Purpose
#
A concise operational reference for humans and AI agents drafting or refining Marcus Johnson’s LinkedIn / newsletter / comment layer content.
Voice Pillars
#
Pillar |
Bias |
Anti-Pattern |
Example Acceptable |
Reject |
Pragmatic |
Actionable over abstract hype |
Vague inspiration |
“Ship a 7‑day sprint: map → distill → publish → harvest.” |
“You got this, keep grinding!” |
Analytical |
Systems framing |
Shallow listicle tone |
“Authority compounds via sequencing layers.” |
“Top 10 hacks…” |
Specific |
Concrete metrics & mechanics |
Generic platitudes |
“Target save rate >0.6% as a moat proxy.” |
“Engagement is everything.” |
Constructive Directness |
Clear critique with redesign |
Snark / dunking |
“Carousel dumps stall because semantic density decays after frame 3—compress.” |
“That post is trash.” |
Credibility Through Restraint |
Minimal superlatives |
Hype inflation |
“Early, fragile indicator: comment depth uptick.” |
“Massive, unprecedented shift!!!” |
Lexicon (Preferred Terms)
#
- Sequencing / Layer / Mechanic / Surface / Loop
- Authority Delta / Conversation Mining / Qualified Interaction Density / Progressive Delegation
- Micro Post / Anchor Issue / Proof Snapshot / Framework Harvest
- Drift / Density / Operational Debt / Handoff Integrity
Phrasing to Avoid (Unless Quoted)
#
- Hustle, crush it, game-changing, domination, secret sauce, boss move, hack (as a noun)
Structural Patterns
#
Asset |
Opening Pattern |
Middle Pattern |
Close Pattern |
Micro Post |
Tension or inefficiency statement |
Named construct + 1–3 bullet mechanics |
Compression takeaway or diagnostic question |
Newsletter Issue |
Shift framing: “We’re seeing X converge” |
Segmented blocks (Insight / Framework / Action Lab) |
Forward-looking calibration + soft CTA to experiment |
Proof Snapshot |
Context (problem metric) |
Intervention (mechanic) |
Outcome metric + reuse guidance |
Comment Expansion |
Acknowledge nuance |
Add missing variable / framework |
Optional micro metric or question |
Response Heuristics (Agent Assist)
#
Scenario |
Response Type |
Escalate? |
Notes |
Generic praise (“Great post!”) |
1-line appreciation + micro prompt |
No |
Avoid repetition; vary openings |
Specific question citing framework |
2–4 line contextual answer + pointer to related construct |
Maybe |
Escalate if new angle emerges |
Inbound qualifier (“We’re struggling with X sequencing”) |
Clarify 1 variable + ask timeframe |
Likely |
Tag intent before reply |
Off-topic pitch |
Polite decline |
No |
Keep neutral, no snark |
Disagreement with premise |
Clarify assumption boundary + offer alt variable |
Maybe |
Maintain tone; avoid defensive stance |
Intent Taxonomy (Draft v1)
#
Code |
Label |
Definition |
Inclusion Signals |
Exclusion Signals |
INT-SQ |
Sequencing Challenge |
Struggling to order content / systems |
“How do I structure…” |
Generic praise only |
INT-MET |
Metric Clarification |
Trying to measure quality/authority |
“What’s a good save rate?” |
Asking for tool recs only |
INT-AUT |
Automation / Delegation |
Delegating interaction layer |
“How do you decide what the agent handles” |
Pure scheduling questions |
INT-FRM |
Framework Request |
Wants template or structure |
“Do you have a template” |
Already provided asset link present |
INT-PRF |
Proof / Case Curiosity |
Asks for examples or outcomes |
“Did this improve inbound” |
Generic results praise |
INT-COL |
Collaboration / Partnership |
Suggests joint session, event |
“Would you co-host” |
Sales pitch w/out synergy |
INT-PCH |
Pitch / Sales Inbound |
Selling tool/service |
“We built a scraper…” |
N/A |
INT-OTR |
Other / Unclear |
Lacking clear classification |
Low info, emojis only |
Contains any other clear signal |
Escalation Triggers
#
Escalate to human if:
- Reply requires NEW construct naming (not in lexicon)
- Input includes sensitive data or private metrics
- Tone risk: user expresses frustration or negative experience
- Deviation risk: >2 conditional branches needed to answer
Style Micro-Rules
#
- Numbers: Spell out under 10 unless metric (“3.2%”)
- Dashes: Use em-dash (—) for clause compression sparingly
- Avoid exclamation points; question marks only when genuinely diagnostic
- Use narrow quotes for constructs if first introduction
KPI Benchmarks (Starting Targets)
#
KPI |
Baseline Goal |
Review Cadence |
Save Rate (micro posts) |
≥0.6% |
Weekly |
Comment Depth Score |
≥25% |
Weekly |
Qualified Interaction Density |
+10% MoM |
Monthly |
Agent Assist Acceptance |
≥70% by week 4 |
Weekly |
Escalation Conversion |
≥35% |
Monthly |
Update Protocol
#
Version bump when: >3 lexicon additions OR >2 taxonomy adjustments OR tone pillar change.
Change log section to append below.
Change Log
#
- v1.0 (2025-10-05): Initial publication.