The hero block gets roughly 3 to 5 seconds to earn the scroll or lose the visitor — and analysis across thousands of A/B tests puts headline rewrites among the highest-impact changes you can make, with reported lifts of 27 to 104% (ALM Corp, 2026). These 15 prompts focus only on above-the-fold copy: the headline that says what you sell and why it matters, the subhead that closes the comprehension gap, the CTA that turns intent into a click, and the trust line that takes the edge off.
TL;DR
- Run these on Claude Opus 4.7 when you need a held brand voice (it resists padding and false enthusiasm) or GPT-5.5 when you want persuasion structure and social-proof scaffolding. Both cost $20/mo (Claude Pro / ChatGPT Plus) as of June 2026 — you do not need a $49/mo wrapper.
- Keep the desktop H1 under ~8 words; keep mobile at ≤ 7. Long headlines lose comprehension inside the 3 to 5 second window.
- Match the hero to the ad hook that drove the click. A Conversion Rate Experts case study attributes a 212% lift to fixing message match alone.
- Test in pairs (headline only, then CTA only) before testing 3+ variables, so you can attribute the lift.
Who this is for
CRO leads, landing-page designers, performance marketers tuning hero blocks for paid traffic, Shopify operators redesigning home and PDP pages, and founders rewriting their own first hero.
When not to use these prompts
Skip these for long-form content pages where the hero is editorial in tone, or for app-store listing variants that follow platform-specific copy rules (Apple and Google cap subtitle and promotional-text lengths separately).
Which model to run these on (June 2026)
You do not need a dedicated copywriting SaaS to get great hero copy — those tools largely wrap the same foundation models behind a template UI. Run the templates below directly:
| Model | Best for hero copy | Price (June 2026) |
|---|---|---|
| Claude Opus 4.7 | Brand-voice fidelity; spare, direct registers; B2B credibility | Claude Pro $20/mo ($17 annual) |
| Claude Sonnet 4.6 | Fast bulk variant generation at lower cost | Included in Pro; API $3/$15 per 1M |
| GPT-5.5 | Persuasion structure, urgency, social-proof scaffolding | ChatGPT Plus $20/mo |
| Gemini 3.1 Pro | Long-context competitor teardowns (1M context) | Google AI Pro $19.99/mo |
Practical split most teams land on: draft variants in Claude Opus 4.7 for voice, then ask GPT-5.5 to stress-test the same set for “where would a skeptical buyer bounce.” Paste competitor pages into Gemini 3.1 Pro when prompt 15 needs you to load several full pages at once.
Prompt anatomy / structure formula
A hero copy prompt should always carry six elements:
- Role: who the AI plays (luxury copywriter / Amazon listing strategist / DTC brand voice / paid-ads hook writer).
- Context: product, brand voice, target buyer, platform, price tier, season — anything that shifts copy.
- Goal: one concrete deliverable — 5 bullets, a 150-word hero, 13 tags, 10 hook lines, a refund reply.
- Constraints: must / must-not (FTC claims, banned words, character limits, tone, no emoji, no superlatives).
- Output format: numbered list, table, JSON, or labeled blocks so you can paste straight into the seller backend.
- Examples / signal: 1-2 reference lines you like, or anti-examples (“not like this competitor”).
Best for
- Paid-traffic landing page hero
- Home page above-the-fold
- Product detail page hero block
- Pricing page hero
- Hero A/B test variant generation
15 copy-ready prompt templates
1. Headline + subhead default scaffold
Works for 70% of DTC and SaaS heroes; iterate from here.
You are a landing-page copywriter for {brand}. Write 5 hero block variants for {product / page}. Each variant: 1 headline (max 10 words, names the buyer outcome), 1 subhead (max 18 words, says how we deliver it differently), 1 CTA button copy (max 4 words). Mobile-first.Variables to swap: brand, product, primary outcome, differentiator
Optimization: If variants overlap, add: “Each variant must lead with a different buyer hesitation: time, money, trust, status, ease.”
