Introduction: The AI Panic

The anxiety around generative AI is palpable. Every day brings a new headline predicting the obsolescence of search engines, the death of web publishing, and the end of digital marketing as we know it. The common fear is that tools like ChatGPT are a black hole for attention, swallowing the user interactions that once powered the open web.

But while the changes are massive, the reality is more nuanced—and far more surprising—than the doomsday predictions suggest. The web isn’t dying, but the fundamental rules of engagement are being rewritten before our eyes. This article cuts through the noise to deliver a new strategic framework for success based on five counter-intuitive truths about AI’s real impact.

1. Myth Busted: ChatGPT Isn’t Killing Google—It’s Expanding the Pie

The common assumption is that generative AI is a zero-sum game: every minute a user spends on ChatGPT is a minute stolen directly from Google Search. This idea, known as the “Substitution Hypothesis,” suggests that AI is simply replacing traditional search behavior.

However, a recent Semrush study challenges this narrative with a competing theory: the “Expansion Hypothesis.” After analyzing an incredible 260 billion rows of clickstream data, researchers looked at user behavior on Google Search before and after they started using ChatGPT. The core finding was staggering: there was no statistically significant change in Google Search usage after users adopted ChatGPT.

This data strongly supports the “Expansion Hypothesis.” Instead of substituting one tool for another, people are simply engaging in more total information-seeking behavior across multiple platforms. What we’re actually witnessing is the emergence of a multi-modal customer journey, where users leverage traditional search for quick retrieval of known facts (‘Know’ and ‘Go’) and turn to conversational AI for complex, generative tasks (‘Doing’).

2. A New Breed of User: We’re All “Doers” Now

This expansion of information-seeking is driven by a fundamental shift in how we use these powerful new tools. For decades, search behavior was defined by three core intents: Do (transactional), Know (informational), and Go (navigational). AI has introduced entirely new categories of user intent, creating the engine for the “Expansion Hypothesis.”

According to major studies from OpenAI and Anthropic, user interactions with large language models (LLMs) fall into three new buckets: “Asking” (49% of use), “Doing” (40%), and “Expressing” (11%).

The “Doing” category represents the most transformative shift. These generative tasks are net-new activities, not substitutes for traditional search queries. Users are now drafting text, writing code, creating business plans, and even building entire websites with AI assistants—work that search engines were never built for. This is precisely why the pie is expanding.

Specialized bots are already emerging to serve these new “Doing” intents. For example, an analysis shows that coding accounts for 33% of work-related conversations on Claude, whereas it makes up only 4.2% of all messages on the more generalist ChatGPT. This signifies a monumental change: AI hasn’t just altered how we find information; it has created an entirely new, mainstream user behavior centered on generation and creation.

3. The Broken Deal: Traffic Is a Lie, and Clicks Are Worthless

For over two decades, the internet has operated on an unwritten deal: publishers create valuable content, and in exchange, search engines send them traffic they can monetize through ads or subscriptions. That deal is now officially broken.

AI tools scrape massive volumes of publisher content to train their models and generate summaries, often without sending any meaningful traffic in return. The data on this is shocking. In a now-viral LinkedIn post, Chris Dicker, CEO of Trusted Reviews’ parent company CANDR MEDIA GROUP, revealed that OpenAI scraped his site 1.6 million times in a single day. The result of that activity was a paltry 603 site visitors—a click-through rate of just 0.037%.

News SEO expert John Shehata perfectly encapsulates the breakdown of this long-standing arrangement:

“The original model was Google: ‘Hey, we will show one or two lines from your article, and then we will give you back the traffic. You can monetize it over there.’ This agreement is broken now. It doesn’t work like before.”

The first-ever user experience (UX) study of Google’s AI Overviews confirms the grim takeaway: we have shifted from a “click economy to a visibility economy.” The primary goal is no longer to earn a click, but to be seen, cited, and trusted within the AI’s answer itself.

4. The New Gatekeeper: Trust, Not Just Relevance

In this new “visibility economy,” the process by which users decide what to believe has been completely turned on its head. In the past, relevance was king. Users searched, scanned for the most relevant-looking blue link, and then clicked.

A groundbreaking UX study on AI Overviews led by analyst Kevin Indig reveals that users now apply a rapid, two-step filter when evaluating search results, and the order is critical:

• Step 1: “Do I trust this source?” (This is based on brand recognition and other authority signals.)
• Step 2: “Does this result answer my question?” (This is traditional relevance.)

This reversal happens because trusting a known brand is a cognitive shortcut—it’s more convenient than critically evaluating the relevance of ten different unknown sources. As expert Barry Adams astutely notes, in the face of information overload, convenience often trumps truth.

“People don’t really care about truth. They care about convenience.”

The implication is profound. If a user doesn’t recognize or trust the brands cited in an AI-generated answer, they become skeptical. If they do recognize a brand, they are primed to accept the information. In a world of AI-generated answers, your brand’s reputation precedes its content.

5. The Survival Playbook: Build a Brand People Search For

Given that AI is expanding information-seeking into generative tasks, breaking the old traffic-for-content deal, and making trust the primary filter for visibility, the strategic imperative is clear: brand is the only durable asset. In a world of probabilistic AI answers and dwindling referral clicks, it is the only competitive advantage that remains.

The consensus among experts like Barry Adams, Harry Clarkson-Bennett, and John Shehata is unanimous. You cannot simply optimize for an algorithm anymore; you must build a brand so strong and authoritative that both users and AI models are compelled to seek you out directly.

“I think too many publishers try to be everything to everybody and therefore are nothing to nobody. You need to have a strong brand voice.” — Barry Adams, Editorial SEO Consultant

This new reality demands a new set of metrics. As analyst Harry Clarkson-Bennett argues, the key performance indicators (KPIs) of the AI era are no longer just raw traffic and rankings. Instead, success must be measured by tracking branded search volume (as a proxy for brand equity and trust), AI visibility (the new measure of on-SERP dominance), and user sentiment (the ultimate indicator of brand reputation).

Conclusion: A New Internet

The web isn’t dying, but the terms of engagement have been fundamentally altered. The era of optimizing for clicks is over. We have entered a new phase where visibility, trust, and brand authority are the currencies that matter. The platforms and publishers that thrive will be those who understand that they are no longer just competing for a click—they are competing for a user’s trust and a machine’s citation.

In an age where AI provides the answers, what question is your brand uniquely qualified to own?