Knowles Knowledge Centre•Virtual Meeting Cues

A Video Grid Destroys
Most of What You Read

A practical framework for the cues the medium removes, and for replacing them rather than guessing.

Virtual meeting non-verbal cues, in the sense used here, are mostly not there. A grid of tiles removes mutual gaze entirely, hides hands and posture, flattens the millisecond timing that turn-taking depends on, and substitutes artefacts of the network — a frozen frame, a nod arriving half a second late — that look exactly like information about a person. What follows is a question of which cues this medium has actually removed, why reading the remainder is worse than not reading it, what explicit rule replaces each lost cue, what you should be saying because your own cues are absent, and whether anybody is still using the rule a month later.

Evidence-led guide•Singapore context•Updated September 2026

Explore the five-stage method

About this guide

Published byKnowles Training Institute
FrameworkKnowles Stop-Reading Framework — an editorial synthesis, not an official standard
Evidence baseSkillsFuture Singapore • IAL • WEF • ILO • OECD
VIRTUAL CUES NAMETHELOST STOPREADING SWAPINARULE SAYYOUROWN CHECKITHELD
Four professionals discussing workplace skills around a meeting table in a Singapore office overlooking Marina Bay

The conventional response to a flat virtual meeting is to read harder. Watch the tiles for engagement, look for folded arms, notice who has gone quiet, learn the signs of withdrawal on camera. All of that treats the grid as a slightly degraded version of a room. It is not a degraded room; it is a different channel that has deleted particular cues and invented others, and a reader who does not know which is which will be confidently wrong about colleagues who are perfectly engaged and looking at a second monitor.

This guide explains the five stages of the method, how it relates to Singapore’s Critical Core Skills, how it differs from camera presence, cue reading and meeting etiquette, and the case in which using it is the wrong move.

The practical answer

The practical answer at a glance

For practical use, the method can be organised into five stages, applied to one recurring meeting at a time:

1

Name the lost

Which cues has this medium actually removed?

2

Stop reading

Which inferences should nobody be drawing from a tile?

3

Swap in a rule

What explicit rule replaces each lost cue?

4

Say your own

What must you say because your own cues are absent?

5

Check it held

Is the rule still being used a month later?

About this framework

This five-stage structure is the Knowles Stop-Reading Framework used in this article. It is an editorial synthesis for workplace application—not an official government framework, accreditation standard or universal definition.

Four traditions supply the method, and none of them is a camera technique

The account of mutual gaze as a channel in its own right comes from the social psychology of interaction; the account of turn-taking as a system running on millisecond timing comes from conversation analysis; the account of a listener’s responses shaping the speaker’s telling comes from the study of face-to-face dialogue; and the account of a system that fails to show you the effect of your own actions comes from design.

Michael
Argyle

Michael Argyle established gaze as a channel rather than a by-product: how long people look, when they look away, and above all mutual gaze, in which each party knows the other is looking. A video grid removes the last of these completely. A camera sits beside the screen, so somebody who appears to be meeting your eyes is looking at a lens and somebody looking at you appears to be looking down. Mutual gaze is not degraded here; it is absent, and every inference that depends on it is unavailable to everybody.

Harvey Sacks and
colleagues

The conversation analysts around Harvey Sacks, Emanuel Schegloff and Gail Jefferson described turn-taking as a system of remarkable precision, in which speakers hand over the floor and listeners come in through gaps measured in fractions of a second. The relevant consequence is arithmetic rather than social. Network delay of a few hundred milliseconds is enough to disable the mechanism, which is why virtual meetings produce simultaneous starts and long dead pauses between the same people who manage a room effortlessly. Nobody has become less socially skilled.

Janet Beavin
Bavelas

Janet Beavin Bavelas showed, from close study of face-to-face dialogue, that a listener is not a passive recipient: their responses shape how the speaker tells the thing, and a narrator deprived of visible listeners tells it worse. Applied to a grid, this is why speaking into fifteen muted tiles is unexpectedly difficult and why the result is flat. The speaker is not nervous and the audience is not disengaged; the loop that ordinarily guides a telling has been opened, and both parties are working without it.

Donald
Norman

Donald Norman described what he called the gulf of evaluation: the gap between acting and being able to tell what your action did. A well-designed system closes it by showing you the effect. A video grid of thumbnails, several of them cameras off, is close to the worst case: a speaker takes an action and receives almost nothing about its effect. The practical implication is that the feedback has to be supplied deliberately, by rule, because the interface will not supply it and no amount of attentiveness invents it.

