VZ editorial frame
Read this piece through one operating lens: AI does not automate first, it amplifies first. If the underlying decision architecture is clear, AI scales clarity. If it is noisy, AI scales noise and cost.
VZ Lens
In VZ framing, the point is not novelty but decision quality under uncertainty. The Impact of Zero-Click Search and AI Overviews on Web Traffic, Publisher Revenue, and SEO from 2019 to 2026. Data-Driven Analysis of Threats and Opportunities. The real leverage is in explicit sequencing, ownership, and measurable iteration.
[!abstract] Summary This document summarizes the impact of zero-click searches (when a user does not click on any results) and Google AI Overviews (AI-generated answers at the top of search results) on web traffic, publisher revenue, and the future of search engine optimization (SEO). The data covers the period from 2019 to early 2026.
In the lobby of a hotel in Bratislava
The laptop screen glows in the dimly lit lobby. A coffee cup sits on my desk, its steam illuminating slowly swirling dust particles in the morning sunlight. Outside, through the glass, the city is still sleepy. My finger glides across the touchpad, scrolling, stopping. Search results line up: concise answers, graphs, summaries—everything is here, on a single page. No need to go further. Sipping my coffee, I watch as the knowledge I seek is no longer a gateway, but a destination. Information settled on the screen, a closed loop. Sitting in front of the laptop, I wonder what happens when the answer is no longer a guide, but a final stop.
1. What is zero-click search?
Zero-click search means that someone types a question into Google, gets the answer directly on the results page—and doesn’t click through to any website. Think about it: if you search for “what’s the temperature in Budapest,” Google immediately shows you the weather. You don’t have to click on anything.
This used to work only for simple questions (weather, currency conversion, sports scores). Today, however, thanks to AI-powered answers, Google answers even complex questions directly—which means websites are getting fewer and fewer visitors.
The Core of the Problem
Search → Results page → [Answer appears] → User is satisfied → DOES NOT click
↓
Website: 0 visitors, 0 revenue2. Zero-click statistics: The story behind the numbers (2019–2026)
2.1 Rand Fishkin / SparkToro Research
Rand Fishkin, founder of SparkToro—one of the most renowned experts in search marketing—has been tracking for years what percentage of Google searches end without a click. They use Datos’ (a Semrush subsidiary) clickstream panel—browsing data collected from tens of millions of devices—for their analysis.
| Year | Zero-click rate (US) | Zero-click rate (EU) | Source |
|---|---|---|---|
| 2019 | ~50% | n/a | SparkToro / Jumpshot |
| 2020 | ~65% | n/a | SparkToro (Jumpshot discontinued) |
| Q1 2024 | 58.5% | 59.7% | SparkToro / Datos (Semrush) |
| Q1 2025 (March) | 60–65% | ~60% | Semrush / Datos / SparkToro |
| Mid-2025 | ~65% | n/a | Onely (aggregated) |
| November 2025 | ~58% | n/a | Superprompt.com (US desktop) |
| 2026 forecast | 70–75% | n/a | Industry estimate |
[!warning] Measurement discrepancies Zero-click rates can vary significantly depending on the measurement methodology. SparkToro/Datos uses a desktop click data panel (millions of devices), Semrush analyzes the entire sample of US searches, while others (e.g., Onely, Superprompt) produce their own aggregates. The 2020 figure of 65% was the last measurement taken before the Jumpshot panel was shut down; the 2024 re-measurement (58.5%) was conducted using a different methodology. The bottom line: the trend is clearly upward.
2.2 Trend visualization (ASCII chart)
Zero-click rate (%) — U.S. Google searches
80% | ▓▓ ← 2026 AIO
| ▓▓ searches: 83%
75% | ╔══════╗
| ║ 70%+ ║ ← 2026 forecast
70% | ╚══════╝ (total searches)
| ┌────┐
65% | ┌────┐ │65% │
| │65% │ └──┬─┘
60% | └──┬─┘ ┌────┐ │
| │ │58.5│ ┌──┴──┐
55% | │ └──┬─┘ │ 60% │
| ┌────┐ │ │ └──┬──┘
50% | │50% │ │ │ │
| └──┬─┘ │ │ │
45% | │ │ │ │
|─────┴──────┴──────────┴────────┴──────────►
2019 2020 2024 2025 20262.3 What These Numbers Mean
Out of 1,000 Google searches:
- In 2019: ~500 searches ended without a click → 500 clicks were received on websites
- In 2024: ~585 searches ended without a click → 415 clicks were received
- In 2025: ~600–650 searches ended without a click → 350–400 clicks were received
- In 2026 (forecast): ~700–750 searches end without a click → 250–300 clicks
In other words, the number of clicks on websites could drop by half over the course of 6 years.
