78% of Marketers Are Using AI — But the Ones Actually Making Money From It Are a Different Crowd

A few days ago, a friend of mine who runs an e-commerce business came to me with a question.

He said his team had bit the bullet at the end of last year and rolled out a full AI stack — content, ads, email, the works. Six months in, content output had doubled and CPA (cost per acquisition) was clearly trending down. But he couldn't sleep.

Why?

Because the company next door — half his size — was churning out more than he was. He asked me if he was doing something wrong.

I didn't rush to answer. I showed him a few numbers. After reading them, he went quiet for a moment and said: "So it's not that I'm doing it wrong — it's that I've only done half of it."

Today, I want to share those numbers with you too.

First, the most basic question: how mainstream has AI really become in marketing?

What does "mainstream" even mean?

Not "some people are using it" — "most people are using it."

HubSpot's State of Marketing 2026 has a number for this: 78%. In other words, of every 10 marketers worldwide, nearly 8 now have AI in their daily workflow.

And that's not all. Salesforce puts it this way: since 2023, AI adoption on the marketing side has grown 3.2x. E-commerce and retail are the most aggressive — 87% are already using it — with B2B SaaS close behind at 82%.

Even 61% of CMOs, in Deloitte's survey, directly called AI "essential."

That is the backdrop of marketing today.

So when you look back at that competitor who still hasn't adopted AI, they're not waiting and seeing. They're gambling. Lose that bet, and they become the "other crowd."

So — is anyone actually making money from it?

This is the question that matters most. Just using it isn't enough — what counts is the return.

Let me do the math with you.

McKinsey Digital 2026: companies using AI in marketing see an average ROI lift of 35%. Forrester's TEI (Total Economic Impact) study is even more direct — for every dollar poured into AI tools, you get back USD 5.2 on average.

What about by stage?

Content production — Content Marketing Institute measured a 63% speed-up. A blog post that used to take 8 hours without AI now gets done in 3.

Ad campaigns — Google Ads' performance report shows CPA cut by an average of 41%. Before AI: USD 52 to acquire one customer. After AI: USD 31.

Email marketing — Mailchimp's benchmark report says personalized emails see open rates 28% higher. Plain emails get a 21% open rate; AI-personalized ones can hit 27%.

Real money.

Not "the future looks promising" — already in the bank.

Think about it: content twice as fast, acquisition 40% cheaper, emails opened a quarter more — when you add those three things together, what does it mean?

It means the same team can fight several more battles.

Where is the money going?

The interesting thing here is that everyone says budgets are tight, yet the money flowing into AI keeps rising.

Gartner's CMO Spend Survey: an average of 19% of marketing budgets now flows to AI. IDC's Worldwide AI Tracker puts the year-over-year (YoY) growth rate at 28%.

By category, the distribution looks roughly like this:

Content generation and copywriting — 28% of AI budgets, up 34% YoY.

Ad optimization and bidding — 22%, up 18% YoY.

Personalization and CRM — 20%, up 42% YoY — one of the fastest-growing segments.

Data analytics and insights — 16%, up 25% YoY.

Chatbots and conversational AI — 14%, up 31% YoY.

See the pattern?

Money is surging in two directions: letting content grow itself (generation), and making sure everyone sees something different (personalization).

Those two happen to be exactly what AI is best at — and where you can see immediate impact.

On the SMB (small and medium business) side, Forrester puts the average monthly spend at USD 1,800, roughly floating between USD 900 and USD 2,700. Larger companies go harder — firms with more than 500 employees invest on average USD 13,500 to USD 50,000 per month.

People are genuinely willing to shell out that kind of money for this.

Which means somebody has run the numbers.

What tools is everyone holding in their hands?

I'm guessing you can already name the top of the list.

HubSpot's stats: ChatGPT — 72% of marketers use it regularly; Canva AI — 58%; Claude — 41%; Midjourney — 36%; Jasper — 24%; Gemini — 22%; HubSpot's own AI — 19%; Perplexity — 17%.

Note one number — Chiefmartec says the average marketer is juggling 4.3 AI tools.

4.3 of them.

Not one or two — a whole little toolbox.

And that toolbox is still expanding fast. Chiefmartec's marketing technology landscape shows the number of AI marketing tools on the market has gone from 1,200 in 2024 to more than 3,800 in 2026. Three times in two years.

The fastest-growing category is video. Wistia's State of Video report says tools like Sora, Runway, and HeyGen saw a 340% jump in usage between 2025 and 2026.

Gartner has another angle: 67% of AI tool usage is generative (text, image, video), and only 33% is analytical.

In other words, people are using AI first to "make things," not to "look at things."

