Is AI really a shortcut to good marketing?

Does anyone still feel a little bit naughty when they use AI to help them with their work? Well, you shouldn’t! It’s not cheating. I use AI tools all the time. Most senior marketers do. They save me time, refine and even challenge my thinking and help me produce more work, quicker.  

But how much does AI move the dial on quality?

A 2025 survey by McKinsey on AI, found that only 27% of organisations using generative AI have someone check all AI-produced content before it is used. A similar percentage check 20% or fewer outputs before they go out the door.

In other words, most businesses are publishing or sharing AI-generated work without anyone experienced enough to know whether it’s actually good checking it.

This is the problem with relying on AI. What it produces is only ever as good as what you put in and what you're capable of judging.

In order to write a prompt that gets genuinely useful output, you need to know what good looks like. You need to understand an audience, the market and the company nuances that you’re working within, the brand, the competitive position etc. You also need to really understand what the business is trying to achieve and what will or won’t work. All that takes years of experience and know-how.

And then when reviewing output, it can’t be just a quick read through. You need to know if the information is correct, if its on brand, if it will land well with a CFO or CEO or if a client will find it engaging. You need to feel comfortable that you can put your name to it.  

These aren’t questions an AI tool can answer. But they are questions that experienced, commercially savvy people will instinctively know the answer to because they have sat in enough meetings, navigated the politics, made all the mistakes and have the scars to prove it!  

AI can speed up execution, but it can’t replace real expertise and experience.

If you think you can do marketing using AI without expert human oversight, you’re taking a major risk.