UX metrics: How to choose user experience metrics

UX metrics: How to choose user experience metrics

Do you track UX metrics (user experience) for your product/service pages?

If not:

  • How do you know your business has delivered the most effective customer experience?
  • Did the design change result in a good or bad user experience?
  • How can you budget for UX if you don’t know how to measure it?

There are many UX metrics to track for a website. But you will go crazy if you follow all these indicators.

Instead, focus on a few key UX metrics.

Typically 6 important indicators I will introduce to you below.

However, which UX metric should you choose to match your website SEO goals?

This article is the answer.

6 Important UX metrics

1. Satisfaction


Basically, a bad experience will not bring satisfaction to customers. You can ask users how satisfied they are with each specific feature, how was their experience today, and what they have in common about your product/service.

Because in fact, we often talk about disappointment rather than satisfaction. Therefore, the best approach is to give the user experience rating on a scale of 5-7 points, from “very dissatisfied” to “very satisfied”.

Based on surveys, and feedback in apps and website forms, you can get the most effective user experience satisfaction level.

Because you will find that most of these indicators not only indicate the level of satisfaction but also explain why users choose low or high satisfaction.

2. Introduction

Like the “Satisfaction” metric, “Referral” is also a great UX metric.

Simply, if the user has a good previous experience, it is very easy for them to recommend your product/service to others to use.

Therefore, it is not surprising that calculating referrals has become a fever in the business market. This is mainly reflected in the NPS (Net Promoter Score) – an indicator that measures the level of customer satisfaction.

Determine NPS (Net Promoter Score)

NPS is very simple. By asking the following question, you can measure the NPS score that shows customer loyalty.

The idea sounds appealing, but the NPS isn’t quite as perfect as people think it is.

In fact, this indicator is very easy to lead to mistakes in measuring UX metrics.

A referral probably doesn’t mean the customer had a good experience, maybe they’re a die-hard fan of your brand.

And users with a good experience are not sure they will recommend it to others.

Ask users to rate their ability to recommend products/services on a 10-point scale.

What is the difference between a 2-point scale and a 3-point scale? Or a 5-point scale and a 6-point scale?

What if I refer a friend instead of a colleague?

Or do I recommend it for one case and not another?

Worse still, the NPS only classifies customers into 3 groups: Critics (0-6), passive, non-praise (7-8), and promoters (9-10).

The NPS classifies customers into 3 groups
The NPS classifies customers into 3 groups

How to calculate the NPS score

NPS score is calculated by:

NPS = % Promoters – % Critics

Neutral people (no praise, no criticism) have no problem. But the problem is that this is too simple.

Level O is more noticeable than level 6 and includes degrees 2 and 6.

That is, even if the UX changes, it will not affect the NPS score.

For example, changing a user’s rating from 0 to 6 (or skyrocketing from 0 to 6) will not affect the NPS because they are still in the review group.

That’s why I say this metric is very easy to lead to mistakes in measuring UX metrics.

Instead, you should ask simpler questions like:

Instead of NPS, use a 5-point scale that measures the likelihood of referrals.

This gives users fewer choices and easier explanations for their choices.

If you still have to report NPS, you should use graphs or data visualizations to gauge the likelihood of recommending a product/service on a scale rather than just relying on NPS scores.

For products/services, you might consider asking users to rate them if they have recommended them to friends and colleagues before. Because these honest recommendations are better than just making assumptions.

3. Usability

Usability may not be the differentiating factor, but it is extremely important to the UX metrics of the product/service.

A product/service that is difficult to use will not bring the best user experience.

The best way to get a sense of the usability of a product is to ask users how they describe the product, from “very difficult to use” to “very easy to use”.

The System Usability Scale (SUS) is a popular way to measure the usability of a product, consisting of 10 questions about usability testing. The order is placed randomly to reduce the change in the bad direction.
Can be compared with the product before and after the change or with the same product/service.


Because the SUS score is usually in the 100s, it is often seen as a percentage.

Many of you think that 50 points are equivalent to 50% of users using this interface.

However, it is also a common mistake. SUS scores are not percentages, they are relative scales. Therefore, you must be careful when expressing the SUS index.

4. Ratings

You can see the rankings everywhere from Amazon to Apple’s app store.

Because customer ratings are the most obvious factor in evaluating the quality of your business’ products/services.

You can suggest customers to rank in general and rank each criterion and feature of the product/service.

You can ask customers to rate each feature.

As I mentioned above, you should use a 5-point scale and ask the customer for a reason for choosing that scale.

5. User Tasks

Actions are the most important element of UX metrics. Because the product cannot support the user’s actions, it will not be able to provide a good user experience.

You should track metrics that show user actions: sessions, user behavior, etc.

In fact, you can measure through the following metrics:

  • Completion Rate: Percentage of users who completed their task.
  • Error Rate: Percentage of users who made an error while performing a task. For example, wrong navigation to the website.
  • an Average number of errors: The number of errors the user made while performing the task.
  • Time taken: The amount of time it takes for the user to complete the task. This is really useful when measuring user performance.
  • Ease of execution: How easy it is to complete the task. SEQ (Single ease question) – Simple question is an effective way to know this factor.

6. Product Description:

How would you describe the Lamborghini sports car (pictured below)?

“Fantastic, Powerful, or Charming?”

