FOUNDATIONS

Why AI makes your marketing generic — and how to fix it

Here's an opinion most of the marketing world won't say out loud: used the way most people use it, AI makes your marketing worse — not because the writing is bad, but because it's average. And average is the one thing marketing cannot afford to be.

Average is baked in

A language model is, at its core, a machine for predicting the most likely next word. That's not a knock — it's genuinely miraculous. But "most likely" and "distinctive" are opposites. When you ask a model to write your homepage with nothing to go on, it gives you the most likely homepage: the statistical center of every homepage it has ever seen.

That's why AI copy has a smell. "We're passionate about delivering innovative solutions that empower our customers." Every business could have written it, which means it says nothing about this business. It's not wrong. It's just nobody.

Marketing is the opposite job. Marketing is how a specific person realizes this one is for me. You cannot get there from the average. You get there from the edges — the specific belief, the specific customer, the specific reason to choose you.

The fix isn't a better prompt

The instinct is to reach for prompt tricks. "Act as an award-winning copywriter." "Make it punchy." These help at the margins and then hit a ceiling, because you're still asking the model to guess what makes you different. It will guess plausibly and generically, because that's all it can do without information.

The real fix is to stop making it guess.

Give the model the three things it can't infer:

  • What you actually believe — the opinion you'd defend that others in your industry wouldn't. This is the source of every distinctive sentence you'll ever write.
  • Who you're actually for — not a demographic, a person. The one for whom your thing is exactly right.
  • What choosing you actually feels like — the before and after, in plain language, for that person.

Feed that in and the same model that produced beige suddenly produces something that sounds like a real point of view. Not because the AI got smarter. Because you gave it something to be un-average about.

AI doesn't replace knowing who you are. It amplifies it. Which means if you don't know who you are, it amplifies the void.

A quick test

Take a piece of AI-written copy about your business. Delete your name and your product from it. Now read it and ask: could a competitor paste their name in and use it unchanged?

If yes, it's average, and average won't move anyone. That's not the model failing. That's the model faithfully reflecting that you never told it what makes you you.

Why this is actually good news

It means the moat isn't technical. Everyone has the same models. The difference between generic and magnetic isn't a secret prompt or a bigger budget — it's clarity about who you are and who you serve. That's available to a solo founder and unavailable to a giant company that has never bothered to figure it out.

Do that work once, and every tool you own gets sharper. Skip it, and you've bought the world's most efficient machine for sounding like everyone else.

Want the foundation done with you?

We help people start and grow a business with AI–figuring out who you are, who it's for, and how the right people find you.

START WITH A CONVERSATION
← All posts