Written by: Ira Berg
In 1992, a prominent political strategist famously distilled Bill Clinton's entire campaign strategy into four words: It's the economy, stupid. It was blunt by design, a way to sharpen focus and force action.
For marketing in the age of AI, I'd offer a parallel: It's the prompt, stupid.
AI is genuinely transforming how brands are built, campaigns are run, and content is created. And yes, for certain tasks, DIY works. The tools are powerful, accessible, and increasingly affordable. But knowing when to use them, and how, is where most people underestimate the learning curve.
Everyone has attempted a DIY project. You watch a YouTube tutorial, feel confident, and get started. Early on, things go surprisingly well. The shelf looks level. The tile is mostly even. You're saving money and feeling capable. Then something shifts. A measurement is off, the wall turns out not to be plumb, or you realize mid-project that what you're actually dealing with requires a different set of tools entirely. That's when most people quietly pick up the phone and call a professional. And then there's the power washer. In much of the country, spring has arrived, and with it comes the annual ambition of tackling the outside of the house. You rent one on a Saturday morning, full of confidence. You start with the side of the house and the deck. Before you know it you've stripped the paint off the siding and splintered the wood, and what started as a weekend project has become a contractor conversation on Monday morning.
AI in marketing works exactly the same way.
Need a logo for a family reunion? A kid's bar mitzvah? A sweet sixteen? AI delivers. It's fast, it's good enough, and the stakes are appropriately low. The same logic applies to a quick first-pass logo for a newly launched fund, something clean with a mountain, a bridge, or a river. For contained, relatively generic tasks, the DIY approach works, and there's no shame in using it.
But building a fully integrated brand system, one that captures a firm's DNA, resonates with a sophisticated institutional audience, and translates consistently across a pitch book, website, investor letters, and event materials, is a different project entirely. That's the moment the YouTube tutorial runs out, and the wall turns out not to be plumb.
Here's why. AI is only as good as the brief you give it. And the brief is only as good as the strategist behind it.
Think of it this way. If you develop a rash that's lasted a few days, you can certainly search your symptoms online. But a good physician doesn't just glance at a symptom and react. They examine. They question. They review your history, because all of it matters in arriving at the right diagnosis and course of treatment.
Marketing is no different. Knowing which questions to ask, when to push back on an output, when to reframe the prompt entirely, and how to connect the results to a broader brand narrative, that's not something a first-time user figures out quickly. Like the power washer, the early results can look promising. It's when the complexity increases that the damage shows up.
And while the prompt is where it all starts, it is really just the surface. Like most things in life, there is considerably more underneath. The choice of tools, the sequencing of inputs, the ability to recognize a good output from a mediocre one, and knowing how to iterate when the results fall short, all of that matters just as much. And then there is something deeper still, a layer that most casual users never reach. It is a space we have been exploring with real curiosity, and one worth understanding. This series, on how firms should think about AI usage and implementation across their businesses, is just getting started. Where we are headed will change how you think about AI entirely, and have you questioning whether it makes sense to have a professional help guide you through it.
This isn't an argument against AI. Far from it.
At Ext. Marketing, we've leaned into the technology, hiring a CTO focused on AI integration, building it into our workflows, and using it to deliver sharper work, faster, at better value for our clients. We're not threatened by the tools. We use them every day.
What we've found is that the firms getting the most out of AI are the ones with experienced professionals guiding the process, setting the parameters, refining the outputs, and ensuring the end product actually reflects the firm's positioning rather than a well-phrased but generic approximation of it.
For clients who want to build that capability in-house, we offer what we've been calling an AI Integration Officer (AIIO) engagement, a focused, hands-on program where we assess the tools a firm already has, connect them in ways that create real operational leverage, and help teams develop the prompting discipline to use AI effectively over time.
The technology is remarkable. But in marketing, as in medicine, the quality of the outcome depends on having a qualified professional involved. One who knows how to diagnose in the case of the doctor, and how to create in the case of the marketer, by asking the right questions from the first conversation to the final deliverable.
It's the prompt, stupid.