Is ChatGPT Killing Search Traffic? What the Data Actually Shows

ChatGPT and Google Are Not Competing for the Same Users

Google Sends 190 Times More Traffic Than ChatGPT Does

Your Google Search Console traffic dropped. Someone blamed ChatGPT. A consultant called SEO dead. Before you restructure anything, look at the actual numbers. SparkToro and Datos analysed US clickstream data and found Google processed more than 5 trillion searches in 2024. That is roughly 14 billion per day. ChatGPT handled an estimated 37.5 million search-like prompts daily during the same period. The gap is 373 to 1. ChatGPT is not killing search traffic at the macro level. Something else is doing it — and identifying the real cause determines whether your response actually works.

Ahrefs’ analysis found ChatGPT processes about 12% of Google’s total query volume when counting all prompts. Its click-through rate to external websites is 96% lower than Google’s. For every 190 visitors Google sends to a website, ChatGPT sends 1. The platform is large in usage and nearly invisible in referral output. Understanding that distinction changes every decision you make about SEO and content investment. The problem is not ChatGPT pulling traffic away from you. The problem is Google keeping traffic on its own pages.

Run This Diagnostic on Your Own Data First

Before reading further, run this checklist against your Google Search Console and GA4. It identifies which problem you actually have — so you address the right one.

1. Search Console impressions are rising while clicks are flat or falling. This signals AI Overview interference, not ChatGPT cannibalisation.
2. Click-through rate for informational keywords dropped sharply around March 2025. This matches Google’s broader AI Overviews rollout, not ChatGPT growth.
3. ChatGPT contributes less than 0.5% of total GA4 sessions. Ahrefs’ analysis of 44,421 sites found ChatGPT drives just 0.19% of web referral traffic on average. You are in the majority.
4. Branded search volume has risen while non-branded organic clicks fell. AI is influencing discovery upstream and leaving no visible referral in your analytics.
5. You have not created a custom GA4 channel group to isolate traffic from ChatGPT, Perplexity, Gemini, and Claude. You cannot manage what you cannot measure separately.
6. Highest-traffic informational pages lost 20% or more in CTR while rankings stayed stable. The Semrush 10-million-keyword AI Overviews study documented this pattern broadly.
7. Commercial pages — product, service, pricing — held traffic while blog posts collapsed. Transactional queries are more protected from AI Overview interference than informational ones.

0 to 2 checked: Your traffic is likely stable. Monitor monthly but no overhaul is needed yet. 3 to 4 checked: AI Overviews are your primary issue. Focus on AI citation optimisation and commercial-intent content. 5 or more checked: You face the compound effect of AI Overviews, zero-click growth, and shifting behaviour. You need a dual-track strategy: maintain traditional SEO while building generative engine visibility simultaneously.

Google Is Keeping Clicks, Not Losing Them to ChatGPT

Google Search Volume Grew as ChatGPT Usage Rose

The counterintuitive finding from 2024 is that Google did not shrink when ChatGPT grew. The SparkToro and Datos research confirmed that Google search volume grew 21.64% from 2023 to 2024. Google’s own disclosure supports this: 5 trillion searches in one year. The AI narrative claimed Google was losing users. The data showed the opposite in query volume terms. More people searched on Google in 2024 than in any previous year. Growth in searches does not mean growth in website clicks, however. That distinction is where the real story lives.

WAV Group’s 2025 analysis quantified the referral gap: Google sends approximately 16 billion clicks to publishers each day. ChatGPT sends roughly 25 million. That is a 640-to-1 ratio in daily referral output. The ChatGPT-as-traffic-killer narrative persists at conferences as if it were settled fact. The data does not support it. What the data shows is more precise: Google is answering more queries on its own pages, sending fewer users to yours. ChatGPT is a small, growing, commercially interesting channel — but not the cause of the traffic you are missing.

AI Overviews Drive the Actual Traffic Decline

Grow and Convert’s analysis of multiple client accounts found a consistent pattern: rankings improved while clicks fell. The inflection point was March 2025. That was when Google expanded its AI Overviews rollout broadly across query types. The timing rules out ChatGPT as the cause. AI Overviews appear for informational queries. They deliver the answer without requiring a click. The result is a decoupling of impressions from traffic that SEO professionals had not seen at this scale before.

Ahrefs’ December 2025 study examined 300,000 keywords. It found that the presence of an AI Overview correlates with a 58% lower average CTR for the top-ranking page. Seer Interactive found organic CTR fell between 49% and 65% for AI Overview queries. Authoritas documented a 47.5% reduction. Multiple publisher disclosures corroborated the same range. The mechanism is straightforward. Google answers the question on the search page. Users who find the answer satisfactory do not click through to any website. Queries without AI Overviews showed slight CTR improvement in some studies — confirming the feature, not broader AI adoption, is the driver.

