LLOS.ai
How Things Work
● How Things Work · Module 3 · Feedback

Feedback

You press the button. Nothing lights, nothing clicks. So you press it again — and again.

You know it’s a control, and you know which one to press — Modules 1 and 2. Now the quietest question of all: did anything happen? Every button, every tap, every ‘submit’ owes you an answer. When it stays silent, you can’t tell success from failure — so you jab it again, or give up, or break something twice.

Feedback is the system telling you what it just did. Do it well and a single flicker turns doubt into trust. Do it badly — too slow, too quiet, or too loud — and people over-press, panic, or stop believing the interface at all. You’ll feel that on a dead lift button, then a silent wait, then an interface that will not stop shouting.

~9 min3 felt labsthe two Gulfs10-question bankglossary · FAQ
The journey — how things work
signalsaffordance
mappingmatch the world
feedbackdid it work?
modelsthe story
constraintsno wrong moves
errorforgive
Same promise, step three: when a design confuses you, it’s the design’s fault — not yours.
What you’ll be able to do
01 · Did it hear me?

The dead button

You press the lift button. Nothing lights, nothing clicks. Did it register? You press again. And again. This is the most common feedback failure alive — a control that works but never says so. In the silence you assume it didn’t hear you, so you over-act.

Call the lift on both. First the silent button, then one that answers with a single flicker. Feel how silence makes you jab, and one flicker makes you trust.

The dead button
Press to call the lift. One button answers; one says nothing at all.
▶ Start on Silent and feel yourself jab. Then switch to Answers — one flicker and you press once.
presses: 0
02 · The Gulf of Evaluation

The silent wait

One press earns one flicker. But what about the seconds after — while the thing actually happens? You tap ‘Send,’ and now… nothing. Is it uploading? Did it freeze? Should you tap again? That anxious gap between what the system is doing and what you can see has a name: the Gulf of Evaluation.

Send the same photo twice — once into silence, once with a bar that shows every step. Same four seconds, opposite feelings.

The silent wait
Tap send. One version tells you nothing until it is done; one shows every step.
▶ Start on No signal and feel the doubt climb. Then Progress — the same wait, now calm.
📷
Ready to send.
doubt:
03 · When everything shouts

The screaming interface

So more feedback is always better? No. When every icon badges, every action toasts, and everything buzzes at once, the signal drowns in its own noise — and you learn to ignore all of it, including the one alert that actually mattered.

Your payment just failed. Find that alert — first in an app where everything screams, then in one that speaks only when it must.

Find the real alert
Your payment failed. Click the Bank alert — as fast as you can.
▶ Start on Everything shouts and try to find it. Then Only what matters.
time to find it:
The craft of talking back

Good feedback is immediate, honest, and quiet

Feedback sounds like a nicety until you meet a system that withholds it. Named clearly, it is one of the quiet forces that decides whether a product feels alive in your hands or dead.

Every action deserves an answer.

The first rule is immediacy. A control must react the instant you touch it — a press-state, a highlight, a click. Under about a tenth of a second it feels instant; past that, doubt creeps in. If the real work takes longer, acknowledge the touch now and report progress as it goes.

The second is honesty and information. Feedback should say what actually happened, not just that something did. ‘Sent’ beats a shrug; ‘Uploading 60%’ beats a bare spinner; ‘Card declined — check the expiry’ beats a red box. The slower or more consequential the action, the more the feedback must explain.

The third is proportion and place. Match the signal to what matters — a whisper for the routine, a clear alarm for the rare emergency — and never let everything shout, or people go deaf to all of it. And put the response where the action was: the error by the field, the confirmation by the button, the change on the thing that changed.

So the working test is a question: after every action, could the user answer “did it work, and what happened?” without guessing? The closer to yes, the more the interface feels like a conversation — and a conversation is a thing people trust.

Cast your vote

Which feedback is clearer?

Five face-offs. Same moment, two responses — one leaves you guessing, one tells you exactly what happened. Pick the clearer, then see why.

Round 1 of 5Score 0
Seen in the wild

Feedback done right

Good feedback is almost invisible — you only notice it when it is missing. Here are two you rely on without a thought.

Saved
Toast confirmation
A quiet yes

A small ‘Saved’ that appears and fades. It never blocks you, but it answers the only question you had: did that work? Silence would have left you re-checking.

Uploading 68%
Progress bar
Progress, not a spinner

A spinner says only ‘working.’ A bar says how far and how long — so a four-second wait feels like a countdown, not a freeze.

Where feedback leads next

Feedback tells you what happened. But why did you expect something else? Because you were running a story of how the thing works — and when that story is wrong, feedback alone won’t save you. That is the next bridge: conceptual models, the story a design puts in your head, and how to make it the right one.

Take it further

The real references

Feedback and system status are among the oldest, best-studied ideas in the field. Four worth your time. (Opens in a new tab.)

The vocabulary

The Feedback glossary

Six words that turn “this feels unresponsive” into a precise fix. Tap a card to flip it.

The nuance

Questions you actually ask

Busting the myths

What people get wrong

Myth
“More feedback is always better.”
No — when everything badges, buzzes and pops, the signal drowns in noise and people learn to ignore all of it, including the one alert that mattered. Proportion is the skill.
Myth
“A spinner is enough for a wait.”
A spinner only says ‘working.’ For anything past a second or two, show progress — how far, how long — or the user can’t tell a slow success from a silent freeze.
Myth
“Errors should be loud pop-ups.”
Put the error where the problem is — beside the field, in plain words, with the fix. A modal that says ‘Error’ and blocks the screen tells the user nothing and costs them their place.
Myth
“If it worked, no news is good news.”
Silence reads as failure. People re-tap, re-submit, or give up. A quiet confirmation of success is as important as a warning of failure — both close the loop.
Train the eye

Read the feedback

Ten questions in seven formats, tagged by thinking level — each with layered hints: a nudge, the reasoning, then a deeper connection. A wrong answer opens a door, never a dead end. Served from the question bank, never hardcoded.

Loading the question bank…
Question 1 of 10
Multiple choice
Keep going

Where this connects

▶ Play the labs