Bad Work That Looks Good

I’ve been getting asked a lot lately whether AI is replacing marketers. My answer usually surprises people: I use AI every day.

Not because I think it can do my job. Quite the opposite. It’s become part of how I research, pressure-test ideas, organize information, challenge assumptions and move faster. After decades of doing this work, I’ve learned that it can help me explore further and sometimes see connections I might not have gotten to as quickly on my own.

But there’s a pretty important qualifier in that sentence: I know what good looks like.

And increasingly, I think that may be one of the most overlooked parts of the conversation about AI and marketing.

We keep talking about what these tools can do. Write a positioning statement. Build a marketing strategy. Develop a campaign. Analyze competitors. Create customer personas. Draft a website. Generate a hundred social posts before lunch. Technically, yes. But just because a tool can produce something doesn’t mean the thing it produces is good.

Garbage in. Garbage out. Only now it looks prettier.

That phrase has been around since the early days of computing for a reason. AI can only work with what you give it, and what you give it is much bigger than a clever prompt. It’s the context you bring, the knowledge you already have, the questions you know to ask, the gaps you notice and the assumptions you challenge.

That’s experience. And experience matters just as much on the other end.

I’ve had AI give me answers that looked completely credible: nicely reasoned, well organized and delivered with absolute confidence. There was just one problem. They were wrong.

I knew enough about the subject to stop and say, essentially, No. That’s not right. Go back and check it. And suddenly the answer became some version of, You’re right. Sorry about that.

Well...that’s awkward.

But it also raises a much bigger question. What happens when the person on the other side of that answer doesn’t know it’s wrong? What happens when the output sounds smart enough, looks polished enough and arrives quickly enough that no one questions it?

That’s the part of AI adoption in marketing that worries me far more than whether AI is going to “replace marketers.” Because the real danger isn’t bad work that looks bad.

It’s bad work that looks good.

The first answer usually isn’t the work.

One of the biggest misconceptions I see is that using AI well means writing one great prompt and getting one great answer. That’s rarely how I use it.

I argue with it. I tell it where I think it’s wrong. I feed it more information. I push it to defend its reasoning. I ask it to take another angle. I go back to the source material. I question assumptions. I rewrite and reconsider. And sometimes I throw the whole thing out.

Because strategy has never been about accepting the first answer. AI doesn’t eliminate judgment. It makes judgment more valuable.

The tools change. That part doesn’t.

I’ve been doing this long enough to watch desktop publishing, the internet, search engines, social media, marketing automation and now AI all arrive with the promise that everything’s about to change. And to be fair, things do change. Sometimes dramatically.

But one thing never does: the tool is almost never the competitive advantage. The person using it is.

So I don’t get too caught up in debates about which AI platform is best. The better question is whether the person using it actually knows what good looks like.

Give AI to someone with years of knowledge, judgment and pattern recognition, and you can dramatically expand what that person can accomplish. Give the same tool to someone who doesn’t yet have that foundation, and you can dramatically expand their ability to produce work that looks like they do.

Those aren’t the same thing.

Experience still matters. Maybe more than ever.

Clients don’t hire me because I know how to use AI. Millions of people know how to use AI. They hire me because I know what great strategy looks like. I know when something holds together, when it falls apart, when something doesn’t smell right and when the answer needs another round of questions.

That judgment came from doing the work over and over again. Seeing what worked, what didn’t and learning to recognize the difference. AI gives me leverage on all of that experience, and that’s incredibly powerful.

So yes, use AI. I certainly am. But don’t confuse access to the tool with mastery of the work.

AI can help you get there faster. You still need to know where you’re going.