2.4 The decline in organic clicks
According to Semrush’s 2025 data, the organic click-through rate (organic CTR—the percentage of users who click on unpaid search results) is also declining:
| Period | Organic CTR (US) | Organic CTR (EU/UK) | Change |
|---|---|---|---|
| March 2024 | 44.2% | 47.1% | — |
| March 2025 | 40.3% | 43.5% | -8.8% / -7.6% |
3. Google AI Overviews (AIO) — Rewriting the Rules of the Game
3.1 Historical Overview
Google AI Overviews (AIO — AI-generated answer summaries) appear at the top of search results and directly answer the user’s question without requiring a click on any webpage.
| Date | Event |
|---|---|
| May 25, 2023 | Google I/O: Introduction of Search Generative Experience (SGE), experimental phase begins (Search Labs) |
| August 2023 | SGE expands to India and Japan |
| November 2023 | SGE available in 120+ countries |
| May 14, 2024 | Google I/O: SGE renamed to AI Overviews, exits beta in the US market |
| August 2024 | AIO expansion: UK, India, Japan, Brazil, Mexico, Indonesia |
| January–March 2025 | AIO appearance rate rises sharply in the US |
| July 2025 | AIO appearance peak: ~25% of all searches |
| November 2025 | AIO appears in ~60% of searches (xponent21 estimate) |
| End of 2025 | Google AI Mode launches — even more comprehensive AI responses, 93% zero-click rate |
3.2 How often does AIO appear?
Data varies by source because different methodologies are used to measure it:
| Source | Period | AIO appearance (% of searches) |
|---|---|---|
| Semrush | January 2025 | 6.49% |
| Semrush | February 2025 | 7.64% (+18%) |
| Semrush | March 2025 | 13.14% (+72%) |
| BrightEdge | March 2025 | ~35% (US desktop) |
| Ahrefs | July 2025 | ~25% (peak) |
| xponent21 | November 2025 | ~60% (US) |
[!info] Navigational searches are also affected The proportion of navigational AIOs (i.e., when someone searches for a specific website, e.g., “Facebook login”) was below 1% in January 2025 but had risen to over 10% by November. AIOs are therefore spreading not only in informational searches but in virtually every type of search.
3.3 The Impact of AIO on Click-Through Rates (CTR)
A Seer Interactive study from fall 2025—which analyzed 3,119 informational search terms across 42 organizations, based on 25.1 million organic and 1.1 million paid impressions (June 2024 – September 2025)—revealed dramatic results:
| Metric | WITHOUT AIO | WITH AIO | Change |
|---|---|---|---|
| Organic CTR | 1.76% | 0.61% | -61% |
| Paid CTR | 19.7% | 6.34% | -68% |
| Zero-click rate | ~56% | ~69% | +13 pp |
However, if your website is cited as a source in AIO:
| Metric | Not cited | Cited in AIO | Difference |
|---|---|---|---|
| Organic CTR | baseline | +35% | More clicks |
| Paid CTR | baseline | +91% | Significantly more clicks |
This is the key insight: being cited in AIO is the new “ranking first”.
3.4 Google AI Mode — The Next Level
Google’s AI Mode (full-screen AI response, conversational interface), launched in late 2025, shows even more dramatic zero-click rates:
| Search Interface | Zero-Click Rate |
|---|---|
| Traditional Google Search (without AIO) | ~60% |
| Google Search with AIO | ~69% |
| Google AI Mode | ~93% |
In the case of AI Mode, 75% of users never leave the Google interface during their search. This effectively spells the end of website visits for a significant portion of information searches.
