Your Traffic Is Down 60% — Here’s Where It Went

Your traffic didn’t disappear. It moved.

If your organic traffic has been quietly declining for the past two years, you’re not imagining it — and it’s not your fault. A new Chartbeat dataset published by Axios on March 17, 2026 tells the story clearly:

The drop is real, it’s accelerating, and it’s structural — not a Google algorithm penalty you can recover from by fixing your H1 tags.

What the Chartbeat data actually shows

The Chartbeat numbers deserve context because they reveal a pattern that most site owners misdiagnose. The traffic decline does not correlate with Google algorithm updates — sites that lost 60% of traffic did not receive manual penalties, and their ranking positions for target keywords remained largely stable. What changed was not where they ranked, but whether anyone clicked.

The decline correlates precisely with the rollout timeline of AI search features:

• Q1 2024: Google begins widespread AI Overviews testing. Search traffic decline accelerates from −15% annually to −25% annually for small publishers.
• Q3 2024: ChatGPT Search launches publicly with OAI-SearchBot. Informational queries — the bread and butter of content marketing — are now answered directly in ChatGPT without a click.
• Q1 2025: Perplexity crosses 100 million monthly queries. Google AI Overviews expand to 80% of informational queries in the US.
• Q1 2026: Matomo confirms AI bots account for up to 50% of web traffic. The traffic is there — it is just not human.

The implication is clear: your traffic loss is not a ranking problem. It is a format problem. Your content still ranks. People just are not clicking because the AI already answered their question.

Where did the traffic go?

It went into AI answers.

Google AI Overviews, ChatGPT, Perplexity, and Gemini now answer the same informational queries your blog posts used to capture. A user asks “how do I reduce churn in my SaaS” — they get a four-paragraph AI answer. They don’t click your article. You ranked #1. You got zero traffic.

Meanwhile, Matomo’s 5.8 release (March 18, 2026) found something that will reframe how you think about your analytics dashboard:

That’s not future-state speculation. AI crawlers from OpenAI, Anthropic, Google, Perplexity, and others are scanning your site right now. Your Google Analytics doesn’t count most of them. Your traffic is down on paper — but your site has never been busier.

The AI crawler traffic breakdown

Matomo’s data identifies the five companies responsible for 84.5% of all AI crawler traffic:

1. OpenAI (GPTBot + OAI-SearchBot) — The most active AI crawler. GPTBot collects training data. OAI-SearchBot powers ChatGPT Search real-time results.
2. Google (Google-Extended + Googlebot-AI) — Google-Extended feeds AI Overviews and Gemini training. Separate from the traditional Googlebot used for search indexing.
3. Anthropic (ClaudeBot + anthropic-ai) — Feeds Claude’s knowledge base and real-time citation capabilities.
4. Microsoft (Bingbot-AI) — Powers Microsoft Copilot across Windows, Office, and Edge browser.
5. Perplexity (PerplexityBot) — The most transparent about its crawling behavior. Crawls in real-time at query time.

The 4× growth rate (2.6% to 10.1% of all web traffic in 8 months) suggests AI crawler traffic will surpass traditional bot traffic within 12 months. Your site is being evaluated for citation-worthiness more frequently than it is being crawled for traditional search indexing.

What this means for your analytics: Your Google Analytics dashboard is showing you an increasingly incomplete picture of your site’s actual traffic. GA4 filters out most bot traffic, including AI crawlers. So your dashboard shows declining human visits while your server logs show surging AI crawler visits. The AI crawlers are reading your content and deciding — in real time — whether to cite you when a user asks a relevant question.

The implication is counterintuitive

You have more visibility into your site than ever before — just not from humans. AI engines are reading your content, deciding whether it’s citable, and either including it in their answers or skipping it.

If they skip you: a competitor who has structured their content for AI extraction gets cited instead. Your traffic drops. Theirs grows. Neither of you changed your SEO.

If they cite you: you start getting referrals from AI answers. You appear in ChatGPT recommendations, Perplexity sources, and Google AI Overviews. That traffic is growing — eMarketer projects AI search referrals will account for 12–18% of web traffic by end of 2026.

