How to write a creative brief that gets you ads worth running

Most weak ads were weak before anyone opened a design tool. They were weak at the brief. The designer (or you, on the day you're your own designer) was handed "we need a few new creatives for the Q3 push, here's the brand kit," and from that nothing, produced something. The something looked fine. It scored a 4 on Hook and a 5 on Offer Alignment and it quietly died in the auction two weeks later.

A creative brief isn't a formality you do because the process says to. It's the single highest-leverage 20 minutes in the entire creative cycle, because it's the only point where fixing a problem costs a sentence instead of a re-shoot. This is the one-page brief that pre-answers the questions a high-performing ad has to answer. Fill it out before any work starts, and the work comes back usable the first time.

Why the brief is where ads are won or lost

A creative review at the end catches problems after you've paid to create them. A brief catches the same problems before anyone spends an hour in production. Same problems, two very different price tags.

The teams that ship work that compounds aren't more talented at the design stage. They're more specific at the brief stage. They hand over a brief that has already made the hard decisions: who this is for, what single thing it has to prove, what the hook angle is. The designer's job becomes execution, not mind-reading. And when the work comes back, the review is fast, because the brief already defined what "done" looks like.

When you brief vaguely, you outsource the strategy to whoever happens to be in the design tool. That person then guesses. Sometimes they guess well. Usually they default to "make it look nice," which is how you end up with a beautiful ad that sells nothing. We wrote separately about why your favorite ad is often your worst performer, and a vague brief is the most common way that ad gets made.

The one-page brief, seven fields

A good brief fits on one page. If it runs longer, you're writing a strategy doc, not a brief, and the designer will skim it. Seven fields, each one a single decision.

1. The one job

What is this specific ad for? Not the campaign. This ad. Pick one: stop a cold scroller, re-engage a warm visitor who didn't convert, drive a trial, recover a cart. One job per creative.

The most common brief failure is asking one ad to do three jobs. An ad that tries to introduce the brand, explain the product, and close the sale does none of them well, because each job wants a different hook, a different length, and a different call to action. If you have three jobs, you have three briefs.

2. Who it's for, and what they already know

Name the viewer and name their awareness level. Problem-unaware, problem-aware, solution-aware, product-aware. This single field decides more about the ad than any other, because it sets where the copy has to start.

Copy written for a solution-aware buyer reads as confusing to a problem-unaware scroller, and copy written for a problem-unaware scroller reads as patronizing to someone already comparing you to a competitor. If the brief doesn't pin the awareness level, the messaging dimension is a coin flip.

3. The one thing to prove

Every ad worth running makes one claim and proves it. Not five benefits. One claim, stated as something falsifiable. "Cuts your reporting time from a day to ten minutes" is a claim. "Streamlines your workflow" is a vibe.

Write the single claim in the brief, in the words the viewer would use, not the words your product team uses. The designer should never have to invent the value proposition. That's a strategy decision, and it belongs to you.

4. The hook angle

The first 1 to 2 seconds decide whether the rest of the ad gets seen, so the brief has to have a point of view on the open. You don't have to write the final hook. You do have to name the angle: a named pain, a pattern interrupt, a specific contrast, a numbered claim, a direct callout to a narrow audience.

If you leave this field blank, you'll get an establishing shot or a logo, because that's the safe default when nobody decided. We broke down what actually stops the scroll in the 3-second hook; pick the angle from there and write it in.

5. The proof you're handing over

Persuasion that isn't tied to something verifiable reads as noise. So the brief should hand over the actual proof the ad gets to use: the review count, the named customer, the specific number, the before-and-after, the certification. Not "add some social proof." The exact artifact, ready to drop in.

If you can't name a single concrete proof point, that's the brief telling you the ad will lean entirely on the offer to do the persuading, which is a known weak spot. Find the proof before production, not after.

6. Format, placement, and the constraints

Name the exact placement this ad runs in, and the constraints that come with it. Vertical 9:16 for Reels and Stories. Square with captions for feed. Sound-off readable, always. Safe zones for platform UI. Aspect ratio is not a detail you fix later; it's a shaping decision that changes how the whole ad is composed.

The most common version of this failure is a 16:9 horizontal video stuffed into a vertical placement with white bars top and bottom. That ad was briefed without a placement, and it looks like it.

7. What "done" looks like

End the brief with the bar. The clearest way to set it is the rubric you'll review against: the eight dimensions of a creative, each scored 1 to 10. Tell the designer the target before they start. "This needs to clear a 7 on Hook, Messaging, and Offer Alignment" is a brief that defines success in advance, instead of relying on a gut reaction at the end.

If you're not scoring against a fixed rubric yet, that's the thing to fix first, because without it "done" means "whenever someone in the meeting stops objecting." We laid out the full rubric in the 8 dimensions of a high-performing ad creative.

What a finished brief looks like

Here's the whole thing, filled out, for a single ad:

One job: stop a cold scroller in our ICP (operations leads at 50 to 200 person SaaS) and earn a click to the trial. Who, and what they know: problem-aware. They know reporting eats their Mondays. They don't yet know a tool fixes it. One thing to prove: monthly reporting goes from a full day to under an hour. Hook angle: named pain, opening on the Monday-morning dread, then the contrast. Proof handed over: "used by 4,000 ops teams," plus the named quote from Priya at [customer]. Format: 9:16, sound-off readable, captions burned in, CTA in the top half. Done looks like: clears 7 on Hook, Messaging, Offer Alignment. Trust at 6 or above.

That brief took ten minutes to write and it has already made every decision that matters. The person who builds from it cannot accidentally produce a logo-open, three-benefit, horizontal-letterboxed ad, because the brief closed every one of those doors. That's the entire job of a brief: close the doors that lead to weak ads, before anyone walks through one.

Briefing yourself when you're the whole team

Solo marketers and small teams hit a specific trap here: when you're briefing yourself, you skip the brief. It feels redundant to write down what's already in your head. It isn't, for two reasons.

First, the brief in your head is fuzzier than you think. Writing the seven fields forces the decisions you've been quietly deferring, like which single claim the ad makes. The act of writing "one thing to prove" and being unable to pick one is the brief earning its keep.

Second, you are the worst-positioned person to judge your own hook, because you already know the product. You can't un-know it, which means you'll skim past the exact 2 seconds a cold viewer needs in order to understand why they should care. The brief is how you force yourself to design for the stranger, not for the version of you that already gets it.

Close the loop: brief, build, score, ship

The brief defines the bar. The rubric checks the work against it. Those two halves are the same system, and a brief without a scoring step on the other end is just a wish.

So when the creative comes back, score it on the eight dimensions before it goes live, and compare the scores to the targets you wrote in field seven. If Hook came back a 5 against a target of 7, you know exactly what to fix and why, and you have a brief that already named the angle you can point back to. The conversation is "this didn't hit the hook angle we briefed," not "I don't love it."

If you want the scoring half handled in about 30 seconds, score the creative free on Adverdly before you spend a cent promoting it. You'll get all eight dimensions, scored 1 to 10, with the priority fix surfaced at the top, which is the fastest way to know whether the work delivered on the brief or quietly drifted off it.

A brief doesn't guarantee a great ad. Nothing does. What it guarantees is that the weak ads die on paper, in a sentence you can rewrite, instead of in the auction, after you've paid to make and run them. That's the trade, and it's the best trade in the entire creative cycle.