In May, Google announced a feature called AI Brief. It’s a Gemini-powered interface that lets advertisers steer AI Max campaigns in plain language: what the ads should say, what they should never say, which searches to pursue, which audiences matter, and how to speak to them differently.
Most agencies have an AI budget line now, with the tools to match. What far fewer have is an input process that can successfully feed those tools.
Deloitte’s State of AI in the Enterprise 2026 report found that 37% of organizations are using AI at a surface level, with little or no change to existing processes. If you work at an agency, you’ve probably seen this firsthand.
What’s Breaking In Agency AI Workflows
Think about how briefs used to work. You wrote “millennial homeowners, premium positioning, avoid discount language,” and a media buyer spent two weeks turning that into keyword lists, placements, and copy variants. Every gap in the brief got filled by a person making judgment calls.
But with AI, whatever you leave out of the brief, AI fills in on its own. If you’ve noticed AI-driven campaigns producing output that feels generic or off-brand, this is usually why. The brief that was good enough for a human handoff isn’t good enough for a machine that takes you at your word.
Why The Brief Carries More Weight Than It Used To
Google’s AI Brief accepts three kinds of guidance:
- Messaging guidelines covering what ads should and shouldn’t say
- Matching guidelines setting boundaries on the searches to capture or avoid
- Audience guidelines describing how to serve tailored messages to the audience you want
Most agencies have answers to those questions somewhere. But they live in a kickoff deck, or in the head of the person who ran the pitch, rather than in a document the campaign itself consumes.
What Belongs In A Brief Written For AI
A few things separate briefs that shape output from briefs that get ignored:
Write the negative space, not just the positive. Most briefs cover what the brand wants to say, but AI needs clearer boundaries, because it will confidently fill whatever space you leave open.
Describe the audience as a person, not a segment ID. “Adults 25 to 54, household income over $100,000″ tells AI almost nothing about what to say. “People who have already researched three competitors and are looking for a reason to stop looking” gives it more to work with.
Define success in business terms. A brief declaring “improve return on ad spend” invites generic output, because it names the scoreboard instead of the game. A brief that specifies which customers are worth acquiring, what they’re worth over time, and which conversion event correlates with revenue gives the system something more to optimize toward. The strongest briefs tend to come from teams who have done propensity and lifetime value work first, because they know which customers matter before they start writing.
Treat the brief as a living document. The best agencies revise briefs against what the system produced last month, the same way they’d revise a bidding strategy. A brief written once at kickoff and never revisited is educated assumptions before you had evidence.
Why Brief Quality Is Also A Data Quality Problem
Before you write the brief, there are three questions you need to be able to answer.
First, can you describe your audience by what they do, not just who they are? AI Brief lets you write audience guidelines in plain language, but it can’t segment by device or tell you whether the same person visited on their phone Tuesday and converted on their laptop Thursday. If your data doesn’t connect those dots, you’ll brief toward separate visitors instead of one real customer. Ask your data team whether identity is resolved at the person level before you write.
Second, do you know what a customer is truly worth? If your team can’t connect a conversion event to a number finance recognizes, the brief will default to “improve return on ad spend” and the system will optimize toward form fills that never become sales. Get alignment on which events count before the brief goes live.
And finally, do you know what you don’t want included in the brief? Think of the customers, claims, and tone you want to avoid.
AI Brief is a better control surface than anything agencies have had before. But it depends upon your guidance and your data. Make sure both are solid before you use the solution.
How Agencies Should Build The Capabilities
The good news is if your people have spent the past two years learning to prompt AI, they’ve already been practicing the skills they need for brief writing in this new era.
Give the brief an owner senior enough to push back when a client’s stated goal and the client’s data point in different directions. Review briefs the way you’d review a media plan. And train by role. Strategists need to get comfortable writing boundaries and describing audiences in behavioral terms. Account leads need to know which questions to ask in kickoff so the brief has something to work with in the first place.
None of this needs new technology, and you don’t have to do all of it at once. Pick the account where a thin brief is costing you the most right now, rebuild it properly, and let the results make the internal argument for you.
The Client-Facing Stakes
Clients are already comparing AI output across their agency partners, and they can tell which one had something specific to say about their business.
An agency that briefs well looks like it understands the client’s business.
The agencies pulling ahead right now are the ones who decided the document at the front of the process deserved to be treated like strategy, gathered the right data, and staffed it accordingly. That’s a change any agency can make, and it costs nothing but attention.
As Head of Agency Growth at Adswerve, Kelly Hayes focuses on helping agency partners win. With a career spanning over a decade, she is responsible for driving revenue growth and retention while fostering deep-rooted relationships with senior stakeholders across the agency landscape. In addition to curating premium inventory with publishers and supply-side platforms, she ensures her agency partners have a distinct competitive edge. Kelly’s career is defined by her ability to turn complex challenges into commercial successes. During her tenure at Outbrain (now Teads), she was instrumental in architecting incentive-based agreements for agency partners and high-growth DTC brands. A collaborative leader known for her “get-it-done” mentality, Kelly is recognized for her ability to scale partnerships through innovative thinking and impactful storytelling.