Gaze
Timing
The listener’s loop
Feedback
False signal

The recurring pattern matters more than the labels. Across the social psychology of interaction, conversation analysis, the study of dialogue and design, the same five ideas reappear: mutual gaze is absent rather than reduced; turn-taking depends on timing the network cannot deliver; a speaker deprived of visible listeners tells things worse; the channel supplies almost no evidence of effect; and what the channel does supply looks like information about people and is frequently information about bandwidth.

The last of those disposes of the usual remedy. Somebody who finds virtual meetings opaque will reliably respond by watching the tiles more carefully, and so will their organisation, because reading cues is the skill the subject appears to call for. Two managers can watch with equal attention, and the one who introduced a rule about who speaks next is the only one whose meetings improved.

The five stages of the Knowles Stop-Reading Framework

01

1. Name the lost

The first stage lists which cues this particular setup has removed. It is specific to a setup rather than to video in general, and it takes one meeting to compile.

It includes:

  • Noting that mutual gaze is absent, not reduced.
  • Noting whether hands are ever visible.
  • Noting whether posture or orientation is visible.
  • Noting how many tiles have cameras off.
  • Noting the typical delay before somebody is audible.
  • Noting what the layout shows the speaker while speaking.
  • Distinguishing what is absent from what is merely small.

A person has completed this stage when the list is specific to this meeting. “It is hard to read people on video” is not that list. “Nine of fifteen cameras are off, no hands are visible on any tile, and the speaker view hides the audience entirely” is one, and each item has a different replacement.

For example, a weekly team call is described as flat and disengaged. The list shows that the platform’s speaker view, used by everybody out of habit, shows the presenter nothing but themselves while they talk. The flatness was a layout setting, and it had been in place for two years.

02

2. Stop reading

The second stage is a prohibition rather than a technique: stop drawing conclusions about people from a tile. It is the stage that most improves judgement and the only one that consists of doing less.

This is where most harm in the subject is done. A read of somebody’s engagement from a thumbnail feels like evidence, is held with confidence, and is not correctable by the person it is about — who may be looking at the shared document on a second screen.

They include:

  • Treating a tile as insufficient evidence about a person.
  • Not reading a camera-off as withdrawal.
  • Not reading a delayed nod as hesitation.
  • Not reading a frozen frame as anything at all.
  • Not reading downward gaze as disengagement.
  • Asking rather than inferring, where it matters.
  • Recording nothing about anybody’s apparent attention.
  • Declining to raise a read with a third party.

The test is whether you can name what you would have had to see to be sure. In a room, disengagement is a cluster of signals over time, and on a grid almost none of that cluster is available. One ambiguous frame is not a thin version of the same evidence; it is a different and much weaker thing, and the confidence attached to it does not shrink accordingly.

There is a second-order effect worth naming. The people most likely to be misread are those with the worst connections, the smallest rooms and the least control over their background, which in most organisations correlates with seniority in the wrong direction. A culture of reading tiles therefore produces systematically unfair conclusions, and produces them invisibly.

03

3. Swap in a rule

The third stage replaces each lost cue with an explicit rule. This is the stage that does the work, and it feels bureaucratic for about a fortnight and then feels normal.

They include:

  • Naming the next speaker instead of pausing for one.
  • Using a hand-raise rather than a gap to come in.
  • Going round by name where everybody must contribute.
  • Asking a direct question instead of reading a face.
  • Agreeing that silence means agreement, having said so.
  • Putting decisions in the chat as they are made.
  • Leaving two seconds after a question, deliberately.

Each rule replaces a specific cue rather than improving the meeting in general. “Who wants to go next” relies on a turn-taking gap the network has disabled. “Priya, then Wei Lin” replaces it outright, costs four words, and removes both the simultaneous start and the dead pause.

For example, a meeting in which the same two people always speak is treated as a confidence problem. The round-by-name rule reveals that four other participants had been waiting for a gap that the delay never produced. Nothing about anybody’s confidence had been established by two years of evidence.

04

4. Say your own

The fourth stage supplies, in words, what your own absent cues would have conveyed. It is the only stage that works on the speaker, and it works on what they say rather than how they appear.

It includes:

  • Saying that you are following, rather than nodding.
  • Saying when you have finished, rather than trailing off.
  • Saying that you want to come in, rather than leaning forward.
  • Saying why your camera is off, once, without apology.
  • Saying when you are reading something on another screen.
  • Saying what you understood, to close the loop.
  • Not performing attentiveness for the camera instead.
  • Keeping it short enough to be sustainable.
?
?
?

A stated cue can answer:
Are you still with me?
Have you finished speaking?
Do you want to come in?
Did that land as intended?

