/copy-council — The masters review the copy

Convene the council. The greatest copywriters in history sit around the table, each in character, each with their real expertise, each speaking in their own voice. They review the copy on the table and tell us what works, what fails, what they'd change, and whether they would have signed off.

This skill is the natural counterpart to /copy-writing. The writing skill produces. The council critiques. Used together they form the loop — write, council, revise, ship.

This is not the same as /llm-council. That's an anonymised LLM-advisor pressure-test. This is a roleplayed historical council with named experts giving copywriting verdicts.

Step 1 — Identify the copy under review

The council needs a piece of copy in front of it. The piece is usually the most recent copy produced by /copy-writing in this conversation. Use it directly without asking.

If the chat does not contain a clear, recent piece of copy, ask once:

"Paste the copy you want the council to review — and the brief if it's not already in the thread."

Do not convene the council without:

• The copy itself (the artefact under review)
• The brief context — at minimum: channel, audience, goal. The council judges fit-for-purpose, not generic quality. A killer cold email is not a killer landing-page hero.

If only the copy is provided and the brief is missing, infer the channel from the format and ask once for audience + goal. Don't fabricate the brief.

Step 2 — Convene the Inner Five (always speak)

Each of the five always weighs in. Each is introduced by name, lifespan, and signature credential. Each speaks in their own voice. Each issues a verdict.

The five voices and how to write each

*David Ogilvy (1911–1999) — Founder, Ogilvy & Mather. Author of Confessions of an Advertising Man (1963) and Ogilvy on Advertising (1983). Built Rolls-Royce, Hathaway, Dove, Schweppes, American Express.*

Voice: Precise, slightly aristocratic, witty, allergic to vagueness. Talks in maxims and case histories. Reaches for specifics — "At 60 miles an hour the loudest noise in this new Rolls-Royce comes from the electric clock". Likely to say "the consumer is not a moron". Hates jargon and creative-for-creative's-sake. Will compliment a fact-anchored claim and behead a vague verb.

Lens: Specificity, headline strength, respect for the reader, research behind the claim, long copy where warranted.

*Eugene Schwartz (1927–1995) — Author of Breakthrough Advertising (1966). Direct-response copywriter and rare-book publisher. The most expensive copywriting book in history was his.*

Voice: Analytical, almost scientific, methodical. Diagnoses copy by stage of awareness and stage of market sophistication. Speaks in frameworks. Will name exactly which awareness stage the copy is targeting and whether it matches the reader's actual stage. Calm, surgical, never theatrical.

Lens: Awareness/sophistication match, whether the lead channels existing desire vs. tries to create it, whether the headline meets the prospect where their mind already is.

*Gary Halbert (1938–2007) — The Boron Letters, the Coat of Arms letter (millions of copies mailed). Founder of The Gary Halbert Letter. Coined "the starving crowd."*

Voice: Brash, plainspoken, contractions, mild profanity allowed but not theatrical. Reads aloud and tells you when it sounds like a brochure. Doesn't trust you until you've earned it on the page. Will strip out every line that doesn't pull its weight. Calls people "Slick" and "kid". Will reference his son Bond, his time in Boron, the starving crowd, and "the A-pile / B-pile" mailbox test.

Lens: Plainness, conversational rhythm, audience-first vs. brand-first, whether the offer would make it past the mailbox B-pile, the strength of the audience selection itself.

*Claude C. Hopkins (1866–1932) — Author of Scientific Advertising (1923). Lord & Thomas under Albert Lasker. Schlitz beer, Pepsodent, Quaker Oats, Goodyear.*

Voice: Sober, methodical, demonstration-driven. Talks in mechanism and proof. Sceptical of claims unanchored by reason. Will ask "why should the reader believe this?" Famous for preemptive claims — saying the obvious true thing about the product before any competitor bothered to. References Schlitz live-steam-bottle-washing as the archetype. Treats advertising as engineering.

Lens: Reason-why behind every claim, preemptive opportunities missed, demonstration vs. assertion, testability of the claim.

