For more than a year now, we’ve been all-in on AI — testing tools, pressure-testing their limits, and building real client work with them along the way. We’ve shared what we’ve learned about AI in design, AI in video, and how we apply AI across our work. But there’s one truth we’ve been circling in everything we’ve written, and it’s time to say it plainly:
On its own, AI will make you average.
Not bad. Not embarrassing. Average. And in B2B marketing, average is the most expensive thing you can be, because average doesn’t get remembered, and average doesn’t get chosen.
Understanding why that’s true, and what it takes to get past it, is where AI’s real value begins.
Why Average is the Default
Generative AI is, at its core, a prediction engine. It was trained on an enormous volume of what already exists: every landing page, every LinkedIn post, every “we help companies like yours” headline ever published. When you ask it for something, it gives you the most statistically likely answer. The middle of the bell curve. The consensus. That’s why AI output feels instantly familiar. It is familiar. It’s a synthesis of everything everyone else has already said.
This isn’t a flaw — it’s how the technology works, and it’s genuinely useful. AI compresses the distance from zero to a workable starting point better than any tool we’ve ever used.
The risk comes when the starting point gets mistaken for the finish line. When a company types “write a blog post about our supply chain solutions,” and publishes what comes back, they haven’t gained an edge. They’ve published the average of their entire industry, the same average their competitors can generate in the same thirty seconds.
And as more companies do exactly that, the sea of sameness gets deeper. Speed and volume are now table stakes. Distinctiveness is the new scarcity, which means the brands that invest in it stand out more than ever.
Generic In, Generic Out
“Garbage in, garbage out” has been true since the first computer. But with AI, there’s a subtler version that trips up smart teams: generic in, generic out.
The input is everything behind the prompt: a positioning that’s genuinely different and defensible. A real understanding of the audience — not “IT decision-makers,” but what keeps them up at night and what they’ve already heard a hundred times. A brand voice a competitor couldn’t credibly imitate. A clear strategic objective the work exists to serve.
In other words, what comes out of AI is only as strong as the marketing thinking that goes in, and that thinking doesn’t come with the tool.
Most AI output is mediocre not because the tool failed, but because it was never given a reason to be anything else. It answered the question it was asked. We’ve seen it firsthand across a year of hands-on work: thin input gets you something you’d delete. Real strategic depth gets you something worth building on.
Notice the phrasing: worth building on. Not worth publishing. Because even the best input is only half the equation, what you do with the output is the other half.
The First Draft Is Not the Deliverable
Even with great inputs, what AI produces is raw material. Teams who get real value from AI are separate from teams who just get the same output faster.
Producing fully AI-generated video for Mouser Electronics required dozens of iterations, multiple platforms, and constant judgment calls about when to refine, when to pivot, and when to rethink the approach entirely. The final videos look effortless, but the process was anything but, and the results were worth it.
The same is true for the written word, for design, for strategy. Between AI’s first output and work that performs, someone still has to:
- Recognize what’s wrong with it. AI is confidently wrong and confidently bland in equal measure. You can’t fix what you can’t see, and seeing it requires a trained eye.
- Push it past the obvious. The first output is the same one everyone else is getting. The value lives in the third, seventh, fifteenth round — where a person with a point of view asks, “what would make this ours?”
- Connect it to strategy. A well-crafted paragraph that says the wrong thing is worse than no paragraph at all. AI doesn’t know your competitive landscape, your sales conversations, or the campaign this asset fits into. People do.
- Know when to step outside the tool. Sometimes the right answer isn’t a better prompt. It’s a designer’s hand, a writer’s rewrite, or a strategic decision no AI model can make for you.
AI has made the distance from zero to average nearly instant. We use it for exactly that, every day. But the distance from average to exceptional hasn’t gotten shorter, and it matters more than ever. What that final output doesn’t show is the process behind it.
The Question Every Marketing Leader Is Asking
We hear it every week: “With these tools, can’t we just do this ourselves?”
Yes, AI will help you produce the first draft of almost anything. Easily accessible AI tools have made it easier than ever to mistake fast output for outcomes that matter. A deck full of content is not a strategy. A folder of generated images is not a brand.
What makes the output worth anything is the strategic input behind it and the expertise to judge and elevate what comes back. They’re the same things that separated effective marketing from noise long before AI existed. That’s where the right partner earns their place: not by gatekeeping the tools, but by steering them forward with real experience and data, turning the results into something unmistakably yours.
Average is Automatic. Exceptional Is a Choice.
The question isn’t “should we use AI?” The real question is who’s seeding it with genuine strategy, knowing (not guessing) whether what comes back is good, and doing the unglamorous iteration between the first draft and the final one.
AI raised the floor. People still set the ceiling — with strategy, taste, judgment, and the willingness to reject “good enough” a dozen times before landing on great.
To stay up to date on agency news, awards, and blog posts, subscribe to our Spire Wire newsletter.
