Marketers: The Playbook You've Been Running for 20 Years Is About to Change Completely

I was chatting with a friend who works in skincare not long ago. She was anxious.

I asked her — anxious about what?

She said, it used to be that when a customer wanted to buy face cream, they'd open a search engine, type in some keywords, and then pick one from the ten blue links on the first page. Now? The customer opens an AI chat assistant and asks directly: "Which face cream is right for me?" And the AI gives them an answer. One answer.

Just one.

Her fear: what if that answer doesn't include her brand?

This question is keeping more and more marketers up at night. Because the consumer's path to purchase is undergoing the most radical transformation in twenty years.

From "Ten Blue Links" to "One Answer"

Let me start with a data point.

In July 2025, AI-driven traffic to U.S. retail websites surged 4,700% year over year. You read that right — four thousand seven hundred percent. That same year, Gartner made a prediction: by 2026, traditional search volume will drop by more than a quarter.

What does that mean?

It means that for the past twenty years, the starting point for every consumer purchase was the search box. But from here on out, that starting point is being steadily eroded by AI chat assistants. Users are no longer scrolling through links and comparing options themselves — they're letting AI just hand them the answer.

And here's what's really worth thinking about: this is just the first wave. The next wave is AI agents.

What's an AI agent? It's the kind of AI that doesn't just answer questions — it does things for you. It can place orders, compare prices, even run the entire purchase process end to end. Buying everyday essentials could eventually be handed off to AI entirely, automatically.

Think about it: if consumers are delegating even the act of "choosing" to AI, how do you as a marketer make sure your product is the one that gets chosen?

What Do You Look Like Through AI's Eyes?

There's a guy at Brandtech named Jack Smyth who works in a new field called Generative Engine Optimization (GEO) — essentially, figuring out how to make your brand show up in AI responses.

He coined a concept called "Share of Model."

What's Share of Model?

Think about "market share" — how big a slice of the market your product occupies. Share of Model is similar, except it measures something different: when a user asks an AI assistant about your category, how likely is your brand to be mentioned? In what context? What good things are said about you, and what bad things?

A lot of people haven't grasped how important this is. Go ask ChatGPT, ask Claude, ask Doubao (ByteDance's AI assistant): "What's the best face cream?" See what they recommend. Then try a few different phrasings. Ask several times.

If you're not actively finding out how these AIs are talking about you, you're leaving "the answer" entirely to chance.

So how do you influence your Share of Model? Most of the old SEO rules still apply. For instance: your content should be clear and machine-readable, your mobile load time should ideally be under two seconds, you should lean on facts and data, and your cited sources need to be credible.

But there are a few new rules. When an AI assistant answers a user's question, it goes looking for content it can lift directly. So you need to organize your information in formats that are easy to extract — tables, lists, short paragraphs. And then embed the natural-language questions people commonly ask about your category right into your content. Because the way users ask AI is the way AI goes looking for answers.

Here's something else most people haven't considered: images.

AI can now "read" images. If your selling point is great value, but the images you feature look expensive and luxurious, the AI will judge that your product is inconsistent with its own positioning — and it'll hesitate to recommend you.

At its core, AI is the one customer you'll never meet — but it might be making the buying decision for your real customer.

Traditional Marketing Isn't Dead — It's More Important Than Ever

At this point you might be thinking: since AI can generate copy, create images, and produce videos, why not just automate the whole marketing operation with AI?

Not so fast.

Think about it: when an AI assistant recommends a product to a user, what's the basis? The user still has to ask a question first, right? And what the user asks depends on which brands they already carry in their head.

Say you make running shoes. The ideal scenario: the user asks an AI, "Which of your running shoes is good for marathons?" — but if the user doesn't even know your brand name, your brand simply won't appear in the question they ask.

And even taking a step back — say the user doesn't mention your brand, but the AI includes you in its recommendation list. Which one does the user pick? Usually the one whose name sounds familiar.

Brand awareness is the entry ticket to the AI world. Without that ticket, you don't even qualify for a recommendation.

The good news is that AI has indeed sent the cost of producing marketing content off a cliff. Tools like Midjourney and Google Veo 3 have made it possible for ordinary people to produce high-quality video and images. Copy can be drafted by a chat assistant, too.

An AI advertising platform called The Brief ran a campaign for the British Council. They used AI to generate over 1,000 creative assets, cut costs by 70%, and localized the campaign for eight different regions.

But The Brief's CEO Tammy Nam said something genuinely sobering: soon, no one will ask whether an ad was made with AI, because the answer will always be yes. What truly separates brands is the stuff that can't be automated — taste, visual style, a distinctive voice.

In other words, AI is a tool, not a soul. If you have the tools but no soul, everything you produce falls into the same trap — it looks polished but is indistinguishable from everything else out there. Nobody remembers it.

So the right approach is this: start with human creativity to design a distinctive advertising concept. Then use AI to multiply that concept into different versions for different audiences, different regions, different languages. Humans handle the soul. AI handles the clones.

When It's Time to Pay Up, People Still Trust People

Let's talk about something counterintuitive.

AI assistants are incredibly useful, and consumers are indeed growing more dependent on them. But when it comes time to spend real money, people still want to hear it from another human.

Tremendous's Scott Shepard said something I think nails it: "No matter how good AI gets, before buying a mattress, people still want a human to give them a nod of approval."

Not for everyday items — for the expensive stuff, the personal stuff. Electronics, furniture, skincare, travel. In these categories, the influence of word-of-mouth and recommendations has always been exceptionally strong.

Have you noticed that Reddit content has been surging in search lately?

Why? Because Reddit has real people having real discussions. A user gets an AI recommendation, feels uncertain, and goes to Reddit to see what actual humans are saying. And here's the fascinating part: those Reddit discussions then get picked up by AI as source material, influencing the recommendations it gives to other users.

It creates a loop: what humans say feeds the AI; what the AI says drives people to search for what humans say.

There's another side effect of the AI era: as the cost of producing content approaches zero, the internet fills up with more and more stuff — but less and less of it is reliable. Consumers will become increasingly wary and increasingly discerning. When that happens, a recommendation from a friend actually becomes more valuable.

That's what referral programs are for. Let your existing customers bring in new ones. Data shows that customers acquired through referrals have stronger purchase intent, higher loyalty, and lifetime value roughly 16% higher than average.

And today, referral programs themselves can be optimized with machine learning. When to reach out, what incentive to offer, which customer to approach — all of it can be calculated.

Your Next Moves

So here's the thing: this isn't simply "AI is here, everything old is useless."

Three things need to happen simultaneously.

First, figure out what AI is saying about your brand. Actively ask various large language models, track your "Share of Model," and structure your content in formats that are easy for AI to extract.

Second, don't abandon brand-building. AI is an amplifier — but what it amplifies depends on what you feed it. Nail the creative soul first, then let AI multiply it.

Third, double down on word-of-mouth. In an era of AI content saturation, human trust only becomes more valuable.

The marketing playbook of the past twenty years was built around search engines. The next twenty years will be built around AI. In the transition between the two, some panic, and some get excited.

Those who panic see the old tactics failing. Those who get excited see new rules still taking shape — and whoever understands them first reaps the rewards.

Which side are you on?