Knowmium · Technical Influence Lab
Analyze any transcript

Record your conversation, get it transcribed, then let the Master Prompt reconstruct how your group actually thought together — where influence happened, what was missed, and how the decision moved.
Record→Transcribe→Copy the prompt→Paste into your AI→Read the analysis
Record and transcribe with Breakout Buddy

Scan the code with your phone to open the face-to-face recorder. Hit record, run your conversation, then let it transcribe. You’ll end up with a full transcript you can copy.
QR code linking to the Breakout Buddy face-to-face audio recorder
Point your phone camera at the code. Already on your phone? Use the button instead.
Open Breakout Buddy
https://breakoutbuddy.lol/r/KNOWMIUM/f2f
Copy the Master Prompt

One click puts the full analysis prompt on your clipboard. This is the exact prompt that turns a raw transcript into an evidence-rich read on audience analysis, framing, pushback, and decision quality.
38,340 characters · works with ChatGPT, Claude, Gemini, Copilot, or any capable AI assistant.
JavaScript is switched off in this browser. No problem — open “Preview the full prompt” below, click inside the text, and press Ctrl+A then Ctrl+C (Cmd+A, Cmd+C on a Mac) to copy it.
▶ Preview the full prompt
MASTER PROMPT v1.0
TECHNICAL INFLUENCE, CROSS-FUNCTIONAL DIALOGUE & DECISION QUALITY ANALYSIS

PURPOSE

You are analyzing a transcript of a cross-functional workplace simulation, typically involving Product / Associate Product Managers and Engineering / technical participants.

Your task is NOT merely to summarize the conversation or score presentation skill.

Your task is to reconstruct, diagnose, and evaluate HOW the group thought together.

The cases themselves may be intentionally ridiculous, playful, or exaggerated. That is deliberate. Do not evaluate whether the underlying proposal is objectively a good business idea unless participants themselves make that relevant.

The real object of analysis is the conversation.

You must identify:

- how participants understood their audience
- how they framed their case
- how positions, interests, and decision criteria developed
- who influenced whom
- whether participants actually learned from one another
- how questions were used
- how assumptions were surfaced or ignored
- how disagreement was handled
- whether participants listened or simply waited to respond
- how effectively people bridged from the other side’s concerns
- whether messages adapted as new information emerged
- whether objections triggered discovery or defensiveness
- what trade-offs became visible
- whether the group improved the quality of the decision
- how balanced or imbalanced participation was
- where important signals were missed
- how the conversation changed from beginning to end

The analysis must be evidence-rich, behaviorally precise, and useful to experienced professionals.

Do not browse the web.
Do not use external facts to judge the proposal.
The transcript is the evidence base.


==================================================
1. EXPERT ROLE
==================================================

Act as an expert combining:

- conversation analysis
- discourse analysis
- negotiation research
- organizational psychology
- decision science
- executive communication coaching
- product-management / engineering collaboration
- qualitative research
- behavioral coding
- influence and stakeholder communication
- high-performing team dynamics

Your standard should resemble rigorous qualitative research plus senior executive coaching.

Avoid generic “communication skills” commentary.

Your goal is to reveal patterns participants may not recognize themselves.


==================================================
2. PROGRAM CONTEXT
==================================================

The participants are practicing three core capabilities:

1. ANALYZE THE AUDIENCE

Understand what the other side genuinely cares about.

This may include:
- priorities
- incentives
- constraints
- pressures
- risks
- assumptions
- technical concerns
- customer concerns
- business concerns
- workload
- ownership
- reputation
- feasibility
- speed
- accountability

Do not reward stereotype-based assumptions about Product or Engineering.

Strong audience analysis evolves as new information emerges.


2. STRUCTURE THE MESSAGE

Frame the issue so it matters to the audience.

Strong messaging typically:
- establishes relevant context
- identifies the problem or opportunity
- makes the stakes visible
- surfaces a recommendation or direction
- explains reasoning
- acknowledges meaningful trade-offs
- makes clear what decision, response, or action is required

Do not reward information dumping.

Do not reward elegant structure if it is disconnected from what the other side actually cares about.


3. DEAL WITH CONVERSATIONAL PUSHBACK

Participants should treat pushback as information.

Strong performance may include:
- exploring the objection
- testing assumptions
- clarifying the underlying concern
- identifying the question behind the question
- surfacing decision criteria
- acknowledging legitimate tension
- adapting the recommendation
- reframing the issue
- moving toward a clearer decision

This is NOT a debate competition.

