Who’s Losing the Most to Zero-Click AI Results? A Category-by-Category Breakdown for CMOs

Search volume is up. Clicks are disappearing. Both things are happening at the same time, and most marketing teams are only paying attention to one of them.

Zero-click searches jumped from 56% to 69% in the single year following Google’s AI Overviews launch, according to Similarweb. 

But the overall number obscures something more important. This shift is not landing evenly. Some categories are being structurally dismantled. Others are holding their ground. And the difference between them tells you almost everything about where to invest and where to stop pretending the old playbook still applies.

Here is the breakdown by category, who is bleeding, who is surviving, and what CMOs can actually do about it right now.

News and Media: The Category That Saw It Coming and Got Hit Anyway

Publishing is where zero-click does its most visible damage, and the numbers are genuinely hard to absorb.

Organic traffic to news publishers dropped from over 2.3 billion visits at its peak to under 1.7 billion, according to Similarweb data. That is more than 600 million monthly visits gone in under a year. 

Business Insider saw organic search traffic fall 55%. HuffPost’s desktop and mobile sites lost half of their search referrals. The New York Times saw search’s share of site traffic decline from 44% in 2022 to 37% in 2025.

The cruelest part: AI referral traffic from platforms like ChatGPT and Perplexity still is not enough to offset the decline in search-driven visits. 

General search referral traffic to 1,000 web domains dipped from 12 billion global visits in June 2024 to 11.2 billion in June 2025, about a 6.7% decline year over year. AI is reading the journalism, synthesizing it, and sending almost nobody back to the source.

The impact is uneven even within the category. Google has confirmed it does not trigger AI Overviews for hard news queries, meaning media brands with greater reliance on lifestyle and evergreen content are more likely to be pushed down search results.

The soft, high-volume editorial content that funded newsrooms for a decade is going first. Deeply reported, time-sensitive work is hanging on. For now.

News publishers expect search traffic to drop 43% by 2029, according to the Reuters Institute. That is not a crisis being managed. It is a category in permanent restructuring.

Education and How-To Content: The Category AI Was Built to Replace

Healthcare and education content faced nearly 90% AI Overview penetration on relevant queries. 

When someone searches “what are the symptoms of X” or “how do you calculate Y,” AI covers the query before any website enters the picture. The content that built some of the biggest B2B and media brands on the internet – definitive guides, how-to tutorials, explainer posts – is now precisely what AI Overviews are best at summarizing away.

Learning platform Chegg reported a 49% decline in non-subscriber traffic between January 2024 and January 2025, attributing the collapse directly to AI Overviews answering homework and study questions. The company filed an antitrust lawsuit against Google in February 2025. 

HubSpot is the case study the whole industry is still processing. HubSpot’s strategy of targeting low-depth, high-volume topics like “how to optimize blogging for SEO” and “what is a CRM” worked perfectly until AI Overviews eliminated traffic for broad, easily summarizable content.

The lesson from HubSpot is not that educational content failed. It is that educational content built for volume instead of depth became structurally vulnerable the moment AI could summarize it. The distinction matters, and it shows up even more clearly in the next category.

Traffic losses landed between 70 and 80%, a drop that sent shockwaves through the content marketing industry precisely because of the company’s reputation. As HubSpot’s own CEO acknowledged, “organic search traffic is declining globally” and “AI overviews are giving answers, and fewer people are clicking through to websites.” 

Finance and Health: The Surprise Victims

You would expect Google to tread carefully where bad information causes real harm. The data says otherwise.

Our own research at Previsible tells the story clearly: YMYL industries are seeing the largest AI adoption rates, with Legal growing 11.9x, Finance 2.9x, and Health 2.9x. 

These are the categories where Google spent years demanding the highest standard of human expertise and accountability. Now they are among the most aggressively summarized verticals in search.

Finance, weather, and conversion queries carry some of the highest zero-click rates, with certain queries scoring 100% according to Similarweb’s own analysis. 

When someone asks “what’s the current Fed rate” or “what should my emergency fund be,” they are not clicking anywhere. The answer is already there, and your content funded building it.

There is a reasonable question about whether that exchange was ever fair. But fairness is not a strategy, and the CMOs waiting for the ecosystem to self-correct are the ones losing ground fastest.

Sites in health, finance, and education verticals lost 11 to 23 percentage points of organic click share.

The categories that invested most heavily in meeting Google’s own quality standards are now being summarized away by the system those standards were meant to serve.

But this is where the conversation about informational content gets most interesting, and most instructive. A how-to article about mortgage rates is easy for AI to summarize. A financial advisor explaining why your specific situation changes everything is not. 

People searching in finance and health categories are not just looking for information. They are looking for guidance. They want to feel advised, not just answered. That is the opening.  Because the brands winning in these categories right now are not the ones producing the cleanest factual summaries, but the ones that are leading with human perspective, professional judgment, and the kind of nuanced context that turns a general answer into something that actually applies to someone’s life. 

And AI Overview can tell you what an emergency fund is. It cannot tell you whether you should prioritize yours over paying down your specific debt profile, with your income, in your market. That gap is where differentiated content lives and where trust gets built.

