Having a website is an absolute necessity for any business. But it is also a substantial marketing investment that you expect to pay off through new business. Gone are the days where you could simply update your website once and forget about it for months on end. In today’s ever-changing digital landscape it is important to constantly track and measure how well your website performs, as well as make changes to continue achieving your goals.
There are a number of factors that can affect website performance. Quite often they are interconnected making it hard to distinguish which one is failing or contributing to your business goals. But the good news is there is one thing that encompasses all aspects of customer interaction with the website, products and services. User experience (UX).
Measuring user experience does not mean you will have answers to all of your questions. But improving UX will solve some of your marketing problems.
The number one question you are now likely asking yourself. How do I go about measuring the UX of my website?
But before we get into that, did you know that 88% of online customers are less likely to return to a website after having a bad experience?
That’s a huge number to ignore.
Tracking and measuring user experience is important to make sure your website not only meets your business goals but also provides excellent customer experience.
So, where do you start?
Start with setting your goals
A common mistake in UX measurement is to start defining the metrics without setting goals first. Tracking certain metrics just because you can or just for the sake of measuring something doesn’t make sense.
Think top-bottom when defining your UX metrics.
See the big picture in terms of the user experience you want to provide.
For instance, your UX goal should not aim to increase time-on-site as an indicator of high engagement. Instead, you should define what the goal of this webpage or feature is? What experience do you want your users to have on this website?
A good UX goal would sound more like this: ensure that the content on this webpage is useful and intuitively leads users to an action (like browsing another page, downloading an ebook or filling out a form).
To answer your goal you need to define which user behaviors count as a signal for your UX goals. For example, if your goal is to provide good UX by offering helpful content, then the behavioral cue may be users reading all the way to the bottom of the page, or downloading a free template.
Now that you know the goal of your website and the behavioral signals to look for, you can define how to measure each of those signals.
Choosing the right metrics
Analytics tools allow us to measure pretty much anything. They provide an enormous amount of data to work with. However, analytics data by itself does not provide much insights in terms of UX metrics.
This sounds controversial, as analytics allows us to track and measure a number of important metrics, such as;
- Conversion rate
- Bounce rate
- Time spent on site
- Visits to purchase
The problem is that these metrics do not provide any context. They can be quite ambiguous and are very difficult to tie to usability and UX elements.
They answer the “what” but give no information about the “why”.
For example, let’s say you are tracking the bounce rate metric and it appears to be quite low. The low bounce rate could be a positive signal, but it can also be negative. Are your users browsing page after page, not finding what they are looking for and feeling frustrated or are they finding the material they are looking for with ease and then finding other relevant materials they are able to quickly read?
Solely relying on analytics to measure UX is not a good idea. But rather use the data collected to complement and validate user testing results.
Measuring UX with User Testing
User testing is a great way to evaluate the user experience of a website. Using a sample group of real users, you can review their thoughts and opinions as they browse through your website and see their behind-the-screen reactions. But most importantly, user testing provides both qualitative and quantitative data that can be translated into measurable UX metrics.
In a typical user testing, target users will attempt to complete various tasks while observers watch, take notes and record their behavior. Depending on how well they are able to perform these tasks you might uncover major usability issues or find small changes. No matter the outcome they help in determining the overall user satisfaction with your website.
The most common UX metrics to measure
User success rate shows the percentage of tasks that users complete successfully. This is a very simple yet powerful metric. It is easy to collect and provides a lot of insights.
Most often users are able to complete tasks at least partially. In this case, you can quantify the test results by giving partial success credit for the task. If they fail to complete typical tasks regularly you need to figure out what is causing their complete failure and fix before proceeding.
How do you calculate user success rate?
There is an easy formula you can use to calculate user success rate of completing tasks. For our example, let’s say you have tested 5 users. Each of them has performed 10 tasks. As a result, 30 tasks were completed successfully, 5 were partially successful and 15 tasks failed.
The success rate formula is as follows: (completed tasks + partial tasks * 50%) / total number of tasks.
For our example (30+5*50%)/50 = 65%
It is very likely that most website users feel frustrated rather than satisfied as according to Nielsen Norman, most websites score less than 50% for this metric.
Time on task
A good user experience is one that is efficient for the user. The time on task metric measures just that. It shows how much time it takes the user to complete each task.
Depending on the number of users participating in the test, you can calculate the time on task metric as the geometric mean or the sample median of all user results. Typically, the geometric mean tends to be more accurate for smaller samples (fewer than 25 participants) and the mean works better for large scale studies.
This metric answers the question as per why users spend more time on certain tasks. It provides powerful insights into user behavior when paired with the time on task metric.
The error rate shows how many times users make errors while attempting to complete the tasks. It is necessary to distinguish between the two types of errors: slips and mistakes. The user might make a typo while filling in the email address, mistype a password, or pick the wrong month for a reservation. It has nothing to do with the interface design but is a slip and usually does not count as error. A mistake is when the user clicks on a heading or image that isn’t clickable, enters today’s date instead of date of birth or types both first and last name in the same field.
While it’s important to consider that users can make a number of errors per task, which complicates the calculation of the error rate. You cannot simply derive a percentage of errors. Instead, you should do one of the following.
- Track the number of errors per task and determine where users make errors the most.
- Calculate the percentage of errors for each task versus total number of errors.
In both cases, you will identify the task with the most pain points for the user.
Users may complete all tasks easily with 100% success rate, no errors and still not feel satisfied with the user experience. This is where the subjective satisfaction metric comes into play.
Users generally prefer websites with higher usability metrics, however the user satisfaction is not always correlated with better usability. In fact, 30% of the time users prefer websites with usability worse than average.
This means you need to measure user preference alongside other usability metrics. Here’s how to do that.
After going through all of the tasks, provide the users a very short satisfaction questionnaire, which can even consist of a single question: “On a scale of 1–10, how satisfied are you with using this website?”. A simple average score will give you an understanding of users’ subjective satisfaction of your website.