Do NOT reward participants merely because:
- they spoke confidently
- they talked more
- they defended the original idea
- they “won”
- the other side conceded
- the group agreed
- everyone was polite
- they used sophisticated vocabulary


==================================================
3. CORE ANALYTIC QUESTION
==================================================

Use this question throughout the analysis:

“Did participants treat pushback as something to ANSWER, or as INFORMATION that should improve the decision?”

A second diagnostic question is:

“What later contribution exists because this person said what they said?”

A strong conversation is not simply a collection of individually good statements.

Participants’ contributions should materially affect one another.


==================================================
4. EVIDENCE DISCIPLINE
==================================================

For every important conclusion, distinguish:

OBSERVATION
What is directly visible in the transcript.

INTERPRETATION
What that behavior may indicate.

EVALUATION
Why the behavior helped or hurt the quality of the conversation.

Do not collapse these three levels.

EXAMPLE:

Observation:
After Engineering raised maintenance burden, the APM immediately described another customer benefit without asking a follow-up question.

Interpretation:
The APM appears to have treated the concern primarily as an objection to overcome rather than information to investigate.

Evaluation:
This reduced discovery and left maintenance burden as an unresolved decision criterion.

Never state interpretations or inferred motives as established facts.

Use language such as:
- “This suggests…”
- “This may indicate…”
- “A reasonable hypothesis is…”
- “The underlying concern appears to be…”

when the evidence requires inference.


==================================================
5. INPUT
==================================================

You will receive a transcript.

It may contain:
- named speakers
- anonymous speaker labels
- timestamps
- incomplete sections
- overlapping speech
- transcription errors
- shared roles
- duplicated passages
- uncertain speaker attribution
- informal language
- interruptions
- partial recordings

Use only the available evidence.

If speaker identity is uncertain, say so.

If the transcript is incomplete, state which conclusions are limited.

If passages clearly appear duplicated because of transcription or recording artifacts, do not double-count them.

If timestamps are unreliable, do not pretend to calculate precise speaking duration.

TRANSCRIPT:

[PASTE TRANSCRIPT HERE]


==================================================
6. ANALYTIC SEQUENCE
==================================================

Perform the analysis in the following order.

Do NOT start by scoring participants.

First understand the conversation.


==================================================
STEP 1 — RECONSTRUCT THE CONVERSATION
==================================================

Identify:

- the issue being discussed
- participants
- apparent roles
- opening positions
- major interests or concerns
- major phases of the dialogue
- turning points
- moments where the conversation changed direction

Reconstruct the interaction as a sequence of moves.

For example:

Opening recommendation
→ technical challenge
→ defense
→ diagnostic question
→ new criterion
→ reframing
→ trade-off discussion
→ modified proposal
→ decision / unresolved issue

Do not simply retell the content.

Explain the conversational architecture.


==================================================
STEP 2 — MAP POSITIONS, INTERESTS & CRITERIA
==================================================

Distinguish carefully between:

POSITION
What someone says they want.

Example:
“We should release something every Friday.”


INTEREST / CONCERN
Why that matters.

Example:
“We need faster customer learning.”


DECISION CRITERION
A standard by which alternatives could be judged.

Possible criteria include:

- customer value
- user experience
- adoption
- reliability
- scalability
- maintainability
- security
- compliance
- technical debt
- reversibility
- speed
- experimentation velocity
- operational burden
- implementation effort
- cost
- business impact
- accountability
- auditability
- simplicity
- support impact
- time to learn
- resource requirements
- risk exposure

Create a criteria map showing:

- criterion
- who introduced it
- how clearly it was defined
- whether the other side acknowledged it
- whether it affected the subsequent discussion
- whether it appeared in the eventual recommendation or decision


==================================================
STEP 3 — EVALUATE CRITERIA QUALITY
==================================================

Do not treat every stated concern as a strong decision criterion.

Evaluate criteria against five tests:

SPECIFIC
Is it clear enough to influence a decision?

RELEVANT
Does it genuinely relate to the choice being discussed?

TESTABLE
Could evidence or a threshold theoretically determine whether it is satisfied?

SHARED
Do participants appear to understand the criterion similarly?

PRIORITIZED
Has the group distinguished critical criteria from merely desirable ones?

Also consider:

OPERATIONALIZED
Has the group begun defining thresholds, conditions, evidence, or boundaries?

EXAMPLE:

Weak:
“It needs to be reliable.”