That's actually pretty counterintuitive — a lot of people assumed AI's biggest job would be analyzing data. Turns out, marketers first turned it into a content factory.

Is it the same all over the world? No, it isn't.

This one you have to look at by region.

The US leads the pack — Salesforce's data shows 84% of marketers there are using AI. North America as a whole is out in front: tools landed earlier, and budgets are thicker.

Europe is more of a mixed bag. The EU average sits at 52%. But the Nordics paint a different picture: Sweden 71%, Denmark 68%, Netherlands 64% — all well above the EU average. The UK sits at 76% too.

Asia-Pacific is the growth champion. McKinsey's Asia report says the region grew 67% YoY, led by Singapore, South Korea, and Japan.

I took a specific look at the Netherlands, because it's a fairly mature sample.

Combine Statistics Netherlands (CBS) data with EU figures: Dutch marketers' AI usage climbed from 41% in 2024 to 64% today. Up 56% in two years.

Dutch SMBs spend on average USD 1,060 per month on AI marketing tools — 24% more than in 2025.

Amsterdam and Eindhoven are the two hotspots, with adoption rates inside marketing agencies hitting 78% and 71% respectively.

The EU AI Act has also started to bite — 34% of Dutch marketers say they've been affected, mainly in ad targeting and personalization.

That's a signal worth chewing on. The harder you use it, the tighter the regulators watch.

There's another data point I made a note of: Stanford HAI's research shows non-English AI content is on average 12% less accurate than English.

In other words, if you operate in multilingual markets, the German, Japanese, or Chinese that AI writes is more likely to contain errors. In moments like those, having a human look it over isn't "nice to have" — it's a must.

So — what about the future?

This is the part I most want to share with you, because three predictions bear directly on how you deploy your forces.

First, Gartner says that by 2027, 90% of online content will have been generated or edited by AI.

90%. That's basically "almost everything."

The interesting part is that Gartner itself added a caveat: precisely because AI content is everywhere, content that is genuinely hand-crafted and has a human touch will actually become more valuable.

Let that sink in for a second.

Second, the global AI marketing market — MarketsandMarkets predicts it will hit USD 107.5 billion by 2027.

For reference, Grand View Research measures the 2026 market at USD 48.8 billion. In other words, it's set to more than double in about a year.

Third, Forrester says that by 2028, autonomous marketing agents (AI agents that can independently plan, execute, and optimize a campaign) will take over 40% of routine marketing tasks.

Gartner drove it home with another figure: by 2028, about 30% of Google searches will be answered directly by AI — users won't click through at all.

Think about what that means. The entire logic of SEO is about to be rewritten.

And Wistia's prediction: by 2028, 45% of marketing videos will be AI-generated. Juniper Research says that by 2027, 25% of customer interactions will be voice and conversational.

All of this is happening within the next few years.

Back to my friend's question

After I showed him those numbers, he asked me: "So what should I actually be worried about?"

I said, you're worrying about the wrong thing.

You shouldn't be worried about AI taking your job. You should be worried that you're using AI, but with the wrong posture.

Deloitte's survey has a finding that cuts to the bone: 83% of marketers say AI lets them "do more with less budget" — less budget, more output.

But the same survey also notes that the teams reaping the biggest gains are the ones where humans and AI go in together. Fully automated setups deliver mediocre results.

Why?

Because AI gets the work done fast, but it doesn't know what your brand is actually trying to say, or to whom. It can write 100 articles — but picking out the one that moves people still falls to you.

Salesforce's data backs this up: 73% of marketing teams plan to keep expanding their AI use in 2027. But in that same survey, 71% of marketing managers believe AI will change their team's structure within two years.

Smaller teams, but sharper. The people who stay won't be competing with AI on who writes faster — they'll be competing on who judges better.

That's the part my friend hadn't figured out. He was treating AI as "the worker." Really, it's more like a multiplier — it amplifies the judgment you already have.

The sharper your judgment, the bigger the return AI hands you. The blurrier your judgment, the only thing AI does is help you produce more mediocre work, faster.

Finally

After my friend heard all this, he went quiet again for a moment.

Then he said, "Then I'll go back and rethink what we actually want AI to help us say."

I said, "Now you've got it."

I'll keep watching the numbers. 78%, 35%, 5.2x — these are trends; they're the tide. But how high the tide is doesn't decide how far your boat sails. What decides that is the hand on the tiller.

As for whether AI will replace you — I've thought about that one for a while.

My answer is: it will replace the person who only knows how to use AI but has no judgment of their own.

As for the person who has judgment and is willing to treat AI as a partner — this era belongs to them.