Businesses that rely on people clicking through to web pages have faced a difficult two years. Pew Research Center tracked the browsing behavior of 900 U.S. adults and found that when Google displays an AI summary, users click a traditional search result just 8% of the time, roughly half the 15% rate recorded when no summary appears. Links cited within AI answers perform even worse, drawing clicks only about 1% of the time.

The impact on publishers has been significant. Chartbeat data reported by Axios showed that page views from Google Search fell 34% across its publisher network between December 2024 and December 2025. Small publishers have lost roughly 60% of their search referral traffic over two years. Business Insider’s organic search traffic dropped 55% over three years, and some smaller publishers have already shut down. Chatbot referrals, meanwhile, still account for less than 1% of publisher page views, despite growing more than 200% in a year.

The common explanation in publishing circles has been that AI is killing the web. Recent developments suggest the situation is more complex.

Machines are reading more of the web

Similarweb’s 2026 Generative AI Landscape report indicates that although AI platforms send fewer people to web pages relative to the number of answers they generate, the AI systems themselves are consuming web content at an accelerating rate by searching the web on users’ behalf to answer questions. The share of ChatGPT answers containing live web citations grew more than fivefold in under a year, reaching 6.8% of all answers by May 2026. In some categories, including travel, the figure is as high as 22.6%.

Every major AI search product retrieves live pages from search indexes and synthesizes answers from them. As a result, the quality of AI-generated answers depends directly on the health of the underlying content layer. Lily Ray, VP of SEO and AI search at Amsive, said in the Similarweb report that when organic visibility declines, AI search visibility follows, because models are less likely to find that content.

That dynamic has produced a difficult feedback loop. AI answers are built on an information supply chain whose funding model, ad-supported clicks, is being weakened largely by AI answers themselves. For publishers, the issue is not only audience reach but also whether the pages being read by machines can still support the reporting, reviews, guides, and reference material those machines rely on. A central question for the continued viability of the open web is whether an alternative business model can work, and what that model would be.

A replacement economy is forming inside chat interfaces

Early data points to where traffic patterns may be heading. After ChatGPT’s May 7 search update, which introduced prominent clickable brand links inside answers, referral traffic from ChatGPT rose 157% in one week. The composition of that traffic also changed: the share of referrals landing on homepages more than doubled, from roughly 25% to nearly 60%.

Traditional search typically sent users to specific articles and deep pages connected to specific queries. AI referrals increasingly send a pre-informed visitor to a brand’s front door. The chatbot conducts the research and comparison process, and the person arrives with more context. Similarweb’s data shows that AI-recommended brands receive two to four times as many subsequent visits as competitors that were not recommended.

Advertising is beginning to follow that behavior. Sponsored results appeared in 26% of U.S. desktop ChatGPT conversations in June 2026, up from 14% one month earlier, according to Similarweb’s ad intelligence data. Two-thirds of those ads appear after the second prompt and are targeted based on conversational context rather than a keyword. Click-through rates are around 0.50%.

The traditional search engine keyword auction is being challenged by a different model: paid placement inside a conversation, targeted using accumulated context. That creates a direct challenge to the auction system Google has operated and dominated for two decades.

Google faces disruption while participating in it

Google is both the incumbent being disrupted and one of the largest participants in the shift. AI Overviews now appear in a growing share of Google searches, exceeding 40% by May 2026, according to Similarweb. Visits to Google’s conversational AI Mode have also climbed steadily since launch. Rather than ceding the category, Google is cannibalizing part of its own click-based economy.

Google’s core business was already under pressure. eMarketer projects that Google’s share of U.S. search advertising will fall below 50% in 2026, the first time since roughly 2004. The largest portion of the lost share is going to Amazon, whose sponsored product searches count as search advertising and are growing three times as fast as Google’s. Conversational ads remain a small part of the market today, but they open a second competitive front in a business where Google has long held a dominant position.

Additional structural changes are also affecting Google. A federal court entered final judgment in the DOJ search antitrust case in December 2025, imposing remedies that bar exclusive default agreements and require Google to share search data with qualified competitors. Google appealed in January 2026, while the DOJ cross-appealed seeking stronger remedies. Regardless of how the appeals are resolved, the arrangement that made Google the web’s de facto tollbooth through widespread defaults and a closed index is changing as conversational advertising emerges.

The resulting competitive landscape is new. OpenAI, Google, Perplexity, and Microsoft are now competing not only for users but also for the advertising demand that funded the open web. None of them, including Google, controls the new surface in the same way Google controlled traditional search.

The impact of conversational advertising on the open web remains unclear

It remains uncertain whether the new model will help or harm the web. The web as a destination for human attention is shrinking, and ad-supported publishers built for that environment face serious pressure. At the same time, the web as a machine-readable substrate is becoming more important, while a new referral and advertising economy is forming that routes value toward brands rather than content pages.

Publishers may have limited ability to influence the outcome. Ahrefs, analyzing more than a billion data points across its studies, found that 67% of ChatGPT’s most-cited sources are entities marketers cannot influence. Wikipedia alone accounts for nearly 30%. Ahrefs also found that 28.3% of ChatGPT’s most-cited pages have zero Google organic visibility, suggesting the retrieval layer is only partly connected to traditional search and complicating assumptions that SEO success translates directly into AI visibility.

That distinction matters for any business trying to understand whether visibility in AI systems can replace lost search traffic. Being cited as evidence, being recommended as a brand, and receiving a human referral are becoming separate outcomes rather than different points on the same search funnel.

Websites are being found in a different way

Three independent datasets indicate that, in the emerging environment, the pages AI systems cite and the pages AI systems send humans to are often different pages serving different purposes. That represents a major shift, while many teams are still optimizing for the older search model.

Similarweb’s data shows that 65% of ChatGPT-cited URLs are located two or three folders deep in a site, while 58.8% of referral traffic lands on homepages. Ahrefs found a similar split in its own analytics: more than 80% of its AI referral traffic goes to its homepage, product pages, and free tools, rather than its extensive editorial content. A Previsible analysis of 6.77 million AI-referred sessions identified a third destination: 28.8% of ChatGPT referrals land on internal site search pages, a navigation surface many publishers have long neglected because Google searches previously filled that role.

Organizations assessing this shift can begin by auditing search traffic patterns. That includes reviewing AI referral logs and available citation data, then mapping which pages are quoted as evidence against the pages where visitors actually enter. If the data reflects the new pattern, several priorities emerge.

Deep pages, documentation, comparisons, and benchmarks should be structured to be citable, with specific claims, clear headings, and descriptive URLs. Ahrefs found that pages with natural-language URL slugs are cited at a rate of 89.78%, compared with 81.11% for pages without them.

Homepages should also be designed for visitors arriving with context from a conversation rather than from a traditional blue link. Such visitors may already know that the site offers what they need, so the page should help them reach the relevant destination quickly.

Internal search, often an overlooked feature on many sites, is also becoming an acquisition surface that may warrant significant user-experience investment.

Much remains uncertain or in flux. But independent sources examining the shift point to the same underlying change: the click economy is not returning to its previous form, and the organizations that learn how to be quoted by machines and convert the humans those machines send will be better positioned for the next version of the web.