The Click Is Being Dissolved, Not Redirected

The click-through — the atomic unit of digital marketing economics for two decades — isn’t being rerouted to a new destination. It’s being structurally dissolved. Google isn’t building a better front door to the web; it’s bricking over the doorway entirely and answering the question itself, right there on the threshold.

The evidence is no longer speculative. AI Overviews now trigger on the majority of informational queries, and their reach is accelerating down-funnel at a pace that should alarm anyone whose revenue model depends on inbound traffic. A Semrush analysis of more than ten million keywords found that the share of purely informational queries triggering AI Overviews actually dropped from 91.3 percent to 57.1 percent over thirteen months — not because Google pulled back, but because it aggressively expanded into higher-intent territory. Commercial queries triggering overviews more than doubled to 18.57 percent, and transactional queries surged from under 2 percent to nearly 14 percent. The implication is stark: even the queries closest to a purchase decision are being intercepted before a user ever reaches your site.

And when those overviews do appear, they work exactly as Google intends. A working paper from researchers at the Indian School of Business and Carnegie Mellon University found that AI Overviews reduced organic clicks on triggered queries by 38 percent, with zero-click searches jumping from 54 percent to 72 percent — and no measurable decline in user satisfaction. People got what they needed without clicking. That’s not a side effect; that’s the product functioning as designed.

Google’s newer AI Mode deepens the problem. As Search Engine Journal reported, the average query inside AI Mode runs three times longer than a traditional search, follow-up queries are climbing 40 percent month over month, and planning queries grew 80 percent faster than other categories. Users aren’t bouncing between tabs and websites anymore. They’re conducting multi-step research sessions that never leave Google’s interface. Every follow-up question that would have been a separate search — and a separate click to a separate publisher — now stays inside a single, contained conversation.

Perhaps the most revealing admission came from Google’s own Robby Stein, who explained in a Q1 recap that when users don’t engage with an AI Overview, Google may remove it for that query. Read that carefully: the system is calibrated to keep overviews only where they successfully prevent further engagement. It self-selects for containment.

Then there are information agents — the feature that should have dominated post-I/O panic but was oddly underplayed. These agents monitor the web, synthesize updates, and deliver them inside Google without generating a single pageview. As SEO consultant Glenn Gabe warned on LinkedIn, “For publishers, information agents can hit ad revenue big-time as less people will be visiting websites.” Your content gets consumed; you just never see the visit in your analytics. Jeff Bullas described the same dynamic from personal experience, watching his carefully researched articles get scraped, synthesized, and surfaced in AI overviews that answered the question without sending a single visitor back.

This isn’t a future threat to monitor. It’s a structural shift already visible in the data, already compressing traffic, and already repricing the value of the click. The question is no longer whether it’s happening — it’s what you build when the click can no longer be your foundation.

Google’s New Ad Formats Prove the Endgame — Ads Inside the Answer

At Marketing Live 2026, Google didn’t just unveil new ad units — it revealed the logical endgame of a platform that no longer needs you to click anywhere. The new formats announced alongside Google I/O make one thing unmistakable: Google is no longer selling ads next to content. It’s embedding ads as content, woven directly into the AI-generated answers that are replacing the search results page entirely.

Consider the three flagship formats. First, Conversational Discovery units act as direct responses to user queries — if someone asks Gemini for easy ways to make their home smell like a spa, the AI evaluates that prompt and generates a custom ad creative from a brand’s product data, surfacing it as though it were part of the conversation itself. Second, Highlighted Answers ads insert sponsored placements directly within the AI’s list of recommendations. Third, and perhaps most consequentially, a Gemini-powered “explainer” feature pops up within the ad experience to synthesize product or service information, providing what Google frames as additional context to build trust. In practice, this means Google’s AI is writing your ad copy, deciding your placement, and even editorializing about your product — all without the advertiser touching a single creative asset.

This is the quiet confession that the click-through model is terminal. If users aren’t leaving the interface — and Google’s own AI Mode data show queries running three times longer than traditional search, with follow-up queries climbing 40% month over month — then the only way to monetize is inside the answer. Google can’t sell the click if the click doesn’t happen. So it sells the integration instead.

But monetizing inside the answer is only half the architecture. The other half is closing the transaction before anyone leaves. Google’s Universal Cart and native checkout functionality, built through partnerships with Walmart, Wayfair, and Shopify, allow consumers to buy products directly from ads across Google-owned platforms. As Adweek reported, merchants still legally own the transactions, but the experience is designed to eliminate the need to click out to a merchant’s website. Discovery, evaluation, and purchase — the entire funnel — collapses into a single Google-controlled interface. The company’s vice president of merchant shopping, Ashish Gupta, insisted Google is “not a retailer” and “not a marketplace,” but a matchmaker. The distinction is increasingly semantic when the matchmaker controls every surface the customer sees.