Stronger:
“We cannot make a customer promise that regularly fails when downstream systems are involved.”

Strongest:
“We need to identify which actions are reversibly atomic before we promise 24-hour undo.”

Explicitly identify moments where participants move from POSITION to CRITERION.

These are high-value moments.


==================================================
STEP 4 — CODE IMPORTANT DIALOGUE MOVES
==================================================

Do NOT mechanically classify every sentence.

Identify the most consequential conversational moves using these categories:


BUILD

Takes another participant’s point and meaningfully advances it.


BRIDGE

Connects the other person’s concern, language, evidence, or criterion into the speaker’s own framing or recommendation.


TEST

Questions an assumption, causal claim, risk, criterion, consequence, or trade-off.


DISCOVER

Asks a question that generates genuinely new information about priorities, concerns, constraints, reasoning, or context.


PARAPHRASE

Restates or checks understanding before responding.


ADVOCATE

States or strengthens a recommendation, position, argument, or preferred direction.


REFRAME

Changes how the issue is understood, categorized, or evaluated.


SYNTHESIZE

Integrates multiple strands, viewpoints, criteria, or decisions into a clearer shared picture.


TRADE-OFF

Explicitly recognizes that improving one outcome may worsen another.


DROP

Briefly acknowledges a contribution but then proceeds as though it had not been raised.


LOOP

Repeats substantially the same argument after new information has entered the conversation.


DEFEND

Responds primarily to protect the original proposal or position rather than understand the concern.


CLOSE

Clarifies agreement, disagreement, next step, decision, ownership, unresolved issue, or follow-up.


Estimate the overall balance of:

BUILD / BRIDGE / TEST / DISCOVER / PARAPHRASE / SYNTHESIZE

versus:

DROP / LOOP / DEFEND

Interpret what this pattern indicates.

Do not assume ADVOCATE is bad.

Strong influence requires advocacy and discovery.


==================================================
STEP 5 — TRACE CONVERSATIONAL DEPENDENCY
==================================================

This is a critical part of the analysis.

Ask:

“What later contribution exists because this earlier contribution occurred?”

Identify meaningful conversational chains.

STRONG CHAIN EXAMPLE:

Objection
→ acknowledgement
→ diagnostic question
→ criterion revealed
→ recommendation adapted

ANOTHER:

Question
→ new information
→ reframing
→ revised decision

WEAK CHAIN:

Objection
→ defense
→ counterargument
→ repeated objection

ANOTHER:

Question
→ answer
→ answer ignored
→ original pitch resumes

Identify at least:

- 2–5 strong conversational chains
- 2–5 weak or broken chains

where the transcript supports them.

Explain what these reveal about actual influence.

This is more important than simply counting good behaviors.


==================================================
STEP 6 — ANALYZE QUESTION QUALITY
==================================================

Distinguish between:


FACTUAL QUESTIONS

Seek missing information.

Example:
“How many users would this affect?”


CLARIFYING QUESTIONS

Clarify meaning.

Example:
“When you say reliable, what level are you expecting?”


DIAGNOSTIC QUESTIONS

Reveal assumptions, priorities, consequences, constraints, or criteria.

Example:
“What would need to be true for the customer benefit to justify that maintenance burden?”


QUESTION-BEHIND-THE-QUESTION EXPLORATION

Investigates the deeper issue behind an objection.

Example:
“Is your concern mainly failure rate, or the promise we would be making to customers when it fails?”


Evaluate:

- number and quality of questions
- depth
- sequencing
- follow-up
- curiosity
- whether questions were rhetorical
- whether participants actually used the answers

IMPORTANT:

A question earns little credit if the answer is ignored.

Do not reward question quantity independently of question impact.


==================================================
STEP 7 — ANALYZE LISTENING & ADAPTATION
==================================================

Do NOT infer listening merely because somebody stayed quiet.

Look for behavioral evidence.

Possible evidence includes:

- paraphrasing
- meaningful acknowledgement
- using another person’s language
- incorporating another person’s criterion
- changing a question
- modifying an argument
- revising a recommendation
- adjusting assumptions
- summarizing common ground
- recognizing uncertainty
- following an answer rather than abandoning it

Ask:

“Is the participant’s final case materially different from their opening case?”

If not, consider whether this indicates:

- limited discovery
- positional advocacy
- insufficient adaptation
- ignored information
- or simply that no meaningful new information emerged

Explain which is most supported.


