Here is the loop almost every business is stuck in right now. Ads cost more than they used to. Results slip. So you open an AI tool, generate forty new versions of your ad, ship them, watch most of them faceplant, and generate forty more. It feels productive, because the number of ads went up, and up is the direction numbers are supposed to go. But the costs keep climbing anyway, and nobody can quite say why.
Here is why. The problem was never that you had too few ads. It is that the ads are average, and the tool you reached for to fix that is an average-making machine. AI will make your ad. It will make it fast, and it will make it competent. The one thing it cannot do is have the idea, and the idea is the only part that was ever going to save you.
The Machine Pulls Everyone Toward the Middle
To see why, you have to understand what an AI tool is actually doing when it writes your ad. At its core it predicts the most likely next thing. Ask it for an ad and it gives you the most probable, most expected version of one, assembled from everything it has already seen. That is genuinely useful if your work was below average, because it drags you up to competent. It is quietly fatal if your goal is to stand out, because the most probable ad is, by definition, the one everyone else is also getting.
This is not a hunch. A 2024 study published in Science Advances had people write short stories, some with AI-generated ideas to draw from and some without. The AI-assisted stories were rated as more creative and better written on their own. But when the researchers looked at them together, they were noticeably more similar to each other than the stories written by people alone. The tool lifted each individual effort and, at the same time, squeezed the variety out of the whole pile.
Now picture your entire market reaching for the same tool, typing nearly the same prompt, and getting nearly the same ad. Everyone gets more competent and nobody gets more distinctive. The result is a feed full of ads that all rhyme, and a customer who has scrolled past your exact ad, in spirit, nine hundred times already. Average creative does not cure the fatigue that is driving your costs up. Average creative, produced in bulk, is what fatigue looks like.
Your Rising Costs Are an Idea Problem in Disguise
There is an old business novel called The Goal, by Eliyahu Goldratt, and its one big idea has aged perfectly for this moment. A factory is drowning. Orders pile up, nothing ships, and everyone is convinced the fix is more of everything: more machines, more shifts, more effort. Then someone points out that the whole system is being throttled by a single bottleneck somewhere upstream, and that pouring energy into any other part of the line does nothing but make you feel busy.
That is us. We have decided the thing choking our results is media: not enough spend, not enough variations, not enough automation. So we shovel budget and volume at it and feel like we did something. Meanwhile the actual bottleneck is sitting further up the line, and it is not the media at all. It is that the ideas stopped being interesting. No amount of budget buys its way past a boring idea. You can put real money behind a forgettable ad, and all you will get is a more expensive forgettable ad.
An Ad Has Exactly One Job: Get Remembered
It helps to be clear-eyed about what an ad can and cannot do. It cannot reach through the phone and pull a wallet out of anyone's hand. What it can do is make someone feel something strong enough that they remember you exist. Then, later, when their knee starts aching or their mind will not switch off at two in the morning, your business is the one that floats to the top.
That kind of memory runs on two things: emotion and surprise. Something that makes you feel, and something you did not see coming. The average, most-probable ad delivers neither. It is pleasant, forgettable wallpaper. So it does not get remembered, which means it does not get recalled at the moment of need, which means it does not sell, and you quietly pay for that failure on every single impression you bought. The tax on a boring ad is not that it flops loudly. It is that it works just badly enough that you keep funding it.
The One Thing the Machine Cannot Do
So if AI can produce the ad but not the thing that makes an ad worth producing, where does that leave you? In a surprisingly strong position, actually. Because the part it cannot do is the part that was always the real work: the idea. The specific, slightly unexpected angle that makes someone stop.
A prediction engine cannot originate that, because originating it means connecting two things that have never been connected before, and the machine only knows what has already been connected. It remixes the existing. The genuinely new angle has to be dragged out of a human head. That is not a limitation you should feel bad about competing against. It is the last unfair advantage you have, and almost nobody is using it on purpose.
Where New Ideas Actually Come From
The writer Arthur Koestler had a word for the mechanism behind every real creative leap: bisociation. You take two things that have no business being in the same room, force them together, and the idea that falls out of the gap between them is far better than anything you would have gotten by making one more version of what already exists.