2. Outcome-led headline (not feature-led)
Write 8 outcome-led hero headlines for {product}. Each names what the buyer GETS, not what we built. ≤ 10 words. No "Introducing", no "The new way to". Examples: "Sleep through the third feeding." vs "Our new bedside lamp." pick the outcome version.3. Subhead that closes the comprehension gap
For the headline "{paste headline}", write 5 subhead variants that explain HOW we deliver the headline promise. Each ≤ 22 words. Add one credibility signal ({"in under 3 minutes", "without the daily app check-in", "designed with sleep coaches"}). Never repeat the headline.4. Hero with primary + secondary CTA
Write a hero block with TWO CTAs: primary ({4 words max, decision verb}), secondary ({5 words max, low-commit option}). Pair them so the secondary catches buyers not ready for primary ({"See How It Works", "Watch 60-Second Demo"}). Same hero, two paths.5. Trust-line beneath the hero
For the hero block above, write 5 trust-line variants ≤ 8 words each: one numbers-based ({"Joined by 12,400 home cooks"}), one press-based ({"As seen in Wired and Wired"}), one guarantee-based, one founder/credential-based, one social proof rating. All facts must be verifiable.6. Mobile-first 320px hero
For a mobile-first hero (320px portrait, single column), write hero copy that fits without truncation: headline ≤ 7 words, subhead ≤ 14 words, CTA ≤ 3 words. Confirm each line will not break awkwardly at 320px. Use shorter words where possible.7. Hero image direction (not just copy)
For the hero copy "{paste headline + subhead}", write 5 hero image direction briefs ≤ 25 words each. Each direction must show the buyer-outcome moment, not the product on a white background. Direction includes: subject, action, setting, time of day, mood.8. Paid-traffic hero (ad-match)
Below is the paid-ad hook that drove traffic to this page: "{paste ad hook}". Write a hero block that DOES NOT repeat the ad but DELIVERS on its promise immediately. Headline + subhead + CTA. Buyer must feel: "yes, this is the page I clicked for." within 1 second.9. Pricing-page hero
For a pricing page hero, write 5 variants. Each headline names the choice ({"Pick a plan, change anytime", "Three ways to start"}), subhead resolves the price anxiety, CTA points to the first tier card. No "From $X/mo" gimmicks unless that is the real price.10. Hero for skeptical buyer
Write 5 hero variants for a skeptical buyer segment ({"tried similar products before", "burned by previous purchase"}). Each headline acknowledges the skepticism without grovel. Subhead names one specific reason this is different. CTA respects the hesitation ({"Test One Risk-Free", "Read 3 Reviews First"}).11. Hero A/B test matrix (6 variants)
For {product} hero, generate a 6-variant A/B test matrix: 2 outcome-led, 2 differentiator-led, 2 social-proof-led. For each: headline, subhead, CTA, the conversion hypothesis being tested, the metric to watch. Output as table.12. Localized hero (US to JP / DE / UK)
Adapt this US hero block for 3 markets: Japan (politer, more spec-driven), Germany (precise, less emotional), UK (drier, more understated). For each, output headline + subhead + CTA + 1-line note on the cultural translation choice.
{paste US hero}13. Hero rewrite for above-the-fold density
Use when current hero is too cluttered to convert.
Below is my current hero block. It has too many elements above the fold. Recommend: which 3 elements to keep, which to cut, which to demote below the fold. Then write the simplified hero (headline, subhead, CTA, trust line).
{paste current hero}14. Hero for returning visitor
Write 4 hero variants for a returning visitor (browsed before, did not convert). The hero should acknowledge the return without being creepy ({"Welcome back. Ready to pick a plan?"}). Lighter CTA than first visit. Subhead may reference what changed since last visit if true.15. Hero rewrite from competitor-teardown insight
Below are 3 competitor heroes in {category}. Identify the pattern they all share. Write a hero for {my brand} that BREAKS that pattern intentionally. Headline + subhead + CTA. The break must be a genuine differentiator, not just contrarian for contrarian sake.
{paste 3 competitor heroes}Common mistakes
- Headline that names the product instead of the buyer outcome — feature-led heroes underperform across categories.
- Subhead that paraphrases the headline — wastes the one chance to close the comprehension gap.
- Hero with 3+ CTAs — splits intent and dilutes click-through.
- Mobile hero that needs scrolling to see the CTA — kills paid-traffic conversion.
- Trust line that uses unverifiable claims (“loved by thousands”) — buyers smell vagueness.
- Hero image of the product on a white background when the brand promises an outcome moment — copy and image must promise the same thing.
- Letting AI rewrite hero without knowing the paid-ad hook that drove the traffic — message-match failure is the #1 landing-page killer.
How to push results further
- Treat hero as one composition: headline + subhead + CTA + trust + image must agree on the same promise.
- Always design at 320px first; if it does not work on mobile, it does not work.
- Match hero copy to the paid-ad hook that drove the click. Message-match is the single largest preventable conversion killer; one Conversion Rate Experts case study attributes a 212% lift to fixing it alone, and unmatched homepage traffic can convert 4 to 5x worse.
- Strip the navigation bar on a paid-traffic hero. Pages with a nav menu convert roughly 10 to 15% lower than equivalent pages without one — the lowest-effort win on most landing pages.
- Lead the headline with the buyer, not the product — first-person and second-person both outperform feature names.
- Run hero A/B tests in pairs (headline only, CTA only) before testing 3+ variables; isolate the lift driver.
- Keep a hero library by category and traffic source; what wins for warm email differs from cold paid social.
- Refresh hero copy whenever you change the offer, the ad creative, or the target audience — they ripple.
FAQ
- How long should a hero headline be?: Aim for under ~8 words on desktop (high-performing H1s often run ≤ 44 characters) and ≤ 7 on mobile. Longer headlines lose comprehension inside the 3 to 5 second scan window.
- Which AI model writes the best hero copy?: For brand-voice fidelity, Claude Opus 4.7 — it holds a spare, direct register without padding sentences to hit a word count, and tests well on B2B credibility. For persuasion structure and built-in social proof, GPT-5.5. Both are $20/mo as of June 2026, so there is no need to pay for a copywriting wrapper that runs the same models underneath.
- Should the hero have a video?: Video heroes can lift engagement but slow load time, which hurts Core Web Vitals and Quality Score. Use only if the product genuinely benefits from motion (apparel fit, electronics demo) and the file stays under 2MB.
- How many heroes should I test?: Start with 3 variants. Add a 4th only if the first round produces no clear winner. More than 4 splits the traffic so thin that most tests never reach significance.
- Does message-match really matter that much?: Yes. The gap between ad-hook and hero copy is the largest preventable conversion killer; a Conversion Rate Experts case study credits a 212% lift to fixing it, and sending paid clicks to a generic homepage can convert 4 to 5x worse than a matched page.
- Should the trust line be above or below the CTA?: Just below the CTA. Above the CTA pushes the button out of the thumb zone on mobile.