Usability — Examples

Example 1 — Turning “sign-up is usable” into a target

A team keeps saying its sign-up is usable, but nobody has named the user or what finishing looks like. Rewrite the claim as a usability target with the three components measured, so the next change has something to move.

Claim:  "sign-up is usable"
User:   everyone?
Task:   ?
Done:   ?

Solution

User:      first-time buyer, no account
Goal:      create an account and reach the basket
Done when: account exists and the basket shows the items
Effectiveness  completed without help      87%    (illustrative)
Efficiency     median time on task         2m 10s (illustrative)
Satisfaction   rated easy, 1-5             3.8    (illustrative)

Naming the user and the completion definition is what makes the numbers mean anything: the same screen is one result for a returning buyer and another for a first-timer. Leaving the user as “everyone” is the usual mistake, because an average across two populations fits neither and moves for reasons nobody can explain.

Example 2 — Faster for regulars, harder to learn

Two candidate layouts are compared on one task. The evidence behind “B is faster” came only from people who already knew the product. Restate the result so the trade is visible.

Variant A  familiar menu, one extra step
Variant B  shortcut-first, no labels
Claim:     "B is faster"
Evidence:  only returning buyers tried it

Solution

                          First attempt      Second attempt
A  median time on task       1m 40s              55s
B  median time on task       4m 05s              30s
B  first-attempt errors      3 of 6              —          (illustrative)

Reporting the two populations side by side shows what the claim hid: B is efficient once known and slow to learn, so the choice depends on which population the product serves. Collapsing learnability and efficiency into a single “faster” is the mistake, because it hides that the two variants swap places between a first visit and a fortieth.