Product Description - Lamborghini Aventador
Product Description – Lamborghini Aventador

What about Toyota?

“Applicability, Prestige or Simplicity?”

The words you use to describe your product/service can tell you what experience you provide to your users.

The best way to collect product descriptions from users is….

Use tags that indicate user reactions provided by Microsoft.

At this point, the user can choose up to 5 adjectives from the list to describe their feelings about the product/ service (as below).

Product Description
Product Description

In fact, these are only 6 important UX metrics, not the entire index to evaluate the user experience of the product/ service.

So which UX metric should be used?

“It depends on your goals”

Use google’s HEART Framework to determine effective UX metrics selection.

HEART Framework

HEART Framework
HEART Framework


Surveys are often used to measure user attitudes. For example, satisfaction, ease of use, and network advertising scores.


User engagement is measured through behavioral proxies such as frequency, intensity, or engagement over a period of time.

Example: Number of visits per user per week or number of photos uploaded per day per user.

A powerful tool for analyzing your website statistics: Google Analytics


The customer is new to the product or feature. For example, the number of accounts created in the last 7 days or the percentage of Gmail users that use a sticker.


Percentage of users returning to use the product/service. For example, how many users were active before and then still active?

Task success

Includes traditional user experience behavioral metrics, such as performance (e.g. task completion time), efficiency (e.g. percentage of tasks completed), and error rate.

These 5 elements are applied at levels – from the entire product to a specific feature. For example, in Gmail, you might be interested in general product adoption. But it is also possible to apply key features such as labels or archives.

Many people ask me:

“Why measure Adoption and Retention when you can only count unique users?”

Should all 5 elements be applied?

It’s important to count the number of users in a particular period (e.g. active users all week).

But if you measure two more factors of Adoption and Retention, you will easily distinguish new users from old users. From there, you can show the growth or stability of your user database.

This is especially useful for:

  • New products/features
  • Redesigned product

You don’t have to create all of the above 5 metrics. Choose the most important metrics for each project of your business.

The HEART Framework can help you decide whether to choose or ignore a particular element.

For example, the customer “accessing” the product makes no sense in today’s business context. Users use the product every day as a habit.

In this case, the whole team must focus more on “Happiness” or “Task Success”. However… it is still possible to consider using the “Engagement” factor for specific product features.

So how do you determine if the metric is being tracked or performed? In fact, there is no magic “HEART Framework” that can define it for you. The most useful index can only be suitable for each specific product or project”.

So what to do …?

Perhaps the following procedure will help you.

Process Goals – Signals – Metrics

Think about metrics in a long list. But this seems very difficult to choose and prioritize which indicators to use. Basically, you want to use a few key metrics that a lot of people are interested in.

And to define that index, you need:

“Define your goals to measure the whole process against that goal”

However, it is not easy to define the goal of the project and where we are in that HEART Framework.


For Youtube, one of the important goals is the “Engagement” section. You want users to enjoy the videos they watch and keep looking for more videos or channels they want.

But…you need to have separate goals for each project or feature instead of giving a general goal for the product. For a Youtube channel, the essential goal is the “Task Success” factor.

When users type in the search bar, you always want them to easily and quickly find the most relevant videos or channels.

A common mistake here is targeting old metrics.

For example:

“Yes, my goal is to increase website traffic”

Yes, most people hope so, but how does improving the user experience help increase traffic?

Are you interested in increasing the reach of old users or attracting new users?

Almost every team member has their own goals for the project.

Therefore, this process will help you to unify those goals together in the most reasonable way.


How to identify success or failure in goal setting through user behavior/attitude?

For example, a sign of “Engagement” for a YouTube channel is probably the number of videos the user has watched on the channel. Or better yet, know how much time they watch those videos.

The sign of failure in the “Task Success” section for YouTube search volume is that the user types in search but does not click on any results.

There are often a lot of very useful markers for a particular goal.

But once you’ve found some potential candidates for that goal, you should stop searching and start researching and analyzing to choose the most useful sign.

First, you need to know:

  • Is it easy or difficult to track that sign?
  • Is your product documented about the relevant actions?
  • Can you run product surveys on a regular basis?

Second, you should choose signs that are compatible with the design change. If there are already useful signals, you should analyze these data and try to grasp which signals can predict your target.


Once you’ve chosen your signature, you can further tailor the markers to the metrics you’ve tracked and use them for comparison in A/B Test (a comparison testing method between the two versions to increase conversion rate).

For example, in Youtube interaction, I usually implement “Time spent watching videos by users” as well as “Average time spent watching videos per day per user”

In the previous stage, you can get many stats from the signs. You will have to analyze the collected data and decide which metric is most suitable.

And simplifying raw numbers makes them more meaningful. (e.g: give an average or a percentage).


Overall, the Goals – Signals – Metrics process can help you prioritize different metrics that align with your top goals.

Avoid adding other unnecessarily attractive stats to the list.

  • Do you use those data to make decisions?
  • Do you need to keep track of those metrics?

Focus on metrics related to your goals to avoid wasting energy on unnecessary things.

Complete the table below to see if your product/service provides the best user experience.


Try applying the HEART Framework and the Goals-Signals-Metrics process and comment and share your results below this article!

Good luck!

Leave a Reply

This site uses cookies to offer you a better browsing experience. By browsing this website, you agree to our use of cookies.