Campaign briefs case studies: campaigns, numbers and lessons

The brief is the one document a case study never contains. You get the finished content, a description of the strategy written afterward, and some numbers. The instruction that actually shaped the work is missing, along with everything that was rejected on the way.

This page names no brands, no creators and no results, because a case study you cannot verify is not evidence and repeating it would only launder it. It is about what a published case study can honestly tell you about briefing, which is less than it appears, and how to build the evidence you actually need. The general reading method sits in campaign case studies; this is the briefing-specific part.

What to take away

• Published case studies are written after the outcome is known, which makes the strategy section a reconstruction rather than a record.
• You cannot infer a brief from the content it produced, because you cannot see what was rejected.
• The only brief evidence that transfers to your situation is your own, and it takes about three campaigns to become useful.

Why the brief is always missing

Partly it is confidential. Claims policies, mandatories and rate-linked scope are not things a brand publishes.

Partly it is unflattering. Briefs contain the compromises: the section added because a stakeholder insisted, the constraint that was relaxed when the deadline moved, the mandatory that arrived late. None of that survives into a case study, which is written to demonstrate competence.

And partly the brief has stopped being the story. By the time a case study is drafted, the narrative is about insight and outcome. The instruction is treated as an implementation detail, even though it is usually where the campaign was decided.

The consequence is straightforward: a case study is evidence about execution and almost never evidence about instruction. Treat any lesson you draw about briefing from one as a hypothesis.

What you can and cannot infer from the finished content

What you can see	What it might mean	Why you cannot conclude it
A consistent structure across partners	A tightly specified brief	Or a platform format that produces this shape regardless of instruction
Wide variation in tone	Creative freedom was granted	Or the brief was vague and nobody reviewed the results
Identical phrasing in several pieces	Required wording	Or a caption supplied by the brand, which is a different arrangement
The product appearing late	A relaxed placement rule	Or a placement rule that was ignored and not enforced
Highly produced content	A production budget and a specification	Or the creator's normal standard, unrelated to the brief
No visible disclosure	Nothing about the arrangement	Possibly an undisclosed paid relationship, which is a reason to be careful rather than to copy

Every row has at least two explanations and the case study will assert one. Reverse-engineering a brief from published content is guesswork dressed as research, and it is confidently done all the time.

The survivorship problem, in its briefing form

You only ever see briefs that produced content that shipped and was worth publishing about.

Missing from the record: the briefs that were so constraining that good creators declined the work, the briefs that produced content the brand quietly never used, and the briefs that were fine but attached to a campaign whose numbers were poor. That is most of the distribution, and it is invisible.

There is a second layer specific to this channel. A brief that produced excellent work with a skilled partner may produce nothing with a different one, because a great deal of what a brief does depends on how much judgment the creator brings. "This brief worked" is a claim about a pairing, not about a document.

Claims that deserve particular skepticism

"Our brief gave complete creative freedom." Almost never true in the sense implied. There were claims restrictions, a disclosure requirement and a set of mandatories. What is usually meant is that structure and tone were left alone, which is a normal and reasonable arrangement rather than a philosophy.

"The one-page brief was the secret." Length is a symptom. A short brief works when the constraints are genuinely few, and fails badly when the unstated mandatories are numerous, which is why deciding what has to be locked comes before deciding how long the document is, as campaign briefs strategy sets out.

"We let the creator lead and the results speak for themselves." Check whether any comparison exists. Without a holdout, a stagger or a baseline stated in advance, the results speak about the period rather than about the campaign, for the reasons set out in measurement and ROI.

Testimonial quotes from the creators involved. These are collected from people who were paid, often while a relationship is ongoing, and they are frequently supplied for a marketing page. Whether a quote or endorsement in someone's marketing needs to disclose that connection is exactly the kind of question the current FTC guidance on consumer reviews and testimonials addresses, and it is worth reading before you either believe such a quote or publish one of your own.

Any number without a method. Uplift against what, measured how, over what window, compared with which period. If the method is absent the number is decorative, whatever it is.

"Award-winning." An award is evidence that a case study was well written and well entered. It is not evidence about the brief.

Questions to ask when someone presents a brief that worked

Ask to see the actual document, not the summary. If it cannot be shared, the discussion is about a memory of it.

Ask what was in the first version and what changed. The delta between drafts is where the real constraints show.

Ask what the creator asked about after receiving it, since those questions are the map of what was unclear.

Ask what came back that was rejected, and why. A brief whose owner cannot remember any rejected work either did not review anything or is not telling you.

Ask how many partners got the same brief and how the results varied between them. Variation within one brief tells you far more than the average outcome does.

Ask what they would change. The answer is usually the most useful sentence in the conversation, and it is never in the published version.

One thing a published case can be checked against directly is whether the work it describes was itself clearly identifiable as advertising. The FTC's native advertising guide for businesses sets out the reasoning, and a case study that celebrates content designed not to read as an ad is telling you something about the brief.

Build the evidence you actually need

Internal case studies about briefing are cheap and almost nobody keeps them. After each campaign, record the brief that was sent, with its version and date. Record the questions the creators asked. Record the revisions requested and, for each one, whether the brief covered it. That last field is the whole exercise: a revision the brief covered is a delivery problem, and a revision it did not cover is a brief defect.

Do that for three or four campaigns and you have something no external case study can give you: a list of the constraints your organization reliably forgets to state, and a view of which briefs your actual partners respond well to. That record turns campaign briefs from a template into a practice, and it is the only case study evidence that is genuinely about you.

Keep the awkward ones. A campaign where the content was fine and the numbers were poor is more informative about briefing than a success, because it isolates the brief from the outcome.

Bottom line

The brief is the missing document in every case study, and the finished content cannot tell you what it said. Assume survivorship, treat freedom claims and unmethodical numbers with suspicion, and ask to see the real document and the questions it prompted. Then build your own record of what your briefs failed to state, because that is the only briefing evidence that will transfer.

Common questions

Are external case studies worth reading at all?

Yes, for the questions and the vocabulary. Read them for what the team chose to worry about, not for their conclusions.

Can we ask a partner agency for briefs from previous work?

You can ask, and expect redaction. A redacted real brief is still more useful than a case study summary, because it shows the shape of the constraints. Ask alongside it how their review process handled disagreements, which is the content approvals question and the one that predicts what working with them is like.

What about a creator's own account of what made a campaign work?

Valuable, and biased in a different direction from the brand's. Both accounts of the same campaign are worth more than either alone.

Should we publish our own case studies?

If you do, publish the method alongside the result and say what you would change. That is rare enough to be genuinely distinctive, and it makes the document useful internally as well.