How to Use AI in Your Marketing, and Where It Still Needs You

Are you using AI for this? Founders ask me some version of that now on almost every first call, and there's a second question sitting underneath it that nobody says out loud.

Which is fair. You've been pitched AI by every vendor who emailed you this year, most of them louder than they were useful, so by now it's genuinely hard to tell what's real and what's a slide deck. When a marketing shop tells you it uses AI, the honest worry is not whether the technology works. It is whether that means faster work, or whether it means nobody senior actually read what you paid for.

That deserves a straight answer instead of a pitch: where AI genuinely helps in your marketing, where it shouldn't go near the work, and who is accountable for the difference.

What AI is actually good at

AI is good at the middle of the work. Not the thinking at the start, not the decision at the end, but the long grind in between that eats most of the hours on any marketing project.

It can read every competitor page in your market and tell you what all of them are claiming, in an afternoon. Hand it a strategy brief and it'll produce a first draft of a landing page or an ad set, so a writer edits instead of staring at a blank screen. It finds the pattern in six months of campaign data and flags the channel that's draining budget for nothing. And it'll check a hundred pieces of copy for the stale offer, the wrong phone number, the typo in the email about to go out to your whole list. Those are the real benefits of AI in marketing. None of them are exciting to talk about, which is probably why so few people describe them plainly.

You'll also hear the word agent, which just means a piece of software set up to carry a job through on its own instead of waiting for a click at every step. Useful. Also not magic. An agent running a research pass is running a research pass faster. It is not deciding what the research was for.

What AI should never touch

Three parts of marketing should never be handed to AI, and none of them are negotiable.

The first is the call on what your brand should say. That comes out of knowing your market, your customers, and what you can actually deliver on. It's a judgment about your business, not a writing task.

The second is whether a piece of work is any good. AI is confident in a way that has nothing to do with being right. It'll hand you something clean, on brief, and quietly wrong, and it'll sound just as sure as when it's right.

The third is the decision to publish. Somebody has to own what goes out with your name on it.

Take a draft that comes back for a client in a regulated field. It reads well. The structure is right, the argument holds, the grammar is clean. And it makes a promise about results the client isn't legally allowed to make. Nothing in the draft looks wrong. You have to know the client to see it. That's the catch that never gets automated, and it's the reason every piece of work needs a person who knows the account reading it before anyone outside sees it.

Where this goes wrong

The failure is not the technology. It is what happens when the review step disappears.

The temptation is obvious. If a system can produce ten times the output, the cheapest way to make money on it is to sell ten times the output and cut the people who used to check it. The work still looks like work. It arrives on time, it's formatted correctly, and for a while nobody notices that nobody senior has read a word of it. Then a claim goes out that never should have, or the message drifts a few degrees a month until it no longer sounds like the company it belongs to.

That's what business owners are sensing when they say they're nervous about AI in their marketing. It is not a fear of software. It is the fear of paying for judgment and receiving volume instead. The worry is well placed, and pretending otherwise is a worse answer than admitting it.

AI is changing how marketing gets made, and that part is fine. What cannot change is who answers for it at the end.

How VBO uses it

At VBO, a senior strategist reads everything before it reaches a client. That's the rule everything else is built around.

Underneath it, AI carries the parts of the job that used to eat the week: the research pass before a strategy gets written, the first draft nobody was ever going to keep as written anyway, the reformatting of one set of numbers into the report you actually asked for, the second and third checks a tired person misses at 6pm on a Thursday. None of that is where the value of hiring us lives, and all of it is where the hours used to go.

What stays with a person is the direction, the standard, and the sign-off. What to say, whether this is good enough, and whether it goes out.

We had to build it. That meant deciding what each piece is allowed to do on its own, where a person signs off, and what happens when it gets something wrong. Then learning the machinery well enough to keep it that way as the work changes. That part is slow, and it's most of the job.

A review step doesn't survive because everyone means well. It survives because somebody built it in. That's the real gap between a shop that uses AI and a shop that built a system around it, and from outside you mostly can't tell which one you're hiring.

If you want the full picture of how that works, including what we build and where the human review sits, it's all laid out there.

So how should you use AI in your marketing? Put it on the middle of the work and keep an experienced person on both ends. That holds whether you're doing this in house or paying someone else to do it.

Which brings it back to the worry underneath the question. Whether a shop using AI means faster work or worse work depends entirely on who reads what the machine produced, how senior they are, and whether they have the standing to throw it out. That's a fair thing to ask anyone you're considering, including us, and the answer should be specific rather than reassuring.

If you want more of how we think about the work first, there's more of our take on marketing worth a read. And when you'd rather just have the conversation, reach out to hello@vboadv.com. Ask us who reviews what. We will tell you straight.

VBO Advertising is a founder-led marketing consultancy and studio. We set the strategy first, then we run the work that proves it.