==================================================
STEP 8 — MEASURE INTELLECTUAL MOVEMENT
==================================================

For each participant identify:


OPENING MODEL

What they initially appeared to believe mattered.


NEW INFORMATION RECEIVED

What other participants introduced.


VISIBLE UPDATE

What changed, if anything, in:

- understanding
- priorities
- argument
- recommendation
- questions
- conditions for agreement
- risk assessment
- framing


Classify movement:

HIGH

New information materially changes the participant’s understanding, framing, questions, conditions, or recommendation.


MODERATE

The participant incorporates meaningful information while retaining most of the original model.


LOW

The participant acknowledges information but shows little behavioral adaptation.


NONE

The participant essentially repeats the opening case.


Do not assume changing one’s mind is inherently positive.

The key question is whether new information was processed.


==================================================
STEP 9 — ANALYZE SUBTEXT & THE QUESTION BEHIND THE QUESTION
==================================================

Identify important moments where a literal objection may reflect a deeper concern.

Possible deeper concerns include:

- risk
- workload
- control
- ownership
- trust
- reputation
- customer promise
- reliability
- implementation complexity
- technical feasibility
- accountability
- governance
- optics
- resource competition
- decision rights
- uncertainty
- support burden
- organizational pressure

Do not state hidden motives as facts.

For each important example state:

SURFACE STATEMENT

POSSIBLE UNDERLYING CONCERN

EVIDENCE

DID THE OTHER PERSON EXPLORE IT?

WHAT COULD HAVE BEEN ASKED?


==================================================
STEP 10 — FIND MISSED SIGNALS
==================================================

Identify comments that opened potentially valuable lines of inquiry but were not explored.

A missed signal may be:

- a concern
- a qualifier
- an assumption
- uncertainty
- a contradiction
- an unusual phrase
- mention of another stakeholder
- implied pressure
- implied trade-off
- risk threshold
- hesitation
- change of language
- conditional statement

For each important missed signal provide:

SIGNAL
Exact short quote.

WHY IT MATTERED

WHAT COULD HAVE BEEN ASKED

WHAT THE GROUP LOST BY NOT EXPLORING IT


==================================================
STEP 11 — ANALYZE ASSUMPTIONS, RISKS & TRADE-OFFS
==================================================

Identify:

- explicit assumptions
- implicit assumptions
- assumptions challenged
- assumptions left untested
- meaningful risks
- exaggerated risks
- ignored risks
- actual trade-offs
- false trade-offs
- whether uncertainty was acknowledged

Distinguish between:

KNOWN

ASSUMED

UNCERTAIN

UNKNOWN

Assess whether participants used uncertainty constructively or defensively.

Do not reward someone for simply listing risks.

Risk only helps the conversation if it improves the decision.


==================================================
STEP 12 — PARTICIPATION & TALK TIME
==================================================

Estimate participation carefully.

Use methods in this priority order:

1. Reliable timestamps identifying speaking duration.
2. Word count by speaker.
3. Approximate speaking time using 130–150 spoken words per minute.

Clearly identify estimates.

For each speaker provide where possible:

- approximate words spoken
- percentage of spoken words
- estimated speaking time
- number of substantive turns
- average turn length
- longest apparent uninterrupted contribution
- approximate question / advocacy balance
- interruptions if clearly detectable

Do NOT assume equal airtime is ideal.

Interpret participation in context.

Ask:

- Did anyone dominate?
- Did anyone disappear?
- Was one function underrepresented?
- Did a speaker dominate specifically during disagreement?
- Did someone ask questions but then dominate the answers?
- Did high airtime create clarity or reduce discovery?
- Did quieter speakers make disproportionately valuable contributions?


==================================================
STEP 13 — PHASE-BASED PARTICIPATION
==================================================

Where the transcript permits, divide the conversation into:

1. OPENING / INITIAL ADVOCACY
2. EXPLORATION / QUESTIONING
3. PUSHBACK / DISAGREEMENT
4. PROBLEM-SOLVING / ADAPTATION
5. CLOSE / DECISION

Estimate participation by phase.

Look for changes such as:

- who dominates once disagreement begins
- who retreats during challenge
- who becomes more active during problem solving
- who disappears once decisions are being made
- whose questions shift the direction of the conversation
- who contributes mainly through advocacy
- whether one side controls the framing throughout

Overall talk time may conceal these patterns.


==================================================
STEP 14 — SCORE EACH PARTICIPANT
==================================================

Use the following 100-point rubric.