Zero-Click Searches Hit a New Structural Level in 2025

The zero-click phenomenon predates AI Overviews, but the feature accelerated it sharply. Digital Bloom’s organic traffic crisis report documented that 60% of Google searches ended without a click in the first half of 2025. Similarweb’s July 2025 study found searches ending without a click rose from 56% to 69% following the broader AI Overviews launch. That 13-point increase represents billions of daily searches where no publisher receives a visit. Most marketing teams attributed this to ChatGPT. It was Google’s own product design decision.

One important nuance from the Semrush AI Overviews study: the zero-click rate for keywords with AI Overviews declined slowly from January to November 2025. The initial disruption included a novelty effect. Some user adaptation occurred. The zero-click rate remains significantly higher for AI Overview queries than for standard results. But the trajectory is not continuously worsening. Publishers who held their strategy through mid-2025 may be better positioned than those who abandoned content investment entirely in response to the initial data.

Who Is Actually Losing Traffic and Why

Publishers and Informational Sites Bore the Heaviest Losses

The traffic losses are real and severe for specific content categories. AdExchanger’s investigation found Business Insider’s organic search traffic fell 55% between April 2022 and April 2025. HuffPost’s desktop and mobile sites lost half their search referrals over the same period. The New York Times saw search’s share of its traffic decline from 44% to 37% in three years. These are structural losses. AI Overviews answer the broad informational queries that news and media content ranked for. Digital Content Next data covering 19 member companies found news publishers saw a 7% median traffic decline in the first half of 2025. Non-news content brands saw 14%.

The Pew Research Center tracked 68,879 actual Google searches by 900 U.S. adults in March 2025. The findings are specific and measurable. Only 8% of users who encountered an AI Overview clicked on a traditional search result. Without an AI Overview, 15% clicked. That is nearly halving the click probability. Additionally, 26% of searches with AI Overviews ended with zero clicks — versus 16% for traditional results pages. Less than 1% of users clicked on links embedded within an AI Overview itself. For high-volume informational keywords, that halved click probability translates to severe traffic losses at scale.

Transactional Queries Remain More Protected

Not all content categories faced the same disruption level. The Semrush study found Real Estate, Shopping, and Arts and Entertainment had AI Overviews appearing on less than 3% of their keywords. Science and Computers and Electronics exceeded 17% coverage. This reflects commercial logic. Google’s business model depends on users completing transactions. AI Overviews summarise information for informational queries. Google has financial incentives to keep transactional queries clicking through to product pages. The distinction between informational and transactional content is now one of the most consequential strategic decisions a content team makes.

Grow and Convert’s client data showed this difference in practice. Clients focused on bottom-of-funnel, commercial-intent content experienced traffic declines of only 10% to 20%. Publishers of broad informational content reported 40% to 50% declines. One Grow and Convert client ranked for “best call center quality assurance software.” Their brand appeared in an AI Overview citing their article. Users who saw the mention searched the brand name directly on Google and converted. The article traffic declined. The pipeline did not. Content positioned close to purchase intent proved far more durable than content serving awareness queries.

Health, Finance, and Education Face Sharpest Penetration

Health, education, and personal finance content experienced the most severe AI Overview penetration. Advanced Web Ranking’s data showed Health had the highest AI Overview frequency of any tracked industry. The first three desktop positions in Health lost a combined 6.55 percentage points in CTR in Q2 2025 alone. The Semrush study confirmed that queries with eight or more words are seven times more likely to trigger an AI Overview. Long-form informational queries — the foundation of content marketing in those sectors — are the most exposed content type. Practitioners in health, finance, and education cannot treat this as a monitoring situation. Active repositioning toward commercial-intent content or authoritative AI citation inclusion is necessary for 2026 planning.

ChatGPT as a Channel: Small Volume, Higher Intent

ChatGPT Traffic Converts at 31% Higher Than Organic Search

Most practitioners assume low ChatGPT traffic volume means low strategic value. Conversion data complicates that assumption. Visibility Labs analysed 12 months of GA4 data from 94 ecommerce brands. All operated at seven or eight figures in annual revenue. Their finding: ChatGPT referral traffic converted at 1.81% versus 1.39% for non-branded organic search. That is a 31% conversion rate advantage. The study covered 9.46 million non-branded organic sessions against 135,000 ChatGPT referral sessions. The conversion advantage held in 10 of the 12 months studied. Revenue per session from ChatGPT was $3.65 versus $3.30 for non-branded organic — a 10.3% advantage per session.

