A landing page isn’t a sales page. Visitors scan the hero for about 5 to 6 seconds before deciding to scroll or bounce, and Nielsen Norman Group eye-tracking found content above the fold draws 84% more attention than anything below it. These 18 prompts focus the model where that attention lands: hero clarity, scannable benefits, single-action CTAs, and an FAQ that handles real objections instead of marketing fluff.
TL;DR
- 18 copy-ready prompts, one per landing-page section: hero headline, subhead, CTA, value prop, benefits, “how it works”, social proof, pricing, FAQ, final CTA, plus rewrite/audit prompts for an existing page.
- Best model (as of June 2026): Claude Opus 4.7 or Sonnet 4.6 for the lowest-edit, most human-sounding draft; GPT-5.5 when you also want CTR-ranked variants and quick ideation; Gemini 3.1 Pro if you paste in a long page and want it audited in one shot.
- Benchmark to beat: the median landing-page conversion rate is 6.6% (Unbounce, 41,000 pages); strong pages clear 10%, and headline rewrites alone have lifted conversions 27–104% in A/B tests.
- The fix that matters most: copy is downstream of positioning. Lock your one-paragraph anchor first, then run these section by section.
Pick a model first
| Model (June 2026) | Best for here | Note |
|---|---|---|
| Claude Opus 4.7 / Sonnet 4.6 | Lowest-edit, human-sounding hero + benefit copy | Strongest “doesn’t read like AI” voice; 1M-token context for long briefs |
| GPT-5.5 | CTR-ranked variants, micro-copy tables, fast ideation | ChatGPT default; good at “rank these by predicted conversion” |
| Gemini 3.1 Pro | Auditing a full pasted page in one pass | 1M context; handy for the rewrite/audit prompts (#12, #18) |
All three run free tiers good enough to test these. Paid plans only matter once you’re iterating dozens of variants per page. See our AI landing page tutorial for the end-to-end workflow.
What these prompts solve
Most landing pages fail at the hero: vague headline, no mechanism, CTA that says “Get started” but doesn’t say what starts. These templates split a landing page into discrete sections and write each one with the constraints that actually convert — outcome-led headlines, mechanism-completing subheads, micro-copy that reduces risk, FAQs that answer real objections.
Who this is for
Indie devs launching SaaS or apps, founders writing copy themselves before they hire, growth marketers iterating on hero variants for A/B tests, product designers stress-testing copy alongside layout, anyone whose current landing page describes the product instead of the outcome.
When not to use these prompts
Skip them when you don’t yet know your audience and outcome — copy can’t fix positioning. If the hero keeps reading as four different things, you probably need to lock the one-paragraph anchor first using our AI positioning statement workflow before more rewrites. Skip them for content marketing pages (those are blog posts; use blog outline prompts). And don’t paste long product specs into a hero prompt — the model writes feature copy from feature input. Give it the outcome you sell, not the feature list.
Prompt anatomy / structure formula
A landing-page copy prompt should always carry six elements:
- Product one-liner: what it is in 12 words or fewer.
- Target user: a single concrete persona, not “small businesses”.
- Outcome: the after-state, expressed as the user’s gain (“ship a draft in 5 min”, not “AI-powered drafting”).
- Mechanism: how the product delivers — one sentence, the “OK but how?” answer.
- Section type: hero / subhead / benefits / social proof / pricing / FAQ / final CTA.
- Format constraint: variant count, max words, banned phrases (“leverage”, “empower”, “best”, “ultimate”).
