AI in Marketing: How Far Has It Really Gotten? I Dug Into 20 Real Cases

A while back, a friend of mine who works in consumer goods invited me out for tea.

He told me his team had slashed half their outsourcing budget this year — their ad creative was now almost entirely AI-generated.

I had my doubts.

So I went home and spent a week scouring through dozens of reports, digging up every real case of generative AI being used in marketing that I could find.

And what I found rattled me.

Not because any single case was mind-blowing, but because this is already happening everywhere. Big corporations, small shop owners, fashion brands, snack sellers. This isn't some PowerPoint future — it's the balance sheets from last month, the month before.

Today I'll walk you through a few representative ones.

Let's see what's really going on.

One Ad, 90% of the Cost — Gone

Let's start with the most direct example.

An electronics company needed to shoot an ad for a new device.

The traditional playbook: hire a director, build sets, shoot, post-production — the whole process measured in months, the budget in millions.

They tried a different approach: the script was written by AI, the storyboard laid out by AI, the background visuals generated by AI.

The result?

Time and cost were cut by ninety percent.

What surprised me even more was that viewers couldn't tell which parts were AI-made.

What does this mean?

It means that for a whole category of ads, "should we even spend that money?" is no longer the question. The question has become "where do we reinvest what we saved?"

Think about it: when the most expensive step in an industry suddenly drops to a tenth of its cost, what happens?

Price wars will come. Industry shakeouts will too.

Seven Million Jars, Each One Different

Here's another one that really made me think.

A brand that sells snack sauces — a decades-old product whose jar label design hadn't changed much in ages.

One day, someone in marketing had an idea: use AI to give every single jar its own unique label.

Not ten versions. Not a hundred versions.

Seven million.

Every jar had a different pattern, a different color scheme — but you could still tell at a glance that it was their product.

The result? Within a month, they sold out.

Seven million jars.

I mulled this over for days.

We used to talk about personalization as "segmenting users and pushing different content to different groups." This case shows that personalization can go all the way to "every single product is one of a kind."

What consumers are buying is no longer just sauce. They're buying that little spark of pride — "I'm the only person in the world with this pattern."

When a mass-market product becomes a limited edition, it goes from feeding people to expressing who they are.

That shift is profound.

A Small Shop That Doubled Its Revenue in a Year

The first two were big companies. Let me tell you about a small business.

There's a one-person online sticker shop in the US.

The owner had to write product descriptions, come up with hashtags, design her store page. She either pulled all-nighters doing it herself or paid someone else to do it.

Then she handed it all over to AI. Product descriptions — AI wrote them. Hashtags — AI generated them. Her store's code — AI typed it out line by line. She even coined a term for this approach: "vibe coding."

In 2024, her revenue more than doubled.

Without hiring a single person.

You might think, well, that's just saving on labor.

But picture this: for a one-person business, doubling revenue means what? It means she might have gone from barely scraping by to actually putting money away. It means she might not have to go back to punching a clock.

For big companies, AI is a tool for cost reduction and efficiency gains. For small businesses, AI might be the lever that changes someone's life.

That difference is enormous.

Customer Service, Inventory, Copywriting — Three Problems Solved at Once

There's also a formal-wear distributor, a small operation with about a dozen people.

They used AI to tackle three things.

First, writing product descriptions. What used to take half a day per listing — AI cut that time by sixty percent.

Second, managing inventory. AI uses historical data and seasonal trends to predict how much stock to order for each style. Overstocking dropped by forty percent.

Third, customer service. Seventy percent of inquiries are now handled by AI chatbots first.

None of this is earth-shattering on its own.

But add all three together, and a dozen-person company managed to absorb a workload that used to require an entire team.

Put plainly: AI won't magically turn you into a different company, but it can make every person you have do the work of three.

When AI Draws Ketchup and It Always Looks Like One Brand

Now that we've covered "what it can do," I want to talk about something I found genuinely fascinating.

A heritage condiment company ran an experiment.

They asked AI to "draw ketchup" — any scenario, any setting.

The result: no matter how it drew it, the ketchup bottles the AI produced always looked like theirs.

When word got out, it blew up.

Some saw it as the ultimate proof of brand power: even machines think ketchup should look like this.

Others felt a chill: if an AI model's understanding of the world already has built-in preferences, then when it helps everyone draw images, write copy, and make decisions, is it quietly homogenizing everyone's aesthetics?

I won't draw a conclusion.

But this case made me realize something: AI isn't a blank slate. It carries all the data it's ever seen — the biases and preferences baked into that data — into every single thing you create.

You think you're using a tool. The tool is also shaping you.

The Controversy: Will Consumers Buy It?

Since we're on this topic, I have to be honest.

Not every AI-driven marketing campaign was met with applause.

A beverage giant used AI to produce an entire holiday ad campaign last year. Thousands of assets, produced in days, plastered across several cities.

It got roasted.

Some said the visuals looked fake. Some said it had no soul. Others questioned: this is a brand that's been around for over a century — holiday ads are one of the most treasured rituals between them and their customers — and you just slapped some AI together to fake it?

Another toy company made a one-minute brand film using AI, telling the founder's story.

That also polarized people. Some thought the technology was impressive. Others felt that taking a real person's emotional life story and handing it to a feelingless model to "perform" was inherently uncomfortable.

And there you have it — the real question facing marketers in 2025.

Technology is no longer the bottleneck.

The bottleneck has become: will your consumers accept a machine speaking on your behalf?

The answer to that question isn't in any tech report. It lives in that hard-to-define trust between every brand and its own customers.

A Few Judgments I've Arrived at Myself

After digging through all these cases, I summed things up into three takeaways for myself.

First: For AI in marketing, "can it be done?" is no longer the question — "can it be done well?" is where the dividing line now sits. Anyone can use AI to churn out a hundred pieces of creative. Which one can truly move people? That still comes down to understanding human beings. AI can't replace that.

Second: The money and time saved won't automatically become a better business. It just gives you a chance to think about questions you never had time for before: What does my brand actually stand for? What are my customers really paying for? Where should I reinvest these resources to make my brand more valuable?

Third — and this is the one that puts me most on guard: When everyone is using the same batch of AI tools, "the standard answer" will be everywhere, and "being different" will get more and more expensive. Those clumsy, inefficient things that can only be made by one specific person — those will become the scarce, valuable ones.

So you ask me, should you use AI for marketing?

Of course. Not using it would be stupid.

But while you're using it, don't forget to ask yourself one thing: Am I using AI to amplify my own uniqueness, or am I using AI to turn myself into everyone else?

That question is worth every marketer thinking about — honestly, alone, at 3 AM when no one's around.

Maybe once you've thought it through, you won't have just saved some ad spend.

You might just turn your brand into something no one else can replace.