3.5 Featured Snippets vs. AI Overviews
The Featured Snippet used to appear at the top of Google’s results page, highlighting the answer from a single source. The AI Overview replaces this, but synthesizes the answer from multiple sources.
| Feature | Featured Snippet | AI Overview |
|---|---|---|
| Appearance | Answer highlighted from a single source | AI-synthesized answer from multiple sources |
| Number of Sources | 1 | Multiple (usually 3–6) |
| CTR Impact | Moderate decrease for other results | Dramatic decrease for all results |
| If Featured | High CTR (highlighting attracts users) | Moderate CTR increase (+35%) |
| Appearance rate | Previously ~20% | Replaced by AIO (13–60%) |
According to Amsive’s analysis, when a Featured Snippet and an AI Overview appear simultaneously for a search query, the average CTR drop reaches -37%.
Ahrefs’ July 2025 analysis of 1 million SERP pages found that AI Overviews have practically replaced Featured Snippets.
4. Alternative AI Search Engines: The Fragmenting Search Market
4.1 The Big Picture: Who Searches Where? (March 2026)
Traditional search remains Google’s domain, but AI chatbots are carving out an increasingly larger share:
Market share of traditional search engines (global, early 2026):
| Search Engine | Market Share |
|---|---|
| 89.87% | |
| Bing | 4.31–4.43% |
| Yandex | 1.92% |
| Baidu | 0.7–0.9% |
| DuckDuckGo | 0.6–0.7% |
[!important] The Hidden Change These figures are misleading because they do not include AI chatbot searches (ChatGPT Search, Perplexity, Claude). If we include these, Google’s “true” share of all digital searches (queries) is actually lower.
AI Chatbot Market Share (March 2026):
| Platform | Market Share (AI Chatbot) | Monthly Active Users |
|---|---|---|
| ChatGPT | 64.5–68% | ~800M–1B weekly |
| Google Gemini | 13.5–25.2% | ~750M monthly |
| Microsoft Copilot | 7.22–14.3% | n/a |
| Perplexity AI | 6.4–8.03% | ~33–45M monthly |
| Claude (Anthropic) | 2–4.5% | ~18.9M monthly |
| DeepSeek | ~1–2% | n/a |
4.2 Perplexity AI — The Rise of the AI-Native Search Engine
Perplexity AI is an AI-native search engine (answer engine) that generates direct, source-cited answers. It is not a traditional search engine, but rather a “research assistant.”
| Metric | Data | Period |
|---|---|---|
| Monthly active users | 33–45 million | Second half of 2025 |
| Monthly search queries | 780 million | May 2025 |
| Search growth | +239% | Aug. 2024 → May 2025 |
| Total searches (2025–2026) | ~6–7 billion | Estimate |
| Valuation | $9 billion → $18 billion | March 2025 |
| Daily active users | 19% of users | 2025 |
| Desktop vs. mobile ratio | 82.5% desktop / 17.5% mobile | June 2025 |
[!note] Perplexity context Perplexity’s market share in the AI chatbot segment is 6.4–8%, which seems small—but it is the fastest-growing AI search engine (524% increase in queries in 2024). Its estimated share of the traditional search engine market, however, is still only a fraction of a percent.
4.3 ChatGPT Search
ChatGPT launched the ChatGPT Search feature at the end of 2024, which enables real-time web searches when generating AI responses.
| Metric | Data |
|---|---|
| Weekly active users (ChatGPT total) | 800M–1 billion (early 2026) |
| Daily queries | 2.5 billion |
| Monthly web visits | 5.7 billion |
| AI chatbot market share | 64.5–68% (declining — 82%+ in early 2025) |
| Traditional search market share | ~1% [NOT VALIDATED] |
ChatGPT is still a marginal player in the traditional search market, but it dominates the AI chatbot segment—and an increasing share of search intent is being fulfilled here.
4.4 Claude and Other AI Platforms
Claude (Anthropic) positions itself primarily as an AI assistant rather than a search tool, but an increasing number of people are using it for research and information retrieval.
| Metric | Data |
|---|---|
| Monthly active users | ~18.9 million |
| AI chatbot market share | 2–4.5% |
| Valuation | $380 billion (February 2026) |
| Corporate customers | 300,000+ |
| Enterprise AI market share | ~29% |
| Claude Code revenue (run-rate) | ~$2.5 billion |
[!note] A unique feature of Claude Claude is not necessarily a competitor to traditional search—however, with a 29% market share in the enterprise segment, business research and analysis tasks are increasingly being performed by AI assistants, not Google.
