Conversational Blind Spots: Glaser's Coaching Lens

Conversational blind spots are gaps between what you think happened in an exchange and what you have actually checked. You may remember giving a clear direction, treat a nod as agreement or interpret a question as resistance. Coaching becomes useful when it makes that account open to correction.

Judith E. Glaser developed this theme within Conversational Intelligence. Her work offers a vocabulary for examining communication. It does not give a coach access to an absent colleague's thoughts or establish someone's hormones from their words.

This guide explains the source distinction, then completes an original meaning-check case. A director discovers that her instruction left the tradeoff unspecified, receives a genuine capacity objection and corrects a faulty summary. The final output is an accepted planning note, with a remaining relationship concern and an honest limit on what has actually been delivered.

Read Glaser's blind spots as a perspective, not a diagnosis

In her 2015 CreatingWE article, Conversational Blind Spots, Glaser describes five themes: assuming shared perceptions; missing emotional filters; difficulty taking another perspective when upset; treating interpretation as accurate recall; and assuming one's intended meaning has reached the listener. Her earlier 2013 Entrepreneur essay illustrates three problems rather than presenting this same five-part list. The accounts are practitioner explanations, not a diagnostic classification established by this guide.

The 2015 article's 85/15 telling-to-question figure comes from her reported two-week role-play work with 17 sales executives. It is not a universal leader average or a target ratio. Its business-success story does not establish a controlled causal effect. We use neither its neurological explanations nor a count of questions as a biological assessment. The practice and case below are original applications, not reproductions of a proprietary assessment or training handout.

A blind-spot hypothesis should be falsifiable in ordinary terms: “I assumed Bram agreed; I have not asked what he understood.” Calling Bram defensive, low-trust or addicted to being right instead replaces one untested story with another.

Separate understanding, agreement, authority and feasibility

These questions are an author-designed distinction aid. They are not four stages that every conversation must pass through in order. Sometimes the person understands perfectly and declines. Sometimes the leader has communicated an impossible combination of priorities. More explanation does not solve either problem.

Start with observable evidence and the person's own account. “She interrupted twice after I mentioned the deadline” describes something discussable. “Her fear shut down her thinking” asserts a mechanism you have not measured. Even a plausible interpretation should stay separate from what was said or agreed.

Choose one episode and a modest purpose

In this fictional case, Sian directs an internal reporting team and works with coach Ada. Their existing agreement permits Sian to share selected, nonconfidential summaries, not company source files or colleagues' private information. Bram, a reporting specialist, has authority over preparation details; Sian can change an internal export but cannot unilaterally cancel customer reporting obligations.

The bounded case runs from 1 to 10 April 2041. Before that period, Sian had told Bram to streamline updates while keeping commitments. He acknowledged receiving the message. She later recalled this as an agreement to stop duplicate reporting. That earlier exchange is a pre-existing fact, not newly performed work in the account below.

Ada spends 15 minutes preparing for their first session. In 40 minutes together, they use five minutes to agree purpose and limits, fifteen to examine Sian's account, fifteen to rehearse a meaning check and five to settle her next request. Sian initially calls Bram obstructive in this private session. Ada does not repeat that label to him or treat it as evidence of his intention.

Ada: What did you explicitly say should stop?

Sian: I said streamline the updates and keep our commitments. I did not name a report.

Ada: What was his actual response?

Sian: He said he had received it. I turned that into agreement about stopping duplication.

Ada: What could you check without asking him to defend a motive you have assigned?

Sian: His understanding, the obligations he was protecting and which change is within my authority.

This is not an exercise in making Sian feel wrong. It makes the next question accurate. Bram may have made a mistake, but they do not know that yet. Sian chooses a short written exchange before a meeting, with no recording, personality score or covert monitoring.

Send a question that allows a different answer

On 2 April, Sian spends 20 minutes checking her original note and preparing and sending this request through their company route. She asks for an answer by 4 April at noon and offers either a written explanation or a later agreed conversation.

I have been describing our exchange as agreement to stop duplicate reporting, but my note does not say what should stop. What did streamline mean to you, and which commitments were you protecting? Please identify any change you could support and anything that needs a decision from me. I am checking my instruction, not asking you to confirm my interpretation. I am also considering a twenty-minute internal planning page; please tell me what would have to change to make room for it.

On 3 April, Bram uses 20 minutes to read the request, check the relevant obligations and send a substantive answer. He understood streamlining as using one evidence pack for the existing reports. He had not agreed to cancel a customer report. The three retained reports require 30, 25 and 40 minutes of preparation. An internal spreadsheet export takes another 35.

He refuses Sian's suggestion of adding a 20-minute planning page without removing something. He can discuss removing the internal export, but not skipping a customer deadline. He offers a fifteen-minute call on 7 April from 14:00 to 14:15 and asks her to read his written explanation first.