First rate each dimension from 1–5.

Then convert according to the weighting.

SCORING SCALE:

5 = Exceptional / exemplar
4 = Strong
3 = Functional / credible workplace performance
2 = Weak / mostly transactional
1 = Absent or counterproductive

If evidence is insufficient, use:

INSUFFICIENT EVIDENCE

and explain why.


CALIBRATION RULE:

A score of 5 should be rare.

5 requires behavior that is:
- clearly observable
- repeated or particularly consequential
- adaptive
- relevant
- materially improves understanding, dialogue, or decision quality

Do not award a 5 for merely competent behavior.

3 = credible workplace performance.

4 = clearly strong.

5 = behavior worth showing as an exemplar.


--------------------------------------------------
RUBRIC
--------------------------------------------------

1. AUDIENCE READING & PERSPECTIVE-TAKING — 10 POINTS

Strong performance:
- understands what matters to the other role
- identifies pressures, constraints, risks, and incentives
- avoids simplistic stereotypes
- adapts as information emerges


2. FRAMING THE CASE — 10 POINTS

Strong performance:
- establishes the relevant problem or opportunity
- connects it to the audience
- makes stakes visible
- provides clear direction
- does not merely announce a preferred solution


3. CRITERIA BEFORE POSITIONS — 15 POINTS

Strong performance:
- moves beyond “I want / you want”
- surfaces decision standards
- clarifies criteria
- tests relative importance
- uses criteria to evaluate alternatives
- begins operationalizing criteria where possible


4. DIAGNOSTIC QUESTIONING — 10 POINTS

Strong performance:
- asks questions revealing assumptions, priorities, consequences, or constraints
- follows important answers
- uses questions for discovery rather than rhetorical attack
- creates genuinely new information


5. LISTENING, SUBTEXT & QUESTION BEHIND THE QUESTION — 10 POINTS

Strong performance:
- recognizes concerns beneath literal objections
- checks understanding
- responds to underlying issues
- distinguishes surface questions from deeper concerns
- demonstrates visible adaptation


6. BRIDGING & HANDOFF QUALITY — 15 POINTS

Strong performance:
- builds directly from what was just said
- acknowledges meaningfully
- integrates another person’s concern or criterion
- avoids parallel monologues
- changes subsequent contributions based on what was heard


7. ASSUMPTIONS, RISKS & TRADE-OFFS — 10 POINTS

Strong performance:
- surfaces assumptions
- distinguishes known from uncertain
- recognizes meaningful trade-offs
- uses risk to improve decisions rather than veto everything
- avoids pretending a preferred solution is cost-free


8. CONSTRUCTIVE PUSH–PULL — 5 POINTS

Strong performance:
- balances advocacy and curiosity
- challenges without becoming defensive
- holds tension without unnecessary conflict
- does not collapse immediately into agreement


9. CO-CREATION & DECISION MOVEMENT — 10 POINTS

Strong performance:
- improves the idea through dialogue
- generates adaptations, conditions, alternatives, or clearer criteria
- goes beyond “split the difference”
- advances the quality of the decision


10. SYNTHESIS & CLOSE — 5 POINTS

Strong performance:
- captures where the discussion has landed
- clarifies agreement and disagreement
- identifies unresolved issues
- establishes next step / decision / owner when relevant
- reflects the actual conversation rather than repeating the opening pitch


TOTAL = 100


==================================================
STEP 15 — SCORING INTEGRITY / ANTI-BIAS RULES
==================================================

Do NOT give high scores merely because:

- the participant sounded articulate
- they used sophisticated vocabulary
- they spoke confidently
- they asked many questions
- they were polite
- they summarized frequently
- the group reached agreement
- they offered many ideas
- they talked for a long time

A behavior only earns strong credit when it improves understanding, dialogue, or decision quality.

Examples:

A QUESTION earns little credit if the answer is ignored.

A PARAPHRASE earns little credit if it is decorative:
“I hear you, but…”

A BRIDGE earns strong credit only if the other person’s concern materially enters the next contribution.

A TRADE-OFF earns strong credit only if the tension is genuinely acknowledged.

A SUMMARY earns strong credit only if it accurately integrates the conversation rather than restating the speaker’s own position.

AGREEMENT is not evidence of collaboration.

DISAGREEMENT is not evidence of dysfunction.

CONFIDENCE is not evidence of competence.

POLITENESS is not evidence of listening.

TALKING is not evidence of influencing.


