Brief a creative for paid social

The single most important line in a creative brief is what differs from what is already running.

If the answer is everything, what comes back is a new concept. That is a legitimate thing to commission, and it is a selection exercise rather than a test: whichever asset wins, nothing transfers to the next brief because there is no isolated difference to carry forward.

If the answer is one thing — this opening rather than that one, this proof rather than that claim, this format rather than the one we have run all quarter — then the result is a finding you can apply. Run a creative test that answers something is the discipline; the brief is where it either survives or does not.

The test for a well-named variable is whether you can write, in one clause, what you would do differently next time if this asset wins.

The job. What this asset is for, in business terms. Filling an intake, clearing stock, reaching an audience that has not heard of you. Not "improve performance".

Who it is for. The person, not the targeting settings. Settings change every few weeks; the description of who you are talking to does not, and it is what lets somebody judge whether a new execution is still aimed at the same person.

The variable. As above. One clause.

What has already been tried, and what it did. The field most often missing and the one that stops a brief re-exploring ground somebody covered in March. It does not need numbers — "openings that lead with the price have twice underperformed openings that lead with the problem" is enough, and it is more useful than a table.

The format, and the placements it has to survive. Formats are not comparable with each other, so an asset delivered in a format you did not plan for cannot be compared with the one it was meant to beat. Say which, and say where it has to work.

The constraints that are not negotiable. Claims the business cannot make, compliance limits, a proof point that must appear, a price that must not. These look like poor judgement to somebody who was not told.

What counts as working. The comparison rather than an absolute: better than the asset it replaces, on the measure you have already decided to read. Adding an absolute target with no basis behind it invites everybody to defend a number nobody sourced.

The creative brief template is these fields as a document, and the campaign brief template is the layer above it when the question is the campaign rather than the asset.

"Make it feel premium." Or fresh, or bold, or native. Every one of these is a description of a feeling rather than a specification, and what comes back will be judged against whatever the reader imagined. If the quality is genuinely the variable, show two references and say which is which.

"Something like what they are running." A brief built on somebody else's ad is a brief with no stated reason. You do not know what that ad is optimised for, who it is aimed at, what it is being measured against, or whether it works at all — and a long-running ad is as likely to be a neglected one as a winning one.

"Give us three options." Three assets made independently to the same brief differ in everything. You will pick one on taste and learn nothing, and you will have paid for three. Three variants of one variable is a different and much better instruction.

Two properties of the machinery that should shape what you ask for.

The name is load-bearing and is decided here. The asset name reaches every report downstream, carried onto the draft and then onto the entity, so the concept and the variable should be named in the brief rather than invented at upload. Name creatives so a report makes sense is that scheme.

Amending a live asset is not a small ask. A creative write swaps the whole specification rather than editing one field of it, so what reads as a tweak in a brief is a wholesale replacement in the account, and undoing it means putting the entire previous specification back. Briefing a new asset instead is usually what you wanted anyway, because it keeps the two periods' numbers apart.

Uploading media is also one of the operations no approval rule may ever clear, so it always waits for a person. The arrangement, stated once: your connections read your accounts, a change is written as a draft on the record with the state it would change from and the checks it passed, applying it is gated on approval — by you, or by a rule you wrote — and no platform write reaches a customer account until that platform has granted write access. Applying a cleared change to a platform is in private beta.

The brief and the test design are the same decision made twice if you separate them.

Before commissioning anything, know roughly how much volume the comparison will need to answer the question, and therefore how long it will run and what it will cost. If the answer is that the account cannot power the test, the brief should be for a bigger difference rather than for a subtler one — which is a better brief anyway.

Size a test before you run it is that arithmetic, and it belongs in the conversation before anybody starts making something.

Give the brief to somebody who has never worked on the account and ask two questions. What is the one thing this asset is trying to do differently. What would make us keep it even if it loses.

If both answers are in the document, the brief will produce work you can learn from. If the second one is missing, you will kill an asset that was telling you something about your landing page rather than about your creative — which is what Read hook rate and hold rate exists to prevent.