Most bad agency work starts with a good agency and a bad brief.
You know the pattern. A kickoff meeting full of energy. A few weeks of silence. Then a presentation that is polished, expensive and somehow completely wrong.
The agency blames unclear direction. You blame the agency. Both of you are partly right, and neither of you gets the money back.
Clients rarely see their own part in it. When the BetterBriefs Project surveyed more than 1,700 marketers and agency staff, 80% of marketers said they write good briefs and only 10% of creative agencies agreed. Both sides guessed that about a third of the marketing budget is wasted on poor briefs and misdirected work.
Switching agencies will not fix that. A one page document written before the first meeting will.
Write down eight things before anyone meets: the outcome, the offer, the audience, the constraints, the proof, the budget, the decision maker and the success metric. Send it ahead, ask the agency to play it back in their own words, and judge every piece of work against it.
Why briefs fail
A brief usually fails in one of three ways.
It describes activity when it should describe an outcome. "We need a new website" or "we want to do more on Instagram" tells an agency what to make. Neither says what the thing is for, so the agency optimises for what it can control, which is how it looks.
It hides the constraints. The budget is secret, the compliance review is never mentioned, and nobody says the founder hates the colour green. Every hidden constraint becomes a revision round you pay for.
It has no finish line. Without a success metric agreed in advance, the work gets judged on taste. Taste changes by the week and by whoever is in the room.
An agency can only be as sharp as the brief it receives. Vague in, expensive out.
Talent does not appear on that list. A brilliant creative team handed a vague brief will produce brilliant work aimed at the wrong target. That can cost you more than mediocre work, because it is convincing enough to get approved.
Gather these before you write a word
Spend an hour collecting the raw material first, and the writing gets fast.
- Your last twelve months of numbers. Leads, sales, average order or job value, and what you spent on marketing. Even rough figures beat none.
- Five to ten real customer quotes. Pull them from reviews and sales calls, and keep the customer's wording.
- Everything you have already tried. What ran, what it cost, what happened. Otherwise you pay the agency to rediscover your failed experiments.
- Your brand assets. Logo files, fonts, colours, photography, any guidelines. Plus anything you are tired of and want gone.
- Your plan for the year. If you have written it down, this brief should fit inside it. If not, a one page marketing plan is worth doing first, so every project serves the same goal.
The eight part brief
Keep it to one or two pages. If it runs longer, you have not decided yet, and the agency will be paid to decide for you.
1. Outcome
What changes in the business if this works. Write a result where you would normally write a deliverable. "Fill the March intake" beats "a campaign." "More quote requests from commercial clients" beats "a new website."
Add a date. An outcome without a deadline is a wish.
If you have more than one outcome, rank them. "More leads and a more premium look" is two briefs pretending to be one, and they can pull in opposite directions. Tell the agency which one wins when they conflict.
2. Offer
What is being sold, at what price, with what guarantee, and why someone should choose it over the alternative. It should fit in a paragraph. When it will not, fix the offer first. No creative team can rescue a vague offer.
Owners skip this section more than any other, because they assume the offer is obvious. To an outsider it rarely is. If yours needs work, our guide to building an offer people cannot refuse will sharpen it before you pay anyone to advertise it.
3. Audience
Who buys, described as a person with a problem. "Women 35 to 54" tells an agency almost nothing. Say what your buyer worries about, what they have already tried and what makes them say no. The most useful thing you can hand over here is real customer language, word for word.
Include who you do not want as well. If a certain type of customer always haggles or costs more to serve than they pay, say so. An agency cannot avoid attracting them if nobody mentioned they exist.
4. Constraints
Everything that limits the work. Deadlines that cannot move. Brand rules. Claims you are not allowed to make. Platforms you will not use. Existing assets that must be reused. Put all of it on the page now, because every constraint discovered later costs a round of changes.
5. Proof
What evidence exists that the offer works. Testimonials, case studies, reviews, results, credentials, photos of real jobs. Hand over the specifics. Without them the agency has to write generic claims, and generic claims are hard to believe.
6. Budget
The real number, with media spend listed separately from fees. Hiding the budget will not get you a better deal. You get a proposal designed for someone else's wallet, followed by a second round of scoping to cut it back down.
7. Decision maker
One name. The person who can say yes, and whose yes is final. List everyone else who will give feedback, and be honest about whether someone senior will appear at the last minute with opinions. A late stakeholder can send finished work back to the start.
