• Science Says
  • Posts
  • Expert view: The 2 ways AI is disrupting marketing

Expert view: The 2 ways AI is disrupting marketing

A special issue by Prof. Stefano Puntoni, where we zoom out for a big picture view on how you should rethink your marketing in the age of AI.

🤖 Welcome to the last issue of our special AI month, where we cut through the hype and noise to give you the AI-related research that will truly shape your marketing decisions. New to Science Says? 📩 Subscribe for $0.

👀 Special takeover: Expert view

This week I wanted to try something a little different (next week we return to our regular 3-minute insights 🙂).

Science Says insights are, by definition, about very specific tactics. That’s because research papers usually go very deep into one specific element of a chosen topic and push our understanding of it. For example, the use of “you” when writing copy, whether to place prices above or below products, or even how to shape lines in a logo.

The problem with that? Sometimes we risk focusing so much on the details and tactics that we lose sight of the big strategic picture.

Strong tactics must go hand in hand with strong strategies. You can’t have just one but not the other.

So I thought, what is the best, most useful way I can give you a big picture, science-based strategic overview of what is happening?

The answer: ask the world’s leading researcher in that space to tell us.

💪 If you like this, let me know! We can make it a regular for all sorts of topics. Reply to let me know:

  1. What you liked about it

  2. What other topics you’d like to hear about (e.g. pricing, GEO/AEO, emotions in ads)

🎓 Our special guest: Prof. Stefano Puntoni

For our first experimental “Expert view”, I invited our Chief Academic Advisor, Prof. Stefano Puntoni, to take the floor. 

Not because he is our advisor, but because I believe he is by far the best person to give us the best, science-based perspective on this topic.

He brings together a unique background and experience:

  • Professor of Marketing and Co-Director of Human-AI Research at Wharton (the #1 business research university in the world).

  • I don’t think anyone has a deeper understanding of AI and human behavior. He started researching AI many years before ChatGPT even existed.

  • He is one of the most prolific, respected, and influential researchers in the fields of marketing and AI (his volume of research in this space is mindblowing).

We bring him here today to ask him one key question: 

How is AI changing marketing?

It’s time for me to hand over to Stefano. I hope you enjoy this special issue 👇

The below was lightly edited for brevity.

🎓 AI is changing marketing on 2 fronts

AI is changing in marketing in two fundamental ways: 1) how people search for information, and 2) how they make purchasing decisions. These two changes will disrupt how marketers think of their current and future strategies.

People can now bypass traditional websites by asking AI chatbots for complete, curated product information. This is starting to happen already, and will eventually drastically reduce web traffic and limit brand visibility.

[Note: In our research insight about this we’ve seen that this is especially true for complex, research-heavy products and those for tech-savvy or young audiences. The shift is happening much slower for other products, but still happening]

💰 Buying

AI agents are increasingly acting as the actual customers by autonomously researching, evaluating, and purchasing products on behalf of human users. Marketers need to shift from focusing on people only, to “bot psychology”, optimizing for how agents decide - not only humans.

[Note: for example, in our insight that touches on this, we saw that simply tweaking a product title to optimize it for AI agents could increase by 80.4 percentage points how likely an AI agent powered by GPT-5.1 was to choose it]

📈 What should you do in this new landscape?

Here are the 3 most impactful things you should start thinking and acting on today, to adapt to these changes:

1️⃣ Audit your traffic and change your website architecture

Check how reliant you are on search-based digital traffic. If it’s a high proportion, you are at higher risk of short-term disruption, as people change how they search.

Then, update your website's backend to include structured data (e.g. json files) so AI agents can instantly read your pricing, availability, and product details without needing to navigate visual menus. 

Write comprehensive content that directly answers common queries and uses clear categorization so AI systems can easily contextualize, and cite your information.

2️⃣ Optimize for “bot psychology”

Chatbots have a psychology of their own, which is similar to ours in how they make choices as that’s what they’re trained on. This means that if you know how they make decisions, you can nudge them toward choosing your product. 

Be careful, this moves fast. For example, at the end of last year, most agents generally favored products on the top row (so it was best to place your highest margin products on the first row), Within the first row, ChatGPT heavily favored products in the first position, and Claude those in the middle. Now, the latest models of ChatGPT seem to be doing the opposite, favoring products in the bottom right of a page.

3️⃣ Build communities and a personal brand connection

Double down on what can’t be replicated by AI. Human experiences are irreplaceable. Shift your focus away from purely informational content and build interactive communities, emotional connections, and shared experiences (e.g. webinars, live streaming). 

Thank you for having me.

Stefano Puntoni,

Sebastian S. Kresge Professor of Marketing and Co-Director of Wharton Human-AI Research

👋 A last word from Thomas

Hey Thomas here! 

I hope you found this expert view useful. If you did, I’d love to hear what are some other topics you may be interested in. And if you didn’t I’d love to hear what we could improve 🙂 I read every reply.

P.S. Remember, you can download for free our latest Blueprint in collaboration with Wharton here: The Wharton Blueprint for AI Agent Adoption.