Think of the last time you watched someone give an update on a video call and felt a small pang of sympathy. Their sentences lost their shape. They said "does that make sense?" twice. They trailed off before the point. You probably blamed their nerves.

Try a different explanation. Maybe they were talking to a wall, and you were part of it.

You were listening. Your mic was muted, your face was neutral, and you were nodding inside your own head, where nobody could see it. To the speaker, that looks the same as not listening at all.

The listener is half the speaker

We tend to treat talking as something one person does and listening as something another person receives. Conversation researchers have argued against this picture for decades. The linguist Victor Yngve coined the term "back channel" in 1970 for the little signals listeners send while someone else holds the floor: "mm-hm," "right," a nod, a raised eyebrow, a short "wow." These aren't interruptions. They're a second track of the conversation that runs underneath the first.

The most vivid demonstration I know comes from Janet Bavelas and her colleagues at the University of Victoria. In a 2000 paper called "Listeners as Co-Narrators," they had people tell a personal story to a listener. For some speakers, the listener was secretly given a distracting task, which kept them from tracking the story closely.

The distracted listeners still looked at the speaker and still sat politely. What they gave up was the well-timed, specific reactions. The effect on the speakers was striking. Their stories got worse. They told them less fluently, and the endings in particular fell flat or fizzled out. Speakers with an engaged listener told the same kind of story and landed it.

A speaker needs a continuous stream of evidence that the message is arriving. Without it, they start to doubt, and doubt costs them. They hedge, over-explain, and speed up to get out of the spotlight. That drives much of what we call "bad presenting" in meetings.

Why video calls starve the channel

In a room, the back channel is nearly free. You lean forward, your face moves, and a breath of laughter escapes. The speaker catches all of it in peripheral vision without trying.

Video takes most of that away, in several ways:

  • Muting. The polite default for a group call removes every "mm-hm" and "ha." The audio back channel disappears entirely.
  • Small, distant faces. A nod that would read across a table shrinks to a few pixels in a grid of tiles.
  • Looking at the wrong place. When you watch the speaker's face on screen, you aren't looking at the camera, so you never appear to make eye contact. Looking at the person and appearing to look at them are two different acts.
  • No-one-is-sure-who's-talking effects. In a large call, each listener assumes someone else will respond. The speaker gets silence from everyone.

The result is a speaker staring into a grid of attentive-but-inert faces. Attention is there, but nothing shows it. A neutral face can also be misread as displeasure, which compounds the problem.

Specific beats generic

One detail from Bavelas's work is worth holding onto. Her team distinguished generic responses from specific ones. Generic responses, like a nod or "uh-huh," say "I'm following." Specific responses are tied to the content of the story: a wince at the painful moment, a gasp at the twist, a laugh at the joke, a "no way" when the thing goes wrong.

Specific responses prove something generic ones can't. They show that you're tracking this story, not merely occupying space. They also help the speaker, because they signal which parts landed.

This matters on video because you often have only a few signals available. Spend them on the specific ones. A single genuine reaction at the right moment does more than a minute of steady nodding.

What to do about the muted grid

You don't need to perform enthusiasm. You need to make your listening legible. The goal is to hand the speaker the same evidence they'd get in a room.

A few mechanisms are worth using on purpose:

  1. Use your face deliberately. A real reaction, such as a raised brow at a surprising number or a smile at the good news, is visible even at thumbnail size if it's slightly larger than you'd do in person. Think of it as turning the volume up by a third.
  2. Use the chat as a back channel. A short "that's a useful point on the timeline" or a single emoji lands as a specific response in a medium where audio is off the table. The speaker sees it appear while they're still talking.
  3. Unmute for short sounds in small groups. In a call of a handful of people, an occasional "yeah" or "mm" is welcome. In a call of thirty, stay muted and lean on the face and chat.
  4. Close the loop with a paraphrase. When the speaker finishes, repeat back one specific thing: "So the launch moves to the 14th, and design is the blocker." This is the strongest proof of listening, because it can't be faked.

The fear that you'll look fake

Many quiet, thoughtful people resist all of this. They feel that reacting visibly is theatrical, and that real listeners should be able to sit still.

But sitting still reads as absence, not depth. The speaker can't see your inner life. They can only see what you send. Showing a reaction isn't pretending. It's translating something you already feel into a form the other person can receive.

The discomfort usually fades quickly. Once you notice how much easier people talk when you respond, the habit starts to feel generous.

Your next moves

Try these in your next meeting. Each takes under a minute to set up.

  • Pick one specific reaction per meeting. Before the call, decide that you'll give at least one visible, content-specific response (a laugh, a raised brow, a "that's a good catch") when someone says something you actually react to.
  • Look at the camera for the first sentence or two when you respond. When you say "mm-hm" or answer a question, glance at the lens so the speaker feels eye contact, then return to the screen.
  • Send one specific chat message to a speaker mid-talk. Something like "the budget point helps" works. Keep it to one line so it doesn't pull anyone's attention from the speaker.
  • End one conversation with a one-sentence paraphrase. After a colleague finishes, say back the single most important thing you heard and ask, "Did I get that right?"
  • Check your own tile. Glance at how you look when you're listening, then ask whether a stranger would read you as engaged or checked out. If it reads blank, add a slight head tilt or a small smile.

What this has to do with the post-mortem

Most of what goes wrong in meetings gets blamed on the speaker: they rambled, they hedged, they faded at the end. Some of that is theirs. But some of it is the room, and you're part of the room. The research suggests that how well people speak depends partly on how well they are heard, and being heard has to be visible.

That's the habit we built Meeting Mortem around: a short, honest look back at a meeting, including the part you played as a listener. It asks what you noticed and what you wish you'd said or signaled, so the next meeting goes differently. If you'd like a place to do that, you can find it at meetingmortem.lumenlabs.works. Whether or not you try it, the idea stands on its own: next time someone is talking into a grid of silent faces, be the one who lets them know they're getting through.