Back to Blog
AI SearchMarket ResearchData AnalysisTrends

The State of AI Search in 2025: A Data-Driven Analysis

ChatGPT processes 1B+ searches weekly. 60% of searches end without clicks. Here's what the data says about where search is heading for your business.

PageX Team12 min read

For the first time since 2010, Google's global search share has fallen below 90%. ChatGPT now processes over 1 billion web searches per week. Gartner predicts a 25% drop in traditional search volume by 2026.

These aren't predictions—they're happening now. This analysis compiles the latest data on AI search adoption, user behavior changes, and what the numbers mean for businesses that depend on search visibility.

The Numbers: AI Search Market Share 2025

Google's Eroding Dominance

Google's position has measurably weakened:

MetricStatusChange
Global search share89.54%Below 90% for first time since 2010
General info searches66.9%Down from 73% in 6 months
Queries showing AI Overviews47-60%Varies by study and query type
89.54%
Google's global search share—first time below 90% since 2010Source: Cloudflare 2025 Year in Review

This isn't a collapse—Google still dominates. But the trajectory matters. For informational queries specifically, Google's share dropped 6 percentage points in just six months. That's the fastest erosion in the company's history.

AI Search Platform Market Share

The AI search landscape in December 2025:

PlatformMarket ShareNotes
ChatGPT61-82%Dominant leader, 77.97% of AI referral visits
Google Gemini13.4%Fastest growth—wasn't in top 10 at start of year
Perplexity6.4-8.2%Second-place AI-native player
Claude (Anthropic)3.8%Highest quarterly growth rate (14%)
Microsoft Copilot~5%Enterprise-focused

The Traffic Reality Check

Before over-rotating on AI search, context matters:

  • AI platforms account for 0.15% of global internet traffic
  • Traditional organic search accounts for 48.5%
  • However, AI traffic has grown since 2024

The absolute numbers are still small. The growth rate is extraordinary.

User Behavior: How People Actually Search Now

The Adoption Curve

77%
of Americans now use ChatGPT as a search engineSource: Adobe Express Survey 2025

The behavioral shift is real:

  • 24% of users now choose ChatGPT before Google when starting a search
  • Daily AI tool usage doubled from 14% to 29.2% in 2025
  • ChatGPT usage for general searches tripled from 4.1% to 12.5%
  • 34.8% of people have actively changed their search behavior

The Generational Divide

GenerationAI Search AdoptionBehavior
Gen Z70%+ active usersIncreasingly bypass Google entirely
Millennials70%+ active usersStrategic platform switching
Gen XModerate adoptionUse AI for complex queries
Boomers (57+)Slowest to adoptHigh satisfaction once they start

The generational pattern is important: Gen Z's preference for AI search (and TikTok, and Instagram as search platforms) isn't a trend—it's demographic inevitability. As they become the majority of internet users, their preferences become the default.

Query Style Transformation

How people ask has fundamentally changed:

Old BehaviorNew Behavior
3-5 word keyword queries20+ word conversational prompts
Single search sessionMulti-turn conversations with follow-ups
Click, read, returnGet answer, maybe verify
Linear journeyNon-linear, multi-platform exploration

"Tell me about" queries are up 70% year-over-year. "How do I" searches hit all-time highs, up 25%.

Users aren't just searching differently—they're having conversations with search.

The Verification Paradox

Despite AI's convenience, users don't blindly trust it:

  • 48% always verify AI answers across multiple platforms
  • Only 10% trust the first AI-generated result without question
  • 98% of ChatGPT users still also use Google

People are becoming strategic about platform selection based on query type. They use AI and traditional search in tandem, cross-referencing and verifying.

This creates both challenge and opportunity: if users verify across sources, you need visibility in multiple places.

Zero-Click: The New Reality

The Death of the Click

60%
of searches now end without any clickSource: Multiple industry studies 2025

The zero-click trend has accelerated:

  • When AI Overviews appear, users are 47% less likely to click traditional results
  • Only 8% of visits with AI summaries result in clicks (vs. 15% without)
  • Top-ranking pages affected by AI Overviews see 34.5% traffic cuts

For Google's AI Mode specifically, the numbers are even more stark: 93% of AI Mode searches end without a click—more than twice the rate of AI Overviews.

