The Icon Cried Wolf

The Icon Cried Wolf

Why this matters

The room's verdict on a calendar warning icon was unanimous: the alarm outranked the emergency. The more useful lesson was the sentence the designer said next.

The Icon Cried Wolf Crit found a warning icon louder than the problem it flags. On severity as a volume knob, and what happens when a designer names her constraint out loud. The room's verdict on a calendar warning icon was unanimous: the alarm outranked the emergency. The more useful lesson was the sentence the designer said next. Treat severity as a designed variable: match the alarm to the consequence and reserve the loudest treatment for the rarest state. In crit, name your constraints before feedback starts; the room can only help with the problem it's allowed to solve.

Crit found a warning icon louder than the problem it flags. On severity as a volume knob, and what happens when a designer names her constraint out loud.

Silent crit this week took up insurance verification on the calendar our users spend time with day in and day out. The brief is sound: clinicians need to know a client’s coverage is active before the session starts, so the design surfaces a warning when insurance is inactive, needs review, or needs more information, and stays quiet when everything is fine. Silence for the healthy state is the right instinct. The trouble was the voice it used when it spoke. (things got loud)

The room’s verdict was unanimous, which is rare for us: the warning icon was louder than the problem. A coverage flag is a problem for this week, and the icon said the building was on fire. Severity is a designed variable, a volume knob, and most of us only ship it at ten. Our users know the failure mode better than we do; in clinical settings it’s called alarm fatigue, and the outcome is always the same. When everything shouts, people stop hearing all of it, including the one alarm that mattered.

Severity is a volume knob, and most of us only ship it at ten.

The better lesson was how the designer took it. No defense, no explanation of intent. She agreed the treatment was too strong, and then she named her constraint out loud: “We thought about rebuilding the environment. That’s just not on our timeline right now. Rebuilding would be ideal, but we can’t. We’re working within this.” Most crits die of unspoken constraints; feedback piles up against a version of the problem the designer isn’t allowed to solve. Naming the box converts the room from wishing to helping.

And the room changed on the spot. Rebuild proposals stopped, treatment proposals started: quieter iconography, reserved color, warning states that scale with the consequence. Revisions go to the design channel next, with a September rollout to a small group. The icon will get quieter. The move worth stealing is the sentence that made the feedback useful.

The teachable part

Treat severity as a designed variable: match the alarm to the consequence and reserve the loudest treatment for the rarest state. In crit, name your constraints before feedback starts; the room can only help with the problem it’s allowed to solve.

Filed under CN Crit Notes — Feedback, with the reasoning shown.