Best for
- Hero headline + subheadline pairs
- Value-proposition refinement
- Conversion-focused CTAs and micro-copy
- Benefits sections (3-bullet format)
- “How it works” 3-step blocks
- Social proof and testimonial framing
- Pricing-tier one-liners
- FAQ sections that handle real objections
18 copy-ready prompt templates
1. Hero headline that names the outcome
My product: {one-liner}. Target user: {audience}. Outcome: {after-state}. Write 12 hero headlines: 3 outcome-led, 3 pain-led, 3 "for [audience]" framings, 3 contrarian. Each ≤14 words. No "best", "ultimate", "AI-powered".2. Subheadline that completes the headline
My headline: "{headline}". Write 5 subheadlines that answer "OK, but how?" in one sentence. Each must name the mechanism (not just restate the outcome). Each ≤20 words.3. Hero CTA button + micro-copy
Hero CTA for {product}. Goal: {visitor action}. Write 8 button variants (≤4 words) and 5 supporting micro-copy lines (e.g., "Free for 7 days. No credit card."). Pair them for testing — output as a table of (button, micro-copy, reduces-which-risk).4. Value proposition in 30 words
Product: {paste 200-word description}. Write a single 30-word value proposition with this shape: [who it's for] + [outcome they want] + [how this product gets them there] + [why it works when others don't]. Output 5 variants ranked best to worst with reasoning.5. Benefits in 3 bullets
Write 3 benefit bullets for {product}. Format per bullet: "(action verb) (specific outcome) — (one-line on the mechanism)." No "leverage" or "empower." Keep each bullet ≤16 words.6. “How it works” in 3 steps
Write a 3-step "How it works" section for {product}. Each step: 4-word title, 1 sentence of explanation, 1 micro-screenshot caption. Voice: human, not robotic. End the block with a one-sentence summary of what the user has after step 3.7. Social proof block
I have these customers / numbers / quotes (paste below). Write a social-proof block: 1 stat line, 3 testimonial framings (≤25 words each), 1 logo-row caption. Strip generic praise ("great product!", "love it!"). Pull the most specific phrase from each quote.
{paste}8. Pricing one-liner per tier
My pricing: {tier 1: name + price + who it's for}, {tier 2: …}, {tier 3: …}. Write 3 one-line summaries (≤14 words each) above each tier that nail "who this tier is for and what they get". No "Most popular!" badges.9. FAQ that handles real objections
Top 6 questions visitors have about {product}: {Q1}–{Q6}. Write FAQ answers. Each: direct answer in 1 sentence, then 2 sentences of context. No "Great question!". If the honest answer is unflattering, give it anyway and add one line on what we're doing about it.10. Final CTA section
Write a final-page CTA section for {product}. Include: 1 hero re-statement (rephrased, not copy-pasted), 1 risk-reversal line (free trial / no card / cancel anytime), 1 button copy, 1 alternative low-commitment option (e.g., "Read the docs first"). Voice: confident, not pushy.11. Meta description for the page
Write 5 meta-description variants (≤150 chars) for landing page "{url}" describing {product}. Each: name audience, name outcome, name the mechanism. Avoid "best", "ultimate", "revolutionary". Output as a numbered list ranked by predicted CTR.12. Hero rewrite for an existing weak page
Below is the current hero copy. Audit it against these criteria: outcome-led / mechanism-clear / specific audience / ≤14 word headline. Then rewrite 3 variants that fix every miss. Show before → after for each line.
Current hero: {paste}13. Pre-launch waitlist page hero
Pre-launch waitlist for {product}. Outcome: {what users get when they sign up}. Write hero copy that earns an email without overselling: headline (≤12 words), one-sentence what-it-is, one-sentence what's-different, CTA button + micro-copy that sets honest expectations on timing.14. Conversion-focused CTA rewrite
Current CTA: "{paste}". It feels generic. Rewrite 6 variants that specify the visitor's next action AND the value they get on click (e.g., "Start a 5-minute setup → free workspace"). Each ≤6 words. Output as table with predicted-conversion notes.15. Objection-handling copy block
Top 3 objections visitors have about {product}: {O1, O2, O3}. Write a "Skeptics, this is for you" copy block on the page that handles all three head-on. 3 short paragraphs (≤60 words each), one per objection. No defensive tone — concede where honest, then pivot.16. Audience-switcher section
{Product} serves 3 distinct audiences: {A1}, {A2}, {A3}. Write a switchable section (tabs UI) — one copy block per audience: headline, 3 bullets, one mini-CTA per tab. Keep total length per tab equal. No overlap in bullets across tabs.17. Comparison-table copy (“us vs them”)
Compare {product} vs {competitor} on 6 dimensions. Output a table — 6 rows, 3 columns (dimension, us, them). Each cell ≤8 words. No bashing — when their option is better on a dimension, say so honestly and explain the trade-off.18. Above-the-fold rewrite for a specific outcome
I want visitors to do exactly one thing above the fold: {desired action}. Rewrite my above-the-fold copy (headline, subhead, primary CTA, secondary CTA, single sentence of micro-copy) so every element pushes toward that one action. Cut anything that doesn't. Output before → after as a table.