On 4 April, Sian spends ten minutes reading and replying. She withdraws the claim that they had agreed to stop reports, accepts the offered call and confirms that the internal export is hers to decide. Bram spends five minutes reading that acceptance. No additional acknowledgment is requested or counted.

The response has changed the problem. Sian no longer needs to explain streamlining more loudly. She needs to decide whether a specific internal activity can stop and accept that customer obligations remain.

Make the tradeoff explicit in the actual meeting

Sian spends fifteen minutes preparing from the records on 7 April. Both then attend the agreed fifteen-minute call: six minutes comparing their interpretations, six examining the capacity options and three recording the choice. Bram's attendance is separate from her preparation.

Sian apologizes for treating acknowledgment as agreement. Bram says the repeated pressure to reduce reporting made him feel that the obligations he was protecting were being dismissed. He does not refer to the private obstructive label, which he has not heard. He accepts the apology as a correction but says his confidence in how future concerns will be handled is still limited.

The original modeled preparation is 130 minutes: 30 + 25 + 40 + 35. Adding the page without removal would make 150. Removing E1 and adding the planned page makes 115, leaving fifteen within the same 130-minute task allowance.

This is a proposed recurring workload, not time already spent delivering all those reports. The case separately counts the actual prototype and review. No one claims that a better conversation has created an extra person, changed a customer obligation or completed the retained 95 minutes of reporting.

Bram also permits Sian to tell Ada the limited process result: different meanings, the accepted prototype, the export decision and his continuing concern. He does not permit the underlying reports or a claim that trust is restored. That permission is part of the three-minute closing discussion.

Test the summary and preserve the first failure

During the call's closing three minutes, Sian and Bram also agree three checks: the record must name the preparation owner, preserve the customer obligations and identify exactly which internal export stops. This is a local clarity test by one reader, not a validated instrument or three independent evaluators.

On 8 April, Bram spends twenty minutes drafting and sending the planning page. It names him as owner and preserves the customer deadlines. However, a sentence says all weekly exports will stop. That is broader than the authorized removal of internal spreadsheet export E1.

Sian uses fifteen minutes to read and test the page, record the failed condition and send a precise correction request. Two of the three stated checks are met; the export check is not. She does not mark it accepted because the overall intention seems right.

On 9 April, Bram spends ten minutes reading the correction, repairing the page and sending version 2. The complete corrected planning note is:

Bram will prepare the internal planning page from the existing reporting evidence. Customer reports A, B and C keep their existing obligations and deadlines. Only internal spreadsheet export E1 is removed from the proposed workflow under Sian's authority. The prototype required thirty active minutes including its repair, rather than the planned twenty. Before recurring use is agreed, Sian and Bram need to review that overrun and the resulting workload. This note records the prototype decision and current limits; it does not claim delivery of the customer reports.

Sian spends ten minutes reading the repaired note, checking all three conditions, sending acceptance to Bram and sending the permitted selected process summary to Ada. Version 1 remains identifiable as the rejected earlier record. The same reader has now judged three of three conditions met. That is one repaired output, not proof of lasting understanding across the team.

The prototype took thirty minutes in total: twenty initially plus ten reading and repair work. If those thirty minutes were repeated alongside the retained 95 minutes, the projection would be 125, leaving five of the task allowance. The original fifteen-minute modeled margin has narrowed. Recurrence remains unapproved.

Return substantive coaching feedback and review the mixed result

Ada uses ten minutes on 9 April to read Sian's selected summary and send this response. Sian uses five minutes to read it; no immediate reply is required.

You corrected your claim about agreement and accepted a real limit on the extra page. The repaired record now separates the customer reports from E1. Please keep the prototype's ten-minute overrun and Bram's remaining concern visible. At our review, I suggest examining what you will do when another colleague challenges an instruction, rather than treating this corrected note as evidence that the relationship is repaired.

Ada spends ten minutes preparing for their twenty-minute review on 10 April. Both attend. Sian says the written answer was more useful than another explanation of her intended meaning. She also notices wanting to call the exchange resolved as soon as the note passed.

She chooses to retain the corrected account and ask about the practical constraint before proposing an additional task. She declines a daily communication diary and does not agree to circulate a blind-spot checklist to the team. No later conversation has yet tested her intended practice.

Ada then spends ten minutes recording and sending a final close: the assumed agreement was withdrawn; the prototype note was repaired and accepted; the recurring change, customer-report delivery and wider relationship result remain unverified. Sian spends five minutes reading the close, with no new acknowledgment counted.