4.5 Google Gemini — Is Google’s own product cannibalizing search?
Google’s paradox: its own AI chatbot, Gemini, is also taking traffic away from traditional Google search.
| Metric | Data |
|---|---|
| Monthly active users | 750+ million (Q4 2025) |
| Market share (AI chatbot) | 13.5–25.2% (growing rapidly) |
| Growth | Jan 2025: 14.7% → Jan 2026: 25.2% |
| Enterprise customers | 120,000+ |
| Device reach (Android integration) | 1–5 billion devices |
5. The collapse of organic traffic — Industry impacts
5.1 Publishers’ traffic loss
The Press Gazette’s January 2026 report (based on Chartbeat data) and the Reuters Institute’s 2026 Journalism & Technology Trends report reveal dramatic figures:
| Metric | Data | Source |
|---|---|---|
| Global Google traffic decline (for publishers) | -33% (year-over-year) | Press Gazette / Chartbeat |
| U.S. Google traffic decline | -38% (year-over-year) | Press Gazette / Chartbeat |
| Europe Google traffic decline | -17% (year-over-year) | Press Gazette / Chartbeat |
| Google Discover traffic decline | -21% (year-over-year) | Press Gazette / Chartbeat |
| Business Insider organic traffic decline | -55% | Apr. 2022 → Apr. 2025 |
| Chegg non-subscriber traffic decline | -49% | Jan. 2024 → Jan. 2025 |
5.2 Specific industry impacts
| Industry | Zero-click effect | Traffic loss |
|---|---|---|
| News media / publishers | Highest — AI responses are taking over news content | -33% (global average) |
| B2B / SaaS | High — how-to articles and comparisons summarized by AI | -34% (annual average), 73% of affected websites |
| E-commerce | Moderate — product comparisons via AI, but purchasing still requires a click | -15–25% [NOT VALIDATED] |
| Healthcare content | High — symptom descriptions, drug information answered instantly | n/a |
| Education / Curriculum | Very high — Chegg, Stack Overflow-type sites most severely affected | -49% (Chegg) |
| Local service providers | Medium — Google Maps and Local Pack continue to generate clicks | Minor impact |
5.3 Publishers’ Expectations — Reuters Institute Survey
The Reuters Institute for the Study of Journalism’s 2026 report surveyed 280 media executives (64 editors-in-chief, 64 CEOs, 51 digital leaders) from 51 countries:
| Question | Answer |
|---|---|
| Expected decline in search traffic over the next 3 years | -43% (average) |
| Optimists (expecting a decline of < 20%) | Few |
| Pessimists (expecting a decline of > 75%) | ~20% of respondents |
| Planning to reduce investment in traditional SEO | Majority |
| Would focus on AI platforms (ChatGPT, Gemini, Perplexity) | Growing intention |
5.4 Visualization of AI traffic displacement (before/after)
1,000 searches → WHERE DO THE CLICKS GO?
2019: 2025: 2026 (forecast):
┌──────────────────┐ ┌──────────────────┐ ┌──────────────────┐
│ Zero-click: 500 │ │ Zero-click: 620 │ │ Zero-click: 720 │
│ ████████████████ │ │ ████████████████ │ │ ████████████████ │
│ ████████████████ │ │ ████████████████ │ │ ████████████████ │
│ │ │ ████████████████ │ │ ████████████████ │
│ │ │ ████ │ │ ████████████████ │
├──────────────────┤ ├──────────────────┤ │ ████ │
│ Organic: 360 │ │ Organic: 280 │ ├──────────────────┤
│ ░░░░░░░░░░░░░░ │ │ ░░░░░░░░░░ │ │ Organic: 180 │
│ ░░░░░░░░░░░░░░ │ │ ░░░░░░░░░░ │ │ ░░░░░░░░ │
├──────────────────┤ ├──────────────────┤ ├──────────────────┤
│ Paid: 100 │ │ Paid: 60 │ │ Paid: 50 │
│ ▓▓▓▓▓▓ │ │ ▓▓▓ │ │ ▓▓ │
├──────────────────┤ ├──────────────────┤ ├──────────────────┤
│ Google internal: 40 │ │ Google internal: 40 │ │ AI chatbot: 50 │
│ ▒▒ │ │ ▒▒ │ │ ▒▒▒ │
└──────────────────┘ └──────────────────┘ └──────────────────┘
Website clicks:
2019: ~460 2025: ~340 2026: ~230 (← 50% decrease over 7 years)6. The “SEO is dead” narrative — What is the reality?