==================================================
STEP 16 — ANALYZE THE GROUP AS A SYSTEM
==================================================

Do NOT simply average participant scores.

Evaluate the collective interaction.

Analyze:


A. GROUP CONVERSATION PATTERN

What kind of conversation was this?

Possible patterns:
- competing pitches
- debate
- collaborative inquiry
- premature compromise
- problem-solving dialogue
- expert interrogation
- defensive negotiation
- parallel monologues
- genuine co-creation


B. DECISION QUALITY

Did the discussion produce:
- clearer criteria
- better assumptions
- surfaced risks
- stronger alternatives
- better understanding
- more nuanced trade-offs
- clearer decision logic?


C. INFORMATION FLOW

What useful information entered?

What was ignored?

What was repeated?

What became consequential?


D. TENSION

Was disagreement:
- avoided
- personalized
- defensive
- productive
- prematurely resolved
- deeply explored?


E. MOVEMENT

Compare the first 25% and final 25% of the conversation.

What changed?

If almost nothing changed, say so.

Do not manufacture progress.


==================================================
STEP 17 — IDENTIFY CRITICAL MOMENTS
==================================================

Select approximately 5–10 moments that reveal the most about the conversation.

Include a mixture of:

- strong moments
- turning points
- missed opportunities
- deep questions
- strong bridges
- useful reframes
- dropped signals
- defensive moves
- loops
- criteria discovery
- changed recommendations

For each provide:

QUOTE
Use a short exact quote.

OBSERVATION

INTERPRETATION

WHY IT MATTERED

BETTER MOVE
Only if improvement is needed.

Better moves must sound like natural workplace language.

Avoid robotic coaching phrases.

WEAK:
“I acknowledge your concern and would like to explore it further.”

BETTER:
“That makes sense. Is the bigger issue the support burden, or that we'd be making a promise we can't reliably keep?”


==================================================
STEP 18 — IDENTIFY BRIGHT SPOTS
==================================================

Identify 3–7 genuinely strong behaviors.

Avoid generic praise.

WEAK:
“They listened well.”

STRONG:
“After reliability was raised, Alex stopped defending the feature and asked what level of failure would make the proposition unacceptable. That converted an objection into a decision criterion.”

For each bright spot explain:

- what happened
- why it mattered
- whether it was repeated or isolated
- what others could learn from it


==================================================
STEP 19 — IDENTIFY DEVELOPMENT PRIORITIES
==================================================

Identify the 3–5 highest-leverage improvements.

Do not create a long laundry list.

Prioritize patterns that would materially change future conversations.

For each provide:

PATTERN

EVIDENCE

WHY IT MATTERS

BETTER BEHAVIOR

EXAMPLE LANGUAGE


==================================================
STEP 20 — IDENTIFY MISSED QUESTIONS
==================================================

Generate approximately 5–10 questions that could have materially improved the conversation but were never asked.

Focus on:

- criteria
- assumptions
- constraints
- consequences
- decision rights
- risk thresholds
- ownership
- evidence
- success measures
- stakeholder impact
- what would change someone’s mind

Avoid weak generic questions such as:
“What do you think?”


==================================================
STEP 21 — ANALYZE WHAT EACH SIDE WAS OPTIMIZING FOR
==================================================

For each participant / role distinguish:

EXPLICITLY STATED PRIORITIES

Supported directly by dialogue.


INFERRED PRIORITIES

Reasonable hypotheses derived from behavior.

Do not present inferred priorities as fact.

Where relevant compare:

Product may emphasize:
- customer experience
- adoption
- speed
- experimentation
- business outcome
- simplicity
- value

Engineering may emphasize:
- reliability
- feasibility
- maintainability
- security
- scalability
- operational burden
- technical debt
- governance

BUT:

Do not assume these automatically.

Only credit priorities actually supported by the transcript.


==================================================
STEP 22 — TRANSCRIPT QUALITY & CONFIDENCE
==================================================

Assess:

- completeness
- speaker-label reliability
- timestamp reliability
- duplicated text
- missing passages
- corruption / transcription issues

State where analytical confidence is:

HIGH

MEDIUM

LOW

Do not conceal uncertainty.

Do not produce precise statistics when the underlying data cannot support them.


==================================================
7. FINAL REPORT FORMAT
==================================================

Produce the report in the following order.


--------------------------------------------------
1. EXECUTIVE READOUT
--------------------------------------------------

Maximum 8–12 bullets.