And the data backs this up. Websites with more organic traffic tend to get more mentions in AI Overviews and Perplexity, according to Ahrefs. The brands earning those citations did not abandon informational content. They made it deeper, more specific, and grounded in real expertise rather than keyword volume. AI is not killing informational content. It is killing shallow informational content. 

Users in these categories are still making high-stakes decisions. They are just making them increasingly inside the AI interface. Brands that show up in those responses with genuine human expertise, not just well-structured facts, are earning trust at the highest-intent moment in the funnel. 

Local and Transactional Search: The Relative Survivors

Not every category is losing ground. But the lines are moving faster than most people realize.

The conventional wisdom is that transactional and local searches are AI-proof. Users still need to click to buy, book, or call. That is mostly still true.

Pew Research Center’s analysis of 68,879 actual Google searches found that users clicked on traditional links just 8% of the time when an AI summary appeared, compared to 15% when it did not. 

That drop holds across query types. But the queries where users still need to take an action like make a reservation, call a contractor, or complete a purchase, are proving more resilient because the click is not optional. AI can describe the restaurant. It cannot seat you at the table.

According to Semrush’s analysis of over 10 million keywords, Real Estate and Shopping have the smallest share of keywords impacted by AI Overviews, with less than 3% of all keywords in those categories triggering one. 

Google already satisfies a lot of this demand with specialized high-utility SERP features like local packs and ads, and these industries lean toward local intent where a searcher is not looking for content, they are looking for directions on where to go next. 

That dynamic is starting to shift. Restaurant queries are one of the fastest-growing sectors for AI Overview coverage, jumping from triggering AI results on 10% of queries to 78% between early and late 2025, according to BrightEdge research. Google is inserting itself into the discovery layer of local search, the moment when someone is deciding where to eat, which contractor to call, or which service to book.

And here is the part of the local story almost nobody is telling yet. When Semrush analyzed the same keywords before and after an AI Overview appeared, zero-click rates actually fell from 33.75% to 31.53%.

Users interacting with AI summaries were clicking more, not less. A customer who reads an AI Overview about your restaurant before visiting already knows the menu, the vibe, and the price range. A user who finds your law firm through an AI summary already understands what kind of cases you handle. The click count may be lower. The quality of the person who does click may be significantly higher.

The net revenue impact for local businesses is still being measured. But the more interesting question is not what AI is taking away. It is what kind of customer AI is sending when it does send one. The early signals point toward a more informed, more decisive, more purchase-ready user, which can change the conversion math considerably.

What CMOs Should Do Right Now

Here is the problem with most of the advice on this topic: it treats zero-click like a terminal diagnosis and hands you a first-aid kit. Schema markup. llms.txt files. FAQ structure. Those are implementation details. They are not a strategy.

The bigger shift is simpler and harder at the same time.

Stop optimizing for the question AI can already answer. Start investing in what only your organization can produce.

That is the strategic reorientation everything else flows from. Here is what it looks like in practice.

Know What You Actually Have

Before changing anything, audit your content by a new metric: summarizability. For each of your top organic pages, ask one question: can AI fully answer this without the user ever visiting your site? 

If yes, that page is at structural risk regardless of its current ranking. This audit will tell you more than any traffic report, and it will force an honest conversation about where your content budget has been going.

Invest in Content AI Cannot Flatten

The companies seeing the worst traffic losses built content empires on volume and coverage. The companies still growing built authority on depth and differentiation. The difference is not a content format or a schema type. It is a commitment to producing things that could only come from inside your organization.

Your research. Your data. Your clients’ actual stories. The perspective that comes from fifteen years of doing the work, not from synthesizing what everyone else already published.

AI is an extraordinary summarizer. It is a poor substitute for genuine expertise, and users in high-stakes categories know the difference. Informational content is still worth building. Build it at a level AI cannot flatten into a paragraph.

Expand What You Measure

Traffic volume is now one signal among many, not the headline number. Deep engagement is occurring inside AI interfaces that your analytics will never see. Users are comparing your brand, researching your services, and making decisions without generating a single session in your reports.

The questions worth adding to your core dashboard: Where does your brand appear in AI responses? How is it being represented? Those questions do not have perfect answers yet. The CMOs building that visibility now will be two years ahead of the ones who notice the gap in 2027.

Treat Your Brand Narrative as Infrastructure

AI is increasingly intercepting the moment someone searches specifically for you. Your brand story, ie what you do, who you serve, why you are different, needs to be structured in a way AI systems can accurately read and represent. 

This is not a technical SEO task. This is traditional marketing. If your own website cannot answer the most common questions about your business in plain, unambiguous language, AI will fill that gap with whatever it can find. Make sure what it finds is yours.

The strategic question is no longer whether AI will reshape how your customers find you. It already has. The question is whether your organization is structured to compete in a discovery environment where the first impression happens inside an AI interface, not on your website.

The companies that move now to auditing content for summarizability, investing in genuinely differentiated expertise, and building measurement frameworks for AI visibility are not adapting to a trend. They are setting the competitive terms for their categories. The companies that wait will spend the next three years catching up to a standard someone else defined.

The old playbook rewarded whoever published the most content the fastest. The new one rewards whoever produces the content most worth citing. That shift favors organizations with real expertise, proprietary data, and a point of view which is exactly the advantage most established brands already have, if they stop burying it under content that AI can answer in a sentence.