Outlier Detector
One strange number can throw off a whole report. A typo in a spreadsheet, a one-time bulk order, or a data entry mistake can quietly skew your average and make a clean report look wrong. Paste your numbers below and this tool will point out exactly which values do not fit the pattern, and why.
Results
What This Means
To Improve This
What Counts As An Outlier?
An outlier is a value that sits far away from the rest of your data. Picture a classroom where most students score between 60 and 75 on a test. One student scores 12. That single score is an outlier. It does not mean the score is wrong. It means the score does not follow the same pattern as everyone else, and it deserves a closer look.
In finance and business, outliers show up often. A month with sales ten times higher than usual. An expense report with one entry that is far above the rest. A delivery that took five times longer than every other delivery that week. Spotting these values fast helps you decide whether they point to a mistake, a rare event, or something worth investigating.
Two Ways To Find Outliers
The IQR Method
IQR stands for Interquartile Range. This method sorts your data from smallest to largest, then splits it into four equal parts. It looks at the middle 50% of your data and builds a “normal range” around it. Anything that falls well outside that range gets flagged.
Any value below the lower fence or above the upper fence is an outlier. This is the method most spreadsheet tools and statisticians reach for first, because it does not assume your data follows any particular shape.
The Z-Score Method
A z-score tells you how many standard deviations a value sits from the average. A z-score of 0 means the value equals the average. A z-score of 3 means the value is quite far from the average. This method works best when your data roughly follows a bell curve.
Values with a z-score above 3 or below negative 3 (2.5 in strict mode) are treated as outliers here. A few textbooks use 2 as the cutoff, which flags more values, so this tool sticks with the more common 3 for standard sensitivity.
A Worked Example
A small shop tracks daily sales for ten days: 210, 230, 205, 220, 215, 225, 208, 980, 218, 212. Nine days sit close together. One day jumps to 980, most likely a bulk wholesale order.
| Step | Value |
|---|---|
| Sorted data | 205, 208, 210, 212, 215, 218, 220, 225, 230, 980 |
| Q1 (25th percentile) | 209 |
| Q3 (75th percentile) | 224.25 |
| IQR | 15.25 |
| Lower fence | 186.13 |
| Upper fence | 247.13 |
| Outlier found | 980 |
The 980 sits far past the upper fence of 247.13, so the tool flags it right away. The owner now knows to check that one entry before it throws off the monthly average.
Frequently Asked Questions
No. An outlier is only a signal that a value looks unusual compared to the rest of your data. It could be a typo, but it could also be a real event, like a holiday sales spike or a one-time large refund. The tool flags it. You decide what it means.
Use the IQR method for most everyday data, especially small samples or data that is not symmetric. Use the Z-Score method when your data is larger and follows a roughly normal, bell-shaped pattern, such as test scores or measurement data.
At least four points are required to run the calculation, but the results grow more reliable with ten or more. Very small datasets can flag values that are not truly unusual, since there is little pattern to compare against.
Not automatically. Look into each flagged value first. If it turns out to be a data entry mistake, correcting or removing it makes sense. If it reflects a real event, keeping it and noting it separately usually gives a more honest picture.
A trend is a pattern that repeats or builds over time, like sales rising every December. An outlier is a single point that breaks the pattern once. If the same “unusual” value keeps showing up on a schedule, it may actually be a trend and not a true outlier.
Explore More Statistics Tools
Pair this with other calculators in the Finance Toolkit to get the full picture of your data.