Creative brief template

A creative brief is one row per asset. Not one per campaign and not one per concept: one per thing that has to be made, because each of them has its own format, its own ratio, its own hook and its own due date.

The row carries what the asset has to do — the hook, the one message, the proof behind it, the action it asks for — and what it has to satisfy: the specs, the things that must appear, and the things that must not. The last column is the one that saves the most money.

Write it before anybody opens an editing tool, and after the campaign brief has settled the offer and the audience. A creative brief written first will produce beautiful work for a campaign that has not been decided yet.

Use it again at review. The row is the thing an asset is checked against, which turns a subjective conversation into a short list of yes and no.

A campaign brief decides what the campaign is for. A creative brief tells somebody what to make, and the only sensible unit for that is a single asset.

This sounds pedantic until you watch what happens when it is ignored. A brief covering a campaign's nine assets gets read by an editor who has to infer, for each one, what ratio it is, how long it runs, which of the three messages it carries and what its first frame should be. Nine inferences later there is a set that shares a style and no logic, and the review meeting is spent reconstructing what was intended.

One row per asset removes that. It also makes the work estimable, because a list of rows is a list of jobs, and it makes a result attributable afterwards, because the asset identifier travels from the row to the file to the ad name to the test.

Most viewers see the first moment of an asset and nothing else. That makes the hook column the highest-value cell in the row, and it is the one most often filled with an adjective.

A hook is a description of what actually happens: the product already in use in the first frame with no logo and no title card; the first line naming a cost the reader is already paying; the before state held for two seconds with no music. Written that way, two people briefed from the row would build roughly the same opening, which is the test of whether a hook has been briefed at all.

The usual objection is that specifying the hook constrains the maker. In practice the opposite happens. A specific hook frees everything after it, because the maker knows what the first two seconds must do and can spend their judgement on the rest.

The message column asks for one thing the viewer has to believe. Not three, and not a list of features.

The reason is mechanical rather than aesthetic: nobody stays long enough for the second message. An asset that tries to carry three delivers the first one badly and abandons the others, and it performs worse than any of the three would have on its own. If there are three things worth saying, that is three rows.

The proof column exists because a message without proof is an assertion. Proof is a demonstration, a number with a source behind it, a review, or the process shown rather than described. Write which one this asset uses, because that decides how it is made — proof by demonstration needs a real shot of the thing working, and that is a production constraint, not a creative note.

The specs column is the least interesting one and it is where money is lost.

Ratio, duration, dimensions, file type, safe areas, text limits, caption requirements. Every one of those is knowable before anybody starts, and every one of them causes a re-edit when it is discovered afterwards. The specific failure that recurs is framing: an asset built at one ratio and cropped into another loses its composition, and a subject that was centred in the original is cut in half in the placement that actually delivers.

Write the ratio the asset will run at, not the ratio it is convenient to shoot. Where one asset has to serve two placements, that is either two rows or a spec that says both, with a safe area that works in each.

Almost every expensive re-edit traces to something nobody said out loud.

A competitor's product visible on a shelf. A claim that has not been cleared. A price on screen that changes next month. Music that cannot be licensed for paid use. Stock footage in a brand that does not use it. A logo in a position the platform overlays. None of these are obvious to somebody who was not in the earlier conversation, and all of them are cheap to write down.

So the column is filled in honestly and specifically, with the reason attached where there is one. A must-avoid with a because is a rule somebody can apply to a situation you did not anticipate; one without is a list to be worked around.

The due date is the day the finished asset is needed, and it is not the day the campaign starts.

Between the two there has to be room for review against this row, at least one re-edit, and the upload and approval time of whichever platform it runs on. Teams that set the due date at the run date discover this every time, and the flight starts on last quarter's creative while the new set is in review.

Where an asset is late, the honest options are to start the flight without it and say so in the plan, or to move the start. Quietly running old creative and reporting the flight as planned is how a set of results becomes unattributable.

Seven checks. There is one row per asset. Every hook describes what actually happens. Every asset carries exactly one message with a proof named beside it. The call to action uses the same words as the landing page. The specs include ratio, duration or dimensions, and safe areas. Must avoid is filled in, with reasons. And every due date leaves room for a review and a re-edit before the day the asset is needed.