram-competitor-tracking/).

## Plan your next post in HeyMark.

Keep the idea, review the draft with your team, and see how it performed in the accounts you connected.

[Start free](https://app.heymark.ai) · [See how it works](https://heymark.ai/en/#product)

## Developer and agent resources

- [HeyMark MCP documentation](https://heymark.ai/en/mcp/)
- [llms.txt](https://heymark.ai/llms.txt)
- [Full site content for language models](https://heymark.ai/llms-full.txt)
- MCP protocol endpoint: `POST https://mcp.heymark.ai`
- OAuth protected-resource metadata: [/.well-known/oauth-protected-resource](https://mcp.heymark.ai/.well-known/oauth-protected-resource)
- MCP server card: [/server-card](https://mcp.heymark.ai/server-card)
