We gave AI two marketing briefs. Only one produced a Reel worth watching.

Here is the uncomfortable truth about AI marketing: a faster production system can produce an ordinary idea just as efficiently as a valuable one.

That does not make AI the problem. It makes the brief more important.

This week, the gap between those two things became hard to ignore. Consumer research showed growing fatigue with AI-generated brand content. At the same time, a small-business owner described lying awake worrying about what to post the next day. The market does not have a production shortage. It has a “what is actually worth saying?” problem.

So we ran a simple experiment.

We made two vertical cinematic films with the same Gemini Omni Flash 1.1 model and adult dance-studio subject. The generic production is a 10-second 1080 × 1920 montage built from three broad mood references. The Signal & Story production uses a lighter 720 × 1280 render while adding what a serious creative brief should: an identity-locked starting frame, defined character blocking, explicit continuity direction, a completed participation beat, and a two-second branded close.

The first brief was the kind of broad request that appears in marketing workflows every day: “Create a beautiful, cinematic Reel for an adult beginner dance class.”

The second started with a signal already present in the market: adults want to try dance, but many fear being the only beginner in the room. That tension became the hook, the character problem, and the emotional payoff.

Watch the difference

Two briefs. Two outcomes.

Turn the sound on and watch them in order. Both are cinematic AI productions. Only one begins with something the audience is already feeling.

01The generic AI brief

02The Signal & Story brief

Both films were generated through FAL using Gemini Omni Flash 1.1. The generic cut is 1080 × 1920 and runs 10 seconds. The Signal & Story version is an eight-second story cut from one uninterrupted, identity-locked 720 × 1280 generation, followed by a two-second branded end card with continuous music.

The first Reel is not bad. That is the problem.

It is polished. The movement is graceful, the light is warm, and the class looks appealing. It would probably pass an internal review.

But any dance studio could publish it. There is no character problem, no emotional turn, and no reason for a hesitant beginner to recognize themselves in the story.

AI is not responsible for that sameness. It faithfully completed a brief that contained nothing distinctive.

The research is not anti-AI.

A Gartner survey of 1,006 US consumers found that 65% believe brands are producing too much AI-generated marketing content, while 57% say it has made them less trusting of brand messaging. But Gartner’s conclusion was not to avoid AI. It was to use AI thoughtfully and prioritize usefulness, distinctiveness, and credibility.

That distinction matters. Consumers do not experience your workflow. They experience the result. AI can research, organize, script, edit, render, resize, and accelerate production. What audiences resist is content that feels abundant but says very little.

Do not ask AI to invent your conviction. Give it the signal, the evidence, and the position—then let it help you build.

The second Reel begins before the prompt.

Its starting point was not “make a dance video.” It was a customer truth: many adults want to dance but imagine walking into a room where everyone else already knows what to do.

That changes the creative direction. The newcomer is already inside the class, where we can read the nerves on her face. The question is simple: First class? She hesitantly raises her hand. Then hands rise throughout the room behind her. Nobody changes position or role; one shared gesture carries the entire reveal. The end card closes the loop: You’re not the only one. Everyone starts somewhere.

The technology did not invent the conviction. It amplified a tension worth interpreting:

Now the technology has something meaningful to amplify.

This is the difference between a content generator and a marketing workflow.

A generator begins with a prompt and ends with a file. A signal-first workflow begins with the market and ends with a connected campaign.

1. Find the signal. Notice the questions, objections, market changes, and language gaining energy now.
2. Interpret the tension. Decide what it means for the audience and why it matters.
3. Choose the position. Add the judgment a model cannot responsibly claim on your behalf.
4. Approve the story. Check the evidence, tone, promise, and intended action.
5. Use AI to build. Produce the Reel, supporting post, article, visual, and release plan from the approved foundation.

This is how Signal & Story is designed to work. Signal Intelligence helps find and source the opportunity. Story planning turns it into an approved direction. The production workflow then creates ready-to-review Reels and supporting content without asking the team to reconstruct the strategy in every tool.

Free field guide

The Signal-First Reel Brief

Eight questions that turn a vague content request into an evidence-led Reel brief. No email gate. Print it, share it with your team, or use it before your next generation.

Download the Free Brief

Before you generate, answer these eight questions.

The new advantage is not making more. It is knowing what deserves to be made.

AI has collapsed the distance between an approved idea and a finished asset. That is enormously valuable for founders and lean teams. It is also why weak briefs have become more expensive: they can now produce a whole campaign of forgettable material before anyone asks whether the idea was worth publishing.

The answer is not to slow production back down. It is to improve what enters the system.

AI did not make your content generic. The empty brief did.