Visibility Labs attributes this advantage to intent compression. A user who queries Google is beginning a research journey. A user who clicks a link in ChatGPT has already conducted a multi-turn conversation. They refined their requirements, compared options, and chose a source to visit. By the time that person reaches your product page, they have self-qualified as a buyer. Average order value was 14.3% lower for ChatGPT visitors. But the higher conversion rate produced a net positive revenue-per-session outcome. This is a channel that is small in volume and disproportionately valuable per visit.

ChatGPT Referral Volume Grew 1,079% Year Over Year

The Visibility Labs data documented that ChatGPT referral sessions grew 1,079% from January to December 2025. Sessions rose from 1,544 in January to 18,202 in December. Non-branded organic search grew 17% over the same period. Digiday’s Similarweb data found ChatGPT referrals grew 52% year over year from September to November 2025. Conductor research across 10 industries found ChatGPT drives 87.4% of all AI referral traffic. Ahrefs’ live tracking showed AI platforms growing three to five times faster than Google across tens of thousands of tracked sites. These growth trajectories cannot diverge indefinitely. The direction of travel is clear enough to justify early investment.

Seer Interactive’s B2B case study produced the most dramatic conversion numbers. ChatGPT traffic converted at 15.9% versus 1.76% for Google Organic. Perplexity converted at 10.5%. ChatGPT visitors viewed 2.3 pages per session versus 1.2 for organic search visitors. That near-double page depth signals engaged research behaviour rather than casual browsing. B2B software buyers use ChatGPT for in-depth vendor comparison before engaging any vendor directly. By the time they click through, they have narrowed their options considerably. This makes ChatGPT a mid-to-lower funnel channel — small in volume, high in purchase proximity.

ChatGPT Users Click Out; Google Users Increasingly Do Not

Most practitioners assume Google sends more referral traffic because Google is bigger. That is true in absolute terms. But per-visit behaviour differs in a way that matters strategically. Momentic’s 13-month clickstream analysis found that the average ChatGPT user clicks 1.4 external links per visit. Google users click 0.6 external links per visit. ChatGPT users are more than twice as likely per visit to leave the platform. Google’s pages-per-visit climbed to 10.1 in April 2025 — reflecting more internal queries and less outbound clicking. Google is increasingly the walled garden. ChatGPT is, proportionally, the leakier platform.

This finding directly challenges the dominant narrative. ChatGPT sends a smaller audience to your site. But that smaller audience arrives with more intent and clicks out at a higher rate per individual visit. The practitioner who assumes ChatGPT steals traffic is misreading the mechanism. ChatGPT users click through to sources. Google users increasingly answer their query on the search results page and stop. That distinction shapes every content decision worth making in 2026.

The Dark Funnel AI Creates in Your Analytics

Most AI-Influenced Purchases Never Show as AI Referrals

Here is the problem that makes ChatGPT’s impact even harder to measure accurately. Most AI-influenced purchases never appear as AI referrals in your analytics. Visibility Labs documented this in their ecommerce study. A user asks ChatGPT which product meets their requirements. ChatGPT recommends a specific brand. The user then opens a browser tab, searches the brand name on Google, and completes the purchase. GA4 records this as branded organic search. The ChatGPT conversation that initiated the decision leaves no footprint.

This is not a marginal edge case. Forrester Research found that nearly 90% of B2B buyers use generative AI during the purchase journey, per their 2025 buyer behaviour research. Gartner’s longitudinal data shows 83% of the buyer’s journey occurs before any vendor contact. That pre-contact research increasingly happens in ChatGPT, Claude, and Perplexity. None of those platforms pass standard UTM parameters or referral identifiers consistently. When teams see branded search rising while non-branded organic falls, they often conclude their SEO is underperforming. The reality may be the opposite. Their content is earning AI visibility that drives brand searches. The attribution framework is misreporting the outcome.

Branded Search Lift Is Your Measurable Dark-Funnel Proxy

When ChatGPT recommends your brand, users frequently follow up with a branded Google search to verify or navigate directly. Research on AI visibility tracking found a measurable signal. Monitoring branded search impressions in Search Console after major content updates reveals 15% to 30% branded search lifts. These lifts typically appear within 7 to 14 days of significant AI mention increases. This is a measurable proxy for AI-driven brand awareness. You do not need ChatGPT to pass referral tags to detect this signal. If you optimise a page for AI citation and see no ChatGPT referral change, look at branded Search Console impressions instead. A clear rise there confirms the dark funnel is working. That signal is measurable without any standard referral tag.