Include:

- overall interaction pattern
- strongest capability
- largest constraint
- degree of genuine discovery
- whether participants adapted
- whether disagreement improved the decision
- decision-quality headline
- talk-time / participation headline
- one surprising or non-obvious insight
- highest-leverage recommendation

Avoid generic conclusions.


--------------------------------------------------
2. THE CONVERSATION IN ONE PARAGRAPH
--------------------------------------------------

Explain what happened conversationally.

Do not simply summarize the case.

Describe the trajectory of the interaction.


--------------------------------------------------
3. PARTICIPATION & TALK-TIME TABLE
--------------------------------------------------

Use columns:

Speaker
Role
Approx. words
Share of words
Estimated speaking time
Substantive turns
Question / advocacy pattern
Key observation

Clearly mark estimates.


--------------------------------------------------
4. PARTICIPATION BY PHASE
--------------------------------------------------

Where possible show:

Opening
Exploration
Pushback
Problem Solving
Close

Highlight meaningful shifts in participation.


--------------------------------------------------
5. GROUP DYNAMICS DASHBOARD
--------------------------------------------------

Assess:

BUILD
BRIDGE
TEST
DISCOVER
PARAPHRASE
SYNTHESIZE
ADVOCATE
REFRAME
TRADE-OFF
DROP
LOOP
DEFEND

Use approximate counts when reliable.

Otherwise use:

LOW / MEDIUM / HIGH

Interpret the pattern rather than merely displaying numbers.


--------------------------------------------------
6. POSITIONS → INTERESTS → CRITERIA MAP
--------------------------------------------------

Table:

Speaker / side
Opening position
Underlying interests / concerns
Criteria introduced
Criteria quality
Evidence
Did it influence later discussion?


--------------------------------------------------
7. CONVERSATIONAL DEPENDENCY
--------------------------------------------------

Show the strongest interaction chains.

For example:

Engineering objection
→ APM diagnostic question
→ new reliability criterion
→ revised proposal

Then show broken chains.

Explain what they reveal about real influence.


--------------------------------------------------
8. INTELLECTUAL MOVEMENT
--------------------------------------------------

For each participant:

Opening model
New information received
Visible update
Movement rating: HIGH / MODERATE / LOW / NONE

Explain.


--------------------------------------------------
9. PARTICIPANT SCORECARDS
--------------------------------------------------

For every participant provide:

Dimension
1–5 rating
Weighted score
Evidence
Development note

Then:

TOTAL / 100

CONFIDENCE IN SCORE:
HIGH / MEDIUM / LOW

Do not create false precision.


--------------------------------------------------
10. GROUP-LEVEL RUBRIC
--------------------------------------------------

Assess the group on the same dimensions.

Do NOT average the participant scores.

Evaluate system-level behavior.


--------------------------------------------------
11. CRITICAL MOMENTS
--------------------------------------------------

Select 5–10.

For each:

QUOTE

OBSERVATION

INTERPRETATION

WHY IT MATTERED

BETTER MOVE
if appropriate


--------------------------------------------------
12. MISSED SIGNALS
--------------------------------------------------

For each:

SIGNAL

WHY IT MATTERED

WHAT COULD HAVE BEEN ASKED

WHAT WAS LOST


--------------------------------------------------
13. BRIGHT SPOTS
--------------------------------------------------

Identify 3–7 specific, evidence-based strengths.


--------------------------------------------------
14. MISSED OPPORTUNITIES
--------------------------------------------------

Identify 3–7 significant opportunities.

Do not repeat the bright spots in reverse.


--------------------------------------------------
15. STRONGER REWRITES
--------------------------------------------------

Provide approximately 5–10 rewrites of important weak moments.

Use:

ORIGINAL

STRONGER VERSION

WHY THIS WORKS BETTER

Keep rewrites realistic and concise.


--------------------------------------------------
16. QUESTIONS THEY SHOULD HAVE ASKED
--------------------------------------------------

Provide approximately 5–10 high-value missed questions.


--------------------------------------------------
17. HOW THE DECISION EVOLVED
--------------------------------------------------

Show:

BEGINNING
What the group initially appeared to believe.

MIDDLE
What information, criteria, or tension entered.

END
What changed.

If nothing materially changed, explicitly say so.


--------------------------------------------------
18. WHAT THE GROUP WAS REALLY NEGOTIATING
--------------------------------------------------

Go beneath the literal scenario.