For advertisers, this represents what Neil Patel described as a shift toward a goal-in, AI-executes model: you define business outcomes, and the platform handles the operational work. That framing sounds like liberation. It’s actually surrender. You are no longer crafting a message, designing a landing page experience, or controlling the context in which your brand appears. Gemini generates your creative from your product feed. Gemini decides when you’re relevant. Gemini writes the explainer that tells the user what your product does and whether it’s worth buying. You’ve become a supplier feeding inventory data into Google’s merchandising engine, dependent on an algorithm’s judgment for every impression.

Google calls this automation. A more honest word is dependency. And the brands that recognize the difference will be the ones searching for leverage that exists outside Google’s walled garden — leverage that no AI overview can generate, summarize, or swallow whole.

The Measurement Black Hole Makes It Worse

Even as Google accelerates the dissolution of organic clicks, the tools marketers rely on to understand what’s happening are failing them in real time. Google Search Console — the default instrument for diagnosing search performance — still cannot cleanly separate traffic arriving via AI Overviews or AI Mode from traditional organic clicks. The result is a reporting environment where marketers are asked to adapt to a radically new search architecture using dashboards designed for the old one.

The problem goes deeper than missing filters. When a page appears in both an AI Overview and in the standard organic results below it, Search Console records impressions for both placements. That double-counting artificially inflates the denominator in any click-through rate calculation, making CTR look worse than it already is. A marketer reviewing their data sees what appears to be a catastrophic drop in engagement, but part of that drop is a measurement artifact — phantom impressions generated by a system that surfaced the same URL twice. As HubSpot’s analysis of AI Overview optimization makes clear, the inability to distinguish between these impression sources leaves teams guessing whether their content is genuinely losing relevance or simply being miscounted by a reporting system that hasn’t kept pace with the product it’s supposed to measure.

And that’s only the traffic you can see. The next wave of content consumption won’t involve a human being visiting your site at all. As Neil Patel observed in his breakdown of Google I/O and Marketing Live 2026, Google is building toward agent-driven workflows where AI handles research, comparison, and even task completion on a user’s behalf. When a Gemini-powered agent reads your page, extracts what it needs, and delivers a synthesized answer to the user — all without generating a pageview, a session, or a single line in your analytics — that content consumption is invisible. No referral string. No event fired. No attribution whatsoever. Your content did the work; your dashboard recorded nothing.

This isn’t a temporary gap waiting for an API update to fix. It’s a structural asymmetry. Google controls the surface where content is consumed, the AI that decides which content gets consumed, and the reporting layer that tells you what happened afterward. It sees every query, every impression, every interaction within AI Mode. You see a filtered, delayed, double-counted approximation. The entity making the rules also keeps the scoreboard — and it has no competitive incentive to hand you a clearer picture when the clarity might accelerate your migration to channels it doesn’t own.

The familiar marketing maxim — you can’t optimize what you can’t measure — takes on an existential quality in this environment. If your CTR data is polluted by phantom impressions, your optimization decisions are based on distorted signals. If agent-driven consumption generates zero attribution, your content ROI calculations systematically undercount the value you’re creating and offer no feedback loop for improvement. Jeff Bullas captured this dynamic precisely when he described watching his carefully researched articles get scraped, synthesized, and surfaced in AI Overviews without sending a single visitor back — a firsthand account of content performing exactly as Google’s system intended while the creator’s analytics showed silence.

The rational response is not to wait for better reporting from a platform that benefits from your blindness. It’s to diversify aggressively into channels where you own the data pipeline from first touch to conversion — where impressions, engagement, and attribution flow through systems you control, not systems controlled by the entity that just swallowed your click.

Why “Just Do Better SEO” Is Necessary but Insufficient

There is a legitimate counter-argument to the doom narrative, and it deserves honest engagement before we move past it. Original analysis, primary data, and genuine expertise do still earn citations in AI Overviews. Google itself has made this case explicitly: as Search Engine Journal noted, Google’s own guidance highlights “non-commodity content” as the only type an AI must cite rather than simply summarize. If you’re publishing proprietary research, offering genuinely novel frameworks, or providing the kind of deep expertise that a language model cannot synthesize from existing sources, you still have a path to visibility. Answer Engine Optimization — the emerging discipline of structuring content so that AI systems surface and credit it — represents a real adaptation strategy, not a gimmick.

And traditional SEO fundamentals haven’t evaporated. HubSpot’s analysis makes this point clearly: AI Overviews are heavily influenced by the same signals marketers already optimize for — technical SEO, content quality, and topical authority. The advice to shift success metrics from clicks to brand visibility, mentions, and citations is sound counsel for anyone willing to rebuild their measurement stack around it.

Here’s why that’s a partial answer at best.

