How to use the sus system usability scale calculator
- Enter one line for each person with their ten answers separated by commas or spaces. Each answer is a whole number from 1 (strongly disagree) to 5 (strongly agree).
- Keep the questions in the standard order, because odd and even items are scored in opposite directions.
- Optionally change the comparison score. The default is an example value.
- Read the mean score and spread, the score for each person and the mean of each item.
Formula
Odd items (1, 3, 5, 7, 9): score = answer - 1
Even items (2, 4, 6, 8, 10): score = 5 - answer
SUS = 2.5 x (sum of the ten item scores)
Mean SUS = average of the person scores
The item scores run from 0 to 4, so the sum is 0 to 40 and SUS is 0 to 100. A SUS score is not a percentage.
Four people
One person answers 4, 2, 4, 1, 5, 2, 4, 2, 5, 1. The odd items give 3 + 3 + 4 + 3 + 4 = 17 and the even items give 3 + 4 + 3 + 3 + 4 = 17, so the sum is 34 and the score is 34 x 2.5 = 85.
With three more people scoring 65, 97.5 and 37.5, the mean SUS is 71.3, the median is 75.0 and the standard deviation is 26.18. The mean is 3.3 points above the comparison score of 68.
What the scale is
The System Usability Scale is a ten-item questionnaire that gives a quick view of how usable people find a product, a site or a system. Half the statements are positive and half are negative, which makes people read each one rather than tick the same box down the page. The scoring turns every answer into a 0 to 4 value where higher is always better, and scales the total to 0 to 100.
Because the final number is built from ten answers, it is easy to compute but easy to misread. A score of 70 is not 70 percent of anything. It is a point on a scale, and it means more when you compare it with earlier versions of the same product or with benchmarks from similar products.
Using small samples
Usability studies often have 5 to 15 people. With so few, one very low or very high score moves the mean a lot, as the standard deviation of 26 in the example shows. Report the range and the number of people alongside the mean, and avoid reading small differences between two groups as real.
Item means help to find where the problems are. A low mean on the item about needing technical support or about learning many things first points to a different issue than a low mean on confidence. The table shows both the raw mean answer and the 0 to 4 score for each item.
Assumptions and limits
- Answers are whole numbers from 1 to 5 in the standard item order.
- The standard deviation is the sample deviation and needs at least two people.
- The comparison score is your own choice and is not a rule.
- The statements shown are the common wording of the form. Nothing is uploaded.
Frequently asked questions
Can I skip a question?
No. The score needs all ten answers. Many teams give a neutral 3 for a skipped item, but that is a choice outside this tool.
Is 68 a pass mark?
No. It is only an example comparison. Pick a benchmark that fits your field and your product.
Why are odd and even items scored differently?
The odd items are positive statements and the even items are negative, so a high agreement is good for one and bad for the other.
How many people do I need?
A few can show large problems. For comparing two designs or reporting a firm number you need more, and the spread tells you how steady the mean is.