Facebook Ad Primary Text: The First Line (2027)

Where primary text lives, and why the first line is the whole ad

Primary text is the block of copy that sits above your image or video, just below the Page name in the feed. It is the largest writing space Meta gives you, and the most misused, because the box you type into and the box a scroller sees are two completely different sizes. The field will happily accept a few paragraphs. The feed shows two or three lines, then a "See more" link, and everything after that is folded out of view. Meta's own ads guide recommends 50 to 150 characters of primary text for a Facebook Feed image ad, not because longer copy is forbidden, but because the tighter end of that range is closer to what a smaller screen actually shows before the cut.

Two facts turn the first line from a nicety into the entire job. First, almost everyone reading your ad is on a phone. Shopify, citing Statista, reports that 98.5% of Facebook users access the platform from a mobile device. Desktop, where more of your paragraph would show, is a rounding error you should never design around. Second, the audience is enormous and moving fast. Facebook ads could reach about 2.28 billion people in early 2025 according to DataReportal, and every one of them meets your copy while thumbing past a feed, not while reading a page. The visible sliver of your primary text is competing with the next post, not with your own second paragraph.

So the honest definition of primary text is not "the caption." It is "the one or two lines that show before See more, plus an optional body that only a small minority will ever expand." Write the visible part as though it is the whole ad, because for most of your audience it is. Everything else in this guide follows from that single constraint.

The truncation specs, placement by placement

The frustrating part is that Meta does not publish one exact character count where the See more link appears. The cut shifts by device, font size, and placement, so treating any single number as gospel will burn you. What you can lean on is a recommended length per placement. Sprout Social, which tracks Meta's specs, recommends keeping Facebook Feed primary text to about 80 characters, and allows more room, around 125 characters, in placements like Stories, in-stream video, and Marketplace. The feed, where most of your budget usually goes, is the tightest of the lot, so design to that 80-character feed figure and you rarely lose a word that mattered.

The bigger point is that primary text does not render the same everywhere, so the field has to survive the tightest placement you run, not the roomiest. Here is how the main placements compare.

Sources: Meta Ads Guide (Facebook Feed 50-150 characters) and Sprout Social (per-placement primary text recommendations, including about 80 characters for the Facebook Feed).

Read the table and one rule writes itself: if you are running a single ad across feeds, Reels, and Stories, write for Reels and Stories, then let it look generous on the feed rather than the other way around. It also explains why the headline field exists as a separate, tighter line (Meta recommends just 27 characters for the Facebook Feed headline): in the placements where primary text is squeezed to almost nothing, the headline beside the button becomes the line that has to sell. This page is about the primary text field specifically, but the two work as a pair, and you should never park your whole message in a field that a given placement is going to hide.

The first-line rule: land the hook in the first 80 characters

If the visible window is short and variable, the safe play is to assume you have one line, roughly the first 80 characters, and to make that line do the work of the entire ad. That 80-character mark is both what Sprout Social recommends for the Facebook Feed and about what you can count on showing almost everywhere, including the tightest placements. Some placements give you a little more room, but write the hook to fit the smaller number and you are covered everywhere.

"Front-load" is easy to say and easy to get wrong, so make it concrete. Picture a coffee subscription brand writing primary text for a cold audience. Here is a line that buries the point:

At our company, we truly believe that great mornings start with great coffee, which is why we set out to...

By the time that sentence reaches anything a buyer cares about, the See more link has already swallowed it. Now the same brand, front-loaded:

We roast every bag the day it ships, so your first cup tastes like it came straight from the roastery.

That lands the single most important claim in about 95 characters, and the real promise (fresh, roastery-quality coffee) is visible inside the first 80. The reader knows what the ad is about before deciding whether to slow down. A tighter version pushes it further:

Roasted the day it ships. Your first cup tastes like the roastery, not the supermarket shelf.

That is under 95 characters, leads with the differentiator, and still has room for a contrast that earns the tap. The pattern across all three is the same: the concrete, specific idea comes first, and the throat-clearing ("At our company, we believe...") is cut entirely. A hook is a specific claim, a number, a question the reader is already asking, or an offer, never a wind-up. If your first line would read the same for a competitor's product, it is not a hook, it is filler.