This matters because it means creativity is not a personality trait you were either born with or not. It is closer to a muscle. You can activate it deliberately, on demand, in as much quantity as you want, by doing one simple thing: refusing to start from the obvious.
Here is what that looks like in practice. A sleep-supplement brand could run the ad everyone expects, the one about your tired body and your restless nights, which is precisely the ad your customer has already ignored a hundred times. Or someone could sit down, force the product together with something totally unrelated, land on the strange connection between a sleep gummy and the airplane-mode switch on a pair of headphones, and end up with an ad about the specific relief of total silence. That second ad did not come out of a swipe file or a competitor teardown. It got pulled out of a person's head on purpose. And it is the one worth watching.
Four Games to Break the Average
The good news is you can train this on command, and it takes about five minutes, away from the computer. Before your next ad brief, play one of these. Not with the AI. With your own brain.
The bridge. Take two random, unrelated words and get from one to the other in as few hops as possible, each hop connected to the last. It sounds like a party game because it is one. What it actually does is warm up the exact muscle that connects distant things, which is where fresh angles live.
Product plus the nearest random object. Grab your product and the dumbest, most unrelated thing within arm's reach on your desk. Force them into one idea. Most of what comes out is nonsense. Some of it is the strangest, most memorable ad you will run all quarter.
Emoji-only. Try to describe your product, or the problem it solves, using nothing but emojis. Stripping away the words you always reach for makes you find the feeling underneath them, and the feeling is what the ad is actually selling.
Alien archaeology. Pretend you just dug your product out of an ancient ruin and have to explain, with a straight face, what the people who made it used it for. The wrong answers are where the interesting angles hide.
Why the Weird-but-Clear Idea Wins
You might worry that "different" just means "confusing." It does not, as long as the idea is easy to grasp the instant it lands. There is a well-studied quirk of the brain called processing fluency: we trust and act on things we can decode quickly and effortlessly. The classic demonstration is two shapes, one round and one spiky, and two nonsense words, "bouba" and "kiki." Ask people which is which and about ninety-five percent hand the round shape to "bouba," because the word simply feels round in the mouth. It is not something anyone was taught, and it holds across most of the cultures researchers have tested.
The point is that your customer's brain is quietly reading every small signal in your ad, the setting, the sound, the messiness or tidiness of the background, long before they consciously decide to pay attention. An average, most-probable ad cannot engineer those signals on purpose, because it does not know they are there. A person who understands their customer can. This is also why, on platforms like Meta, your creative has become your targeting: the system reads the arrangement of your ad to decide who should see it, so a more distinctive, more deliberate arrangement does not just get noticed, it gets delivered to the right people for less. Being genuinely creative stopped being a soft, nice-to-have skill. It is now one of the cheapest ways to lower what you pay to reach the people who matter.
What to Do With AI on Monday
None of this is an argument against using AI. It is an argument about which job you hand it. Use the machine to execute: to make the thing quickly, to produce the variations, to take the expensive and slow parts of production off your plate. That is real, and it is a gift. Just never hand it the one thing it cannot do, which is decide what the ad should say in the first place.
So the workflow inverts. The scarce, valuable step is no longer making the ad, because making the ad is nearly free now. The scarce step is the idea, and the idea is also the cheapest thing you own to improve. It costs you five minutes and a willingness to start somewhere other than the obvious. Spend those five minutes before the brief, know your customer well enough to have a real insight about them, and let AI handle everything downstream of the idea. Do that and you stop competing in the one race everyone is losing, the race to produce more average faster.
The Last Unfair Advantage
When the tools are the same for everyone and the cost of making things falls to nearly zero, the only thing left that is genuinely scarce is a good idea. That has quietly become the most valuable thing in your entire ad account, and it happens to be the one thing you cannot buy, automate, or copy. It lives in your head, and it comes out when you make yourself connect things that do not belong together.
We help small businesses build creative that performs, which means we spend most of our time on exactly this part, the angle, not the output. So if you would rather have a partner do the thinking with you, we are glad to help. And if you would rather train the muscle yourself, even better. Either way, we're here.
