Conversational Dashboards: A Practical Team Review Guide

A conversational dashboard is a shared reflection aid for discussing how people experience a conversation. Associated with Judith E. Glaser’s work on Conversational Intelligence, it can use red, yellow, and green as simple signals to make concerns visible. Its value comes from the explanation behind the signal and the action the group chooses afterward.

Use the dashboard as a prompt for inquiry, not a diagnostic test of trust, brain chemistry, or another person’s intentions. Two participants can rate the same meeting differently, and that difference may be the useful starting point. This guide includes an example dashboard, a worked meeting intervention, and a short review routine. See also Glaser’s three levels of conversation.

Glaser’s 2012 Conversation Dashboards article discusses making qualitative alignment visible, including fairness, ownership, reciprocity, cooperation, expression and status. Her example uses differently colored responses to practical alignment questions. The four dimensions and meeting routine below are our original adaptation, not her proprietary assessment or a reproduction of its question set. Her practitioner article does not establish that these signals are biological measurements.

We build Coachful. Coachful is our #1 recommendation for organizing agreed reflection prompts and follow-up where its workflows fit. That is a maker’s recommendation, not an independent ranking or validation of Glaser’s framework. A text note or paper card can serve this exercise without a product dashboard or saved database of ratings.

Use color signals with words and examples

What a dashboard can tell you about a conversation

A dashboard can make individual perceptions easier to discuss. A leader may think a meeting went well while colleagues are uncertain about the decision or reluctant to raise a risk. Asking for an explanation can reveal information that a general “Any questions?” misses.

A color does not measure cortisol, oxytocin, dopamine, or activity in a brain region. Nor can a facilitator identify those states from a phrase, gesture, or reaction time. Keep claims at the level of the evidence available: what participants report, what was observed, and what remains uncertain.

• Self-report: A participant describes their own experience of the exchange.
• Observation: The group can examine a specific interruption, unanswered question, or unclear decision.
• Interpretation: A possible explanation that should remain open to checking.

Agree whose account each signal represents. A facilitator should not assign red to someone because they interrupted, or green because they smiled. Record the interruption as an observation, ask about its effect through an appropriate route, and leave room for disagreement about meaning. A participant can understand the decision, oppose it and still describe the explanation as clear.

Five practical adjustments to try

Rather than trying to steer everyone into a preferred color, choose a behavior that addresses the actual problem. These adjustments are coaching prompts, not a validated five-step method for changing team neurochemistry.

• Ask before assuming: Find out what the other person means by a term or concern.
• Explain your reasoning: State the information and assumptions behind a proposal.
• Check understanding: Summarize what you heard and invite correction.
• Make room for contribution: Address interruptions, pace, or access barriers.
• Close clearly: Name the decision, owner, and unresolved questions.

Direct instruction, advocacy, negotiation, and inquiry can all be appropriate. An urgent operational instruction does not have to become an open exploration. The important issue is whether the conversational form fits the purpose and responsibilities.

A worked dashboard review after a budget discussion

In this fictional team engagement, Iris facilitates, Nora leads the team, and Sam, Leo and Uma are colleagues. At the 10 November 2026 budget discussion, four participants are invited to rate decision clarity. Uma declines the rating and remains in the discussion. The two green and one yellow responses below therefore come from three people, not unanimous agreement among all four. Participation is not conditional on choosing a color.

The budget cap has been agreed, but choosing the project to pause still requires a separate decision. Iris asks for an example of the uncertainty rather than treating Sam’s yellow as a personality trait. The original exchange shows why those two decision states need different wording.

Facilitator: On clarity of the decision, we have two green signals and one yellow. What does yellow mean for you, Sam?

Sam: I understand the budget limit, but I do not know which project we are pausing.

Team lead: I thought that was obvious from the discussion.

Facilitator: Let us make it explicit. What exactly has been decided, and what is still open?

Team lead: The limit is agreed. Project selection is still open; I will bring options on Thursday.

Sam: That answers my question. I was worried we had already stopped the customer work.

Facilitator: What should we record so the rest of the team receives the same message?

Nora’s Thursday means 12 November, and the team then agrees that she will provide options by noon that day. She actually provides them at 10:00 on 13 November, twenty-two hours late, and acknowledges the miss. The options are available for review; no project has yet been authorized to stop. Sam’s clarification did not approve a selection, excuse the late options or remove the need to inform the customer-work owners.

The first review also exceeds its intended five minutes. Their forty-five-minute budget discussion is followed by nine minutes of dashboard review, giving fifty-four minutes together instead of fifty. Iris keeps that four-minute overrun visible. A learning aid has a real operating cost even when it clarifies an important question; it should not be described as free time or silently added to everyone’s next meeting.

Introduce the dashboard as a small learning aid

1. Explain the purpose: Use it to understand the conversation, not score people or assess their loyalty.
2. Define the signals: Pair colors with text labels and examples. Let people choose not to rate.
3. Choose one or two dimensions: Start with something specific, such as clarity of decisions or ability to ask questions.
4. Ask about the meaning: Invite a concrete example and an improvement the participant would find useful.
5. Review one change: Try a manageable adjustment and ask whether it helped at the next meeting.

