How to write the marketing brief your agency can actually execute
Most marketing briefs describe the company instead of constraining the agency's work. Five-section structure that turns a description into a constraint the agency can execute against.
By Stacey Tallitsch | May 19, 2026
You sent the brief on Tuesday. By Friday the campaign concepts came back. You read them on the train and somewhere around the third deck you realized it: the agency did exactly what the brief asked for. The work is not wrong. It is just not useful. You cannot tell the agency to do better because they did what the document said. You cannot write them a sharper brief because you do not know what was actually missing from the one you sent. That is the situation. You wrote a brief. The brief was internally consistent. The output is on-brief. And the output is not going to move the number. This post is the brief structure that does.
The description brief is the default failure mode
Most marketing briefs are descriptive documents. They describe the company, the audience, the brand, the budget, the timeline. They read like the about-us page with a deadline attached. The standard creative-brief templates circulating across the major marketing-blog ecosystem reinforce this posture — project overview, goals, target audience, brand guidelines, distribution, success metrics. Each section asks the founder to describe something. None of them ask the founder to constrain something.
This is not a template problem. This is a posture problem. A brief that describes the company tells the agency what the company is. A brief that works tells the agency what the customer is deciding. Those are different documents.
Harvard Business Review's research on why strategies fail names this exact pattern as the first of four root causes: not understanding the problem. Per a 2021 HBR analysis of why so many strategies fail to translate into execution, the most common failure mode across thousands of strategy engagements is the leader's inability to articulate the problem precisely enough for the executing team to act on it. The marketing brief is where this failure shows up first. The brief looks complete. Every section is filled in. The output is on-brief and off-target.
What the brief is actually for
A creative brief is not a creative document. It is a constraint document. Its job is not to inspire the agency. Its job is to tell the agency, with precision, what it cannot do.
This is the reframe most founders need before they rewrite a single line. The agency does not need your inspiration. The agency has 15 client decks open, two pitches due Friday, and a creative team hired specifically because they generate ideas. What the agency needs from the founder is the boundary. The customer it is solving for. The decision that customer is making. The alternatives that customer is comparing against. The thing the campaign will not be allowed to say. The number this campaign must move.
The default brief gives the agency a field. The brief that works gives the agency a corridor.
This reframe also explains why the "agency is the problem" complaint so often turns out to be a brief problem instead. The companion diagnostic for that situation lives in an earlier post on how to tell if your agency is the problem or your brief is. An agency executing inside a wide field will produce wide work. An agency executing inside a narrow corridor will produce concentrated work. Same agency, different brief, different output.
The five sections that turn a description into a constraint
The brief that works has five sections. Not seven. Not 12. Five. Each section is one to three sentences. The whole document fits on one page.
Section one: the customer's decision
Not the customer's demographics. Not the customer's pain point. The decision the customer is making when they encounter the campaign. "A regional manufacturing CFO deciding whether to renew the existing software contract or evaluate alternatives" is a decision. "Mid-market manufacturing CFOs" is not. The decision is what the output has to influence.
Section two: the alternatives in the room
What is the customer comparing you against, including doing nothing. "A competitor with a 30% lower sticker price and a worse implementation reputation; or staying on the legacy system for another two years." That one sentence rules out half the creative directions the agency would otherwise pitch.
Section three: the structural reason they choose you
Not your features. Not your unique selling proposition. The structural reason a rational buyer in this decision picks you. If you cannot answer this in one sentence, you do not have a marketing problem. You have a positioning problem, and no campaign will fix it. This is the same upstream failure mode covered in an earlier piece on why your marketing problem is probably a pricing problem — a brief that exposes a missing answer here is doing exactly what the brief is supposed to do.
Section four: what was already tried that did not work, and why
This section saves four weeks of rediscovery. List the three most recent campaigns or directions that did not produce. State the theory for why each one failed. The agency does not need a full post-mortem; it needs to know which directions to stop pitching.
Section five: the single metric this campaign must move
Not three metrics. Not a dashboard. One number. The number can be qualified leads, signed contracts, demo-to-close rate, average deal size, or pipeline velocity. The qualification does not matter — the singularity does. A brief that names three success metrics is a brief that has not decided what the campaign is for. This is also the upstream version of the metrics-versus-revenue mismatch covered in how to read your marketing dashboard when revenue is flat — a brief without a single named metric guarantees the dashboard divergence the founder will be debugging six months from now.
What to leave out, on purpose
The brief that works is shorter than the brief that fails. The sections that get cut are usually the ones founders feel the most pride in writing.
Cut the company history. The agency does not need it. If they need it, they will ask.
Cut the long brand-voice description. Send the style guide as an attachment if it exists. If it does not exist, the agency will build a working approximation faster than the founder can describe one.
Cut the channel speculation. The brief is not where the founder decides that this is a LinkedIn campaign or an email sequence. The agency picks the channel after the decision, the alternatives, and the metric are known. Specifying the channel inside the brief writes the conclusion before the analysis.
Cut the inspirational adjectives. "Bold, modern, refined, customer-obsessed" describes nothing. Strike every sentence in the brief that could appear in a competitor's brief without changing meaning. What remains is the brief.
The revision itself is short. Re-read the last brief sent. Cross out every sentence that describes the company. Keep every sentence that names a constraint on the customer's decision. If more than 10 sentences remain on the page, cut to 10. If fewer than three remain, the problem is not the brief — the problem is understanding of the customer, and that work happens before the brief.
A founder running a $4M industrial-services firm worked this exact protocol last quarter when an agency engagement was running off the rails. The original brief was 11 pages. The revised brief was one page. The agency produced new concepts in two weeks. The first concept the founder approved came from a constraint sentence she had not realized was a constraint until she wrote it down: "This campaign will not say anything about price." Once that constraint was on the page, the agency stopped pitching the discount-led directions that had been wasting three months of cycle time. The brief did not need to be better written. It needed to be smaller.
— Stacey Tallitsch, Stronghold CMO
About the Author
Stacey Tallitsch is the President of Stronghold CMO, a Fractional AI CMO service operating under Talisman Capital, Inc. He is a 30-year tech veteran and the author of 21 books on systems thinking, operator-grade decision-making, and personal sovereignty, with more than 30,000 students across his Udemy course catalog.