UX Goals, Signals, Metrics: What Makes a Good UX Metric?

What makes a good UX metric?

This is a three-part series about how to use the goals, signals, metrics framework to define and measure user experience goals. Missed the first two posts? Read part 1: “Align Business Goals with User Goals” and part 2: “Define Signals that Indicate Progress.”

Since you’ve read the first two posts in this series, you know that goals, signals, metrics is a framework for helping your team decide on the metrics that matter the most in achieving your UX goals. If you haven’t already aligned business goals with user goals or taken the time to define the signals that show you are moving towards—or away from—those goals, don’t bother defining metrics. They won’t be meaningful and they will waste valuable time you should be spending to improve your digital product and demonstrating a UX ROI.

Assuming you have done the work required during the goals and signals parts of the framework, let’s move on to what makes a good UX metric—something measurable at the product level that you can track over time.

What makes a good metric?

In the first two parts of this series, we talked about the goal of finishing a triathlon in under three hours. You also have a subset of goals for swimming, cycling, and running. If you’re the average person, your goals probably differ drastically from elite triathletes and they also probably differ from the goals of your next-door neighbor. Similarly, your digital products goals will differ from your competitor’s goals. And, as a result, the signals and metrics associated with those goals will also be unique to your product.

Remember, signals can be both actions and psychological (and even physiological) reactions that tend to fall into one of these categories:

  • An action or series of actions the user takes.

  • How efficiently a user can complete those actions moving towards a goal.

  • How the user felt—either overall or completing a certain action or goal.

Keep in mind that, in many cases, out-of-the-box metrics—like those you find when you add Google Analytics to your website or app—will not be useful in measuring the signals listed above. You will either need to customize what your analytics software is tracking or you will need to come up with different ways to measure.

So what makes a good metric? First, the metric must be something you can compare over time. Because of this, UX metrics are often percentages or averages. For example, “percent of users in Group X who added 3 or more entries this week” or “average number of entries added per user per day.” Second, the metric needs to be something you’ll actually do something about. In other words, you’ll take action as a result of measuring. Otherwise, it’s not worth the effort to measure.

Examples of UX Metrics

Let’s walk through a few examples of how UX metrics are developed based on a specific UX goal:

Example 1: Fixing Usability Issues

Goal: Let’s say that you’ve discovered through a usability study that many users can’t complete critical tasks on your digital product. Your team needs to fix this problem before focusing on any other goal.

Signal: Users can complete critical tasks A, B, and C.

Metric: After each round of improvements, measure the percent of users who are able to successfully complete Critical Task A (or B or C) through a usability study.

Example 2: Encouraging Product-level Behaviors

Goal: Increase subscriptions. However, through user research, you realize users aren’t moving beyond the onboarding because the process is tedious. Your focus is to streamline the onboarding process.

Signal: Users complete the on boarding process and go on add a post (a behavior you’ve uncovered is associated with long-term engagement).

Metric: With each release of changes, compare the percent of new users who complete the onboarding process and go on to add a post. In this example, you’re comparing cohorts, or groups of people who start using your product at the same time.

Example 3: Focusing on How Users Feel

Goal: You’re focusing on making the product easier to use.

Signal: Users think the task is easy.

Metric: You can measure perceived ease-of-use by asking users, after they complete the task, to rate how easy a task was to complete, using a Likert scale, for example. Compare average ease-of-use ratings as you make changes to the product.

Notice in all three of these examples the UX metric is specific to that product at that particular point in time.

To summarize the goals, signals, metrics, framework: first align business goals with user goals. Then figure what signals indicate that you are moving towards—or away from—those predefined goals. A “good” metric is a metric specific to your product and business and it’s measurable at the product level. Keep in mind that what you measure will change over time. As your product changes, your goals, signals, and metrics change along with it.