Everyone's talking about AI for marketing, and LinkedIn is full of takes on prompt engineering and the future of creative work. Sit in on most marketing reviews, though, and you'll see people using AI to make more content faster, with no idea what they're trying to say.
That's a strategy problem wearing a technology costume, and it starts with the one piece of work most companies have never done.
The garbage-in problem
AI multiplies whatever you give it. Feed it clear positioning, meaning a sharp brand essence, distinct messaging pillars and a defined competitive edge, and it produces content that sounds like you and lands with your audience. Feed it vague guidelines and a competitor's messaging with your logo swapped in, and it produces more of that, faster.
I've seen the same pattern across a lot of brands. The product beats the competition, the margins are strong and retention is solid, but the messaging is indistinguishable from the rest of the category: "Crafted with care." "Premium quality." "Trusted by thousands." So they buy an AI content system, and it multiplies the problem.
Now they have ten times as much indistinguishable messaging. AI only becomes useful once someone goes back and excavates what makes the brand different, what customers value and what the market has missed. With that clarity, the same system starts producing content that moves product.
This is the norm. Most organizations are trying to optimize their way past a weak foundation, and it almost never works.
What positioning is
Conversations about branding usually break down because people treat positioning and branding as the same thing.
Branding is the expression: your visual identity, your tone of voice, the colors, fonts and imagery you own. It's valuable work, and it sits on top of the positioning, which is the structure underneath.
Positioning is the irreducible truth about what your brand is, what it believes and why it exists in the market. It's what your best customers say about you when you're not in the room, and everything else gets built on it.
Most brands have never dug for it. They have a tagline, maybe a mission statement, a messaging framework an agency built five years ago and some tone-of-voice guidelines. None of it answers the basic question of what the brand is about.
Once you have that answer, everything gets simpler. The visual identity stops feeling arbitrary. Messaging stops needing consensus meetings, because there's a principle to measure it against. Social content sounds like it came from one brand. And your AI systems can finally do their job.
What clarity makes possible with AI
Building AI systems for brands has taught me that clarity changes what AI can do for you, well beyond making the marketing better.
With clear positioning, you can build content systems that hold the voice at scale and review workflows that catch off-brand work before it goes out. You can generate variations that are unmistakably yours, and you can have a marketer who's new to the brand producing on-brand work in their first week.
It also lets you work in fewer, sharper messages. Five on-brand messages will beat fifty generic ones, and once you know what you're saying, the job of AI is to amplify it. Testing and refining those few messages is what moves the numbers.
Without clarity, you're trying to find the signal by generating noise, and AI is extremely good at generating noise.
The four phases
This is where my methodology comes in. It has four phases, and they take a brand from unclear positioning to scaling with confidence.
Excavate is research and discovery: interviewing customers, reviewing the competitive set and working out what the market needs. You're looking for the gap between what the brand thinks it is and what it is in customers' minds.
Distill turns that research into a position: a brand essence, the pillars that support it, and the one thing you own that nobody else can credibly claim.
Amplify applies the position to everything the brand puts out, including the website, social, email, ads and sales materials, until it all sounds like one brand saying the same thing in different ways.
Industrialize is where most companies want to start. It's where AI holds the clarity at scale, generates variations, tests messaging and adapts in real time, and it only works once the first three phases are done.
Most companies jump straight to phase four, which is why they end up with more content and less clarity.
What this looks like in practice
I've seen this arc enough times to know its shape. A brand has real heritage, loyal customers and a product that outperforms, but its messaging could belong to anyone. It layers AI tools on top and gets exactly what you'd expect: more of the same, faster. The output is grammatical and completely forgettable.
The fix is always to go back to the beginning. Talk to the people who know the brand best, including founders, long-time employees and customers who adore it, and run a proper competitive review. Almost every time, the brand's real truth has little to do with what the marketing says. The brand sells "premium" while customers value trust and longevity, or it sells "innovation" while its real edge is consistency built over decades.
With that clarity, everything shifts: packaging, social voice, website copy and the way the brand talks about itself. The messaging gets grounded in something customers already feel but the brand has never said out loud.
The AI systems that were producing forgettable work can stay exactly as they are, with the same tools and the same team. Once they're fed something true, they start producing work that moves the business.
Why most companies skip this
Positioning work can be awkward, because it sometimes shows that the story you've been telling the market isn't true. Your claimed differentiation may not be real, your customers may value something you've never mentioned, or you may have been positioned wrong for three product cycles.
Most teams would rather not know. It feels easier to take the existing messaging and make it louder and faster with AI than to face the possibility that the foundation is wrong.
I understand the instinct. Excavation takes time and surfaces conflict, and turning up the volume looks cheaper in the short run.
The companies that skip it end up doing it anyway. They make more content, get worse results and spend more money until they realize they built on sand, and by the time they do the positioning work they're two years behind a market that has moved on.
The companies doing well with AI built clarity first and amplified it second.
The path forward
If you're going to invest in AI for marketing, and you should, first make sure you know what you're trying to say.
That means excavating: talking to your best customers, working out what you own in the market, and finding the truth about your brand that competitors can't claim. It's a strategic requirement, and the branding comes after it.
Once you have it, your AI systems start working for you, your team moves faster because everyone is saying the same thing, and your budget stops disappearing into volume with no signal.
