Claude-First Copywriting: Long-Form That Actually Lands
20 Claude prompts engineered for long-form sales copy, nuanced narrative, strategic positioning, and the voice-matching work where Claude outperforms ChatGPT.
Long-Form Sales Copy
5 promptsLong-Form Sales Page (Claude-Optimized)
1/20<role>Senior direct-response copywriter with 15 years of experience</role> <task>Write a 2,500-word sales page for [product]</task> <product_details> - What it is: [describe] - Target: [audience] - Price: [amount] - Unique mechanism: [describe] - Proof: [describe social proof, results] </product_details> <framework>Problem-Agitate-Solution with embedded story</framework> <output_requirements> - Hook that earns the first 30 seconds - 4-6 persuasion passes (each with new angle) - Social proof integrated, not clumped - Guarantee + urgency - Clear CTA </output_requirements> Before writing, outline the emotional arc I want to create in the reader. Then write the page.
Writes structured long-form sales pages using XML-tagged prompts that Claude processes with high fidelity.
Pro tip: Claude responds especially well to XML-tagged prompts. Wrapping sections in <role>, <task>, <constraints> improves output quality 20-30% over prose prompts. Claude was trained on structured inputs.
VSL Script (Conversational)
2/20<role>You are a conversational copywriter who writes VSLs that sound like a friend talking, not a marketer pitching.</role> <task>Write a 10-minute video sales letter script for [product]</task> <voice>Conversational, specific, human. Use "I" and "you." Avoid "revolutionary," "game-changing," "unlock."</voice> <structure> 1. Hook (first 30 seconds) 2. Story of discovery 3. Problem agitation with specific customer pain 4. Solution reveal 5. How it works (3 elements) 6. Proof 7. Offer + bonuses 8. Urgency + guarantee 9. CTA </structure> <constraints> - Spoken cadence — short sentences - No filler - One big promise, many small proofs </constraints>
Scripts conversational VSLs using XML structure for natural spoken cadence and voice authenticity.
Pro tip: VSLs fail when they sound like ads. Claude excels at matching human cadence — feed it 2-3 minutes of your actual speaking style as reference, and it will write scripts that sound like you.
Long Email Sequence (7+ emails)
3/20Write a 7-email launch sequence for [offer]. Each email serves a specific purpose in the narrative arc: Email 1 (Day 1): Introduce the problem via a specific story Email 2 (Day 2): Deepen the problem + introduce the framework Email 3 (Day 3): Share a contrarian insight that reframes the situation Email 4 (Day 4): Case study showing transformation Email 5 (Day 5): Open cart — introduce the offer as the natural answer Email 6 (Day 6): Address the top 5 objections Email 7 (Day 7): Last call with scarcity + recap of transformation For each email: - Subject + preview text - 400-700 word body with consistent voice - Single CTA - Smooth transition to next email's theme Keep the narrative thread across all 7 emails. Use <voice_samples> I provide to match tone.
Writes 7-email launch sequences with narrative arc, per-email purpose, and voice consistency.
Pro tip: Email launch sequences live or die on narrative cohesion. Claude maintains multi-email consistency better than most models — give it the arc up front and it writes emails that feel like chapters, not isolated pitches.
Story-Driven Sales Page
4/20<role>Story-first copywriter who sells via narrative, not tactics</role> <task>Write a 2,000-word sales page that reads like a personal essay</task> <offer>[describe]</offer> <story>[describe the story arc — founder story, customer transformation, discovery moment]</story> <framework> - Open with the moment (scene, not summary) - Build tension with specific detail - Introduce the discovery - Show the transformation - Invite reader into the same journey (the offer as a bridge) - Close with a specific next step </framework> <voice>First person, conversational, specific. Avoid list-format. Flowing paragraphs.</voice> Before writing, identify the three emotional beats you want the reader to feel.
Writes story-driven sales pages with narrative arc, specific sensory detail, and emotional beat planning.
Pro tip: Claude is particularly strong at narrative coherence over 2,000+ words. Where ChatGPT drifts after 1,000 words, Claude holds the thread. For long-form story-driven copy, Claude is often the better choice.