For the wider principle, Acas's challenging-conversations guidance emphasizes preparation, relevant facts and organizational procedures. Its discussion of questions distinguishes clarification from leading someone toward your conclusion. This is practical management guidance, not evidence that this fictional intervention causes trust or replaces a formal workplace process.

Use a seven-field meaning-check record

The following completed record is an author-designed learning illustration of the case. It summarizes the already described work; it is not another entry task or hidden meeting.

A useful record should let another authorized reader distinguish the person's words from your interpretation. Do not fill missing fields with a theory of their personality. The next accurate answer may be that something was not asked or remains disputed.

W3C's writing-accessibility tips favor clear instructions and concise language. Apply that to the task, format and reply route, while checking the actual person's needs. A long verbal explanation is not automatically more accessible than a short written version.

Choose evidence for the next conversation carefully

If the client wants another practice attempt, agree a narrow observation before collecting anything. One option is to note whether an important instruction named the task that would stop. Another is to ask whether the recipient could distinguish a request from a decision. Neither requires recording the entire meeting.

Separate three kinds of evidence in your review. First is the communication artifact: what the instruction or summary actually says. Second is the recipient's reported understanding or experience, obtained through an agreed route. Third is the subsequent work or decision, if it has genuinely happened. A clear artifact does not prove a positive experience, and a positive experience does not establish delivery.

Also choose the stopping point. The client could review one selected exchange, decline a colleague's involvement or decide that the relevant barrier is a policy decision outside the coaching work. Avoid accumulating observations because a dashboard makes it easy. Keep only what serves the agreed purpose and apply the relevant access and retention arrangements.

In Sian's case, another attempt has not occurred. Her intended habit is a future choice, not an improved score. Bram's criticism remains part of the actual result even though the repaired note meets its clarity conditions. The practice becomes more credible when it can preserve those different outcomes together.

Count the receiving work and actual preparation

The whole bounded case uses 330 person-minutes against 395 allocated, leaving 65: Ada 20, Sian 25 and Bram 20. The proposed recurring reporting allowance of 130 minutes is a separate planning unit. It does not replace this coordination account, and no customer receiving or report-delivery work is represented as observed.

Do not treat all of Bram's final ten minutes as pure writing; they include reading Sian's correction. Likewise, Sian's ten-minute acceptance block includes checking, sending the acceptance and preparing and sending the selected coaching summary. No work is counted twice simply because it serves two purposes.

Keep power, privacy and the tool in proportion

A leader asking for an honest answer still holds power. A practical invitation includes an achievable reply window, time to check facts and a route to disagree. Do not require someone to perform vulnerability in front of colleagues or disclose private circumstances to explain an ordinary capacity limit.

The current ICF Code of Ethics, revised May 2026, addresses agreements about roles and information exchange. In coaching, clarify which selected account can leave the session and who may receive it. A learning prompt does not authorize sharing a colleague's private information or uploading workplace records.

Coachful is our #1 recommendation for organizing the agreed coaching conversations and selected follow-up. This is an editorial workflow recommendation, not a measurement of brain activity or a guarantee of trust. Compare Coachful's coaching features with your actual communication needs.

In the case, Sian sends the chosen process account through her existing private client conversation, and Ada actually returns the feedback there. The coach can open the client profile and message icon. Internal notes alone would not establish delivery. An optional worksheet could provide a format for a reflection, but the case does not claim a native form assignment or automatic sharing of private thoughts.

The demo images illustrate preparation and follow-up, not Sian's company records. For other conversation choices, read conversational agility and the three-levels guide. Explaining, advocating and exploring can each be appropriate; constant questioning is not the goal.

Frequently asked questions

Are conversational blind spots a clinical diagnosis?

No. Here the term identifies an unchecked communication assumption. The guide does not diagnose a colleague, measure hormones or assign a neurological cause to disagreement.

Should a leader ask questions 85 percent of the time?

No universal ratio follows from the role-play account. Choose the balance for the purpose. A clear instruction can be appropriate, provided authority, meaning and practical constraints are understood.

What does Glaser's double-click metaphor suggest?

Examine an important word more closely rather than assuming its meaning is shared. Ask for a concrete example or boundary, and remain willing to revise your own account.

Does a nod establish agreement?

Check what it meant to that person. Acknowledging receipt, understanding, agreeing and accepting responsibility are different possibilities. Avoid deciding from body language alone.

When should a formal process take priority?

Use the appropriate organizational route for misconduct, discrimination, harassment, grievance or safety concerns. Do not turn a serious concern into a mutual misunderstanding simply because conversational coaching is available.

Source scope: Glaser's named essays and interview are primary practitioner accounts. Acas, ICF and W3C support the limited practice and ethical distinctions identified. The dialogue, planning note, clarity checks, reflection record and time account are original fictional illustrations.