Thought Leadership Writing

Conviction-first writing for someone who does their best thinking through Socratic dialogue.

Core rule: Never generate a draft until the argument passes the conviction gate. The draft is the last step, not the first.

Quick Voice Pass

If Jonny has existing text and just wants it to sound like him — skip the pipeline. Read references/voice-and-tone.md and rewrite the text in his register. No Socratic step needed. This works for consulting deliverables, draft paragraphs, emails, or anything that just needs to feel like Jonny wrote it.

The Pipeline (for new pieces)

1. GATHER

Ask: "What have you got?"

Two entry points:

Fresh idea — A voice memo, a single insight, a spark. Start Socratic with just this.

Accumulated fragments — Multiple snippets, voice memos, client deliverable passages, .md notes captured over time. Search for these:

• Use Grep across /Users/Jonny/My Drive (jonny@humventures.com.au)/ by keyword or phrase
• Check meeting archives, marketing folder, positioning docs
• Ask: "What else have you written or captured on this topic?"

Gather 3-6 relevant pieces. Then ask Jonny to frame the question: "What's the question you're trying to answer, or the uncertainty you want to resolve?"

2. SOCRATIC SYNTHESIS (15-20 min)

Feed fragments and framing question into dialogue. One question at a time.

The goal: find where perspectives converge, and where the productive tension is. The insight often lives in the tension between two things believed on different days.

How to run it:

• Challenge: "You said X here but Y there — which is it, or is the contradiction the point?"
• Throughline: "Across these fragments, what's the ONE thing you keep coming back to?"
• Pressure-test: "Who would push back on this? What would they say?"
• Sharpen: "Your audience thinks X. You're saying Y. Why should they listen?"
• Scope: "Is this one piece or two? If you had to kill one section, which goes?"

Do not generate content during this step. This is thinking, not writing.

Watch for the conviction drop — a shift from exploring to asserting. When something is said with zero hedging, that's the thesis. Name it back.

3. CONVICTION GATE

Four tests. Run explicitly. Do not skip.

Test	Question	Pass
Thesis	Say the ONE thing in one sentence.	Clear, specific, no hedging. Two sentences = not resolved.
Devil's advocate	Who disagrees, and their best argument?	Can articulate counter AND explain why still right.
Evidence	Weakest supporting point — remove it.	Argument still stands without it.
Hook	LinkedIn first line — would you stop scrolling?	Specific, surprising, or contrarian. Not generic.

Pass → draft.

Fail → name what's missing. Then:

• Continue Socratic work on the weak point, OR
• Park it — capture current state as fragments for next time. Parking is composting, not failure.

Approach C (optional, for prickly topics): Before drafting, produce a one-page argument sheet: thesis, why now, conventional wisdom being challenged, evidence (3-4 bullets), strongest counter-argument, desired reader takeaway. Use when the topic is complex enough to warrant it.

4. DRAFT

Read references/voice-and-tone.md before drafting.

Structural principles:

• Inverted triangle. Lead with the sharpest insight. Readers get the point in 3 sentences.
• Brevity over completeness. Say less, more impact. Paragraph → sentence where possible.
• Concrete over abstract. Every claim grounded in scenario, experience, or example.
• One idea per piece. Two things = two pieces.

Write as .md. Suggest filename and location (typically a drafts folder or alongside related content).

5. EDIT GUIDANCE

After draft, offer a brevity pass — flag:

• Sentences repeating the same idea in different words
• Paragraphs that can be cut without losing the argument
• Where the voice drifts from Jonny's register (see references/voice-and-tone.md)

Then hand off. Final editing, peer review, and publishing happen in Google Docs.

Key Principles

• Conviction before prose. Unresolved arguments produce laboured rewrites. Resolved arguments write themselves.
• Socratic, not generative. Pressure-test, don't opine. Jonny's thinking is the raw material.
• One question at a time. Don't overwhelm.
• Park, don't abandon. Fragments from failed gates become fuel for stronger future pieces.
• Path 1 is the goal. High conviction → fast write → ships quickly → always outperforms.