Usability is a measurable property, not an evaluative adjective

What it is

Usability is not an adjective like "good to use" but a set of measurable, quantifiable properties: completion rate, error counts, task time, help requests, satisfaction scale scores. Treating it as an adjective keeps the discussion at the level of positions; treating it as a measurement object moves it onto evidence.

Why it happens

"Measurable" means the components can be operationalized: effectiveness maps to completion rates and error types, efficiency to time and steps, satisfaction to standardized scales. Once operationalized, baselines and thresholds exist—"raise completion from 62% to 90%" is a testable claim, while "make it better" cannot be accepted. Measurement also exposes trade-offs: a change can raise efficiency while lowering satisfaction, a split invisible to adjectival talk.

Studying it

The standard approach is task-based measurement: define the task and success criteria, have target users perform it, and collect behavioral metrics (time, errors, paths, help requests) plus subjective metrics (SUS and similar scales). Report sample sizes and dispersion, not just means; re-measure before and after iterations with the same task and criteria so metric changes attribute to the change. Field observation supplements lab measurement for contextual effects it cannot see.

Where it stops holding

Measurable is not the same as accurate: lab tasks simplify real goals, and the weighting chosen among metrics shapes conclusions. Satisfaction scales drift with expectation and the rising baseline of comparable products. Some properties—trust, long-term reliance—lack short-task instruments, and measurement cost limits granularity: not every design decision earns a test round.

Applying it

• Define acceptance metrics and thresholds per release, such as core-task completion rate, first-use time, and a SUS floor.
• Re-measure before and after redesigns with identical tasks and criteria, reporting metric changes with sample sizes; "feels smoother" is not delivery evidence.
• When metrics move in opposite directions (efficiency up, satisfaction down), report the split so the trade-off enters the decision explicitly.

Related

• Same group: B5.01.1 Usability is the extent to which specific users achieve specific goals in specific contexts · B5.01.2 Usability is meaningless to discuss apart from users, goals, and context
• Nearby: B5.02 Effectiveness, efficiency, and satisfaction · Q2 Usability Testing
• Search terms: usability metrics · system usability scale · task success rate