Cold Sales Letter (Legacy-Style)
5/20<role>Old-school direct-response copywriter in the tradition of Eugene Schwartz, Gary Halbert, and David Ogilvy</role> <task>Write a 3,000-word cold sales letter for [product]</task> <target>[audience — sophistication level]</target> <sophistication_level>[1-5 — 1=naive, 5=jaded market]</sophistication_level> <mechanism>[unique mechanism or angle]</mechanism> <structure> 1. Arresting headline (one that stops on page) 2. Deck copy that extends the headline 3. Opening story 4. Big promise 5. Mechanism revealed 6. Proof elements woven throughout 7. Offer 8. Guarantee 9. Close with P.S. </structure> <constraints>Every sentence must earn the next. No padding. Rhythm matters.</constraints>
Writes classic long-form cold sales letters using XML-structured prompts rooted in Schwartz/Halbert tradition.
Pro tip: Direct-response copy has a rhythm that modern AI tends to flatten. Claude can preserve it IF you give it structural and stylistic guardrails. Tell it to channel specific copywriters — the output improves dramatically.
XML tags are just the start. Learn the full Claude workflow.
A growing library of 300+ hands-on AI tutorials covering Claude, ChatGPT, and 50+ tools. New tutorials added every week.
Positioning & Strategy
5 promptsBrand Positioning Doc
6/20<task>Build a positioning document for [product/company]</task> <inputs> - What we sell: [describe] - Who we sell to: [describe] - Competitors: [list] - Our unique insight: [describe] - Current positioning (if any): [describe] </inputs> <output_structure> 1. Target customer (specific segment) 2. Alternative (what they'd use instead — status quo or competitor) 3. Key benefit (the transformation we deliver) 4. Category (who we compete with in customer's mind) 5. Unique value (why us vs alternatives) 6. Proof points (evidence supporting the claim) 7. Messaging pillars (3-5 repeatable talking points) </output_structure> Think step-by-step before finalizing. Flag any weak inputs.
Builds structured positioning docs with target, alternative, benefit, category, unique value, and messaging pillars.
Pro tip: Claude's strength is strategic thinking. A positioning doc isn't just copy — it's the framework everything else flows from. Get positioning right before writing any landing page or ad copy.
Message House / Talking Points
7/20<task>Build a "message house" — the structured set of talking points every piece of content reinforces</task> <brand>[describe]</brand> <positioning>[paste positioning statement]</positioning> <output> - Umbrella message (one sentence that captures what we stand for) - 3 pillars (supporting themes) - Per pillar: 3 proof points with specific examples - Per pillar: 2 customer pain points addressed - Per pillar: 5 specific talking points / phrases to weave into all content </output> <constraints> - Language that sounds human, not corporate - Phrases that could land verbatim in a podcast, tweet, or sales call </constraints>
Builds message houses with umbrella, pillars, proof points, and verbatim-usable talking points.
Pro tip: Most brands lack message discipline — every piece of content says something slightly different. A message house forces coherence. Use it to brief writers, align sales teams, and QA content at scale.
Competitive Differentiation Brief
8/20<task>Analyze competitive positioning and identify differentiation angles</task> <us>[describe our product, target, key features]</us> <competitors>[list 3-5 with brief descriptions]</competitors> <output> 1. A 2x2 matrix with dimensions that matter in our market 2. Where each player sits on the matrix 3. 3 differentiation angles we could own 4. For each angle: how to position, proof points, messaging implications 5. Angles to avoid (already owned by stronger competitor) 6. The positioning we should own (recommended) with reasoning </output>
Analyzes competitive landscape with 2x2 matrix, differentiation options, and recommended positioning.
Pro tip: Differentiation is what you don't do, more than what you do. The strongest positioning often sacrifices broad appeal for narrow dominance. Claude is good at helping identify the sacrifices worth making.
Customer Voice Mining
9/20<task>Extract customer voice patterns from [source — reviews, interviews, sales calls, support tickets]</task> <raw_content>[paste or describe]</raw_content> <output> 1. Top 5 pain points in customer words (verbatim where possible) 2. Top 5 desires / outcomes in customer words 3. Objections and concerns 4. Language patterns (specific phrases customers use) 5. Emotional triggers (fear, shame, pride, hope) 6. How customers describe the "before" vs "after" states 7. Words they use to describe competitors (negative + positive) </output> Deliver this as a copywriting reference I can draw from.
Mines customer voice from reviews/calls/tickets for verbatim language, triggers, and transformation narratives.
Pro tip: The highest-converting copy uses customer language, not marketer language. Mining reviews and calls for actual phrases is the single most valuable copywriting exercise. Claude is excellent at extracting patterns from large amounts of text.
