The short answer
The best Facebook ad copy examples come from ads that kept getting budget. We read the primary text of 2,435 live Meta ads with $50k+ in estimated spend. The median runs 407 characters, 37% run 500 or more, and the primary texts that open with a question reach the top tier only 6.4% of the time, against 10.1% for everything else.
What the primary text of winners looks like
Primary text is the copy above the creative in the feed. It's the part of a Facebook ad that copywriters control most and that most guides fill with invented examples. Here are real ones.
We read the primary text of every ad on the Ad Radar board on October 9, 2026: 2,443 live Meta ads, each with $50k or more in estimated spend. 2,435 of them have primary text (eight run with none). We capture the first 500 characters of each.
407
median characters of primary text
37%
of winners run 500+ characters
15%
run under 125 characters
41%
use at least one emoji
Long copy is the norm, not the exception. The single largest group of winners, 901 ads, fills our full 500-character capture and keeps going. Short copy is common too, though, and it behaves very differently depending on the format.
| Primary text length | Video ads | Top-tier rate | Median est. spend | Image ads | Top-tier rate | Median est. spend |
|---|---|---|---|---|---|---|
| Under 125 chars | 313 | 10.5% | $292K | 56 | 8.9% | $193K |
| 125 to 299 | 497 | 7.8% | $244K | 84 | 7.1% | $193K |
| 300 to 489 | 492 | 7.9% | $226K | 92 | 17.4% | $146K |
| 490 or more | 661 | 9.4% | $217K | 240 | 14.6% | $177K |
In video ads, short primary text does best. Video ads under 125 characters have the highest top-tier rate and the highest median estimated spend of any length. The video does the selling, and the primary text just frames it.
In image ads, long primary text does best. Image ads with 300 or more characters reach the top tier at 14.6 to 17.4%, about twice the rate of image ads with short copy. A static image can't tell a story, so the text has to. We look at this in depth in long-form ad copy.
Which copy patterns show up in winners
We checked every primary text for common patterns. Each row compares ads with the pattern to ads without it:
| Pattern in the primary text | Ads | Top-tier rate | Without it | Median est. spend |
|---|---|---|---|---|
| First line in first person ("I", "My") | 140 | 12.9% | 9.5% | $197K |
| Opens with a customer quote | 166 | 11.4% | 9.5% | $215K |
| Checkmark list | 351 | 10.8% | 9.5% | $177K |
| Emoji anywhere | 994 | 10.4% | 9.2% | $208K |
| Percentage or "% off" | 444 | 10.4% | 9.5% | $249K |
| "You" or "your" in the first line | 1,017 | 9.3% | 9.9% | $241K |
| A word in ALL CAPS | 573 | 8.2% | 10.1% | $224K |
| "Guarantee" | 120 | 6.7% | 9.8% | $201K |
| First line is a question | 326 | 6.4% | 10.1% | $209K |
None of these is a rule on its own, and several overlap. But two results are strong enough to act on.
Question openers underperform. 326 primary texts open with a question, and they reach the top tier 6.4% of the time. That matches what we see in video hooks (see question hooks). A question asks the reader to do work before they've been given a reason to.
Customer quotes last. Primary texts that open with a customer quote have a median run of 241 days, against 120 for the board. A review at the top is evergreen: it doesn't reference a season, a sale or a trend, so brands keep the ad running.
Eight primary texts from the board, annotated
1. The offer stack
While prices everywhere are going up…
- Est. spend
- $2.8M
- Days live
- 123
- Format
- Image
- Variants
- 4
A discount framed as a stance against inflation, then a checkmark list. In the top tier with four variants.
The full primary text:
- "While prices everywhere are going up…"
- "We're doing the opposite."
- "Reverse inflation? Yep - We're doing it!"
- "For a limited time:" followed by two emoji bullets: "Get up to 37% OFF" and "Plus FREE shipping on every order"
- Three checkmarked lines: "Made in the USA.", "Never raised our prices since 2015.", "And we're not starting now."
What it does well: the discount has a reason. "37% off" alone is a sale. "Prices everywhere are going up, we're doing the opposite" is a position the reader agrees with. The checkmarks then turn the offer into three facts, the last of which ("never raised our prices since 2015") is specific enough to be believed. For more on honest offer pressure, see offers and urgency in ads.
2. The identity opener
Finally, a miracle concealer for your complexion designed for those 50, 60, 70 years and beyond…
- Est. spend
- $3.3M
- Days live
- 133
- Format
- Image
- Variants
- 7
Three ages in the first line select the reader. Same product as the Vibriance ad above, run from a different page with a completely different copy approach. Top tier, seven variants.
The second line does the real work: "It seems like every concealer out there is made for your 20-something granddaughter; we're here to change that." It names the reader's frustration in her own terms and casts the rest of the category as the problem. The third line takes apart the closest alternative: "There are concealers for those over 50, but they only hide imperfections." Only then does the product appear.
Look at the pairing: one brand runs an offer-first text and an identity-first text for the same product, from two pages, and both are in the top tier. That is creative diversification done in copy.
3. The surprising promise
It may sound crazy... but you can become a Grief Coach as soon as next week.
- Est. spend
- $4.7M
- Days live
- 84
- Format
- Image
- Variants
- 15
The objection ("it may sound crazy") comes before the promise. Fifteen variants and in the top tier.
