For three decades, I have watched consumer behavior evolve through television, search, and social media. Each change changed the tactics, but not the basic logic of decision-making.
What I’m watching happening right now is different. And I know I’m not alone. Every marketing professional I talk to, whether they’ve built their careers in offline communications or digital platforms, says some sort of the same thing: something important has changed, and the old playbooks don’t work the way they used to.
This is not just a change of scenery. It’s a psychological thing. For the first time in my career, I’m watching users move from seeking information to seeking certainty, and that difference changes everything.
When Behavior Was Predicted
I remember a time when a celebrity’s face on television was a guarantee. Brand loyalty is closely tracked by fan loyalty. If your brand ambassador had a dedicated following, that following would follow them to your brand. It was a simple, time-tested formula: attention creates engagement, engagement creates purchase. And it worked, consistently, for decades.
💜 for EU tech
The latest talk from the EU tech scene, a story from our genius founder Boris, and some incredible AI art. It’s free, every week, in your inbox. Register now!
When the Internet came along, it digitized that prediction rather than eliminating it. Google and Yahoo have transformed discovery into a structured, keyword-driven system. Users have searched, engines have returned ranked results, and businesses that appear at the top of those results have won customers. For the better part of a decade, with many algorithm updates, with the rise of paid search, with the SEO arms race, the main goal is managed: be seen, and you will be chosen.
Both eras have rewarded the same thing: access. Who can get in front of more people, more often? That question has shaped marketing strategy for nearly three decades.
What Has Really Changed?
The change I am describing is not about which platform wins or loses. It’s deeper than that, it’s about how people make decisions.
The credibility of celebrities has been destroyed like never before. It’s not that people don’t trust celebrities, it’s that modern consumers understand the commercial ecosystem they work within. They know that compromise is work. And with global knowledge at their fingertips at all times, they also know that one endorsement isn’t enough reason to spend.
Younger consumers in particular, Gen Z and late millennials, have flocked almost entirely to first-hand information. Their experience, or that of someone in their immediate circle, their age group, their specific context. Not a famous person. A related person. And yet, they confirm.
The distinction between online and offline has also become more blurred. A shopper who sees a product in a store will pull out their phone before adding it to their cart. A buyer who hears a recommendation from a friend will check it out before buying it. Behaviors that once lived in separate countries, browsing a physical shelf, reading an Internet review, asking a peer, now happen simultaneously, slowly, and constantly.
What Research Has Shown Me
To check if what I was looking at shows broad behavioral patterns, I conducted a personal field survey from the middle of 2025, about 500 people, not a formal academic study, but deliberately diverse: college students, working professionals, homemakers, and retirees in all age groups and economic backgrounds. The results confirmed the pattern I had been seeing.
Among 16- to 20-year-olds, 87% said that their main source of trust in purchasing decisions lies with friends, parents or teachers, people in their immediate circles. In the 21 to 30 age group, 73% combine input from peers and social media to select the people they follow, but 96% of that group said they double-check suggestions before acting on them. Almost everyone. Among 31- to 40-year-olds, 65% show the same verification behavior. Even in the 41 and over segment, 44% now follow the same pattern, a slower recovery, but the same direction.
A common thread across all age groups: trust is no longer accepted. It is earned and confirmed. Consumers of all generations have become active validators, not passive receivers.
Are LLMs a New Idea or a Response to Market Pressure?
If we look at the history of technology, a pattern appears almost every ten to fifteen years: radio gave way to television, television to the Internet, the Internet to search engines, search engines to social media. Each revolution not only created a new platform, it changed the way consumers behave. That is, if you are a marketer trying to understand the AI era, the first question is not “How do I use this platform?“The Icon”How has consumer behavior changed, and why?“
The rise of large language models – ChatGPT, Gemini, Perplexity, and others, is a direct response to the mental shift I’ve been describing. These tools did not create an inherent sense of certainty in modern consumers. They answered it.
Traditional search engines provided a list of options and left the user to sort through competing claims. LLMs include. They combine information from multiple sources and return structured feedback. For a buyer, his instinct is to verify, test, and reach certainty before making a decision, that’s not easy, that’s exactly what they’ve been trying to do, it’s done quickly.
Here’s an insight that I think isn’t being missed: the tech giants that have invested heavily in this space – OpenAI, Google, Microsoft – weren’t just motivated by innovation. They understand something uncomfortable. The audience that sat in their seats was scattered. Attention was divided into social media, e-commerce platforms, and many other channels. LLMs are, in part, a strategic effort to regroup that audience under a single, trusted interface.
They don’t build these tools because they want to. They build them because there are still risks of losing the next layer of the Internet.
And that changes the stakes a lot. Because an LLM that consumers trust enough to make purchasing decisions is an LLM that should remain unbiased. The moment users hear a trade preference in a recommendation, they abandon it, and move on to the next tool that feels neutral. The entire value proposition of these platforms is dependent on being deemed trustworthy.
What This Means for Products
The shift from seeing to believing is not subtle. In the classical paradigm, the symbol that appears frequently enough and is loud enough will eventually be selected. In this paradigm, reflection is necessary, but nowhere near enough. If your product won’t live up to the time a potential customer decides to verify your claims, with an AI tool, peer networks, reviews, or independent sources, you probably won’t stay in the consideration set.
A useful example of this change can be seen in the way consumers now make relatively small purchasing decisions. A user may first find a product via TikTok or Instagram, search for reviews on YouTube, check opinions via Reddit, compare alternatives via Google, and finally ask ChatGPT or Perplexity to narrow down the best option before making a purchase. What matters is not the number of platforms involved, but the behavior itself.
Another example: Take a procurement manager evaluating CX vendors. They might first come across a shortlist through AI Overview, cross-review on Clutch, G2 or Trustpilot, look at case studies on the vendor’s site, scan Reddit or industry forums for unfiltered ideas, and finally ask ChatGPT to compare the top options. A company that has invested in verified reviews, documented news studies, and third-party editorial coverage survives that journey. If you don’t have it, you don’t.
Consumers are no longer dependent on a single source of authority. They build trust through layered validation, and for brands, that behavioral shift has a tangible effect.
Essentially, this means thinking less about impression counting and more about information integrity. Are your claims substantiated? Are you consistent across all the places a user might check – your website, third-party reviews, forum chats, AI-generated summaries? Is there enough legitimate, high-quality information within the trusted ecosystem for LLM to accurately represent your product? These are not marketing questions. They are infrastructure questions.
Many brands are still preparing for the old game: reach, frequency, creative impact. The pullers do something different. They make it easy to trust when a skeptical buyer decides to take a closer look, not by being loud, but by having nothing to hide when someone does.
The Deeper Shift
What I keep coming back to, after everything I’ve seen in my research and three decades of watching the markets move, is that the basic human need has not changed. People have always wanted to feel certain before committing. What has changed is the extent of that confidence, and the speed at which they expect to reach it.
Search is not that important. It has become a decisive factor. Increasingly, users are not just looking to test; they are looking to reduce uncertainty immediately. And if your product is not going to be a part of that moment, in a way that holds up to the test, then at that particular decision moment, your product doesn’t exist.
That’s a harder problem than getting your SEO right. But it’s also very honest, because it forces brands to ask not just “How can I be found?“but”am i eligible to be selected?“
In the age of AI, that’s the only question that matters.
