i

Feedback & the Three Thresholds

STR·06 · 6 of 11

0.1s feels instantaneous; 1s keeps the flow of thought; 10s is the limit of attention — after that, people leave.

documented A named, findable source says this. Why: The 0.1 / 1 / 10 second thresholds are Card, Robertson & Mackinlay and Nielsen.

The same task, built twice

Both panes below are real interfaces. They respond, they can be got wrong, and each is timed from your first move in it. The times are not a study — two panes, one reader, no controls. They are here to make a difference you can feel into a number you can see, and nothing is recorded or sent anywhere.

Your task

Place the order. Both take 2.5 seconds to respond — press the button as many times as you feel you need to.

Silent—

Enamel jug · £22 · delivery Thursday

The request is in flight and the interface has not said so. There is no way to tell a slow system from a dead one, so the reasonable thing to do is press again — and it counts every press.

Answers—

Enamel jug · £22 · delivery Thursday

Same 2.5 seconds. The button takes the press, says what it is doing, and stops accepting new ones. One order, and you knew within 100ms that you had been heard.

What to watch Your own duplicate count on the left. Nobody presses twice out of carelessness; they press twice because the first press produced no evidence it happened.

The principle

OriginResponse-time limits synthesized by Nielsen (1993) from earlier work (Miller 1968; Card et al. 1991).
EvidenceThe quiet constitution behind spinners, skeletons, optimistic UI, and progress bars — each is a different answer to which threshold you're about to miss.
Limits & misuseThe numbers are bands, not physics; context moves them (a payment may be granted patience a search never gets). And faking progress to soothe the 10s band is a loan from the basement.