One more discipline for the first line: write the benefit, not the topic. "A better way to buy coffee" is a topic. "Roasted the day it ships" is a benefit. Topics get scrolled past; benefits and specifics stop the thumb.

A question can hook well, but only a question the reader is already asking themselves. "Tired of coffee that tastes flat by the second week?" works because it names a real frustration in the reader's own words. A vague or self-serving question ("Ready to upgrade your coffee game?") reads as a sales prompt and gets ignored. The same goes for numbers: a specific figure early in the line ("Roasted within 24 hours of shipping") is more believable and more arresting than a rounded superlative ("the freshest coffee anywhere"), because specificity signals that you actually know the thing you are claiming. Whatever device you reach for, the test is the same. Read only the first line, cover the rest, and ask whether a stranger scrolling past would understand what you are offering and why it might matter to them. If the answer is no, the line is not finished, no matter how good the paragraph underneath it reads.

Structuring the body for the few who expand

Most readers never tap See more, but the ones who do are your most interested prospects, so the folded body is not wasted space, it is a second chance at a warmer audience. Structure it for scanning, not for prose. A wall of text below the fold reads as effort; a few short lines reads as a series of reasons.

A body that works usually moves in this order. The first line has already delivered the hook. The next line or two add the proof or specifics the hook implied: what makes the claim true, a concrete number, a detail a skeptic would want. Then comes the offer or the reason to act now, stated plainly. Then a short closing line that points at the next step, even though the call-to-action button technically carries the click. The button says "Shop Now"; the last line of your text can still say "Grab your first bag this week" to make the action feel like a natural continuation of the story.

A few mechanical habits keep that body readable on a phone:

• Use line breaks, not one dense paragraph. Two or three short lines with white space between them are far easier to scan on a small screen than a single block.
• Keep to one idea per line. The reader is skimming, and a line with two competing thoughts loses both.
• Use an emoji only as a signpost, and rarely. One at the start of a benefit line can aid scanning; a scatter of them reads as noise and can cheapen the brand.
• Do not paste the link into the text. The destination is already attached through the button and headline, so a raw URL just clutters the copy and eats visible characters. Spell out the next step in words only when a lead or messaging flow genuinely needs it.
• Watch the policy line. Copy that names a reader's personal attribute ("Struggling with your weight?") is restricted under Meta's advertising standards, and the primary text is where that trap is easiest to fall into. Write to the many, not at the one.

To make the shape concrete, here is the coffee brand's folded body written to that order, under the front-loaded first line:

Roasted the day it ships. Your first cup tastes like the roastery, not the supermarket shelf.

Most beans sit in a warehouse for weeks before they reach you. Ours are roasted to order and posted within 24 hours, so the aromatics are still there when you open the bag.

Your first order ships free, and you can pause or cancel any time.

Try a bag this week and taste the difference in the first cup.

Notice that the hook is the only line most people will ever see, and every line after it earns its place for the reader who expanded: proof of the freshness claim, the offer that lowers the risk, then a plain nudge to act. Nothing is repeated, and nothing depends on the reader having read the line above it in order, because feeds do not guarantee careful reading. The body is where you earn the considered purchase, but it never rescues a weak first line. Get the hook right, then treat the body as support for the minority who wanted more.

Writing 3 to 5 primary-text variants for dynamic delivery

The single biggest upgrade most advertisers can make to their primary text is to stop writing one version. Meta's delivery is built to test copy for you: with dynamic creative you can submit up to five primary text options in a single ad, and Meta's built-in text generation can produce up to five versions of the field from your input. Once several options are live, the system mixes and matches, learns which line resonates with which person, and you can open the breakdown by dynamic creative element in Ads Manager to see which primary text actually earned the results. Supplying more distinct options also feeds Meta's opportunity score, which the company presents as a signal tied to stronger performance.

The catch is the word distinct. Five variants that are the same sentence with a synonym swapped teach the system nothing, because the reader responds to the idea, not the punctuation. The variants that move numbers are different angles on the same product. Take the coffee subscription again and write five primary-text openers that each lead with a different reason to care.