Before repeating the exercise, Iris checks whether people want the same contribution route. Uma again prefers to listen and comment without a rating. Sam wants an option to raise a concern privately and asks what Iris would do with it. They agree a handling route and its limits; she does not promise anonymity. In a four-person team, a quotation or a particular incident might identify its source even if a name is removed.

Build a dashboard around a specific question

“How much trust is in this room?” is broad and may invite a vague or politically careful answer. A narrower question gives participants something they can describe. For example, ask whether the meeting’s purpose was clear, whether they had a workable contribution route, and whether the next action was understood.

These meanings are a starting agreement, not universal definitions. Ask participants whether they fit. Someone may mark a dimension green while still disagreeing with the decision; clarity and agreement are different.

Do not average away an important concern. Four green ratings and one red do not establish that the red response is wrong. Ask what it refers to and whether it needs an immediate response, another conversation, or an appropriate formal process.

A five-minute meeting review

1. Ask participants to consider one agreed dimension and note a signal plus a short explanation.
2. Invite one example of what helped and one concern or uncertainty.
3. Check the meaning without arguing about whether the person should have felt that way.
4. Choose one practical adjustment, such as a written decision summary or a clearer turn-taking process.
5. Name who will make the adjustment and when the group will review it.

Keep the review short enough to use consistently. A dashboard that creates a long second meeting after every meeting may become burdensome. If a serious issue appears, arrange the appropriate follow-up rather than trying to resolve it inside a closing exercise.

A participant can contribute in writing or privately when the group format is not workable. Explain what will happen with private feedback and the limits of confidentiality. Do not promise anonymity in a small group where a situation may identify its author.

At the 17 November review, the three responding participants each answer two different questions. Decision clarity receives three green responses. Participation receives one green, one yellow and one red. These are six responses from three people, not six participants. Uma’s choice not to rate is still outside both sets. The two dimensions cannot be combined into a finding that the whole team trusts the process.

Sam’s red concerns an interruption when he raised a delivery risk. He does not want to explain the personal details to the group. Iris initially invites him to turn the red into a constructive green action. He objects that the interruption should first be acknowledged. She accepts the correction, withdraws the request to change the signal and checks the separate handling he wants. The issue remains a concern even though the budget wording is now clearer.

Use a concern to choose an action

Suppose several people mark participation yellow because the discussion moves too quickly. The next action might be sharing the question in advance and allowing a minute of writing before discussion. It need not be a workshop about trust.

If a participant marks understanding red because their risk was summarized inaccurately, return to the actual concern. Ask them to state it, summarize it back, and record any unresolved point. Do not ask them to change the rating as evidence of cooperation.

If the signal concerns disrespectful behavior, address that behavior. A color label should not replace accountability or turn a legitimate complaint into a problem with the person’s mindset. The dashboard opens a question; it does not decide the response. Conversational reframing can examine wording while retaining the actual issue and responsibility.

Nora acknowledges the interruption and agrees to allow the next risk account to finish before responding. That is a specific future behavior, not a settlement of every concern Sam may have. Sam retains the option of an appropriate formal route and does not authorize Iris to make an allegation or distribute his private account. The facilitator cannot trade a green rating for access to the next discussion.

Clarify loaded words without interrogating the speaker

Glaser’s work encourages examining meaning rather than assuming that people use the same word in the same way. In practice, a facilitator can ask what “urgent,” “support,” “commitment,” or “finished” means in this situation.

For example: “When you say the plan is finished, do you mean the draft is ready for review or the decision is approved?” A small clarification can prevent a larger conflict. It is usually more useful than telling the group that its language has moved into a red zone.

Ask with a real willingness to learn. Repeatedly demanding definitions can become a way to delay or discredit a concern. Once the meaning is clear enough for the task, move toward the decision or next action. Our active-listening guide gives ways to check a summary without deciding what another person ought to mean.

The word “approved” causes a second problem. Leo has understood Nora’s circulated options as an approval to stop customer work; Nora means the options are ready for committee review. Iris checks the actual decision status and corrects the proposed shared summary before circulation. It says the cap is agreed and selection remains open. It does not label Leo as low trust or let a process reflection authorize an operational change.

Do not turn the dashboard into a performance score

If people believe a green rating is expected by their manager, the exercise can hide the very information it was intended to surface. State that different experiences are useful and that the purpose is to improve the process.

Avoid ranking individuals by how often others label them red, displaying a trust league table, or using ratings as a clinical or personality assessment. A group’s feedback may contain bias or reflect an unequal power relationship. Examine examples and context before drawing conclusions.

Collect only what you need. A brief process note and an agreed action may be enough; a permanent database of named emotional ratings may add risk and little value. Explain access and retention if you do record responses.