Search Engine Land’s dark funnel analysis framed the strategic shift. Teams that adapted best in 2025 stopped asking “what got the click?” They started asking “what shaped the decision?” That reframing changes what you measure. It also changes what counts as a successful outcome. Traffic is a downstream metric. Citation frequency, branded search volume, and AI share of voice are upstream signals. Post-purchase surveys asking how customers first discovered the brand remain the most reliable diagnostic. They bypass the attribution model and capture self-reported discovery channels that analytics platforms miss entirely.

ChatGPT and Google Are Used in Sequence, Not Competition

Similarweb’s analysis found that 95% of ChatGPT users still also use Google. Only 23 million users worldwide are exclusive to ChatGPT. Users are not abandoning Google for ChatGPT. They use ChatGPT for research and evaluation, then complete their journey on Google. This sequential usage explains the dark funnel attribution problem completely. The tools operate in sequence, not competition. Optimising for one at the expense of the other is a strategic error that the data does not support. An SEO consultancy like Metrics Rule finds consistently that businesses misreading this dynamic make the largest unnecessary strategy pivots in 2025 and 2026.

A Framework for Responding Without Overreacting

Diagnose the Cause Before Redirecting Any Budget

Most businesses facing traffic decline need to answer a specific diagnostic question first. Is your loss caused by AI Overviews stealing clicks from rankings you hold? Or by actually losing rankings? Or by ChatGPT-driven zero-click behaviour? Each cause requires a different response. AI Overview interference shows as falling CTR with stable impressions. Ranking losses show as falling impressions and clicks together. ChatGPT displacement — the rarest cause — would show as stable Google performance alongside rising branded search and direct traffic simultaneously.

Run these three checks in Google Search Console before attributing any decline to a single source. If AI Overviews are the primary cause, the highest-leverage response is repositioning content toward commercial-intent queries. AI Overviews appear on less than 3% of transactional keywords but more than 17% of informational ones. Structure your existing content to earn AI Overview citation as a second priority. Ahrefs found that 76.1% of URLs cited in Google AI Overviews also rank in the top 10 of standard Google results. You cannot abandon traditional SEO and expect AI visibility to compensate. Lily Ray, VP of SEO Strategy at Amsive, stated this directly in the SEOFOMO 2025 search survey. The best path to AI visibility, she argued, is solid SEO combined with strong brand awareness. They are prerequisites, not alternatives.

Answer-First Content Serves Both Channels Simultaneously

Kellogg School research on AI’s web traffic impact documented a concrete example. One company used AI-powered optimisation guidance to increase appearances in Google’s AI Overviews by 61%. Blog impressions and clicks both improved despite the industry-wide decline. The tactics that increase AI Overview inclusion are the same tactics that improve traditional rankings. Comprehensive topic coverage, answer-first structure, structured data markup, and authoritative sourcing all serve both systems simultaneously. Answer-first platforms like Reddit and Wikipedia win disproportionate ChatGPT citation share because they address user queries directly. Content that answers the question first and positions the brand second is what both Google AI Overviews and ChatGPT’s retrieval systems favour. Conversion-first content — “schedule a demo” — gets passed over by both.

For businesses where AI traffic already exceeds 1% to 2% of organic volume, deliberate GEO investment is justified now. This is especially true in legal, healthcare, higher education, and hospitality. First Page Sage’s study across 160 client companies found conversion rate lifts of 1.7 to 3.4 percentage points from AI-referred visitors in those sectors. For businesses where AI referrals remain below 0.5% of organic volume, traditional SEO delivers the higher return on investment. Treat AI optimisation as a secondary track rather than a budget reallocation away from organic search foundations.

Set Up AI Traffic Tracking in GA4 Before Anything Else

The first tactical step is creating a custom channel group in GA4. Use a regex pattern matching source names including ChatGPT, Perplexity, Gemini, and Claude. Without this segmentation, you cannot distinguish AI referral sessions from direct traffic or standard organic. You are managing a problem you cannot see. Once the segment exists, monitor it weekly. Track volume, conversion rate, pages per session, and engagement rate. Compare those metrics against non-branded organic as a baseline. If AI traffic converts above your organic baseline, that signals an investment case for content that earns AI citation.

A data-first SEO consultancy like Metrics Rule can audit your GA4 configuration and search visibility at the same time. That means identifying which content earns AI Overview and ChatGPT citation. It also surfaces which terms are losing clicks to AI summaries and where branded search lift signals dark-funnel influence. The analytics infrastructure question is not separate from the SEO question. In a multi-platform search environment, they are the same problem. Traffic you cannot measure correctly is traffic you cannot recover or grow with any confidence.