Facebook and Instagram feeds move fast, and most ads get skipped in under a second. The ads that stop the scroll rarely look like traditional advertising at all โ they read like a post from a friend, not a brand. This prompt is built to produce that native, in-feed tone across three distinct angles, so you have real variety to test instead of three versions of the same idea.
Write 3 ad copy variations for:
Product/offer: [WHAT YOU'RE ADVERTISING]
Target audience: [WHO YOU'RE TARGETING โ demographics, interests, or behavior]
Campaign goal: [e.g. cold traffic awareness, retargeting, direct conversion]
Offer/promo: [ANY DISCOUNT, DEADLINE, OR SPECIAL OFFER โ or "none"]
Landing destination: [WHERE THE AD SENDS PEOPLE]
For each of the 3 variations, write:
– A primary text (under 125 words, scroll-stopping first line, native "in-feed" tone rather than sales-y)
– A headline (under 40 characters)
– A description line (under 30 characters)
– A call-to-action button suggestion (from standard Meta options: Shop Now, Learn More, Sign Up, Get Offer, etc.)
Make the 3 variations distinct in approach:
1. Problem/solution angle
2. Social proof or testimonial-style angle
3. Direct offer/urgency angle
Write in a tone that feels native to the Facebook/Instagram feed โ conversational, not corporate, no excessive emojis.
Why This Prompt Works
Meta’s ad algorithm rewards engagement, and engagement comes from copy that doesn’t feel like an interruption. By explicitly instructing the AI to write in a “native, in-feed tone” and avoid a “sales-y” voice, this prompt steers away from the stiff, brochure-style copy most AI tools default to when asked for “ad copy.”
Requesting three distinct angles โ problem/solution, social proof, and urgency โ is also deliberate. Ad performance is unpredictable; you rarely know in advance which emotional angle will resonate with your specific audience. Testing three genuinely different approaches from day one gives you real signal much faster than testing three near-identical variations of the same idea.
How to Use This Prompt
- Fill in your targeting details as specifically as you can โ “women 25-40 interested in home fitness” performs better in the prompt than “women.”
- Paste into your AI tool and generate the three variations.
- Check character counts before uploading โ Meta truncates headlines and descriptions that run long, so trim anything over the limit.
- Launch all three as separate ad sets or within a dynamic creative test, and let the data โ not your gut โ decide the winner.
- Use the UGC script prompt below to turn your best-performing angle into a video ad.
Bonus Prompt: UGC-Style Video Script
Video ads that look like organic user content consistently outperform polished brand videos on Meta platforms, especially for cold traffic. Once you know which angle is working in your static ad copy, use this prompt to adapt it into a scroll-stopping video script.
– A hook line for the first 3 seconds (this is the most important line โ it has to stop the scroll)
– The body of the script in casual, first-person spoken language
– A natural-sounding call-to-action at the end
Keep the total script under 100 words so it fits a 30-second read-aloud pace.
Tips for Better Results
- Front-load the hook. The first line of your primary text (and the first 3 seconds of video) decides whether anyone sees the rest โ spend the most editing effort there.
- Match copy to campaign objective. Cold traffic ads should educate and hook; retargeting ads can be more direct about the offer since the audience already knows you.
- Keep emoji use minimal and purposeful. One or two well-placed emojis can add warmth; a string of them reads as spammy and can hurt trust.
- Write like you’re texting a friend, not presenting to a boardroom. If a line sounds like it belongs in a press release, rewrite it.
- Always test creative and copy together. Even great copy underperforms with the wrong image or video โ treat copy and visuals as one variable, not two.
Common Mistakes to Avoid
The most common mistake is writing to “everyone” instead of a specific audience segment โ vague targeting produces vague, underperforming copy. Another frequent error is leading with the brand name or logo instead of the problem the viewer is experiencing โ nobody scrolls past their feed to learn about your company first. Finally, avoid changing more than one variable per test; if you change the copy, the image, and the audience at the same time, you won’t know what actually drove the result.
FAQ
How many ad variations should I test at once?
Three to five distinct angles is usually enough to find a clear winner without splitting your budget too thin. More than that and each variation may not get enough spend to reach statistical significance.
Should ad copy match the landing page exactly?
The tone and core promise should match, but the copy doesn’t need to be identical โ mismatched messaging between ad and landing page is one of the biggest silent killers of conversion rate.
Is UGC-style copy still effective if I don’t have real customers on camera?
Yes โ many brands use actors or creators to deliver UGC-style scripts. What matters most to viewers is the casual, authentic tone, not necessarily whether the person is a “real” customer.
How often should I refresh ad copy?
Watch your frequency metric โ once an ad’s frequency climbs past 3-4 for a given audience, performance typically starts to decline and it’s time to refresh the copy or creative.
A Quick Example of How the Output Should Read
Imagine you’re advertising a meal-prep delivery service to busy working parents. A weak, corporate-sounding ad might open with “Introducing the future of family nutrition.” The version this prompt is built to produce opens instead with something like “The 6pm ‘what’s for dinner’ panic is real โ here’s how I stopped having it every night,” written in first person, as if a parent in the target audience is talking directly to another parent scrolling their feed. That tonal shift โ from brand announcement to peer conversation โ is usually the difference between an ad that gets scrolled past and one that gets a comment.
If a draft still opens with your brand name or a formal announcement tone, ask the AI to “rewrite the opening line as if a customer is telling a friend about this, not as if a brand is announcing a product.”
Editing the AI Draft Before You Launch
Check every claim against Meta’s advertising policies before submitting โ health, finance, and “before and after” claims in particular face strict review, and copy that reads fine on your own website can still get an ad account flagged. Also verify character counts precisely; Meta will truncate primary text, headlines, and descriptions that exceed platform limits, which can cut off your call-to-action mid-sentence if you don’t check. Finally, make sure the copy’s promise matches what’s actually waiting on the landing page โ a mismatch between ad copy and landing page content is one of the fastest ways to tank your Quality Ranking and drive up cost per result.
FAQ (continued)
How long should I let an ad run before judging performance?
Most ad accounts need to exit the learning phase before performance stabilizes, which typically requires a meaningful number of conversion events per ad set โ judging results too early, before the algorithm has optimized delivery, often leads to killing ads that would have performed well with more time.
Do I need different copy for Instagram Stories versus the main feed?
Yes, ideally โ Stories and Reels favor even shorter, punchier text since the format is vertical, full-screen, and fast-moving, so consider asking the AI for a condensed version of your winning feed ad specifically for these placements.
Building a Repeatable Testing Habit
The real value of this prompt shows up over repeated use, not a single campaign. Keep a simple swipe file of which hooks, angles, and CTAs actually performed for your account, organized by campaign objective and audience segment, and feed a few of your past winners back into future prompts as examples of “tone that has worked for us before.” Over a few months, this turns AI-assisted ad copy from a generic starting point into something that increasingly reflects what your specific audience responds to โ which is ultimately more valuable than any single high-performing ad. It’s also worth noting which angles fatigue fastest for your account; problem/solution hooks often have a longer shelf life than urgency-driven ones, which tend to lose effectiveness quickly once an audience has seen them a few times.
Paid social rewards fast iteration more than perfect first drafts. Use this prompt to generate real variety quickly, let the algorithm and your data tell you what’s working, and keep testing new angles from there.