Current ATF: {paste}Common mistakes
- Hero copy that describes the product, not the outcome. “AI-powered scheduling” tells a visitor nothing they care about; “10 hours back in your week” does.
- CTAs without micro-copy. A “Get started” button alone leaves visitors uncertain about cost, commitment, or what happens next.
- FAQ stuffed with marketing. “Why is your product the best?” is not a question anyone has. Pull from support tickets and sales calls instead.
- Stacking 5 CTAs on one page. Pick one primary action; secondary CTAs only when they reduce a specific risk.
- Benefits written from inside the company. Internal feature names (“smart sync”) instead of outcomes (“never lose a draft again”).
- Comparing yourself only on dimensions you win. Honest comparisons (template #17) build more trust than rigged ones.
- Hero changes without changing the underlying offer. Copy is downstream of positioning; rewriting alone won’t fix bad fit.
How to push results further
- Pass the target visitor’s job-to-be-done in every prompt, not just the product description. The same product writes different copy for different jobs.
- Demand variant counts (5, 8, 12). One option lets the model settle for “fine”; ten forces real range.
- Pair every CTA with a risk-reversal micro-copy line generated in the same prompt. They convert better as a unit than individually.
- For hero copy, generate before-the-fold AND first-scroll copy together. Visitors who scroll past hero need the next beat to land, not a generic features grid.
- Compose hero + subhead as pairs (template #2). Asking for them separately makes them not match.
- For FAQ, pull real questions from support tickets, sales calls, or churn surveys (template #9 + faq writing prompts). Invented FAQs read like marketing.
- After generating, rate each variant against the 6 anatomy elements (audience / outcome / mechanism / format / variant count / banned phrases). The model often skips one, usually mechanism.
- Write the page for one traffic source, not all of them. Unbounce data shows email visitors convert about 77% better than paid-search visitors when the page is built specifically for that campaign. Tell the model which source the visitor arrived from and what they already know.
FAQ
- Which model should I run these in? As of June 2026, Claude Opus 4.7 or Sonnet 4.6 give the most human, lowest-edit hero and benefit copy. GPT-5.5 is stronger when you want variants ranked by predicted CTR or quick ideation across sections. Gemini 3.1 Pro is handy for auditing a long pasted page (#12, #18) in one pass. All have a free tier good enough to test.
- What’s a good conversion rate to aim for? The median across 41,000 landing pages is 6.6% (Unbounce). Anything above 10% is strong; 15%+ usually means warm email traffic or a tightly targeted offer. Mobile typically converts at 60–70% of the desktop rate, so test the hero on a phone.
- What’s the difference between this and FAQ writing prompts? FAQ-writing prompts (here) cover help-center and product-page FAQs in depth — long-tail, multi-page. Landing-page FAQ is one section, 5–8 questions, optimized for conversion.
- How many hero variants should I A/B test? Two pairs of (headline, subhead). More variants split traffic too thin for meaningful signal in a typical indie-scale audience.
- Should I write the hero before the rest of the page? Yes. Everything below the fold derives from the promise the hero made. Writing benefits first leads to misaligned copy.
- Do I need testimonials to use the social-proof template? No. The template works with stats, customer logos, or specific use-case quotes from beta users. Generic praise is the only thing that doesn’t work.
- Can these prompts handle a “long-form” sales-letter style page? No — those need different rhythm and length. Use sales copy prompts for that pattern.
Related
- FAQ writing prompts — fill the FAQ section with real, objection-handling questions
- Sales copy prompts — when the page needs a long-form sales letter, not a scannable landing
- Product description prompts — for product-detail pages on e-commerce, not landings
- Landing page hero copy prompts — deeper drill on the hero block alone
- Landing page section prompts — section-by-section variations
- AI landing page tutorial — full workflow from positioning brief to ship-ready page
- Landing page copy use case — end-to-end example with before/after
- About Page Copy Prompts: 12 Templates for Honest, Human Pages
- Generate Landing Page Section Ideas With AI