DEV Community

draftkit
draftkit

Posted on

10 AI Prompts for Marketing Copy That Converts (Tested on Real Landing Pages)

10 AI Prompts for Marketing Copy That Converts

I've written marketing copy for 5 digital products in the last three weeks — landing pages, tweets, cold emails, feature descriptions, and the inevitable "about" section that everyone hates writing.

Here are the 10 prompts that actually moved the needle. Each one follows a four-part structure I'll explain at the end. Every example output below is real, edited only for length.

1. The Pain-First Headline Writer

Most headlines fail because they describe the product instead of naming the pain. This prompt forces pain first.

Role: Senior direct-response copywriter.
Context: I'm writing a landing page headline for [PRODUCT].
It helps [AUDIENCE] who struggle with [SPECIFIC PAIN].
The product's one concrete mechanism is [MECHANISM].
Constraints: Max 9 words. No verbs in the -ing form.
Lead with the pain or the cost of inaction — never the product name.
Output: 8 headline options numbered 1-8. No subheads, no rationale.
Enter fullscreen mode Exit fullscreen mode

Real output (for a developer launch pack, audience = indie devs shipping their first product):

  1. Launch week without a launch checklist.
  2. Your product launches once. The copy doesn't.
  3. Stop writing launch copy from a blank page.
  4. The tweet that announces you to the world.
  5. Ship the product. Borrow the launch words.
  6. First launch, last blank page you'll stare at.
  7. Every launch asset, one prompt library.
  8. Don't rewrite launch copy. Generate it.

Pain-first beats feature-first on every test I've run. Option 3 became the headline I shipped.

2. The Benefit Ladder (Feature → Outcome → Identity)

Feature lists don't sell. Identity does. This prompt walks a feature up the ladder to who the buyer becomes.

Role: Marketing strategist trained in jobs-to-be-done.
Context: Take this feature: [FEATURE DESCRIPTION].
Constraints: Produce exactly 4 rungs. Each rung is one sentence
under 14 words. Move strictly from feature → benefit → outcome → identity.
Do not skip the outcome rung.
Output: A labeled ladder.
Enter fullscreen mode Exit fullscreen mode

Real output (feature = "50+ ready-to-use prompt templates"):

Feature: 50+ ready-to-use prompt templates
Benefit: You stop writing prompts from scratch every time
Outcome: You ship marketing copy in a quarter of the time
Identity: You become the person who ships, not the person who drafts

The identity rung is where you discover whether the feature actually matters. If you can't write it, the feature probably doesn't belong on the page.

3. The Objection Crusher

The fastest way to find what's killing your conversion is to list every reason a buyer would say no, then answer each one.

Role: Skeptical buyer with 12 years in [AUDIENCE'S INDUSTRY].
Context: I'm selling [PRODUCT] for [PRICE] to [AUDIENCE].
Constraints: List exactly 5 objections in the buyer's own voice —
casual, slightly snarky, the way they'd actually think it.
Then answer each objection in one sentence that does NOT use
the words "trust", "quality", or "premium".
Output: 5 numbered objection/answer pairs.
Enter fullscreen mode Exit fullscreen mode

Real output (product = $49 prompt pack, audience = SaaS founders):

1. "Forty-nine bucks for text files?"
Answer: It's $49 for the 40 hours of writing you won't do this week.

2. "I can just ask ChatGPT myself."
Answer: You can — and you'll spend the next hour rewriting the output. These prompts are pre-tested.

3. "How do I know they actually work?"
Answer: Every prompt has an example output you can read before you buy. (Link the free sample.)

4. "This is probably recycled from Reddit."
Answer: It's not — every prompt was tested against a real task and rated by a human before it made the pack.

5. "I'll buy it later."
Answer: Later is the objection that kills more launches than price. If it helps now, now is when it's cheapest.

The objection/answer pairs become FAQ copy, sales-call rebuttals, and ad creative — three assets from one prompt.

4. The Cold Email That Earns a Reply

Cold email isn't dead — bad cold email is dead. The prompt below forces the one thing that gets replies: a specific, falsifiable reason for reaching out to this person.

Role: B2B sales writer who has booked 200+ meetings.
Context: I'm emailing [RECIPIENT ROLE] at [COMPANY TYPE].
The one specific trigger that makes them worth emailing now is [TRIGGER].
I'm offering [OFFER] with no obligation.
Constraints: Subject line max 5 words, lowercase, no punctuation.
Body max 90 words. No "I hope this finds you well".
The first sentence must reference the trigger, not introduce me.
End with a one-question call to action — never a meeting link.
Output: Subject + body.
Enter fullscreen mode Exit fullscreen mode

Real output (recipient = head of growth at a Series A SaaS, trigger = "just raised and hiring 3 marketers"):

subject: saw the three marketing hires

Hi [name] —

Three new marketers means three people who'll each need onboarding into your messaging. Most teams handle that with a Notion doc that goes stale by month two.