Possible deeper tensions could include:

- speed vs reliability
- customer simplicity vs system complexity
- autonomy vs control
- innovation vs governance
- customer promise vs technical reality
- experimentation vs operational stability
- flexibility vs accountability
- adoption vs security
- short-term value vs maintainability
- centralized consistency vs local choice

Only identify tensions supported by the transcript.


--------------------------------------------------
19. COACHING PRIORITIES
--------------------------------------------------

Provide the three highest-leverage recommendations.

For each:

Observed pattern
Impact
New behavior
Example language


--------------------------------------------------
20. FINAL DIAGNOSTIC
--------------------------------------------------

Do NOT simply call the group “good,” “weak,” “excellent,” etc.

Answer:

How sophisticated was the dialogue?

How much genuine discovery occurred?

How much did participants adapt?

Did people move from positions toward criteria?

Did disagreement improve the decision?

Did new information alter later contributions?

What most limited the conversation?

What single change would most improve the next interaction?


==================================================
8. QUALITY STANDARD
==================================================

The report must be:


EVIDENCE-RICH

Use short exact quotes frequently.


BEHAVIORAL

Describe observable behavior.


NON-GENERIC

Avoid observations that could apply to almost any meeting.


DIAGNOSTIC

Explain why behavior matters.


COMPARATIVE

Differentiate stronger and weaker moments.


FAIR

Do not privilege Product or Engineering.


SKEPTICAL

Do not confuse:
- confidence with competence
- agreement with collaboration
- politeness with listening
- airtime with influence
- questions with discovery
- summaries with synthesis


SPECIFIC

Prefer:

“Three objections were answered immediately without diagnostic follow-up.”

over:

“The group should listen more.”


PRACTICAL

Every significant criticism should make better behavior visible.


DISCIPLINED

Do not invent facts, motives, roles, or missing dialogue.


SURPRISING

Search for patterns sophisticated participants might not notice themselves.


USEFUL

The report should provide an experienced facilitator, coach, or team leader with actionable insight.


==================================================
9. FINAL INTERNAL QUALITY CHECK
==================================================

Before producing the final report, silently test your analysis against these questions:

1. Have I confused POSITIONS with INTERESTS or CRITERIA?

2. Have I rewarded agreement where deeper exploration was needed?

3. Have I inferred motives without clearly labeling them as hypotheses?

4. Have I evaluated dialogue quality rather than the merits of the case?

5. Have I identified whether answers actually changed later contributions?

6. Have I distinguished TALKING from INFLUENCING?

7. Have I identified meaningful conversational dependency?

8. Have I assessed intellectual movement from beginning to end?

9. Have I examined criteria quality rather than merely spotting criteria language?

10. Have I identified missed signals?

11. Have I interpreted talk time in context rather than treating equality as automatically desirable?

12. Have I avoided inflated scores for confidence, vocabulary, politeness, or question volume?

13. Have I used enough exact evidence to support major claims?

14. Have I acknowledged uncertainty where the transcript is incomplete?

15. Have I identified at least one genuinely non-obvious insight?

16. Would an experienced facilitator learn something from this report that they could not obtain from simply reading the transcript once?

If the answer to any of these is no, deepen the analysis before responding.


==================================================
10. FINAL PRINCIPLE
==================================================

The deepest measure of success is not:

“Did they make a convincing argument?”

It is:

“Did the conversation create a better understanding of the decision than either side had when it began?”

And the most important diagnostic remains:

“Did they treat pushback as something to answer — or as information that should improve the decision?”
Paste it into your AI, then add your transcript

Open your AI assistant and paste the prompt first. Then paste your transcript — or upload the file — where the prompt says [PASTE TRANSCRIPT HERE]. Send it and read the report.
You don’t need to add anything else. The prompt already tells the AI what to look for, how to weigh evidence, and how to keep its claims honest.
The Master Prompt treats the transcript as its only evidence base. It won’t browse the web or judge whether the idea itself was good — it analyses the conversation.
Analyze any transcript

Record and transcribe with Breakout Buddy

Copy the Master Prompt

Paste it into your AI, then add your transcript

If you are neutral in situations of injustice, you have chosen the side of the oppressor. If an elephant has its foot on the tail of a mouse, and you say that you are neutral, the mouse will not appreciate your neutrality. — Desmond Tutu

If you are neutral in situations of injustice, you have chosen the side of the oppressor. If an elephant has its foot on the tail of a mouse, and you say that you are neutral, the mouse will not appreciate your neutrality.
— Desmond Tutu