6.1 The Gartner Prediction
According to Gartner’s February 2024 forecast:
“By 2026, traffic to traditional search engines will decrease by 25% due to AI chatbots and virtual agents.”
This forecast received significant media attention. In March 2026, we are in the forecast’s “target year”—and based on available data, the 25% decline is close to reality, though the volume of traditional search engines alone has not decreased by that much: rather, a portion of searches is shifting to AI platforms.
6.2 The reality is more nuanced
The claim that “SEO is dead” is an exaggeration — but SEO is undergoing a fundamental transformation. Here is the reality:
| Claim | Reality |
|---|---|
| ”Google is dead” | False. Google remains dominant with an 89.87% market share. The number of searches isn’t decreasing; it’s increasing. |
| ”SEO is unnecessary” | False. SEO is more important than ever — but different types of SEO are needed (GEO, AEO). |
| ”No one clicks anymore” | Partially true. CTR is dropping drastically for informational searches. But searches driven by purchase intent, comparison, and navigation still generate clicks. |
| ”AI is replacing websites” | Partially true. Informational content (articles, how-tos, definitions) is receiving less and less traffic. But transactional content (webshops, services, bookings) is holding its own for now. |
| ”SEO is going back to the way it used to be” | False. The change is structural and irreversible. |
6.3 The new search engine optimization: SEO → GEO + AEO
SEO isn’t dead—but it has split into three branches:
| Type | Description | Purpose |
|---|---|---|
| SEO (Search Engine Optimization) | Traditional search engine optimization | Organic results on Google/Bing |
| GEO (Generative Engine Optimization) | Generative engine optimization | Appearing in AI chatbot responses (ChatGPT, Perplexity, Gemini) |
| AEO (Answer Engine Optimization) | Answer engine optimization | AI Overviews, Featured Snippets, Google AI Mode quotes |
Gartner also notes: GEO and AEO are synonymous in many contexts—the key is for your content to become a source of AI-generated answers.
[!tip] The 2026 triple-threat strategy All three optimizations are necessary for an effective online presence:
- SEO — for people browsing (who are comparing and buying)
- AEO — to appear in Google AI Overviews and AI Mode snippets
- GEO — appearing as a source in ChatGPT, Perplexity, and Claude responses
6.4 What’s Really Dead
It’s not SEO that’s dead, but “bad SEO”:
- Mass-produced, AI-generated content (created solely for ranking)
- Keyword stuffing and manipulative link building
- “Thin content” (superficial, short articles that offer no added value)
- Clickbait headlines with no real content behind them
What works in 2026:
- Unique, expert content (E-E-A-T: Experience, Expertise, Authoritativeness, Trustworthiness)
- Structured data (schema markup) — easily readable by AI
- Primary research, proprietary data — which AI cannot gather elsewhere
- Brand loyalty and direct traffic (direct traffic)
7. The Financial Impact — Revenue Loss and Adaptation
7.1 Shift in Advertising Revenue
The impact of AI search on the advertising market is twofold:
| Impact | Details |
|---|---|
| Publishers’ advertising revenue declines | Fewer visitors = fewer impressions = less advertising revenue |
| CPM rises | As “free” traffic declines, paid traffic acquisition becomes more expensive |
| Google’s advertising revenue (for now) increases | Google also displays ads within AIOs (Jan 2025: 3% → Nov: 40%) |
| Subscription models are coming to the fore | 76% of publishers view subscriptions as their main source of revenue |
7.2 Industry Adaptation Strategies
According to a Reuters Institute survey, publishers are planning the following steps:
- Subscriptions and membership (76%) — the primary revenue focus
- Display ads (68%) — but with declining expectations
- Native ads (64%) — content that is harder to summarize using AI
- Partnerships with AI platforms — licensing fees for AI use of content
- Building direct traffic — newsletters, apps, communities
7.3 The Future of Web Traffic — Aggregated Forecast
| Period | Expected organic traffic change (vs. 2024) | Main reason |
|---|---|---|
| End of 2025 | -20 – -30% | AIO adoption, rise of zero-click |
| End of 2026 | -30 – -40% | AI Mode, growth of Perplexity/ChatGPT |
| 2029 | -43% (publisher estimate) | Reuters Institute survey |
8. Future Scenarios for the AI Search Ecosystem
8.1 Scenario A: “Google Retains Control” (40% probability)
Google successfully integrates AI into search (AIO, AI Mode) and shifts advertising revenue to the new format. Alternative AI search engines (Perplexity, ChatGPT Search) remain niche.
