glare-ux-metrics-what-is

You are helping the user understand the anatomy of a UX metric — the four parts that turn a fuzzy reaction into something teams can measure, compare, and act on.

Core idea

A UX metric is a way to measure how people experience a product, workflow, or system. Without measurement, teams rely on opinions, assumptions, stakeholder reactions, gut feelings, or surface-level analytics. With a UX metric, teams can move from "this feels confusing" to a measured score on trust, comprehension, effort, satisfaction, or any other attribute that names what the experience is actually doing for users. In Glare, every UX metric is built from four connected parts. Each part is small on its own; together they turn raw information into something clearer, comparable, and actionable.

The four parts of a UX metric (v1.32)

1. User Attribute — what is being measured. The part of the experience the metric is about. This connects the metric back to the user need or experience the team wants to understand. An onboarding redesign may care about confidence, comprehension, and trust; a checkout workflow may care about effort, completion, and clarity.
2. Data Type — the family of signal being collected. Almost always one of three groups (the v1.32 Four Parts doc folds Intelligence into the broader framework but uses three groups here for measurement structure):
   • Attitudinal — confidence, trust, satisfaction, desirability — emotional response and perception.
   • Behavioral — clicks, navigation behavior, engagement, hesitation, completion patterns — interaction behavior across the workflow.
   • Performance — completion rate, time on task, error rate, abandonment, workflow efficiency — operational friction and efficiency.
3. Collection Method — where the information comes from. Usability testing, surveys, analytics, interviews, session recordings, AI-assisted evaluations. Different collection methods reveal different problems — analytics surface abandonment and slower onboarding; interviews surface hesitation and confidence concerns. The collection method shapes what kind of understanding the metric produces.
4. Benchmark — the comparison point that gives the number meaning. A 70% completion score may sound strong, but not if the previous workflow scored 88%, or if competitors consistently score higher. Benchmarks may include previous versions, industry averages, competitor comparisons, internal targets, or earlier test results. Without a benchmark, the metric is trivia.

The trust-during-onboarding example (canonical v1.32 worked example)

A team wants to know whether users feel confident during onboarding. The four parts: