Conversational Intelligence Levels: Choosing a Meeting Mode

Conversational Intelligence levels distinguish exchanging information, advocating a position and exploring possibilities together. Judith E. Glaser calls them transactional, positional and transformational. Use the distinction to choose what an agenda item needs, rather than expect every contribution to work in the same way.

We build Coachful. Coachful is our #1 recommendation for organizing selected coaching reflections and agreed actions when its access and sharing controls fit the engagement. This is our maker's preference, not independent evidence that a platform improves meeting quality, trust or organizational performance.

The original fictional example follows chair Amal, coach Nia and colleagues Theo, Ruth, Esme, Ivo and Mara. It includes a design discussion, three trial meetings and a review with mixed results. These are illustrative people and figures, not customer outcomes. Dates are in 2026 and times use Europe/London.

Attribute the framework without turning it into a diagnosis

In her MindTools interview, Glaser describes telling and asking, advocating and inquiring, and sharing and discovering. She also says all three levels have a place. Her 2013 CreatingWE article, Build Trust Through Conversational Intelligence, discusses openness to influence. These are sources for the named framework, not independent trials establishing its effects.

The labels describe conversational patterns, not levels of human worth or a ranking of staff. A factual exchange can be useful and respectful. Strong advocacy can bring expertise into a decision. Joint exploration requires something genuinely open to influence; asking an apparently open question does not make a predetermined decision collaborative.

This guide uses observable communication and explicit choices. It does not infer a participant's hormones, brain activation or psychological state from a phrase. A framework label is not a physiological reading, a validated trust score or evidence that a person needs to be moved into a superior level.

Match three modes to the task

These task checks are original applications. A safety instruction, settled legal requirement or simple date update does not need to become an extended exploration. Conversely, repeating an instruction will not necessarily resolve an uncertain design problem. Name both the relevant expertise and the actual decision authority.

For another conversational move, see the reframing guide. A mode change and a reframing are different choices. You can clarify the interaction's purpose without replacing the substance of a person's concern.

Signal information, advocacy, exploration and closure

• Information: state what is known, its source, what is uncertain and what the recipient needs to do.
• Advocacy: identify your recommendation and reasoning, then say which concern you want tested.
• Exploration: identify what remains open and ask about a genuine uncertainty.
• Closure: confirm the decision, owner, next action and review conditions.

“Let us establish the facts first” can signal a useful transition. “I recommend this approach” can make advocacy more honest than disguising it as a question. “What would make this unworkable?” is useful only if the answer can influence the proposal or identify an issue needing a response.

Clarify authority before inviting input. Amal can decide the weekly agenda format within her existing role. She cannot use a coaching exercise to change contracts, staffing authority or safety obligations. Her six-person team can help design the format while she retains responsibility for the decision.

Work through a nine-turn meeting-design conversation

Five earlier weekly meetings lasted 58, 65, 68, 72 and 70 minutes. Four exceeded the scheduled hour. Amal initially assumes people are talking too much. During a 45-minute coaching session on 2 November, Nia asks what the meeting is supposed to accomplish and what the records actually show. Those timings alone do not identify the cause of the overruns.

Amal invites the six team members, including herself, to a twenty-minute design discussion on 5 November. She explains that the hour remains the normal weekly limit and that the open questions are which updates need live discussion and how preparation can fit working time. The following nine turns sit inside that discussion, not an additional meeting:

Amal: “Four of the last five meetings exceeded an hour. We need a trial that retains useful decisions and questions.”

Theo: “I recommend moving routine updates into a written note. They take much of the meeting.”

Ruth: “Some updates need questions. A note alone could move work into more messages.”

Amal: “Which updates need live discussion?”

Ruth: “A decision, a risk or a dependency that needs input.”

Theo: “Then the note could identify those, and the agenda could use them.”

Amal: “What would make that practical?”

Ruth: “A simple template and time to read it during work.”

Amal: “We will trial that for three meetings. I own the agenda. We will check decisions, preparation and what we miss.”

The discussion contains facts, a recommendation, a challenge, joint design and a decision. It does not show that everyone agrees enthusiastically. Ruth retains a concern about preparation burden. Ivo contributes a written suggestion for a clear question route before the discussion closes; quiet speech is not treated as assent or disengagement.

Theo accepts responsibility for the template on 5 November and delivers it on 6 November. Amal checks it and sends the agreed version to the team that day. The template separates a routine fact, a live decision or risk, its owner and the response needed. Reading it is work, not a request to prepare in unpaid evening time.

Plan the conversation in seven stages

1. Name the result: a usable weekly format, not proof that a communication theory works.
2. Establish facts and uncertainty: four overruns are known; their complete causes are not.
3. Hear recommendations: Theo's written-update proposal and alternatives need reasons and challenge.
4. Specify what is open: the format and preparation route, within Amal's actual remit.
5. Compare criteria: necessary decisions, inclusion, preparation effort, time and important questions.
6. Confirm the decision: a three-meeting trial, named responsibilities and a route for exceptions.
7. Review usefulness: examine what happened and whether to revise, stop or continue.

A related three-level matrix can help compare contributions. Keep labels light enough that they do not distract from the task. Participants may prefer update, recommendation, exploration and decision. They do not need to learn the framework's vocabulary to contribute.

Fill a seven-field meeting worksheet

Separate a proposal from an accepted responsibility. “Theo could draft it” becomes a delivery commitment only after he accepts the task and date. An open design question can coexist with a fixed requirement. State which part can change so that input has a meaningful destination.

Test three meetings and retain the failed cue

All six colleagues attend each trial. On 9 November the meeting lasts fifty-five minutes. Routine facts stay in the note while an unresolved dependency gets live discussion. Amal closes by confirming two decisions and their owners. That first meeting supports continuing the trial; it does not establish a reliable time saving.

On 16 November Ruth raises a risk that the new note's short format has obscured. Near the planned end, Amal says, “We are getting stuck in Level II.” Ruth objects that the risk still needs an answer. Amal withdraws the label, summarizes the substantive concern and uses the agreed work route for a decision. The meeting ends at sixty-five minutes. The five-minute overrun remains in the record.

A person can advocate strongly because the evidence matters. Ruth's concern is not disproved by calling it positional. Nor does asking a collaborative-sounding question discharge Amal's responsibility to address the risk. Serious conduct, retaliation or safety concerns need the relevant organizational process, not a requirement to stay in a preferred conversational mode.

Theo revises the template to include an explicit risk and dependency field on 18 November, with Amal's approval. Esme checks the revised wording on 19 November and says it makes her uncertainty easier to identify. That is one person's feedback on a field, not a validated measure of team trust.

The 23 November meeting lasts fifty minutes. Amal moves one unsettled item to a named follow-up rather than declare it solved. Theo accepts responsibility for obtaining the relevant work information by 26 November. He returns information that day, but a separate resource decision remains with its authorized owner. There is no claimed full project resolution.

The three meetings total 170 minutes against 180 scheduled minutes. One of three overruns contrasts with four of five in the earlier record, but the periods and agenda demands differ. This small comparison does not establish that the framework caused improvement or that the team's decisions were better.

Preparation figures are the six participants' combined active minutes for each trial, including Amal's agenda work and everyone's reading. They total 205. The team also holds its planned twenty-minute review on 30 November with all six present. Across the design discussion, trials and review, measured participant live time is 6 × (20 + 170 + 20) = 1,260 minutes. Adding the defined trial preparation gives 1,465 measured participant minutes.

That is a bounded account, not whole labor or a saving. It excludes design preparation, correspondence, other administration and coaching; comparable earlier preparation is unknown. Do not compare 1,465 person-minutes with 180 calendar minutes. A shorter meeting can move substantial work elsewhere, and a useful exception can make a meeting longer.

Respond to repeated positions and keep advocacy available

When a thread repeats, summarize the strongest version of each concern. Ask what evidence, criterion or authority decision would help. Repetition may mean the concerns remain unheard, the options genuinely conflict or someone has no authority to accept the proposed change. Do not assume a lower conversational level explains it.

For this team, meeting time and preserving questions are both legitimate concerns. “Which updates need live discussion?” emerges from that tension. If the options still conflict, use the agreed decision process and preserve the reservation. Joint exploration can inform Amal's choice without producing consensus.

Keep a specialist's clear recommendation available. They can state what they advise and why, then invite the relevant challenge. Requiring everyone to speak only in tentative questions can hide important expertise or make accountability harder to locate.

Use writing and participation routes deliberately

A written update needs a purpose too. Label information, a recommendation, an open question and a decision separately. For a proposal, include its reason, alternatives considered and the response needed. For joint design, give enough context and working time for meaningful input rather than ask for immediate approval.

If a written thread repeats positions, consider a short live discussion with a defined question. If someone needs time to think or an accessible format, written input may serve them better than another meeting. The mode serves the work across channels; it does not belong to one app or medium.

The HSE Control standard includes employee influence over pace, breaks and work patterns where possible. That employment context supports examining whether preparation time is actually available. It does not establish that an individual can solve every workload or power problem through a better question.

Acas consultation guidance, updated in August 2026, distinguishes good-practice consultation from processes required by law. It explains that consultation informs an employer's decision, while collective bargaining is different. This meeting lens does not replace required employee or representative processes, and co-creation is not a blanket transfer of decision authority.

Offer practical routes: advance material, a pause, a structured turn or written input. Ask how an invitation is received when the leader has signaled a preferred answer. Do not count a quiet participant as supporting the proposal, diagnose reluctance or treat criticism as a failure to collaborate.

Use five questions to examine the result

1. What became clearer? Identify a specific responsibility, assumption or unresolved question. A general report that the room felt better may matter to the speaker, but it does not identify what changed in the work.
2. Which information changed the discussion? Ruth's obscured risk affected the template. Record the concern and the response, rather than award the group a higher level because it eventually agreed to listen.
3. What was actually decided? Distinguish the format trial, a work decision made during it and a later request still awaiting authority. Do not mark all three complete from one meeting's closure.
4. What did participation require? Include time, readable information and usable contribution routes. Ask people whether those routes worked for them rather than infer accessibility from the fact that a document was sent.
5. What remains contested? Preserve the reservation and its review condition. A person may accept a bounded trial while continuing to question whether its preparation cost is justified.

Choose the evidence appropriate to each question. A duration can answer how long the meeting ran. A readback can expose differing understandings. A participant's account can describe their experience, while remaining that person's account. An accepted owner field establishes the responsibility they agreed to take; it does not verify that the later work succeeded.

Make a small review sufficient for the decision at hand. This team needs to decide whether to retain or revise a meeting format. It does not need to collect private personal narratives or turn every contribution into a score. If the evidence is incomplete, retain the uncertainty and decide what is worth checking next.

Review the result and correct the final account

At the team review on 30 November, Ruth says the explicit risk field helped but preparation remains burdensome. Ivo wants written questions retained. Amal keeps a mixed format: routine figures written, decisions and risks on the agenda, urgent matters through their appropriate route. She does not announce a permanent organization-wide policy.

In a separate thirty-minute coaching review that day, Nia calls the trial evidence of greater trust. Amal objects that they did not measure trust and that Ruth still has a workload concern. Nia withdraws the claim. Delivered coaching totals seventy-five live minutes with the earlier forty-five-minute session; that coach-client time is outside the team's measured participant account.

Amal approves the exact final team note on 30 November at 15:00 and sends it to the other five colleagues at 16:00. It records the three durations, the overrun, the risk-field correction, Ruth's reservation and the continuing resource decision. By the requested receipt point of 1 December at 17:00, Theo, Ruth and Esme acknowledge. Ivo acknowledges at 10:00 on 2 December, seventeen hours after that point. Mara's receipt remains unconfirmed.

That is three timely acknowledgments, one later acknowledgment and one unknown among five recipients. Neither sending to five nor chair approval proves all five read or accepted the note. The 2 December record retains those distinctions. A further review on 7 December is planned; future adoption, whole-work savings and downstream business results are unknown.

Keep coaching records separate from company discussion

The company's meeting notes and work approvals remain in its authorized work system. For coaching, Nia grants an original preparation worksheet and sends Amal its private client access link on 30 October. Amal opens it on 1 November and chooses only the prompts relevant to her chairing question. No Program enrollment or universal My Coaching task placement is presumed from that standalone grant.

Coachful's community views can support appropriately scoped exchanges, but do not copy confidential company minutes into a coaching community merely because it has a discussion feed. Inspect the actual audience and every eligible access route. A manually selected membership list does not automatically exclude people who qualify through other configured grants or moderator access.

Unshared client-owned journal or takeaway pages remain separate; deliberate use of an enabled Share with coach control exposes the selected page. Amal declines to share her private reflection with colleagues. The final team note uses only the exact content she approved. Internal coach review notes are team-private and excluded from the client's response PDF; saving them does not deliver a message.

Session recap delivery depends on the workspace automatic or approval setting and any session-type override. Review recipients and actual settings instead of assuming every recap waits for permission. The current ICF Code addresses agreements, information exchange and competence for ICF professionals. It does not grant colleagues or sponsors access to private coaching records.

Explore Coachful's coaching features for agreed preparation and action review. A customizable reflection is not an official Conversational Intelligence assessment, automatic trust measurement or proof that the team has changed its culture.

Frequently asked questions

Are the levels a ranking of people?

No. They describe patterns within Glaser's framework. A person can use several within one meeting. Do not use a label to dismiss a colleague's expertise, criticism or right to raise a concern.

Is Level III always best?

No. Choose the interaction for the task. Clear information, reasoned advocacy and joint exploration can all matter. A fixed requirement should be explained honestly rather than presented as open co-creation.

Can I identify a mode from a phrase alone?

Not reliably. Purpose, context, openness and the actual response matter. Ask what the speaker is willing to reconsider and what happens when someone disagrees. A question can still be an attempt to secure compliance.

Does co-creation require consensus?

No. State how input will be used and who decides. Record unresolved reservations and use the proper authority or required process. A useful discussion can lead to a contested but clearly owned trial.

Can the framework measure trust or hormones?

No. A label is not a physiological test or validated trust score. Use explicit observations, defined measures and participant feedback with their limits. Amal's durations and acknowledgments do not establish a hormonal change or greater trust.