Ask ChatGPT to "write a landing page for my product" and you'll get a neat wall of claims that could sell almost anything. The model isn't broken. It just has nothing to work with: no voice of your customer, no real proof, and no single job for the page. So it fills the gaps with the average SaaS pitch.
This post is for solo founders rewriting a one-page offer, where the visitor is supposed to buy, book, or try. It walks through what to collect before you prompt, then three prompts you can copy: one that drafts hero variants from real customer language, one that turns that language into benefit blocks and an honest FAQ, and one that roasts the draft the way a skeptical buyer would. No invented conversion stats anywhere.
Why AI landing pages all sound the same
Most people don't read web pages word for word. Nielsen Norman Group's research has said this for decades: visitors scan, pick out phrases that match what they came for, and leave when the page doesn't answer them quickly. A page full of "seamless," "powerful," and "next-generation" fails that scan, because nothing on it is specific enough to grab onto.
When founders paste a vague brief into ChatGPT, three failure patterns keep showing up:
- Feature soup. The hero lists what the product has instead of what the buyer gets.
- Invented proof. "Join 10,000 happy customers" or "increase conversions by 40%" appears even though you never said anything like it.
- No single job. The page tries to educate, sell, collect a lead, and book a call at once, so the visitor never knows which action to take.
The fix is to change the inputs, not to hunt for a cleverer prompt. Give the model the words your buyers already use, ban claims you can't back up, and give the page one primary action.
What to gather before you open ChatGPT
Fifteen minutes of prep beats an hour of regenerating the same generic draft. Have these ready:
- A one-sentence offer. Who it's for, what they get, and the primary action (buy, book, or start).
- Real customer language. Quotes from sales calls, support tickets, Reddit threads, competitor reviews, or DMs. Paste the exact phrases, not your polished paraphrase of them.
- Three objections you've actually heard. Price, "I can do this myself," timing, trust. Skip the hypothetical ones.
- Proof you can name. A named result, an outcome you could screenshot, a short testimonial with a real first name. If you don't have any, leave this blank. Blank is better than fake.
- Traffic context. Where visitors come from (an ad, a tweet, a newsletter), so the headline can match the promise that brought them.
One note on benchmarks: Unbounce's Conversion Benchmark Report, built from tens of thousands of landing pages, puts the median conversion rate across industries somewhere in the mid-single digits. That's useful as a sanity check for your own analytics later. It is not something the model should be writing into your copy.
Prompt 1: Load real language, then draft hero variants
Paste your transcripts or quotes first. Then ask for hero options that reuse the buyer's words instead of your feature list.
You are helping me rewrite the hero of a one-page landing page for solo founders.
Primary action: [BUY / BOOK A CALL / START FREE SAMPLE]
One-sentence offer: [WHO IT'S FOR + WHAT THEY GET]
Traffic context: [e.g. reply to a tweet about cold outreach / Google "X"]
Real customer language (exact quotes; do not paraphrase before using):
[PASTE 5–15 SHORT QUOTES OR CALL NOTES]
Constraints:
- Do not invent customers, results, percentages, awards, or "trusted by" logos.
- If I gave no proof, do not invent a proof line — omit it.
- Prefer the visitor's verbs and nouns over marketing adjectives.
- Keep each hero scannable: headline ≤12 words, subhead ≤35 words, CTA ≤5 words.
Output:
1) 5 phrases from the quotes that should appear (almost verbatim) on the page
2) The single job of this page in one line
3) 5 hero variants. For each: Headline / Subhead / Primary CTA / 1-line risk of that angle
Label each variant: (A) outcome, (B) problem, (C) "for people who…", (D) contrast with DIY, (E) traffic-match.
What good output looks like: headlines you could imagine a buyer saying out loud, and CTAs that name the outcome ("Get the pack") rather than "Submit." If any variant invents social proof, throw it out.
The five labeled angles are there for a reason. Each one comes with its own one-line risk, so you aren't just picking the variant that sounds nicest. You're picking the trade-off you're willing to make.
Prompt 2: Benefit blocks and an objection FAQ, without invented proof
Once you've chosen a hero angle, build out the middle of the page from the same quotes. Features stay in the page, but as supporting detail under the benefits, not as the headline act.
Continue from the chosen hero below.