The seventh of those bullets is the one worth dwelling on. Much advice in this area amounts to performing attentiveness at a lens — exaggerated nodding, a fixed camera stare. That supplies the appearance of the missing cue without its content, and it is tiring, slightly dishonest and useless to a speaker who cannot see nine of the fifteen tiles anyway. Four words of speech do the job.

05

5. Check it held

The fifth stage checks whether the rules are still in use. They decay, quietly and quickly, and nothing in a meeting announces that a protocol has lapsed.

Checking it held includes:

  • Naming a date a month out to check.
  • Asking whether the naming rule is still being used.
  • Counting how many people spoke, against how many attended.
  • Asking the quietest participants whether they tried.
  • Reinstating a rule rather than inventing a new one.
  • Dropping any rule nobody used, and saying why.
  • Accepting that a lapsed rule is the usual outcome.
“

A protocol that replaced a lost cue will lapse within about a month, and the meeting will go back to being read from tiles. Nothing announces this, which is why a date has to be written down.

This fifth stage connects the other four. The list says what is missing; the prohibition stops the guessing; the rules supply what the channel cannot; saying your own closes the loop for everybody else. All of that produces a meeting that works without non-verbal cues — and a meeting that works is exactly the one in which the rules start to feel unnecessary, which is how they lapse and how the guessing returns.

How the five stages work together

The stages are one sequence applied to a recurring meeting, not five habits to adopt generally.

Consider an illustrative Singapore workplace scenario: a regional team’s weekly call is reported as flat and disengaged, camera-on is made mandatory, and nothing changes except that people are tired.

The sequence runs:

01


Name the lost:

the setup is listed. Mutual gaze is absent for everybody. No hands are visible on any tile. The speaker view everybody uses shows the presenter nothing but their own image. The usual delay to the regional offices is around half a second.

02


Stop reading:

the conclusions in circulation are examined. Two participants have been described as disengaged on the evidence of downward gaze; both turn out to have been reading the shared document on a second screen, which is what they were asked to do.

03


Swap in a rule:

the lost turn-taking gap is replaced. The chair names the next two speakers each time, a hand-raise replaces coming in through a pause, and decisions are typed into the chat as they are taken.

04


Say your own:

participants begin saying the things their absent cues would have shown: that they have finished, that they want to come in, that they are reading something on another screen. It sounds stilted for a fortnight.

05


Check it held:

a date is set a month out. On that date the naming rule has lapsed, the hand-raise has survived, and the number of people speaking has fallen back towards the original two. The naming rule is reinstated rather than replaced.

The team’s first instinct was mandatory cameras. Cameras on supplied more tiles to misread, cost everybody the energy of being watched, and did nothing about a half-second delay that had disabled turn-taking for two years.

The stages depend on one another in a specific way: a prohibition on reading tiles with no rule to replace the lost cue produces a meeting in which nobody infers anything and nobody knows whose turn it is; and a set of rules with no check produces a month of improvement followed by a quiet return to guessing, which most organisations experience as the method not working.

Four situations should be named before anybody adopts this. Where somebody’s camera is off because of their circumstances — a shared room, caring responsibilities, a condition they have not disclosed, a connection that cannot carry video — nothing here is a reason to require it, and an organisation using this method to justify mandatory cameras has inverted it. Where a colleague is deaf or hard of hearing and relies on seeing a face, the cues are not optional and the remedy is the opposite of a grid: a stable layout, captions, a named speaker and whatever adjustment is owed. Where somebody is genuinely disengaged, the evidence for that is their work rather than their tile, and this method will not supply it. And where a meeting is flat because it has no purpose, protocol will make a purposeless meeting better organised and no less flat.

Is this the same as camera presence?

No. The four overlap, and they differ in what they assume about the channel.

Term

Primary emphasis

Examples

Camera presence

Appearing better on video than you otherwise would

Framing, lens height, lighting, eyeline, background

Cue reading

Interpreting other people’s signals more accurately

Baselines, clusters, engagement and withdrawal signs

Meeting etiquette

Agreeing how people should behave in a meeting

Muting, punctuality, interrupting, cameras, chat use

Virtual meeting cues, as used here

Replacing the cues the channel has removed

What is lost, what not to read, the rule, what to say, whether it held

Researchers have a name for the confusion that a family of overlapping labels creates: the jangle fallacy, the assumption that two things must be different because they carry different names. “Virtual presence,” “online communication,” “remote facilitation” and “video etiquette” describe substantially the same territory with different emphasis, and nothing in this method depends on preferring one of them.