If that person cannot give the project an hour a week, name someone who can and give them real authority.
8. Success metric
The number that decides whether this worked, and when you will measure it. Cost per qualified lead. Booked appointments. Revenue from the campaign. Agree it now, in writing, so the final report cannot quietly swap it for impressions.
When results are soft, reports tend to drift toward reach, views and engagement, the numbers that always look healthy. We have written about why vanity metrics quietly drain budgets. A metric fixed in the brief is your protection.
A worked example
Here is a hypothetical brief, compressed, so you can see what good looks like.
Say you run a Newcastle accounting practice that wants more small business clients. A weak brief says: "We need a new website and some Google ads to grow the business." A strong one reads like this.
- Outcome. Twenty new small business clients on monthly packages by 30 June.
- Offer. Fixed monthly bookkeeping and BAS packages from a stated price, with a free first quarter review.
- Audience. Trades and hospitality owners doing their own books at night, scared of the tax office, burned by an accountant who never called back.
- Constraints. No claims about guaranteed refunds. Existing logo stays. Launch before the end of financial year rush.
- Proof. Forty Google reviews, three client stories the partners can tell, and a named contact for every client.
- Budget. A stated agency fee, with a separate monthly media budget.
- Decision maker. The managing partner, with the office manager giving feedback on process.
- Success metric. Cost per signed monthly client, measured at 30 June.
Same business, same agency. The second version gets a very different proposal, because the agency now knows what it is building, who it is for and how it will be judged.
How to use the brief
Send it before the first meeting. A good agency will come back with questions, and those questions show you where your own thinking has gaps.
Be wary of an agency that asks nothing and simply agrees. Either they did not read it, or they plan to do what they always do.
Then ask them to play it back. You want a short written response, in their own words, covering the outcome and the success metric. When their version does not match yours, you have just saved weeks of work heading in the wrong direction.
Keep the brief as the reference point for the whole project. Put it at the top of every status meeting. When a new idea appears halfway through, hold it up against the outcome and the success metric. An idea that serves them goes in. Anything else waits for the next project.
When you brief several agencies for a pitch, send them all the same document. The differences in their questions will tell you more than the differences in their slides.
The mistakes that undo a good brief
- Briefing the solution. Give them the problem and the outcome. If you have already decided it must be a video campaign, you are buying execution and the strategy is yours.
- Feedback by committee. Consolidate comments before you send them. Five conflicting notes from five people amount to a new brief.
- Changing the brief without saying so. Priorities shift. When they do, update the document and send it again. Silent changes are how scope creep starts and how invoices grow.
- Judging on taste. Go back to the outcome and the metric before you react to the creative. Ask whether it will work before you ask whether you like it.
Write it this week
Before you sign with anyone, or before the next project with the agency you already have, open a blank page and write the eight headings.
Fill in what you know. Mark what you do not. The gaps are your homework, and doing it yourself is far cheaper than paying an agency to do it for you through revisions.
Bigger gaps than you expected are worth knowing about too. Sometimes the real job is working out the offer and the metric before any creative work begins. That is the kind of thinking our consulting sessions are for.
A clear brief will not guarantee great work. A vague one makes it very likely you will pay for the job twice.
Frequently asked questions
What should a marketing agency brief include?
A strong agency brief covers eight things: the business outcome you want and by when, the offer being sold, the audience described as people with a problem, every constraint, the proof you have, the real budget with media separate from fees, one named decision maker, and the success metric that decides whether it worked. Keep it to one or two pages.
Should I tell an agency my budget?
Yes. Hiding the budget will not get you a better price. You get a proposal designed for a different client, followed by a second round of scoping to cut it back down. Share the real number and split media spend from agency fees, so the agency can recommend what is achievable with what you have.
How long should a creative brief be?
One or two pages is enough for most projects. If your brief runs much longer, it usually means key decisions have not been made yet, and the agency will end up being paid to make them for you. Length matters less than clear decisions on the outcome and the success metric.
How do I stop scope creep with an agency?
Agree the outcome and success metric in writing before work starts, and keep the brief at the top of every status meeting. When a new idea appears, test it against the outcome. When priorities change, update the brief and resend it so both sides see the change. Silent changes are where scope creep and bigger invoices begin.