Bill Bernbach (1911–1982) — Co-founder of Doyle Dane Bernbach (DDB) in 1949. The Creative Revolution. Volkswagen "Think Small" and "Lemon," Avis "We try harder," Levy's Rye Bread.

Voice: Concise, conceptual, allergic to formula. Famous internal memo warning against "rigid rules" killing creativity. Will spot the unexpected angle the writer missed. Defends wit. Defends self-deprecation as a strength. Writes in short, declarative sentences. Will say things like "It's not what you say, it's how you say it" and mean it.

Lens: The unexpected angle, truth told entertainingly, whether the piece earns the reader's attention or demands it, the tonal sin of taking yourself too seriously.

Format for each Inner Five entry

[Full Name] ([birth]–[death]) — [signature credential, one line]

[2–4 sentences in their voice, judging the piece. Should sound like them, 
not like a generic critic. Quote a specific line of the copy if it 
illustrates the point.]

What he'd change: [one concrete, actionable revision — specific to a 
line, a claim, the headline, the CTA, the awareness stage, etc.]

Verdict: APPROVE / REVISE / KILL

The verdict is binary-ish — three states. APPROVE = would have signed off as-is. REVISE = the bones are right, the execution needs the named change. KILL = the strategic frame is wrong, start over.

Step 3 — Call the Honourables (selective)

Not every honourable speaks for every piece. Calling all five every time turns the council into a mob and dilutes the signal. Call 2–3 honourables based on the channel and audience of the piece.

Who speaks when

If the channel doesn't fit any of the above, call the two whose lenses apply most directly to the brief. If genuinely no honourable adds signal, call zero — the Inner Five plus a strong synthesis is enough.

The five honourables and their voices

*Joe Sugarman (1938–2022) — Founder of JS&A. BluBlocker sunglasses. Author of The Adweek Copywriting Handbook. Coined "the slippery slide."*

Voice: Mid-century-mail-order energy, conversational, charming, very practical. Reads each line for whether it pulls the reader into the next. Will quote his own slippery slide rule and test the copy against it line by line. Genuinely warm.

Lens: Whether each line earns the next, opening hook, sentence-to-sentence momentum.

*Helen Lansdowne Resor (1886–1964) — J. Walter Thompson. The first major female copywriter in American advertising. "A skin you love to touch" — Woodbury Soap, 1911.*

Voice: Refined, observant, warm but steely. Speaks to the reader's emotional reality before the product's spec sheet. Will catch a piece that treats a feeling-based purchase like a feature comparison. Cares about whether the reader sees herself in the copy.

Lens: Emotional reality of the buyer, whether the piece flatters spec over feeling, gendered tonal misfires.

*Dan Kennedy (b. 1954) — Direct-response marketing consultant. Founder of GKIC. Author of The Ultimate Sales Letter, The Ultimate Marketing Plan. Coined "message-to-market match."*

Voice: Plainspoken, blunt, no-nonsense Midwestern direct-response. Will ask whether the exact message in front of him matches the exact market it's aimed at. Will catch a 10% mismatch and call it a 50% miss. Hates fluff. Hates "brand voice" that won't sell. Will compliment specificity in the offer.

Lens: Message-to-market match, offer clarity, whether the piece would actually sell to a real human in a real moment.

*Drayton Bird (b. 1936) — Direct marketing legend. Ran Ogilvy Direct internationally. Author of Commonsense Direct & Digital Marketing. Ogilvy called him "the man who knows more about direct marketing than anyone in the world."*

Voice: British, dry, mildly amused, ruthlessly practical. Will defend the CTA above almost everything. Will quote response-rate experience. Has the patience of someone who has tested everything and seen which versions actually got the reply.

Lens: The CTA — its clarity, placement, language, primacy. Whether the piece has one ask or smuggles in a second.

*Bond Halbert (b. ~1970s) — Gary Halbert's son. Co-author of The Boron Letters. Active teacher of his father's methods.*

Voice: His father's son — same plainness, slightly more measured. Will quote his dad. Edits ruthlessly. Catches every padded line.

Lens: Editing — every line earning its place, length discipline, "would my dad have crossed this out?"

Format for each Honourable entry

Same as the Inner Five — name + dates + credential, 2–4 sentences in voice, one concrete change, verdict.

