glare-ux-metrics-performance
You are helping the user select, structure, and apply Performance metrics — the family that measures how efficiently a workflow actually functions during real usage.
Canonical metric list (recommend ONLY from these eleven). Abandonment Rate · Bounce Rate · Click-Through Rate · Completion Rate · Drop-off Rate · Error Frequency · Error Rate · Recency · Session Duration · Time on Task · Visit Frequency. Do not invent or borrow performance metric names from other frameworks — if the user's question doesn't match one of these eleven, route to
glare-ux-metricsor another type-level skill instead.
Core idea
Performance metrics measure how efficiently a workflow, system, or experience actually functions — how long tasks take, where users drop off, where errors increase, how much effort workflows require, whether systems perform reliably under real usage. Product teams often focus heavily on whether users completed a task, but completion alone rarely explains how difficult the experience actually was. Users may finish onboarding while struggling the entire time, complete checkout after repeated errors, use AI recommendations while constantly double-checking results, or finish tasks successfully while moving much slower than expected. Performance metrics measure the cost of friction.
What performance metrics catch that other types miss
Behavioral metrics explain what users do; attitudinal metrics explain how users feel; performance metrics explain how efficiently the system actually functions underneath the experience — time, effort, errors, reliability, workflow efficiency, operational breakdowns. A workflow may technically perform well while users feel frustrated, hesitate repeatedly, lose confidence, or struggle to understand what to do next.
The eleven performance metrics (v1.32)
reference.md carries the per-metric summary table — one row per metric with definition, when to use, common mistakes. The eleven metrics:
Abandonment Rate · Bounce Rate · Click-Through Rate · Completion Rate · Drop-off Rate · Error Frequency · Error Rate · Recency · Session Duration · Time on Task · Visit Frequency