The “create content AI can’t synthesize” directive is strategically correct but economically devastating for the majority of digital marketing operations that exist today. Think about what actually powers the content marketing industrial complex: the how-to posts, the comparison articles, the “best X for Y” listicles, the FAQ pages that funnel informational queries toward affiliate links and display ad impressions. By Google’s own definition, all of that is now commodity content — the exact category that AI Overviews are designed to absorb and render without a click. The field experiment data cited by HubSpot bear this out starkly: on queries where AI Overviews appear, organic clicks dropped 38% and zero-click searches surged from 54% to 72%.

That 38% number isn’t a rounding error in someone’s analytics dashboard. It’s an extinction-level event for business models built on volume clicks from informational queries. Affiliate marketers, content arbitrage operations, and ad-supported publishers don’t have the luxury of pivoting overnight to “content AI can’t synthesize.” Their entire economic model depends on ranking for precisely the queries Google is now answering itself. Telling them to produce more original research is like telling a fast-food franchise to become a Michelin-starred restaurant — technically possible, but a category change, not an optimization.

The people warning that SEO isn’t dead and the people warning about collapsing traffic economics are, as Search Engine Journal’s post-I/O analysis put it, both correct simultaneously. SEO as a discipline survives. SEO as a traffic acquisition channel for commodity content does not — or at minimum, it delivers dramatically diminished returns. AEO can keep your brand name visible inside an AI-generated answer, but visibility without the click is brand awareness, not demand capture. It’s a survival tactic for maintaining mindshare; it is not a traffic recovery plan, and no amount of schema markup or structured data will change the underlying math.

This is the gap that native advertising is uniquely positioned to fill — not by replacing SEO, but by providing the economic engine that SEO can no longer reliably power on its own.

Native Advertising on the Open Web — The Last Unmediated Channel

Here is the core argument, and it requires no hedging: native advertising on the open web is the last channel where marketers retain full sovereignty over the creative, the context, the landing page, and the data trail — and that sovereignty is exactly what Google’s AI ecosystem is methodically stripping away.

Consider what happens inside Google’s new advertising paradigm. As Adweek reported, Google is now launching “Conversational Discovery” ad units inside AI Mode where Gemini itself evaluates user prompts and then generates a custom ad creative from a brand’s product feed. The advertiser doesn’t write the headline. The advertiser doesn’t choose the image. The advertiser doesn’t decide the framing. Gemini does. Alongside these units, “Highlighted Answers” ads embed sponsored placements within AI-generated recommendation lists, and a new Gemini-powered “explainer” feature synthesizes product information on the advertiser’s behalf — ostensibly to build trust, but functionally to eliminate yet another reason for a user to leave Google’s surface. Add the Universal Checkout Protocol, which lets consumers complete purchases without ever visiting a merchant’s site, and the picture crystallizes: inside Google’s AI ecosystem, the advertiser is a data input. Your product feed, your brand signals, your merchant profile are raw material that Gemini processes, reshapes, and surfaces according to its own logic.

The strategic implications extend well beyond creative control. As Neil Patel observed, Google Ads is migrating toward a “goal-in, AI-executes” model in which advertisers define business outcomes while the platform absorbs nearly all operational and creative decision-making. Measurement quality, Patel argued, is becoming a competitive advantage precisely because automation is absorbing so much of the execution layer — but that advantage only holds if you actually own the measurement infrastructure. Inside Google’s walled garden, the measurement infrastructure belongs to Google.

Native advertising on the open web inverts every one of these dynamics. When you place a native ad on a publisher’s site through an open-web network, you write the headline. You select or approve the thumbnail. You choose contextual alignment — running alongside editorial content that matches your audience’s mindset, not alongside whatever conversational thread an AI model happened to generate. Most critically, you own the landing page. The click delivers a human being to a destination you built, where you control the narrative, capture first-party data, set attribution pixels, and measure the full journey from impression through conversion without depending on a platform’s self-reported metrics.

This is not a marginal distinction. It is the difference between being an advertiser and being an ingredient.

The open web still functions on the original architecture of digital marketing: a publisher creates content that attracts an audience, an advertiser places relevant creative adjacent to that content, a reader clicks because the creative earned their attention, and the resulting visit generates measurable, attributable value. No algorithm is deciding whether your brand deserves to appear. No AI is rewriting your value proposition. No checkout overlay is capturing the transaction before the customer reaches your site.

None of this means native is without challenges — viewability standards, content quality variance, and scale limitations are real. But those are optimization problems, not structural surrenders. In a landscape where Google is reshaping discovery around Gemini-powered AI experiences that absorb more of the user journey with each product cycle, the channel that still lets a marketer own the entire relationship between ad and action isn’t a nostalgic relic. It is a strategic necessity — and increasingly, the last unfair advantage available to brands unwilling to cede their identity to an algorithm.