Don’t report descriptive statistics like success rates and averages unless you ran a quantitative study. Reported numbers must be qualified with statistical information such as confidence intervals or statistical significance.| Nielsen Norman Group
A confidence-interval calculation gives a probabilistic estimate of how well a metric obtained from a study explains the behavior of your whole user population.| Nielsen Norman Group
In addition to being expensive, collecting usability metrics interferes with the goal of gathering qualitative insights to drive design decisions. As a compromise, you can measure users' ability to complete tasks. Success rates are easy to understand and represent the UX bottom line.| Nielsen Norman Group