Chosen hero: [PASTE HEADLINE + SUBHEAD + CTA]
Offer: [ONE-SENTENCE OFFER]
Customer language: [SAME OR EXTRA QUOTES]
Proof I can actually use (or write "none"): [NAME / QUOTE / RESULT — FACTS ONLY]
Objections I've really heard: [LIST 3]
Write:
1) Three benefit blocks. Each: H3 title (benefit, not feature) / 2–3 sentences / one concrete "what you get" bullet. Use customer language where it fits.
2) A "How it works" section with exactly 3 steps, each one sentence, ending at the primary action.
3) An FAQ of 5 questions drawn from my objections. Answers: honest, short, no invented guarantees. If I wrote "none" for proof, do not add testimonials.
Banned words unless they appear in my quotes: seamless, cutting-edge, revolutionary, leverage, unlock, game-changer.
Do not invent numbers. If a claim needs proof I didn't give, cut the claim.
Why this works: the model stays anchored to language you gave it. The FAQ handles objections in the visitor's own words, instead of turning into a legal-sounding wall of fine print. And the explicit "none" option for proof gives the model permission to leave testimonials out rather than make them up.
Prompt 3: The skeptical-buyer roast
Before you publish, have the model attack the draft the way a tired buyer with too many tabs open would.
Act as a skeptical solo founder who has been burned by generic AI tools and has 20 tabs open.
Landing page draft: [PASTE FULL DRAFT OR HERO + BENEFITS + FAQ + CTA]
Do this in order:
1) Roast: list every vague claim, every place that sounds like every other SaaS page, and every invented or unearned proof line. Be specific — quote the phrase.
2) Message match check: in one sentence, what promise did the traffic source make, and does the hero keep it?
3) Scan test: in 10 seconds, what three phrases would someone notice? Are they the right ones?
4) Rewrite: return a tighter version of the hero + the weakest benefit block + the weakest FAQ answer. Keep my facts. Cut fluff. No new claims.
If the page asks for more than one primary action, say so and recommend which action to keep.
Run this pass once. Then go through what's left and cut anything you couldn't defend if a buyer asked "says who?" on a sales call.
The scan test in step 3 ties back to how people actually read pages. If the three phrases someone would notice in ten seconds aren't the ones that explain your offer, the page has a problem no amount of extra copy below the fold will fix.
Match the traffic, then ask for one action
If someone clicked a tweet about rewriting cold emails, the first line of your hero should echo that promise. It shouldn't pivot to a generic "AI toolkit for founders." Message match is simple: the headline should read like the next sentence of the ad, post, or search result that brought the visitor there.
Then pick one primary action and repeat it. Secondary links, like other guides or an FAQ anchor, are fine lower on the page. Competing buttons in the hero ("Buy," "Book a call," and "Download a sample" all at once) usually mean the visitor does none of them. Solo-founder pages win by being obvious, not by offering every possible path.
After you publish, watch the first screen. If people bounce without scrolling, the problem is the hero or the traffic match, not the FAQ. Change one thing per week so you know what actually moved.
Common traps
- Polishing before grounding. Nice prose on top of a vague offer still performs like a vague offer.
- Letting the model invent social proof. If you don't have a named result yet, say what the offer includes and what happens after payment. Don't add fake logos.
- Five CTAs. One primary action, repeated. Secondary links can exist, but they shouldn't compete in the hero.
- Writing for yourself. If the headline only makes sense to people who already know your product name, rewrite it for the stranger arriving from a tweet or a search result.
- Skipping the roast. The third prompt is where most of the "sounds like AI" lines get caught. Budget five minutes for it.
The short version
- Collect real quotes and three real objections before you prompt.
- Use Prompt 1 to draft heroes that reuse the buyer's language, and pick one job for the page.
- Use Prompt 2 for benefit blocks, three steps, and an honest FAQ with no invented proof.
- Use Prompt 3 to roast and tighten, and delete every claim you can't back up.
- Ship, watch where people bounce, and change one section at a time.
ChatGPT can draft fast. What makes a landing page worth scanning is still the specific stuff only you have: your customer's words, your real proof, and one clear ask.
Originally published as the full guide: ChatGPT Prompts for Landing Page Copy That Doesn't Sound Generic.
Top comments (0)