The 35% of Marketers Still Not Using AI Are Quietly Being Left Behind

Translator's note: Liu Run is a Chinese business consultant and writer. The article is written in his voice, addressed to a friend.

A few days ago, a friend who runs a brand complained to me.

He said, "Liu Run, I know AI is the trend. But my plate is already full — copywriting, ad scheduling, email, chasing data — I'm half-dead. You want me to add another layer of AI tools on top of that? Can I even learn it all?"

I said, "You're asking the wrong question."

He paused.

I said, "What you should really be asking is — the peers who already use AI, how much more are they earning than you right now?"

I laid out a few cases I'd come across over the past few months. Today, I'll lay them out for you too.

One Jar of Nutella, Seven Million Packages

Let's start with the one that least sounds like an AI story — a jar that holds chocolate.

Nutella. That hazelnut cocoa spread. The Italian one.

It did one thing: it had AI scrape social-media discussions, consumer preferences, what colors and flavors were trending right now — all of it. Then it generated seven million unique jar designs, each one mapped to a specific audience segment.

How many did it sell? More than the regular edition.

You see, packaging used to be designers' gut calls. Now AI takes one look at what the whole internet is saying — and it's more accurate than a designer. Designers get tired and have moods. AI doesn't.

A Lingerie Brand Saw Ad Clicks Jump 20%

Cosabella. An Italian lingerie brand — not a big one.

How did it used to run ads? A few people from marketing would sit in a room — write copy, pick images, run A/B tests. One round took a few weeks.

Then they let AI take over. AI did two things: first, it dug through their historical customer data to find "which kind of person is moved by which kind of copy"; second, it generated personalized ad creative and email content directly, distributed by audience segment.

The result: click-through rate climbed 20%.

Twenty percent doesn't sound like much. But think about it — clicks that used to cost a dollar now cost eighty cents. Over a year, the savings add up to real money.

Volkswagen Slashed Its Ad Budget by 15%

Volkswagen. The German automaker.

It handed ad placement over to an AI predictive model. What did the AI do? It watched the historical data and market swings, then worked backward to figure out "where the next dollar should go" — and kept adjusting in real time based on live engagement data.

How much did it save? It cut ad spend by 15%.

Fifteen percent. That's not small change. A multinational automaker's annual ad budget runs into the billions. Fifteen percent is hundreds of millions of dollars.

Money saved like that could fund a whole new car model.

A Mattress Company Compressed Content Production from "Months" to "Days"

Tomorrow Sleep. They make smart mattresses.

What's the most painful part of content marketing? It's slow. A single piece of in-depth content — from topic selection, research, writing, to review — easily takes a month.

Tomorrow Sleep let AI take over content production. AI reads what users are searching for, asking about, caring about — then automatically produces targeted content.

The effect? Engagement rose 30%, and website traffic rose with it.

But more important than that number: a content matrix that used to take months to roll out collapsed into a few days.

Move faster, and your cadence quickens. Cadence quickens, and competitors can't keep up.

JP Morgan Let AI Write the Copy

JP Morgan Chase. One of the largest banks in the United States.

It handed its ad copy over to AI — specifically, a natural language processing model.

Writing ad copy used to mean hiring senior copywriters, going through endless revisions, waiting for the creative director's sign-off. Now AI spins out a few lines in seconds — pick one and you're done. You could see click-through rates from customers climb.

Think about that. Even banking — the most conservative industry there is — has dared to hand "talking to the outside world" over to AI. And you still think your industry is special, that AI doesn't apply?

The 35% of Companies Not Using AI Are Being Left Behind

I told these stories to my friend.

He was silent for a moment after hearing them, then said, "These case studies — I hadn't seen a single one before."

I nodded. That's normal.

Because the people who haven't seen them are usually that 35%.

Ascend2 has a tracking report — put the 2023 data next to the most recent numbers, and you'll see: marketers who view AI as a "key ad attribution tool" rose from 13% to 29%. More than doubled.

But flip that around — another 35% of companies haven't put AI to use at all.

Thirty-five percent. Almost a third.

Are your competitors in that third? Quite possibly.

And that's exactly where the gap is opening up.

Email Open Rates Climbed as Much as 40%

Let me give you one more number.

ItsOn. A telecom company. It used AI to personalize emails — sending different content based on each user's behavior and preferences.

The result: email open rate climbed 40%.

Forty percent.

There's a stat in the industry: personalized emails lift open rates by an average of 26% and click-through rates by 14%. Eighty-seven percent of marketers plan to maintain or increase their email spend.

Why? Because AI turned "one-to-many" into "one-to-one."

You think you blasted a hundred thousand people. But from AI's perspective, it sent a hundred thousand different people each a letter meant just for them.

A DTC Brand Cut Customer-Service Response Time by 80%

One last one — the one that surprised me most.

A direct-to-consumer (DTC) retail brand wired its customer-service process into AI. Response time dropped by more than 80%, and average handling time shrank by four minutes.

Four minutes doesn't sound long.

But do the math: one agent handles 50 tickets a day, saves 4 minutes each — that's 200 minutes a day, three and a half hours. The hours one agent saves in a year is the equivalent of hiring an extra person and a half.

And what's a customer-service agent's annual salary?

This is what AI adoption looks like at its most practical — it doesn't talk big, it just shows you the math.

So — Should You Get On?

After I finished laying out the numbers, my friend asked, "So should I go buy an AI tool tomorrow?"

I said, don't.

Look at these case studies — not one of them adopted AI because "everyone else is using it, so I should too." Each one started with a specific pain point: packaging wasn't selling, ads were too expensive, content was too slow, customer service was too slow, no one was opening emails. Then they let AI solve that problem.

AI isn't the goal. AI is the sharpest knife you can grab.

A knife is for cutting. What do you need to cut?

Figure that out first — then get on.

But don't drag out the figuring-out part too long.

Because that 29% is already on the road. And that 35% — if you don't move now, it really will be too late.