The brief is the most important document in marketing — and the least thought about. Most teams rush through it to get to execution. AI doesn't fix that instinct. But used correctly, it can force more rigorous thinking before anything is written.
The problem isn't the copy — it's the brief
When AI-generated marketing content disappoints, the usual diagnosis is 'the AI isn't good enough.' The real diagnosis, almost every time, is 'the brief wasn't specific enough.' AI tools are remarkably capable at generating content when given the right inputs. What they cannot do is compensate for a vague or underdeveloped brief. 'Write a campaign for our new product targeting young professionals' produces generic output not because the model is bad at copywriting — but because that brief is bad.
Invert the workflow
The teams getting the best results have flipped the process. Instead of using AI to generate content directly, they use it to pressure-test the brief before a single word of copy is written. The prompt isn't 'write this campaign.' It's 'here's our brief — what's missing? What assumptions am I making? What questions would a senior creative director ask before approving this?'
That single shift produces better input — which produces dramatically better output downstream. It also surfaces strategic gaps before you've committed to a creative direction.
Three prompts that improve any brief
- "You're a sceptical creative director. What's the weakest part of this brief, and what would you need to see to approve it?"
- "What are three different angles we could take on this positioning? List the risks and opportunities of each."
- "Our target audience is [X]. What's the most common objection they'd have to this message — and how would we address it?"
What this looks like in practice
At L'Oréal, we introduced a required AI interrogation step before creative review. Teams share their brief with an AI tool, run it through three to five critique prompts, and incorporate the feedback before presenting to the creative team. The result: fewer revisions, faster approvals, and noticeably more focused creative work.
The time invested in AI interrogation is recovered many times over in reduced back-and-forth later. The brief improves. The creative improves. The client relationship improves.
The skill that's actually in demand
What's emerging in marketing isn't 'AI skill' in isolation — it's the ability to think more rigorously about inputs. Prompting well is really brief-writing well. The teams outperforming their peers aren't using AI more; they're using it to hold themselves to a higher standard before they start. That's a skill worth developing intentionally.
The AI literacy gap in marketing isn't about knowing which tools to use. It's about knowing what to ask — and being precise enough about the problem that the answer is actually useful.