The practical consequence is twofold. Treating the subject as replacement rather than as reading invites the question this article asks — which cue is missing, and what rule replaces it? — which cue reading does not invite, because cue reading assumes the signal is there to be read.

Conversely, the vocabulary should not be stretched over matters it cannot carry. This is not a method for looking better on camera, and it is not a code of conduct.

Camera presence, for example, remains a genuine craft rather than a cue replacement. Somebody may well raise their laptop, light their face and look at the lens when saying something that matters, and all of it helps. None of it tells a speaker whether fifteen muted tiles have followed them, and no amount of it restores mutual gaze, which the geometry of a camera beside a screen makes impossible.

For a fuller decision guide, see virtual meeting cues vs camera presence, cue reading and etiquette.

8

How does this relate to Singapore’s Critical Core Skills?

SkillsFuture Singapore identifies 16 Critical Core Skills that support workers’ ability to develop and use technical competencies. They are arranged into three clusters:

Thinking Critically

Creative Thinking, Decision Making, Problem Solving, Sense Making and Transdisciplinary Thinking.

Interacting with Others

Building Inclusivity, Collaboration, Communication, Customer Orientation, Developing People and Influence.

Staying Relevant

Adaptability, Digital Fluency, Global Perspective, Learning Agility and Self Management.

Conceptual overlap — not an official equivalence

Stop Reading→ Thinking Critically

Swap In a Rule→ Interacting with Others

Check It Held→ Staying Relevant

Two things in that mapping are worth stating here rather than leaving to the crosswalk. The first is that Name the Lost is reached by no named skill at all. Sixteen skills describe thinking, communicating and keeping up; not one describes establishing which channel a medium has deleted, which is the question everything else here depends on. The second is that Digital Fluency maps to Swap In a Rule rather than to anything about the platform — the useful competence is not operating the software but noticing that a four-word naming rule replaces a turn-taking gap the network has disabled. The frameworks should not be presented as interchangeable. SkillsFuture’s Critical Core Skills are an official Singapore framework; the five-stage structure is a simplified editorial model.

This crosswalk is not a claim that a Knowles Training Institute programme is SSG-accredited, WSQ-approved or SkillsFuture-funded.

Why does the replacement matter now?

Hybrid and distributed working have made the grid the default setting for a large share of workplace conversation, and the advice offered has stayed what it was for a room: read people better. The channel has not got worse over time; what has changed is how much of an organisation’s talking now happens inside it.

World Economic Forum 2025 employer survey

Analytical thinking was the most widely cited core skillIdentified as essential by around seven in ten surveyed employers

Employers also gave prominence to:

Resilience, flexibility and agility
Leadership and social influence
Technological literacy
Creative thinking
Curiosity and lifelong learning

International employer survey — not a Singapore-specific forecast.

Singapore’s Critical Core Skills similarly combine transferable cognitive, interpersonal, adaptive and digital capabilities with technical competence.

Requiring cameras on will not by itself change what anybody understands, because more tiles supply more things to misread and nothing about the half-second delay.

9

Can this be learned?

Yes, and the gains come from the rules rather than from the reading. A working cycle contains:

1

A list of what this setup actually removes

2

A stated prohibition on reading tiles

3

One explicit rule per lost cue

4

Spoken substitutes for your own absent cues

5

A date to check whether the rules survived

6

A count of how many people spoke

Evidence from Singapore’s Institute for Adult Learning

762worker-learners

43educators and facilitators

8organisations

The Institute for Adult Learning’s Adult Learning Collaboratory examined future-oriented teaching methods across 762 worker-learners, 43 educators and facilitators, and 8 organisations, studying 86 learning interventions.

The work focused on emergent scenarios — which ask learners to respond to situations that evolve, change or remain uncertain — and generative dialogue, which asks learners to challenge assumptions and build understanding collectively rather than absorbing information individually.

The findings do not prove that every course using scenarios will produce the same result. They support contextual practice, dialogue and reflection over information delivery alone.

Which part should be fixed first?

Start with the turn-taking rule—not with the cameras.

1

Which cues has this setup actually removed?

2

What is anybody inferring from a tile?

3

Is the obstacle the channel, the purpose or somebody’s circumstances?

4

What rule replaces the gap people were waiting for?

Example: mandatory cameras, unchanged meetings

The cause may be a half-second delay, a speaker view that hides the audience, and two people misread from downward gaze—not a shortage of engagement.

The response may require a naming rule, a hand-raise and decisions in the chat, rather than a requirement that everybody be visible.