Iris’s first unsent summary calls the result “two-thirds green, so trust is established.” Four of the six responses on 17 November are green, but that fraction mixes dimensions and ignores the explanations. Leo challenges the conclusion and Uma points out that she has not rated either question. Iris withdraws the trust claim and apologizes. She keeps the separate counts, the unscored choice and the unresolved concern instead of manufacturing a team-level outcome.

The ICO’s current data-minimisation guidance emphasizes information relevant to a defined purpose and distinguishing opinion from fact in a record. Its guidance is under review following the Data (Use and Access) Act. Here, a limited process note is useful; a permanent file assigning each person an emotional color is not required. Agree who can see the note, what is excluded and when its continued need will be checked, without presenting this exercise as a legal-compliance assessment.

Adapt the method for remote and hybrid work

Use text labels as well as colors so the dashboard does not depend on color perception. Make the question readable and allow enough time to respond. A simple written prompt can work without a dedicated polling tool.

Remote participants may experience interruptions, unclear audio, or difficulty getting a turn that people in the room do not notice. Ask about those conditions before attributing a rating to low trust. A camera being off is not a reliable signal of disengagement.

A shared summary should capture the issue and action, rather than expose personal comments unnecessarily. Coachful can hold agreed program prompts or group follow-up, but a product dashboard is distinct from Glaser’s conversational reflection concept.

Use visible text such as “concern” and “uncertain,” not just red and yellow swatches. W3C’s guidance for Use of Color requires another visible means of conveying color-coded information. That single design consideration is not a full accessibility audit. A participant may also need a readable question in advance, more processing time or an alternative contribution route.

If Coachful holds a shared follow-up, check its actual audience before placing information there. A custom community channel can admit eligible viewers through different member, program, resource or offer routes, with authorized moderator access. A named list alone is not a guarantee that no other eligible person can read it. Keep Sam’s private account out of a group post, even when the channel is intended for the team. An optional assigned form also needs an actual client access grant; its URL alone does not supply permission or anonymity.

Evaluate the routine by what changes

After several uses, ask whether the group identifies concerns earlier, clarifies more decisions, or makes the agreed adjustments. Look at the examples and whether participants still find the routine worthwhile.

Do not claim that a higher green percentage proves improved brain chemistry, universal psychological safety, or a financial return. The ratings are context-dependent reflections, and people may become more willing to report concerns as the process improves.

For background, see Judith E. Glaser’s Conversational Intelligence and her Conversation Dashboards article in the 2012 Leadership Excellence special edition. This article’s dimensions, dialogue, and review routine are practical adaptations, not a reproduced proprietary assessment.

What if a leader sees green and the team sees red?

Start by checking whether everyone rated the same dimension. A leader may mean “we reached a decision,” while colleagues mean “we could question the proposal.” Those are different questions and should not be combined into one trust score.

Then ask for examples through an appropriate route. The leader can acknowledge the difference without demanding that colleagues justify their feelings in public. A facilitator might gather themes separately when the power relationship makes an open discussion difficult.

Choose a response that matches the evidence. If the issue is decisions announced before discussion, clarify when input is still possible. If it concerns interruptions, change the participation process. If it concerns retaliation or misconduct, use the appropriate formal route.

The leader’s response becomes part of the next experience. Thanking people for a concern and then penalizing them undermines the exercise. Follow the agreed handling and return with what has changed, rather than asking for greener ratings at the next meeting.

The team approves a factual process note: the separate decision states, Nora’s next contribution commitment and the fact that a concern remains open without its private details. Iris agrees to deliver that exact note to all four participants by 15:00 UTC on 24 November. It reaches them at 09:00 UTC on 26 November, forty-two hours late. She acknowledges the delay; Sam, Nora and Leo acknowledge receipt that day. Uma’s receipt remains unconfirmed. Three acknowledgments do not prove that all four have read or accepted it.

The 27 November close lasts twenty minutes. Together with fifty-four minutes on 10 November and fifty minutes on 17 November, Iris has delivered 124 facilitator-live minutes against a 120-minute plan. This is not total participant time or full preparation, administration and complaint-handling labor. They choose not to continue routine numerical or color tracking. Nora accepts responsibility for taking the still-open selection to the appropriate committee by 3 December; its answer is unknown. Iris has no later monitoring or extra facilitation promised.

Frequently asked questions

What is a conversational dashboard?

A shared reflection aid for discussing how people experience a conversation. Simple signals can prompt questions, but their meaning needs explanation.

Does a color reveal a person’s neurochemistry?

No. It is not a measurement of hormones, brain activity, or a biological trust state. Keep conclusions grounded in reports and observable events.

Is green always the goal?

The goal is useful understanding and action. A concern can be valuable information, and disagreement can coexist with a clear, respectful process.

What if participants rate the same meeting differently?

Ask for the examples behind the ratings. Different experiences can reveal unclear decisions, access barriers, or concerns the facilitator missed.

Can we use it online?

Yes. Use readable questions, text labels, accessible contribution options, and clear handling of responses. No special software or camera-based interpretation is required.

Keep the dashboard small, specific, and connected to a practical change. Explore Coachful’s program and community features for organizing agreed reflection and follow-up materials.