Quantitative UX Research: Tips to Calculate the System Usability Scale
Summary:?In this article, I share tips on calculating the System Usability Scale (SUS) and provide a free Google Sheets template for download.
What is the System Usability Scale??The SUS is a tool to measure the #usability of a product, service, or system. It’s calculated using a 10-question questionnaire with a Likert scale, where responses range from “strongly agree” to “strongly disagree.”
My Journey to the System Usability Scale (SUS)?Our organization recently embarked on creating a new Digital Workplace to meet the diverse needs of our almost 6,000 employees who work across 29 countries. I was assigned as Lead Product Designer, and one of my key responsibilities was to coordinate all user research activities and provide our Product Manager with valuable data for decision-making.
Our design team spent months on standard qualitative research, including one-on-one interviews and co-creation exercises like card sorting and tree tests. However, once we had developed a Minimum Viable Product (MVP), it was time to launch it to a subset of users and gather quantitative data to measure usability without the direct observation of our researchers.
The question arose: which metric should we use to evaluate the MVP?
I thought we could try the System Usability Scale. Even though I hadn’t used it before, I knew it was broadly used in the industry, so we gave it a try.
Challenges with the SUS?I was surprised to find that there was little information online about calculating the SUS. While some sites explained what it is and why it’s useful, few detailed the calculation process, and I found no free templates without requiring payment or account creation.
Even after learning the steps and formulas, there were still several smaller steps and tricks needed for the calculations. That's why I'm sharing here the detailed steps that I couldn’t find online, hoping to document them to ease the learning curve for the future. There may be easier and better ways to make these calculations. If you know them, please share in the comments for everyone’s benefit.
Step 0: ?Collect answers.
I used Microsoft Forms, but you can use any survey system you prefer. The first thing you need to do is to set up the Likert scale with the statements from the SUS.
In the picture below, notice how the language was changed to make it more specific and relatable to the respondent by replacing the generic term “system” with the actual product name, “IDB Hub.”
Once the answers are collected, the survey system will likely provide a visualization similar to the one below. Ideally, odd-numbered answers should lean towards the right (strongly agree), and even-numbered answers should lean towards the left (strongly disagree). This pattern occurs because half of the statements are positively worded, while the other half are negatively worded. This approach keeps users attentive and prevents them from defaulting to the same response automatically.
However, the calculation of the SUS has little to do with this visualization.
Here is how to begin:
Step #1: Export the data
Whichever system you have used to collect your answers will allow you to export the data to an Excel spreadsheet.
?
Step #2: Turn text-like answers into numbers.
The SUS is a numeric score, but we have collected answers in the form of text. So, we need to convert them as follows:
?
Watch out:
By the time you replace “Agree” with “4”, your entire set will be affected. Answers that previously read “Strongly Disagree” now will read “Strongly Dis4”.
After going through all the replacements, you will a numeric set instead of words. You can begin calculations.?
?
Step #3: Create space for the calculations
You need to create space between the columns (you have one per question right now) to make room for the calculations.
领英推荐
Another suggestion here is to color-code all Odd vs. Even questions, because they are calculated differently.
?
Step #4: Calculate the points for each question
Note: like I mentioned before, Odd questions are calculated differently than Even questions.
Let’s begin with the Odds: 1, 3, 5, 7, and 9.
[User Rating] – 1 = ___ points
?
Now you can do Even questions: 2, 4, 6, 8, 10.
5 – [User Rating] = ___ points
??
?
Step #5: Get the total points per user
Add up all the points for the ten questions for the first user:
[Question 1: points] + [Question 2: _ points] + … [ Question 10: ___ points]??= ____ total points from user 1
?
Step #6: Get the SUS per user
Now you have a total number of points per user.
Multiply them by 2.5, this way we can get scores from 0 to 100.
[Total points from user 1] x 2.5 = User #1 score
?
Step 7: Get the total SUS
Repeat steps 1-3 for all users, and then?average all user’s Scores together to get a SUS score:
User #1 score + user #2 score + user #3 + [etc.]? / # of users?= SUS Score
?
After obtaining the SUS for a digital product, the next step is to communicate it to the product team. While updated benchmarks are scarce, this useful site offers a qualitative interpretation of the scores by correlating SUS with Active Ratings.
?
?
To close, I want to share that using the SUS had a positive effect on our product and team. It provided a quantitative metric that complemented and supported the qualitative data gathered during the MVP launch (see picture below).
As designers, we tend to work well with interviews and qualitative observations, but having a numerical metric, a score, can be invaluable for engaging stakeholders who are less involved in day-to-day design operations, such as executives and engineers. This objective data helps bridge the gap and facilitates more informed decision-making.
?
Access and download the template here.