Step 4 — The Synthesis

The synthesis is the most valuable part of the council. Five named voices is colour. The synthesis is signal. Will needs to know: what do I do now?

Format:

THE SYNTHESIS
─────────────

Consensus: [1–2 sentences — where the council clearly agreed. The strongest signal.]

Conflict: [1–2 sentences — where they disagreed and why. Often a strategic 
choice the writer must make: e.g. Halbert wants to cut, Ogilvy wants more 
specifics. Name the tradeoff explicitly.]

Must-do changes (in priority order):
1. [the single most important revision, tied to the dominant verdict]
2. [the second]
3. [the third — only if it actually moves the needle]

Council ruling: SHIP / REVISE / RESTART

Ruling logic

• SHIP — at least 4 of the Inner Five issued APPROVE, no Honourable issued KILL, and the must-do list is genuinely cosmetic. Rare.
• REVISE — the dominant verdict is REVISE, and the must-do list is finite (≤3 changes) and actionable. This is the most common outcome.
• RESTART — 2 or more of the Inner Five issued KILL on the strategic frame (wrong awareness stage, wrong audience, wrong goal). The bones are wrong; rewriting won't fix it.

State the ruling in plain language with one sentence of justification. Do not soften it. The whole point of convening a council is to get the truth.

Step 5 — Deliver

Display in chat. No save unless Will asks. Format the response cleanly:

CONVENING THE COUNCIL
On: [one-line description — channel, audience, brief in a phrase]

──────────────────────────────────────────
THE INNER FIVE
──────────────────────────────────────────

[Ogilvy entry]

[Schwartz entry]

[Halbert entry]

[Hopkins entry]

[Bernbach entry]

──────────────────────────────────────────
THE HONOURABLES (called for this piece)
──────────────────────────────────────────

[Selected 2–3 entries]

──────────────────────────────────────────
THE SYNTHESIS
──────────────────────────────────────────

Consensus: …
Conflict: …

Must-do changes:
1. …
2. …
3. …

Council ruling: [SHIP / REVISE / RESTART]. [One-sentence justification.]

Do not wrap the council output in a code fence — Will may want to copy parts of it.

Step 6 — Optional iteration loop

If the ruling is REVISE, end the response with a single line:

Want me to apply the council's must-do changes and produce v2?

If Will says yes, do not re-trigger /copy-writing from scratch — that would re-do discovery and re-pick a voice. Instead, take the existing draft, apply the three must-do changes, display v2 inline, and ask whether to re-convene.

If the ruling is RESTART, end the response with:

The strategic frame needs to change before we write again. Want to revisit the brief?

If the ruling is SHIP, end with no follow-up question. The piece is done.

Non-negotiables (recap)

• Each council member is addressed by full name, lifespan, and signature credential — every time. No "Mr. Ogilvy says…" with no context.
• Each Inner Five member always speaks. Skipping one is not allowed.
• Honourables are called selectively (2–3 per piece) based on the channel/audience table. Calling all five every time dilutes signal.
• Each member speaks in their own voice — characterised, not caricatured. Halbert can swear; Ogilvy cannot. Hopkins demands evidence; Bernbach defends wit. Real critique through real voices.
• Each member ends with one specific, actionable change and a binary-ish verdict (APPROVE / REVISE / KILL).
• The synthesis is mandatory — consensus, conflict, top 3 must-do changes, one ruling. No floating critiques without a takeaway.
• Never fabricate a quote attributed to a real person. Channel their style, paraphrase their known principles, but do not invent a sentence and claim they said it. Per the never_fabricate_facts memory.
• The ruling must be stated plainly. SHIP / REVISE / RESTART. No "well, it depends."

Related

• .claude/skills/copy-writing/SKILL.md — the writing skill this council reviews
• /llm-council — the AI-advisor council (different tool — anonymised LLM panel for decisions, not roleplayed historical experts on copy)
• Memory: never_fabricate_facts.md — applies to attributed quotes; channel voice, do not invent quotations
• Memory: feedback_software_ideation_directness.md — the council exists to be direct; soft rulings defeat the purpose