Traffic Will Disappear. Your Brand Doesn't Have To.

CH.01

Two in Three Searches Will Never Click to Anyone Again

SparkToro’s US browser-panel analysis for January–April 2026 reports roughly 68% zero-click Google searches and 232 open-web clicks per 1,000 searches. These are US panel estimates, not a diagnosis of your site’s traffic.

SparkToro’s June 8, 2026 report separately describes an 8 percentage point decline, about 22%, in Google’s share of traffic to Ahrefs’ opt-in website panel from June 2025 to May 2026. This measures the share of website visits by referral source, not open-web clicks per Google search. The 2024 Datos and 2026 Similarweb search panels cover different users and devices, so they are not a continuous cohort.

Notice one thing: this curve was already climbing long before ChatGPT showed up. Zero-click isn't an AI invention: Google has been nibbling away at clicks with featured snippets, knowledge panels, and "People also ask" for a decade. What AI Overviews did was turn the nibbling into a feeding frenzy, and turn a chronic condition into an emergency.

CH.02

What three CTR studies measure

Pew tracks clicks after search-page visits, Ahrefs estimates desktop position-one CTR, and Seer follows monthly changes in a specific query cohort. Their periods, samples and measures differ. They do not establish one common effect or a decline of at least half in every study.

Pew’s July 22, 2025 study tracked browsing by 900 US adults in March 2025 and collected the corresponding search results on April 7 to 17. A traditional-result click followed 8% of visits to search pages with an AI summary, versus 15% without one. A cited-source click followed 1% of visits to pages with a summary. These are shares of search-page visits, not a finding that 1% of source links were clicked. The study did not measure brand influence.

Publishers got hit hardest. Similarweb reports that organic traffic to news sites fell from 2.3 billion monthly visits in mid-2024 to under 1.7 billion by May 2025 to 600 million visits evaporated in a year. Business Insider lost 55% of its organic search traffic over three years, then laid off 21% of its staff in May 2025. This isn't a forecast. It already happened.

• →Semrush: AI Overviews trigger rate doubled in three months (6.49% → 13.14%, Jan to Mar 2025) and keeps expanding.
• →SparkToro: in 2026, AI Overviews already appear in more than 20% of searches; when they do, CTR drops by nearly 60%.
• →Google I/O 2026: AI Mode crossed 1 billion monthly active users, with query volume doubling every quarter: answer engines aren't a trend, they're the status quo.
• →Seer’s November 4, 2025 retrospective cohort study reports paid CTR of 19.7% in June 2024 and 6.34% in September 2025 for informational/educational queries classified as AIO-prone in September 2025, with a publisher-reported decline of 68%. This is not a same-month comparison of pages with and without an AIO.

CH.03

The Field Playbook: The Vacuum After AI Overviews Lands in Your Market

Bain, Semrush, SparkToro, Pew: every zero-click study you've read is built on US or EU data. What about your own market? In May 2025, Google pushed AI Overviews into far more languages and regions. And then, almost no one measured what actually happened on the ground, market by market.

Scattered field tests from under-served markets point to a consistent range: after AI Overviews go live locally, site clicks typically fall 10 to 30%, across healthcare, beauty, e-commerce, and B2B alike. No systematic industry research, no local CTR benchmark, no zero-click annual report for these markets: it's a data vacuum. And a data vacuum has always been the early mover's window of advantage.

There's a second, badly underrated structural opportunity in under-served language markets: corpus scarcity. On the same topic, a mature English market has hundreds of authoritative pieces competing for the AI's citation; a smaller-language market often has only a handful of qualified sources. When an AI engine searches a thinner corpus for a credible answer to cite, the competitive density can be a tenth of what it is in English. Put another way: in the US, winning an AI citation is a marathon with a hundred runners; in an under-served market, there's almost no one on the starting line yet.

• →AI Overviews expanded to 200+ countries and 40 languages alongside Google I/O in May 2025.
• →Field tests: clicks down 10 to 30% (industry aggregate, 2025): the same shock curve as the English-language market, with the magnitude catching up fast.
• →Citation competition for high-quality professional content in under-served languages is far lower than in English, and the shelf life on that asymmetric advantage is shrinking as early movers arrive.

CH.04

The Invisible B2B Battleground: Shortlists Form Where You Can't See Them

Every zero-click study talks about consumers. But for B2B brands, the damage hides somewhere deeper: your buyers are building their vendor shortlists with AI, and that process leaves no trace on any dashboard you own.

Bain's survey sketches the outline: 68% of LLM users use them for research and information summaries, and 42% for purchase recommendations. Drop that behavior into a 6 to 18-month B2B sales cycle and the meaning changes completely: the early-stage questions of "what options are on the market" and "how do I choose between A and B" are exactly the kind of prompt AI is best at and buyers lean on most. The moment a member of the buying committee asks ChatGPT "what HR systems are good for a mid-sized company," the shortlist sets in those 30 seconds.

The brutal part: none of this stage sends a single signal your way. No impression, no referral, no form fill. You'll only notice three months later that inbound has thinned and your name isn't on the RFP invite list, and then chalk it up to "the market cooling off." Neil Patel calls this the "quiet erosion" of brand presence; for B2B, the sharper way to put it is: you're losing a race you didn't know was being run.

CH.05

Bain Says Kill Gated Content. Patel Says Keep It. Who's Right?

There's an unspoken, head-on contradiction in the zero-click literature. Bain argues that PDFs and gated content are "relics in the AI ecosystem" that should be retired; Neil Patel lists gated content as the core of a first-party data strategy that should be kept. Both are right, because they're talking about two different layers of content.

Bain's logic: AI crawlers can't read what's behind a form, so gated content is invisible to AI, and AI is becoming the primary interface for discovery. Patel's logic: traffic is scarce in the zero-click era, so every visitor willing to leave an email is a rare asset, and first-party data is the only moat left. Stack the two logics together and the answer isn't either/or: it's layered.

Figure 2.1 is fully public and free to quote. The complete seven chapters and the general whitepaper library's full text are directly readable; only the 2026 flagship report requires an email.

CH.06

With No Clicks Left, How Do You Answer to the CFO?

Every source shouts "change your metrics," and not one of them hands you a complete operating system. NP Digital's scoring model (1 point for a featured snippet, 2 for an AI citation) is a start, but take that to your CFO for budget and they'll ask: how does a score turn into money?

The root of the problem: exposure metrics (search impressions, AI visibility scores, brand mention volume) describe "presence," while budget decisions need a "chain of causation." A measurement system for the zero-click era has to run all the way from presence to revenue: break any link, and the narrative falls apart.

• →The weekly report watches layer one (citation-rate shifts → rewrite queue), the monthly watches layer two (answer share → strategy review), the quarterly watches layer three (pipeline → budget).
• →Set alerts, not a wall of dashboards: "citation rate on the core prompt set has fallen two weeks running" beats 20 charts.
• →Don't treat GSC impressions as an outcome metric: surging impressions + collapsing CTR is the textbook symptom of the AI Overviews era (Semrush case: impressions doubled, CTR fell from 1.5% to below 0.5%).

CH.07

The 90-Day Zero-Click Response Playbook

Understanding is over; now you build. This playbook assumes you're starting from zero: a B2B team with content assets but no systematic GEO work yet. Ninety days from now, you'll have a baseline, a first batch of cited content, and a set of numbers you can take into a leadership conversation.