What Zero-Click Means

Zero-click doesn't mean zero value. When your brand is cited in an AI answer:

  • Users see your brand name (awareness)
  • You're positioned as authoritative (trust transfer)
  • Users may search for you directly later (delayed conversion)
  • Brand recall increases for future purchase decisions

The shift is from clicks as the goal to visibility as the goal. You're optimizing to be cited, not just to be clicked.

The Conversion Quality Offset

Here's the counterintuitive finding: while clicks decrease, the clicks that do happen are more valuable.

Adobe's research on AI referral traffic shows:

  • Visitors from AI search stay 8% longer
  • They explore 12% more pages
  • They're 23% less likely to bounce

The traffic is more qualified. Users arriving from AI recommendations have already been filtered—they're further in their research and more likely to convert. Our AI search conversion rates study quantifies how these engagement differences translate into revenue impact across different e-commerce categories.

What Users Actually Ask AI

Understanding query distribution helps prioritize optimization:

Query CategoryShare of AI Queries
General knowledge40.3%
Writing assistance19%
Product recommendations14.8%
Local business recommendations11.6%
Coding/technical help8.4%

For e-commerce, that 14.8% in product recommendations represents the direct opportunity. But the 40.3% in general knowledge is where you build authority—answer the questions that lead to purchase decisions.

AI Overviews: Google's Response

Rollout Data

Google's AI Overviews have been volatile:

Month (2025)Queries with AI Overviews
January6.49%
February7.64% (+18% MoM)
March13.14% (+72% from Feb)
July24.61% (peak)
November15.69% (pulled back)

Google aggressively tested in mid-2025, then scaled back—likely based on user data and quality concerns. The volatility suggests they're still calibrating.

Query Type Distribution

AI Overviews heavily favor informational queries:

  • 88-91% trigger on informational keywords
  • Navigational queries grew from under 1% in January to 10%+ by November
  • Science (25.96%), Computers & Electronics (17.92%), and People & Society (17.29%) see highest saturation

The Trust Problem

Despite high prevalence, trust remains an issue:

  • 58% of US adults have encountered at least one Google AI summary
  • Only 8.5% always trust AI Overview information

This trust deficit is a major hurdle—and an opportunity for brands that can become the trusted source AI cites.

Perplexity: The Challenger to Watch

Explosive Growth

Perplexity's trajectory deserves attention:

MetricValueGrowth
Valuation$18 billionUp from $14B mid-2025
Monthly active users22-45 million+47% YoY
Monthly queries780 millionTripled from 230M mid-2024
Daily queries30 millionUp from 3,000/day in 2022
Revenue (ARR)$100M+Up from $20M in 2024 (500% YoY)

Why Perplexity Matters

Perplexity has the best crawl-to-refer ratio among AI platforms:

PlatformCrawl-to-Refer Ratio
Anthropic25,000:1 to 100,000:1
OpenAIUp to 3,700:1
PerplexityBelow 400:1, often under 200:1

Lower ratios mean more efficient traffic generation per crawl—Perplexity sends more actual visitors relative to how much they crawl. This matters for site operators concerned about AI crawling overhead.

Business Model Evolution

Perplexity is building multiple revenue streams:

  • Premium subscription: $20/month
  • Email Assistant: $200/month (enterprise)
  • Publisher revenue sharing: 300+ publishers receive ad revenue when cited
  • Search API: Developer-focused product

The publisher revenue-sharing model is particularly interesting—it creates aligned incentives between Perplexity and content creators.