10

Common mistakes with virtual meeting cues

1. Treating the grid as a slightly worse room

A grid has deleted particular cues and invented others, which is not the same as showing less of the same thing. Mutual gaze is absent rather than faint, and a delayed nod is an artefact of bandwidth rather than a hesitant one. Reading harder produces more confident errors rather than fewer.

2. Making cameras mandatory

Requiring cameras supplies more tiles to misread, costs everybody the energy of being observed for an hour, and overrides reasons people are entitled not to explain. It also leaves the actual losses — mutual gaze, hands, turn-taking timing — exactly where they were, since none of them returns with a camera.

3. Waiting for a gap that the network cannot produce

Turn-taking runs on fractions of a second, and a few hundred milliseconds of delay disables it. The result is simultaneous starts and long pauses among people who manage a room without difficulty, and it is routinely attributed to confidence, hierarchy or culture.

4. Performing attentiveness at the lens

Exaggerated nodding and a fixed camera stare supply the appearance of a cue without its content, exhaust the performer, and are invisible to a speaker who cannot see most of the tiles. Saying “still with you” takes three words and actually arrives.

5. Reading a camera-off as withdrawal

A camera is off because of a connection, a room, a household, a condition or a preference, and the inference drawn from it falls hardest on whoever has least control over those things. It is the single least reliable and most consequential read available in this medium.

6. Introducing rules and never checking them

A protocol that replaces a lost cue lapses within about a month, and nothing in a meeting announces it. The meeting then returns to being read from tiles, and the organisation concludes that the rules did not work rather than that they stopped being used.

7. Using the speaker view by default

Many platforms default to a layout that shows a presenter almost nothing while they present. A speaker with no sight of their listeners tells things worse, which is a documented effect of the dialogue being one-way rather than a failure of preparation, and it is a setting rather than a skill.

Turning a framework into a working practice

A practical development cycle contains six steps:

1List what is missingSpecific to this setup: gaze, hands, timing, cameras, layout.
2Stop reading tilesDraw no conclusion about a person from a thumbnail.
3Name the next speakerFour words that replace a gap the network has disabled.
6Say what is not visibleThat you have finished, that you want in, that you are reading.
5Put decisions in writingTyped as they are taken, so nothing rests on having heard.
4Diary the checkA month out: is the naming rule still being used?

This cycle prevents the organisation from treating a channel problem as an engagement problem. It also creates a clearer basis for deciding whether anything here is a development need at all, or whether what is needed is a different layout and a four-word rule.

Organisations that decide facilitated development is the right response can review the Virtual Meeting Non-Verbal Cues Course in Singapore, a commercial programme option for facilitated development across the five stages.

Course format, facilitator suitability, customisation and expected workplace application should be evaluated against the organisation’s identified needs.

Frequently asked questions

?
What are virtual meeting non-verbal cues, in this sense?

Mostly cues that are not there. The method establishes which this setup has removed, prohibits drawing conclusions about people from tiles, replaces each lost cue with an explicit rule, supplies in words what the participant’s own absent cues would have conveyed, and checks a month later whether any of it survived.

?
How is it different from cue reading?

Cue reading assumes the signal is present and seeks to interpret it better. This method establishes that particular signals are absent and replaces them. The two are not complementary in the usual way: where a cue has genuinely been deleted, better reading produces more confident error, which is worse than the original uncertainty.

?
Does this mean cameras are pointless?

Not pointless, and much less informative than they appear. A camera restores a face and restores nothing about mutual gaze, hands or turn-taking timing. It is worth having on by choice for the first few minutes of a conversation among people who do not know each other, and it is not worth requiring, because the inference it invites is unreliable and the cost of being watched is real.

?
What should be measured?

Not engagement. The cues this setup removes, how many people spoke against how many attended, whether the naming rule was used, how many decisions were typed as they were taken, and whether the quietest participants report having tried and failed to come in.

?
When is this the wrong method?

Where somebody’s camera is off for reasons that are theirs, and nothing here justifies requiring it; where a colleague relies on seeing a face and the adjustment owed is the opposite of a grid; where somebody’s work rather than their tile is the evidence of disengagement; or where the meeting has no purpose. In those cases protocol is not the constraint.

Conclusion

Virtual non-verbal communication is usually treated as a reading problem and is better treated as a replacement problem.

A participant needs to know which cues this setup has deleted, which inferences to stop drawing, what rule replaces the gap people were waiting for, what to say because their own cues are invisible, and whether any of it is still in use next month. Watching the tiles more carefully answers none of those, which is why a team can make cameras mandatory and hold exactly the same meeting.

For employers, the first step is not to require cameras. It is to sit through one recurring call, list what the setup removes, and count how many of the people present actually spoke.