The copy repeats its key number twice on purpose: "in just 40 hours", then "That's right: you only need 40 hours." The repetition isn't sloppy. It is the claim the reader will doubt, so the text says it again in plainer words. Fifteen variants means this text has been tested against a lot of alternatives and kept.
4. The post-purchase diary
I ordered TheraPet on Monday night...
- Est. spend
- $946K
- Days live
- 296
- Format
- Image
- Variants
- 3
A day-by-day account told one line at a time, starting from buyer's remorse. Live 296 days and in the top tier.
This is one of the most unusual primary texts on the board. It starts after the purchase: "And then I panicked. Did I just throw away $75? What if it doesn't work?" Then: "If you just ordered it... Or are about to order it... Let me tell you exactly what happened. Day by day."
It speaks to two readers at once: the person who already bought (reassurance, fewer refunds) and the person about to buy (a preview of the experience). The format is one short line per thought, which makes a 500+ character text easy to scan on a phone. The headline reinforces it: "What really happens after you plug in TheraPet…"
5. The review on top
It's time to retrieve all those files you thought you lost.
- Est. spend
- $2.5M
- Days live
- 1442
- Format
- Video, 57s
The primary text is a verified-buyer review, the reviewer's name, and one offer line. Live 1,442 days, almost four years, in the top tier.
The whole primary text: a quoted review ("This is what I need! Very easy to use, just plug it in..."), "Raymond M (Verified Buyer)", and a gift-emoji line: "30% Off Sale Ends Soon". That's it. The video demonstrates the product, so the text only has to add proof and a reason to click now.
This is the pattern behind the 241-day median for quote openers. A plain review doesn't age. Note the details that make it credible: a specific use (transferring files off an old hard drive), a first name and initial, and "Verified Buyer".
6. The repeatable template
Nobl, a carry-on luggage brand, runs the same primary text structure across many of its winning video ads. One version, on an ad live 241 days and in the top tier:
- A customer quote with stars: "Laptops, medicine, jewelry, camera gear… I used to worry about everything I had in my bag. Now I don't even think about it. Five stars" followed by five star emojis.
- One positioning line: "The TSA-approved, zipperless design of NOBL makes travel stress-free"
- A feature list with an emoji per line: laptop pocket, charging, hidden cup holder, phone stand, lifetime warranty.
- A scarcity line: "As soon as inventory arrives in our warehouse, we sell out in a week."
Quote, claim, list, scarcity. The brand swaps the quote and keeps the skeleton. When a primary text works, you don't need a new one for every video. You need a template you can rotate quotes through.
7. The one-liner over a long video
WINPRO Pet's top-tier video runs 207 seconds and has an estimated spend of $2.67M. Its primary text opens with two sentences: "For every dog that's starting to go a little gray. And for every person who loves them just the same."
No product, no offer, no claim. The video carries all of that. The primary text sets the emotional frame for the viewer and selects them (owners of aging dogs) before they press play. This is the short-copy video pattern from the table above: under 125 characters for the opening, and the creative does the rest.
8. The question that answers itself
Question openers underperform, but some work. True Classic runs this one on a top-tier video: "Baggy, rough, too expensive? Forget it. True Classic tees have a fit that flatters your figure and are soft from day one!"
The question isn't really a question. It's three complaints with a question mark, answered in the next two words: "Forget it." The reader isn't asked to think. They're told the brand already knows what's wrong with their T-shirts. If you use a question, answer it in the same breath.
What the winners have in common
Across these eight, and the hundreds of others we read, a few habits repeat:
- The first line works alone. "I ordered TheraPet on Monday night..." and "While prices everywhere are going up…" both make you want the second line. On mobile, the first lines are often all anyone sees.
- One thought per line. The long texts that win are broken into short lines and lists, not paragraphs. 61% of primary texts on the board have three or more line breaks.
- Specifics instead of adjectives. 40 hours. $75. Since 2015. Four business days. 50, 60, 70 years. Not "fast", "affordable", "trusted".
- Proof near the top. Reviews and quotes lead; they don't close.
- The offer has a reason. Inflation, overstock, a launch. "Limited time" alone is weaker.
Four templates to write from
Use these as starting skeletons, then fill them with your own specifics.
Template A: Review on top (video)
- A real customer quote that names a specific use.
- Customer first name, initial, "Verified Buyer".
- One offer line.
Template B: Offer with a reason (image)
- A line about the world ("Prices are going up...").
- Your stance ("We're doing the opposite").
- The offer, with numbers.
- Three checkmarked facts, the last one the most specific.
Template C: Identity callout (image)
- Name who it's for, specifically ("for those 50, 60, 70 and beyond").
- Name their frustration with the category.
- Take apart the closest alternative in one line.
- Introduce the product.
Template D: Diary (image or video)
- "I [did the thing] on [day]..."
- The doubt, in the reader's words.
- Day-by-day what happened, one line each.
- The result, then the link.
For headlines to pair with these, see Facebook ad headlines. For how the classic formulas (PAS, AIDA, BAB) map onto texts like these, see ad copywriting formulas. And for whole ads rather than copy alone, Facebook ad examples has a broader set.
If you want to search primary texts yourself, the Ad Radar board indexes headlines, body copy and video transcripts for every ad, so you can look up a phrase like "verified buyer" or "since 2015" and see which winners use it.
Figures marked as estimated spend come from Ad Radar's model of engagement on public Meta Ad Library ads. They are estimates, labeled as such, and are best used to rank ads against each other.