How to Write Facebook Ad Copy That Converts

Good Facebook ad copy does three things: it earns the first line, it speaks to one specific person, and it makes one promise you can keep. Everything past the first 125 characters is read only by people the opening already convinced.

That constraint is the whole craft. You are writing a headline that happens to have paragraphs attached.

What makes good Facebook ad copy?

Specificity, and a reason to keep reading before the "See more" cut.

Weak copy sounds like a brand describing itself. Strong copy sounds like a person describing a problem the reader recognises. The difference is usually one level of concreteness.

• Weak: "Premium skincare for healthy, radiant skin."
• Better: "Three weeks of retinol and my skin got worse before it got better. Here is why that happens."

The second one is not cleverer. It is more specific, it names a real experience, and it creates a small open loop the reader wants closed.

Every sentence in the opening has one job: buy the next sentence.

What structure should I use?

There is no magic template, but this sequence fails least often.

Notes on each part.

Hook. A problem, a claim, a number, or a contradiction. Not a greeting, not a rhetorical question, not your brand name.

Context. This is where trust is built. Show that you know what the reader has already tried and why it did not work.

Turn. The product should feel like the obvious next line, not a swerve into a pitch.

Proof. Specific beats impressive. "412 reviews" is stronger than "loved by thousands". An honest limitation ("it takes about six weeks") often converts better than another superlative.

Ask. One action. Shop now, read the guide, get the sizing chart. Not three.

How long should the copy be?

Long enough to earn the click, short enough that the first line is not wasted.

Both short and long copy win. What loses is medium copy that says little.

The honest rule: cold traffic rarely reads long copy. Retargeting audiences do, because they already have a reason to care.

How does the audience change what I write?

Awareness level changes the job of the copy entirely.

• Problem unaware. Lead with the symptom they would recognise. Do not mention the category yet.
• Problem aware. Name the problem precisely and explain why the usual fix fails.
• Solution aware. Compare. Say what makes your version different in one concrete way.
• Product aware. Remove the objection. Shipping, returns, sizing, price framing.

Running the same copy across all four is the most common reason a good ad stops working when it scales. The audience widened, but the message stayed written for the people who already knew.

What should I cut?

Cut anything that could appear in a competitor's ad without changing.

Words to delete on sight: premium, innovative, cutting-edge, solution, journey, empower, transform, curated, bespoke. Not because they are ugly, but because they are unfalsifiable. A claim that cannot be wrong cannot be persuasive.

Also cut:

• Greetings. "Hey there!" costs you a quarter of your visible line.
• Your brand name in the first sentence. It is already on the ad.
• Three exclamation marks doing the work of one specific fact.
• Emoji bullets used as decoration rather than structure.

Headlines follow the same discipline, and the patterns worth reusing are collected in Facebook ad headlines.

How do I find copy patterns that already work?

Read ads that have been running for a long time.

The Meta Ad Library at facebook.com/ads/library shows every ad currently running from any advertiser worldwide, free and with no account. Each card shows a "Started running on" date.

That date is the useful part. Nobody keeps paying to run an ad that loses money, so an ad still running after three months is a tested piece of copy, not a guess. That is the closest thing to a performance signal the library exposes, and it is explained further in Facebook ad longevity.

Read fifty long-runners in your category and the patterns show up fast: which objection everyone handles first, whether the category leads with price or with proof, how long the winning copy tends to be.

The library keeps no history, so ads vanish when they are paused. Our free Meta Ad Library Downloader exports a whole search into one ZIP with a searchable swipe board and a CSV containing every ad's full text, which makes copy patterns searchable rather than something you scroll past.

A structured way to break those ads down is in how to analyse competitor ad copy, and finished examples worth studying sit in Facebook ad copy examples.

How many versions should I test?

Three to five hooks against one body, rather than five completely different ads.

Changing everything at once tells you which ad won but not why. Holding the body and the creative constant while the first line changes gives you a reusable answer: this audience responds to price framing, or to the time-to-result claim, or to the objection being named upfront.

Then take the winning hook and rebuild the body around it. Hook strength is measurable in video through hook rate, and the same logic applies to statics through the click-through rate on the first line.

The short version

Write for one person. Put the strongest specific thing in the first 125 characters. Prove you understand the problem before you sell the fix. Cut every word that could belong to any competitor. Then go read ads that have been running for six months and copy their structure, not their sentences.

FAQ

How long should Facebook ad copy be?

Short copy works for visual, obvious products; long copy works for expensive or high-scepticism purchases aimed at warm audiences. What consistently fails is medium-length copy with a weak opening line.

What makes good Facebook ad copy?

Specificity and a first line that earns the second. Copy that names a real problem in concrete terms beats copy that describes the brand in flattering, unfalsifiable adjectives.

Should I use emojis in Facebook ad copy?

Sparingly, and as structure rather than decoration. One arrow or checkmark can help a list scan, while rows of sparkles signal low-trust advertising in most categories.

How do I write a Facebook ad hook?

Start with the problem, a surprising number, or a contradiction the reader recognises. Write ten versions of the first line and pick the most specific one, not the most polished.

How many versions of ad copy should I test?

Three to five hooks against a single body and creative. Changing everything at once tells you which ad won but not which idea was responsible.

Can I copy a competitor's ad copy?

Copy the structure and the angle, not the words. Reusing someone's actual sentences carries legal risk and rarely works anyway, because their copy was built around their proof and pricing. This is not legal advice.