I put together a prompt library that lets a new hire write on-brand landing copy in their first week — copy your senior people would otherwise have to rewrite.

Want me to send the free sample pack? It's the fastest way to see if it fits.

— [name]

The lowercase subject line, the trigger reference in sentence one, and the question-instead-of-a-link are the three levers that compound.

5. The Social Proof Rewriter (No More "Game-Changer")

Testimonials full of adjectives don't convert. Specifics convert. This prompt turns a vague quote into a specific one — ethically, by expanding on what the customer actually said, not inventing facts.

Role: Editor who has written case studies for B2B SaaS.
Context: Here is a raw testimonial: "[VERBATIM QUOTE]".
Constraints: Rewrite it in the customer's voice but add one specific
number and one specific before/after comparison that are implied by
the quote — never anything that contradicts it. Keep it under 35 words.
Do not use the words "amazing", "game-changer", "revolutionary",
or "seamless".
Output: The rewritten testimonial, in quotation marks.
Enter fullscreen mode Exit fullscreen mode

Real output (raw quote = "These prompts saved me a ton of time, honestly the best purchase I've made for my workflow this year"):

"I stopped rewriting launch copy from scratch. What used to take me a full afternoon now takes about 20 minutes, and the output reads better than what I was writing myself. Easily my best workflow purchase this year."

The rewrite adds a specific duration (20 minutes) and a before/after (full afternoon → 20 minutes) that were implied by "saved me a ton of time" — without fabricating anything.

6. The Ad Creative That Doesn't Sound Like an Ad

Ad copy that sounds like ad copy gets scrolled past. This prompt writes ads that sound like a recommendation from a coworker.

Role: Someone who just discovered a tool and is texting a friend about it.
Context: The tool is [PRODUCT]. The friend is [AUDIENCE] and has
been complaining about [PAIN]. The one concrete thing that surprised
you is [SURPRISE].
Constraints: Write it as a text message. Max 3 sentences.
No exclamation marks. No emojis. No "you have to try this".
Output: The text message.
Enter fullscreen mode Exit fullscreen mode

Real output (product = prompt pack, friend = freelance developer):

found this prompt pack that has the launch tweet template i've been looking for. wrote my announce tweet in like five minutes instead of staring at the compose box for an hour. the tweet generator alone was worth it for me.

That's an ad that doesn't read like one. It reads like a Slack message — which is exactly how people recommend things to each other.

7. The Feature Page That Skips the Marketing-Speak

Feature pages are where conversions go to die. This prompt forces a structure: what it is, what it isn't, and what it replaces.

Role: Technical writer who respects the reader's time.
Context: Document this feature: [FEATURE].
Constraints: Use exactly three subsections.
"What it does" — one sentence, no adjectives.
"What it doesn't do" — one sentence naming the honest limitation.
"What it replaces" — one sentence naming the manual work it eliminates.
Output: Three labeled lines.
Enter fullscreen mode Exit fullscreen mode

Real output (feature = "show HN title checker"):

What it does: Scores your Show HN title against the patterns of posts that hit the front page.
What it doesn't do: Write the title for you — you still have to decide what your post is about.
What it replaces: The 20 minutes you'd spend rewriting the title in a text editor and second-guessing each version.

"What it doesn't do" is the most underrated section in product copy. Buyers trust you the moment you name a limitation.

8. The Email Sequence That Has a Plot

Welcome sequences die in the middle. Email three goes off the rails because there's no through-line. This prompt builds the whole sequence around a single transformation.

Role: Email strategist who has written 40+ welcome sequences.
Context: I'm writing a 5-email welcome sequence for [PRODUCT].
The reader's starting state is [STARTING STATE].
The reader's desired end state is [END STATE].
Constraints: One transformation per email, in order.
Email 1: orient — confirm what they signed up for.
Email 2: first small win — use the product once.
Email 3: the hidden cost of not using it — name the alternative.
Email 4: case study — one specific person's result.
Email 5: the ask — one clear next step.
Each email's subject line max 6 words.
Output: 5 subject lines + one-sentence summaries.
Enter fullscreen mode Exit fullscreen mode

Real output (product = prompt library, starting state = "downloaded the free sample, hasn't bought"):

1. subject: the pack you grabbed
One line: "Here's the free sample you asked for, and the one prompt I'd start with."

2. subject: try this one first
One line: "Use the launch tweet prompt on your current project — it'll take five minutes and you'll see the pattern."

3. subject: what it costs to keep drafting
One line: "Every launch you write from scratch costs you about three hours and a worse result — here's the math."

4. subject: how [customer] shipped in a week
One line: "[Customer] used four prompts from the pack to launch their product in seven days — here's what they used and what they skipped."

5. subject: the full library
One line: "The full library is $49, it has every prompt from this sequence plus 20 more, and it's the fastest way to stop rewriting copy."