Each opener is under about 80 characters, leads with a different promise, and would pull a different type of buyer: the frustrated drinker, the follow-the-crowd shopper, the deal seeker, the story lover, the curious skeptic. That is a real test. If one clearly wins on cost per result, you have learned something you can carry into the next round of creative; if two tie, you have two angles worth their own ads. Compare that to five versions of "Great coffee, delivered fresh," which would tell you nothing except that the system had nothing to choose from.

Testing copy this way is cheap leverage, and cheap leverage matters because media is not cheap. Gupta Media's tracker put the blended Meta CPM near $8.19 across 2025, and WordStream's 2025 data put an all-industry Traffic click at $0.70 with a 1.71% click-through rate. You cannot bid your way past those costs, but a first line that earns more clicks at the same CPM lowers your cost per click directly. The copy is the free variable, and variants are how you find the version that pays.

The math makes the point better than the argument does. At a fixed CPM, cost per click moves inversely with click-through rate: the more of the same impressions you turn into clicks, the less each click costs. WordStream's all-industry Traffic numbers are a $0.70 click at a 1.71% click-through rate. Now suppose one of your five primary-text variants pulls the click-through rate up to 2.5%, not because you paid more but because the first line finally named something the reader wanted. On identical spend and identical targeting, those same impressions now buy clicks at about $0.48 each, roughly a third cheaper. Nothing changed except the words above the image. That is the whole case for writing variants rather than one line: the auction price of reach is fixed by forces you do not control, but how many of those impressions turn into clicks is decided by copy you can rewrite for free.

Common primary-text mistakes that quietly get cut or ignored

Most weak primary text fails for one of a handful of repeatable reasons. Knowing them turns editing into a checklist.

The most common is burying the offer. The writer opens with context ("For over a decade, our team has been dedicated to...") and the actual reason to buy sits in sentence three, safely behind the See more fold where nearly nobody reads it. If the point of the ad is not visible in the first line, the ad has no point for most viewers.

The second is writing to desktop. It is tempting to admire your full paragraph on the big screen where all of it shows, but 98.5% of the audience is on a phone seeing two lines. Preview the ad on mobile, or better, in the actual placements you are running, before you decide the copy is done.

The third is the single undifferentiated block: one long run-on with no first-line hook and no line breaks. It shows as a gray smear the eye slides off. Break it up and lead with the sharpest line.

The fourth is keyword stuffing or link-dropping, cramming search terms or a raw URL into the copy in the hope it helps. It does not help delivery, it costs you visible characters, and it reads as spam to a human.

The fifth is the fake variant set: submitting three or five options that are near-identical, then concluding that "testing copy does not work" when they all perform the same. They performed the same because they were the same. Test angles, not adjectives.

The last is ignoring the shortest placement. An advertiser writes a lovely 140-character line for the Facebook Feed, runs it across Reels and Stories too, and never notices that in those placements the caption is clipped mid-thought. Write for the tightest window, then let the roomier placements look generous.

A primary-text checklist you can reuse

Before any primary text goes live, run it through the same short list, in this order.

1. Does the first line, inside roughly the first 80 characters, state the single most important benefit, proof, or offer? If the hook is not visible before See more, rewrite the opening line.
2. Would that first line read differently for a competitor? If not, it is filler. Make it specific to your product.
3. Is the body, if any, broken into short scannable lines, one idea each, ending on a clear next step? Cut anything that does not earn the expand.
4. Is the link out of the text and left to the button? Are there no personal-attribute claims that could trip Meta's standards?
5. Have you written 3 to 5 genuinely different angles, not word swaps, so dynamic delivery has a real choice? You do not have to fill all five slots, but you should fill more than one.
6. Have you previewed it in the tightest placement you are running, not just the desktop feed?

Primary text rewards the same loop as everything else in running a Facebook ad: write to how Meta actually renders and tests it, ship several distinct versions, read which line won, and feed that lesson into the next set. A platform like AdPlay.ai keeps that write-test-read cycle in one place, but the discipline holds with any workflow. The field is small and the fold is unforgiving, yet the fix is entirely within your control: lead with the line that matters, and give the system something worth choosing between.