ChatGPT Search: The New Giant

Massive Scale

OpenAI's search numbers are staggering:

  • 1 billion+ web searches in ChatGPT per week (as of April 2025)
  • 700 million weekly active users (September 2025)
  • 400+ million weekly active users (February 2025 benchmark)
  • Overtook Wikipedia in monthly website visits

The Shopping Expansion

In April 2025, ChatGPT search added:

  • Product recommendations with images and reviews
  • Direct purchase links to merchant websites
  • Trending searches during typing (autocomplete competitor to Google)

This positions ChatGPT as a direct shopping research tool—users can now ask "What's the best moisturizer for oily skin under $50?" and get actionable purchase options.

Roadmap Signals

OpenAI has indicated focus on:

  • Improved vertical search in shopping and travel
  • Integration with o1 reasoning models for deeper research
  • Advanced voice integration for search
  • Canvas integration for complex research tasks

Analyst Predictions: Where This Goes

Gartner's Forecast

Gartner's predictions are the most widely cited:

By 2026:

  • Traditional search engine volume drops 25%
  • Daily queries decline from ~14 billion to 10-11 billion
  • "Generative AI solutions are becoming substitute answer engines"

By 2028:

  • Organic search traffic decreases 50% or more
  • Traditional SEO/PPC gives way to "Agent Engine Optimization"
  • Products need to be machine-readable for autonomous M2M transactions

The Economic Crossover

Multiple sources converge on timing:

  • Late 2027: AI search delivers equal or greater economic value than traditional search
  • 2028: $750 billion in US revenue flows through AI-powered search
  • 2029-2030: AI search definitively surpasses traditional search worldwide

Unprepared brands may see 20-50% decline in traditional search traffic with no alternative channel. For a deeper look at what these trajectories mean for e-commerce strategy, see our AI search predictions for 2026 and beyond.

What This Means For Your Business

The Strategic Imperatives

1. Visibility Over Clicks Optimize to be cited, not just ranked. Your brand mentioned in an AI answer—even without a click—builds awareness and trust.

2. Multi-Platform Presence Users verify across platforms. You need visibility in ChatGPT AND Perplexity AND Google AI Overviews AND traditional search.

3. Conversion Quality Over Volume AI-referred traffic converts better. Fewer clicks might mean equal or greater revenue if those clicks are more qualified.

4. Content for Citation Structure content so AI can extract and cite it accurately. Answer-first formatting, specific facts, proper schema markup.

5. E-E-A-T Investment AI systems cite authoritative sources. Expert credentials, third-party validation, and consistent cross-platform presence matter more than ever.

The Timeline

You're not preparing for a future shift—you're adapting to a current reality:

  • Now: 60% zero-click, 24% choose AI before Google
  • 2026: 25% volume decline, behavioral shift accelerates
  • 2027: Economic value crossover
  • 2028+: AI search becomes primary for majority of users

The brands building AI visibility today will have compounding advantage. Those waiting until 2027 will be playing catch-up against established AI-visible competitors.

See Your AI Search Position

PageX analyzes your visibility across ChatGPT, Perplexity, and Google AI Overviews. Understand where you stand today—and what to fix.

Get Free AI Visibility AuditFree • No credit card required

Frequently Asked Questions

Is Google dying?

No. Google still has 89%+ market share. But growth has stopped and share is eroding for the first time in 15 years. The more accurate framing: Google is becoming one of several search platforms rather than the only one.

Should I block AI crawlers?

Generally, no. Blocking AI crawlers means you can't be cited in AI answers. Unless you have specific concerns about AI training on your content, blocking removes you from a growing discovery channel.

Now. 24% of users already choose AI before Google. The brands building AI visibility today will have compounding advantage as adoption accelerates.

Will AI search replace traditional SEO?

AI search is evolving from SEO, not replacing it. The sites that rank well in traditional search are the sites AI cites. Your SEO foundation matters—AI visibility builds on top of it.

How do I measure AI search success?

Tools like Otterly.AI, Semrush AI Toolkit, and ZipTie track citations across AI platforms. Key metrics: citation frequency, accuracy, sentiment, and share of voice vs. competitors.


Sources

This analysis draws from:

Share this article

Ready to get AI-visible?

See how AI search engines view your site. Get your free AI visibility audit.