The through-line — "from sample to first win to cost-of-inaction to proof to ask" — is what keeps a sequence alive past email two.

9. The Landing Page Section That Tells the Truth

The middle of a landing page is where attention dies. This prompt writes the section that re-engages by being honest about who the product is not for.

Role: Copywriter who would rather lose a sale than mislead a buyer.
Context: I'm writing the "who this is not for" section of a landing
page for [PRODUCT].
Constraints: Three bullet points. Each starts with "This is not for
you if". Each names a specific behavior, not a demographic.
Do not use the word "beginners" or "advanced".
Output: Three bullets, one sentence each.
Enter fullscreen mode Exit fullscreen mode

Real output (product = $49 SaaS marketing prompt pack):

This is not for you if:

  • You write copy by hand and enjoy the craft of it — these prompts will shortcut the part you like.
  • You need fully finished copy handed to you — these are prompts, not a copywriting service.
  • You're looking for a magic button — every prompt still requires you to fill in the specifics about your product.

Honest disqualification is the highest-trust move in copywriting. The people who stay after reading it convert at a higher rate.

10. The Pricing Page That Justifies the Top Tier

Pricing pages lose money at the top tier. Most buyers pick the middle plan by default. This prompt writes the copy that makes the top tier feel like the rational choice, not the aspirational one.

Role: Pricing strategist who has designed SaaS pricing tiers.
Context: Three tiers for [PRODUCT]: [TIER 1], [TIER 2], [TIER 3].
The top tier exists because [REASON TOP TIER EXISTS].
Constraints: For the top tier only, write a one-paragraph justification
that names the specific scenario where it pays for itself.
Do not use "best value", "most popular", or "ultimate".
Output: One paragraph, max 60 words, for the top tier.
Enter fullscreen mode Exit fullscreen mode

Real output (product = prompt packs, tiers = $9 / $49 / $999):

The playbook pays for itself the first time you avoid bringing in a consultant to write your AI rollout plan. It includes the operations system, the transformation playbook, and every prompt from the lower tiers — so you're not rebuying, you're completing the set. For a team rolling out AI across more than one function, this is the one that doesn't get outgrown in a quarter.

The phrase "completing the set" reframes the top tier from a splurge to a rational consolidation. That single reframe is worth the prompt.


The Structure Behind All 10 Prompts

Every prompt above follows the same four-part structure. Once you internalize it, you can write your own.

Role — Tell the model who it's being. Not "act as a marketer" — "act as a direct-response copywriter with 12 years of experience". Specificity in the role changes the output more than any other lever.

Context — State the situation in concrete terms. The audience, the product, the constraint the buyer is under. Vague context produces vague copy.

Constraints — This is where most prompts fail. A constraint is a hard rule: word count, banned words, required structure, a specific first sentence. Constraints are what turn generic output into something you'd actually ship.

Output format — Tell the model the shape of the response. "Eight numbered headlines." "A labeled ladder with four rungs." "Three bullets, one sentence each." Without an output format, you get paragraphs you have to edit down.

How I Tested Each Prompt

I didn't ship any of these on the first output. Each one went through the same test:

  1. Run it three times with the same inputs. If the outputs vary wildly, the prompt is too loose — add a constraint.
  2. Run it on a different product than the one it was written for. If it only works for one product, it's overfit — generalize the context.
  3. Read the output out loud. If it sounds like marketing copy, it failed. Good copy sounds like a person talking.
  4. Cut anything that could apply to any product. "Save time and grow your business" applies to everything and sells nothing. Specificity is the test.

The Full Library

Every prompt above is open-sourced — but it's 10 out of a larger library I use to write all the marketing copy for my products. The full set covers landing pages, email sequences, ad creative, sales pages, pricing pages, and the launch assets (tweet, Show HN post, changelog, release notes).

If you want the whole library rather than copying prompts one at a time:

  • Developer Product-Launch Prompt Pack ($9) — 7 prompts for launch tweets, Show HN posts, changelogs, release notes. The entry-level pack. Get it on Gumroad →
  • SaaS Marketing Copy Prompt Pack ($49) — 25 prompts for landing pages, ad creative, objection handling, email sequences. The one most relevant to the prompts in this article. Get it on Gumroad →
  • Developer Productivity Prompt Library ($49) — 30 prompts for code reviews, PR descriptions, bug reports, test cases, API docs. Get it on Gumroad →
  • AI Startup Operations Prompt System ($199) — 50+ prompts across 10 workflows for running a startup with AI. Get it on Gumroad →
  • AI Business Transformation Playbook ($999) — 100+ prompts and 12 workflows for rolling out AI across a whole company. Get it on Gumroad →

Or browse the full storefront — every product has a free sample you can read before buying.


That's the set. If you found even one prompt here useful, the easiest next step is the $9 launch pack — it's the lowest-risk way to see if the format works for you.

What marketing copy problem would you want a prompt for next? I read the comments.

Top comments (0)