30 August 2026
Most video call etiquette lists are twenty items long and nobody follows them. Three or four habits do almost all the work: muting when you're not talking, leaving a half-second before you speak on a laggy line, and saying who you're addressing before you ask a question. The rest is decoration.
This is the one people get wrong in good faith, not out of laziness. Your brain filters out your own keyboard, your own dog, your own fridge compressor kicking on, because you've heard them a thousand times and they carry no information for you. Everyone else on the call hears them fresh, at whatever volume your microphone's gain happens to be set to that day, layered under whoever is trying to talk.
The fix isn't 'be more considerate.' It's structural: mute by default, unmute to speak, mute again. The habit only sticks if it's cheap, which is why the spacebar-push-to-talk pattern some apps offer beats remembering to click an icon. If your platform doesn't have that, get comfortable with the keyboard shortcut and use it the way you'd use a turn signal — not when you remember, but every time.
On a call with 150ms of round-trip delay, two people who both think it's their turn will start talking at exactly the moment the other one stops — because they can't hear the stop until it's already 150ms old. The result is the two-second stutter-stop-stutter every remote team knows: 'sorry, go ahead' — 'no, you' — 'sorry, go ahead.'
The habit that fixes this costs almost nothing: count one full second after someone appears to finish before you start. It feels slow in the room. It is not slow on the wire, because the gap you're adding is smaller than the confusion you're avoiding. This matters more on international calls, more on cellular connections, and more in groups of five or more, where the odds that two people jump in at once go up with every added person.
'Can you get that done by Friday?' lands differently in a room with one person than in a grid with eight faces. On camera, a question without a name attached produces a two-second silence while everyone individually decides whether it's aimed at them, followed by two people answering at once, followed by an apology loop. Name the person first — 'Priya, can you get that done by Friday?' — and the two-second silence disappears because seven people know immediately it isn't theirs to answer.
This one is almost free and almost never done, because in a physical room eye contact does the job automatically. On video, especially in a grid layout, eye contact does not travel. The name has to do the work eye contact used to do.
Most etiquette lists spend their length on things that don't change the call for anyone but the person doing them. Camera-on is a genuine trade-off between presence and cognitive load, not a moral failing — we've covered that separately. Whether you eat on camera affects your own image, not the meeting's outcome. A tidy background behind you changes nothing about whether the decision gets made.
The test worth applying to any rule before you adopt it: does breaking this make the call worse for someone else, or just for how you're perceived? Muting fails that test in the second person's favor — an unmuted mic degrades the call for everyone listening. A messy bookshelf fails it in nobody's favor. Rules earn their place on a list by clearing that bar, not by sounding responsible.
Nobody joins a call intending to talk over the CFO. The mistakes that keep happening are structural, not moral: someone forgets their laptop mic picks up their mechanical keyboard at a volume they can't hear; someone answers a question aimed at the group because no name was attached; someone jumps in early on a satellite connection because the pause looked like a full stop when it wasn't.
AVAY's transcript makes one of these visible after the fact in a way live listening doesn't: when the same three seconds get transcribed as garbled or overlapping speech more than once in a meeting, that's usually the lag-collision problem, not a content problem, and it's worth raising with the group rather than assuming someone was being rude.
| Assumption | What actually happens | |
|---|---|---|
| Unmuted background noise | I can hear if I'm making noise | You filter your own sounds out; everyone else hears them at full volume |
| Jumping in after a pause | A pause means it's my turn | On a laggy line the pause is often still in flight, not a full stop |
| Asking the group a question | Someone will just answer | Everyone waits for someone else, or two people answer at once |
| Camera off | It signals disengagement | It's usually bandwidth, fatigue, or a bad hair day — no bearing on outcome |
No — mute is for when you're not speaking, not a measure of how much airtime you take. If you're the one talking, stay unmuted; the rule protects the room from noise during everyone else's silence, not from your voice during your turn.
It's less rude than guessing and answering the wrong question. Lag and audio dropouts are common enough that asking for a repeat reads as normal, not as inattention, especially on international or cellular connections.
Not by itself, but naming who you're addressing before you speak removes the ambiguity that causes two people to unmute at once. The collision is the problem, not the order.
Say it plainly and specifically — 'there's a keyboard sound coming through, can you mute when you're not talking' — rather than hinting. Most people genuinely can't hear their own background noise and will fix it immediately once told.
Three habits do the real work: mute when you're not speaking because you can't hear your own noise, leave a beat before you talk on a laggy line, and name who a question is for. Everything else on the long lists is cosmetic.
AVAY is a video meeting platform that transcribes the call itself — no bot joins, because there is nothing to join. Start one at avay.ai, read how each part works in the documentation, or see what it costs.