Offer Architecture
10/20<task>Design an offer architecture for [product/service]</task> <inputs> - Core deliverable: [describe] - Audience sophistication: [describe] - Price objections: [describe] - Competitor offers: [describe] </inputs> <output> 1. Core offer (what they get, structured) 2. Bonuses (3-5 that increase perceived value) 3. Guarantee (risk reversal) 4. Urgency mechanism (real, not fake) 5. Price anchoring strategy (what to compare to) 6. Stack presentation (how to show value) 7. Payment options (single, split, subscription) </output> <constraints>Every bonus must solve a specific secondary problem. No fluff bonuses.</constraints>
Designs offer architecture with bonuses, guarantee, urgency, and stack presentation with no-fluff rule.
Pro tip: Weak offers die even with great copy. Strong offers convert even with mediocre copy. Spend more time on offer architecture than on headlines — leverage is upstream.
Content & Thought Leadership
5 promptsEssay Ghostwriting
11/20<role>Ghostwriter capable of matching a specific writer's voice</role> <task>Write a 1,500-word essay on [topic] in my voice</task> <voice_samples>[paste 3-5 of my best essays, 200-500 words each, for voice reference]</voice_samples> <position>[my thesis / angle]</position> <structure> - Hook anchored in a specific moment - Setup — why this matters now - Main argument with 3-4 supporting points - Strongest counter-argument acknowledged - Practical takeaway - Close that lingers </structure> <constraints> - Match the sentence rhythm of my voice samples - Use my specific quirks (if any) without over-indexing - Avoid AI-tells I flag </constraints>
Ghost-writes essays by matching voice samples with structural scaffolding and anti-AI-tell constraints.
Pro tip: Claude's voice matching is remarkable when fed real samples. Don't just describe your voice — paste it. 3-5 real samples gets 90% of voice right; description alone gets maybe 40%.
Narrative Case Study
12/20<task>Write a 2,000-word narrative case study about [client transformation]</task> <inputs> - Client context (who, industry, starting situation): [describe] - The challenge: [describe] - Interventions: [list 5-7 with specifics] - Results: [describe with numbers] - Key lessons: [describe] </inputs> <framework> - Lead with tension (opening scene or surprising result) - Establish stakes - Move through the interventions as chapters - Include specific moments of friction and resolution - Close with transferable principles for readers in similar situations </framework> <voice>First or third person, journalistic + warm. No corporate case-study template feel.</voice>
Writes narrative case studies with tension-driven structure, friction moments, and transferable principles.
Pro tip: Corporate case studies ("Situation-Action-Result") bore readers. Narrative case studies ("Once upon a time, a founder was stuck...") hold attention. Claude excels at the narrative mode specifically.
Whitepaper Writer
13/20<role>Senior analyst writing for a technical executive audience</role> <task>Write a 3,000-word whitepaper on [topic]</task> <thesis>[your central argument]</thesis> <evidence>[data, research, case studies available]</evidence> <audience>[describe — technical level, concerns, decisions they're making]</audience> <structure> - Executive summary (readable in 60 seconds) - Introduction framing the issue - 3-4 analytical sections with data - Implications for the audience - Recommendations - Appendix with methodology/references </structure> <voice>Analytical but readable. Assertions backed by specifics, not hand-waving.</voice>
Writes technical whitepapers with executive summary, analytical sections, and evidence-based assertions.
Pro tip: Whitepapers win trust when they're rigorous but readable. Claude handles the analytical tone without drifting into academic impenetrability — it can maintain precision while staying approachable.
Interview Content Package
14/20<task>Turn this transcript into multiple content pieces</task> <transcript>[paste — 30-90 min interview]</transcript> <outputs> 1. Long-form article (2,000-3,000 words) weaving insights together with narrative 2. 5 tweet-thread hooks with 3-5 tweets each 3. LinkedIn post (200-300 words) highlighting one idea 4. Newsletter-ready excerpt (500-800 words) 5. Pull quotes (5-10 quotable lines) 6. Suggested titles for each format </outputs> <voice>Match the interviewee's voice in direct quotes; use editorial voice for connecting tissue.</voice>
Turns interview transcripts into multi-format content packages with voice-matched quotes and editorial framing.
Pro tip: One good interview = 10+ pieces of content. Claude is excellent at repurposing long transcripts into multi-format packages. Feed it the transcript + desired formats + voice preservation rules and you get a week of content from one conversation.
Tweet / Thread Rewriter
15/20<task>Turn this essay into a 10-tweet thread</task> <essay>[paste]</essay> <constraints> - Tweet 1: hook that works standalone - Each tweet: single idea, max 280 chars, readable on phone - Build momentum — tweet 5 is the peak insight - Include 1-2 specific examples - Final tweet: CTA (follow, reply, click) - Avoid "but here's the thing" and AI-tell phrases </constraints> <voice>[paste my previous best-performing tweets for voice reference]</voice>
Rewrites essays into 10-tweet threads with voice matching, momentum pacing, and hook optimization.