Impact: Website traffic will decrease by 30–40%, but most of the traffic will remain within the Google ecosystem.
8.2 Scenario B: “Fragmentation” (45% probability)
The search market splits across multiple platforms: Google (60–70%), ChatGPT Search (10–15%), Perplexity (5–8%), Gemini (10–15%), others (5%). There is no single dominant AI search engine—users switch between them depending on context.
Impact: GEO and AEO become critical. Websites must optimize for all platforms. Total traffic decreases, but comes from more sources.
8.3 Scenario C: “The Age of AI Agents” (15% probability)
AI agents take over search: the user does not search, but gives instructions to the AI, which independently researches, summarizes, and decides. The concept of website visits becomes obsolete in the realm of information searches.
Impact: The collapse of the click-based business model (advertising, affiliate marketing, SEO). The value chain will be restructured: content owners will collect licensing fees from AI platforms.
9. Summary: The most important numbers in one place
| Metric | 2019 | 2024 | 2025 | 2026 (forecast) |
|---|---|---|---|---|
| Zero-click rate (US) | ~50% | 58.5% | 60–65% | 70–75% |
| AIO appearance rate | — | — | 13–60% | 60%+ |
| Zero-click rate for AIO searches | — | — | ~69% | ~83% |
| AI Mode zero-click rate | — | — | ~93% | ~95% [UNVALIDATED] |
| Organic CTR (AIO search) | — | — | 0.61% | Declining further |
| Publishers’ Google traffic change | — | — | -33% (year-over-year) | -43% (3-year forecast) |
| Perplexity users (monthly) | — | ~15M | 33–45M | 60M+ [UNVALIDATED] |
| ChatGPT weekly users | — | ~200M | 800M+ | 1B+ |
| Google search market share | ~92% | ~91% | ~90% | ~89% |
| Gartner: decline in traditional search | — | — | — | -25% (forecast) |
10. Methodological notes and sources
Data Sources
Main data sources cited in this document:
- SparkToro / Datos (Semrush) — clickstream panel, millions of devices, desktop
- Seer Interactive — 3,119 search terms, 42 organizations, 25.1M organic impressions
- Semrush — US/EU search data, Q1 2025
- Chartbeat / Press Gazette — traffic data from 2,500+ publisher websites
- Reuters Institute — survey of 280 media executives, 51 countries
- BrightEdge — AI Overview impression data, US desktop
- Ahrefs — 1M SERP analysis, July 2025
- Gartner — February 2024 forecast
- First Page Sage — AI chatbot market share, March 2026
- Similarweb — chatbot market share, January 2026
- DemandSage, BusinessOfApps, Backlinko — platform-specific statistics
Important caveats
[!warning] Data quality warnings
- Zero-click rates are methodology-dependent — SparkToro/Datos uses a desktop panel, Semrush uses a different sample, and aggregate estimates (65%, 70%+) are often extrapolations.
- AIO impression rates vary widely (from 6.49% to 60%) — this depends on the measurement methodology and the time period.
- The “2026 forecast” values are industry estimates, not measured data.
- Market share data may vary by 10–20 percentage points depending on the source (e.g., ChatGPT: 64.5% vs. 81%).
- Data marked [UNVALIDATED] could not be confirmed from a primary source.
Research completed: March 9, 2026 | Search tools used: Brave Search, WebSearch, manual source analysis Next step: [[03_LLM_Traffic_Alternatives]] — How to build traffic in the world of AI search engines?
Strategic Synthesis
- Translate the thesis into one operating rule your team can apply immediately.
- Use explicit criteria for success, not only output volume.
- Iterate in small cycles so learning compounds without operational noise.
Next step
If you want your brand to be represented with context quality and citation strength in AI systems, start with a practical baseline and a priority sequence.