Pro tip: Thread conversion rate depends on tweet 1 and tweet 5. Tweet 1 earns the read; tweet 5 must deliver the peak insight. Claude is good at thread pacing if you tell it explicitly to structure for this.
Editing & Refinement
5 promptsLine Edit for Pace
16/20<task>Line-edit this copy for pace and rhythm</task> <copy>[paste]</copy> <edits_to_apply> 1. Cut throat-clearing ("It's important to note", "Let me explain") 2. Vary sentence length (mix short and long) 3. Remove hedges that weaken ("might", "perhaps", "kind of") 4. Tighten passive to active 5. Remove redundant modifiers 6. Preserve voice — do not rewrite into your voice </edits_to_apply> <output> - Edited version - List of top 10 specific changes with reasoning - Anything cut that might matter (flag for my review) </output>
Line-edits copy for pace, active voice, and removal of hedges while preserving author voice.
Pro tip: Most first drafts are 20-30% too long. Line-editing for pace often doubles readability without losing content. Claude is a disciplined line editor if you give it clear rules — otherwise it rewrites into its own voice.
Voice Audit
17/20<task>Audit this copy against my voice</task> <my_voice_samples>[paste 3 examples of my best work]</my_voice_samples> <draft>[paste the draft to audit]</draft> <output> 1. Voice match score (1-10) 2. Where draft deviates from my voice (specific passages) 3. What's slightly "off" (phrases that don't sound like me) 4. AI-tell indicators (words I'd never use) 5. Rewrite suggestions for the top 5 voice mismatches 6. Strengths to preserve in the edit </output>
Audits drafts against voice samples with match score, deviations, AI-tells, and targeted rewrites.
Pro tip: Even when you draft yourself, Claude's voice audit catches blind spots. Your brain glosses over inconsistencies — a dispassionate audit catches them. Run this on any content going public.
Headline Optimization (10 Variants)
18/20<task>Generate 10 headline variations for [context]</task> <context> - Offer: [describe] - Audience: [describe] - Primary emotion: [describe] - Destination after click: [describe] </context> <output> - 10 headlines using different formulas (curiosity, specific number, contrarian, benefit, question, etc.) - For each: predicted performance category (high/mid/low CTR) - For each: which audience segment it would hook - Top 3 picks with reasoning - Top pick formatted for A/B test </output>
Generates 10 headline variants across formulas with performance predictions and A/B test recommendation.
Pro tip: Always generate 10+ headlines. Your first 5 are obvious; the winning headline is usually #7-15. The only way to beat your first instinct is to force yourself past it.
Simplification Pass
19/20<task>Simplify this copy to an 8th-grade reading level without losing substance</task> <copy>[paste]</copy> <rules> - Keep sentence length mostly under 20 words - Replace jargon with plain terms - Remove compound-clause sentences - Preserve technical accuracy - Flag anything that genuinely requires expert terminology </rules> <output> - Simplified version - Flesch-Kincaid estimate before/after - Top terms replaced with plain equivalents - Anything I should NOT simplify (explain why) </output>
Simplifies copy to 8th-grade readability while preserving technical accuracy and flagging necessary jargon.
Pro tip: Simpler copy converts better — even with expert audiences. Intelligence doesn't equal enjoyment of complex prose. Even C-level executives prefer 8th-grade reading in marketing. Simplification is rarely a mistake.
Cross-Persona Rewrite
20/20<task>Rewrite this copy for a different persona</task> <original>[paste — copy as it exists now]</original> <original_persona>[who it was written for]</original_persona> <new_persona>[who I'm now targeting — role, pain points, decision criteria, tone they respond to]</new_persona> <output> - Rewritten copy in new persona's language - Specific word swaps made - Argument order changes (what matters differently to them) - Proof points swapped (what credibility signals work for this persona) - Anything I should drop vs add for this persona </output>
Rewrites copy for different personas with word swaps, reordered arguments, and persona-fit proof points.
Pro tip: The same product often sells to 3-4 distinct personas with different language, concerns, and proof preferences. Rewriting per persona 2-3× conversion vs generic "all audience" copy.
Frequently Asked Questions
Prompts are the starting line. Tutorials are the finish.
A growing library of 300+ hands-on tutorials on ChatGPT, Claude, Midjourney, and 50+ AI tools. New tutorials added every week.
7-day free trial. Cancel anytime.