The compounding advantage of early AI citation

The economics of AI citation work differently than traditional search rankings. In traditional SEO, ranking positions are relatively fluid — you can overtake a competitor with better content at any time. AI citation builds compounding advantages that are harder to displace:

1. Training data persistence. Once your content is included in an LLM’s training data (Claude, Gemini base models), it becomes part of that model’s knowledge. A competitor cannot retroactively remove your presence from training data.
2. Citation feedback loops. AI engines track which sources they cite and how users respond. Sources that are cited and receive positive engagement signals get cited more frequently in subsequent responses.
3. Brand entity recognition. The more frequently AI engines cite a brand for a specific topic, the stronger the entity association becomes. Over time, the AI develops a default preference — similar to how Google develops PageRank authority preferences.
4. First-mover indexing advantage. For platforms like Perplexity that crawl in real-time, the first comprehensive, well-structured page on a topic often becomes the default citation.

This means the cost of delay is not just “missing traffic today.” It is “allowing competitors to build citation moats that become exponentially harder to breach.”

What decides whether AI cites you?

Three factors matter most:

This is Generative Engine Optimization (GEO) — and it’s the discipline that determines whether the traffic you’ve lost to AI ends up at your competitors or comes back to you as AI citations.

The numbers behind AI search growth

Understanding the scale of the shift helps contextualize why this matters now, not later.

AI search adoption

• ChatGPT: Over 200 million weekly active users as of early 2026, with ChatGPT Search handling an estimated 2+ billion queries per month
• Perplexity: Over 100 million monthly queries, growing at approximately 40% quarter-over-quarter
• Google AI Overviews: Appearing on an estimated 80% of informational queries in the US
• Microsoft Copilot: Integrated into Windows 11 (1.4 billion devices), Microsoft 365 (400 million paid seats), and Edge browser (680 million users)
• Apple Intelligence: Shipping on all new iPhones, iPads, and Macs as of iOS 18 (over 1 billion active Apple devices)

Traffic impact projections

• Gartner: Traditional search engine volume will drop 50% by 2028
• eMarketer: AI search referrals will account for 12–18% of total web traffic by end of 2026
• Chartbeat/Axios (March 2026): Small publishers already down 60%, mid-sized down 47%, large down 22%
• Matomo 5.8 (March 2026): AI bot traffic grew 4× in 8 months, now representing up to 50% of all website visitors

The trajectory is clear: traditional search traffic is declining at an accelerating rate while AI search traffic is growing exponentially. The crossover point — where AI-referred traffic exceeds traditional search traffic for a given site — is approaching faster than most marketers anticipate.

What to do now

Industry-specific impact

The traffic decline is not distributed equally. Some industries are feeling the impact faster than others.

SaaS and B2B technology

Informational queries like “best project management tools” and “how to reduce SaaS churn” are now answered almost entirely by AI. Product comparison queries — historically the highest-converting organic traffic — are being replaced by AI-generated comparison tables. SaaS companies that rely on content marketing for lead generation are seeing the sharpest traffic declines for informational content.

What to do: Optimize product pages and comparison content with Product schema. Add FAQ sections addressing common buyer questions. Ensure your product is mentioned in AI-generated recommendations by building citation-optimized content around your product category.

E-commerce and DTC

ChatGPT’s shopping features are expanding rapidly, with product recommendations appearing for purchase-intent queries. Google AI Overviews show product cards for commercial queries. Shopify stores with proper Product schema are already appearing in these AI shopping results. Stores without schema markup are invisible to AI shopping features.

What to do: Implement Product schema on all product pages. Add FAQ content to category pages. Ensure your Shopify theme renders content server-side (not client-side JavaScript) so AI crawlers can read it.

Professional services

“Best [service] in [city]” queries are now answered by AI with curated lists. Firms without LocalBusiness schema, author credentials, and FAQ content are being excluded from these recommendations.

What to do: Add LocalBusiness schema with complete NAP (name, address, phone) data. Add FAQ sections addressing “how to choose” and “what to expect” questions. Build author credential pages with Person schema.

Content publishers and media

This is the hardest-hit category, and the Chartbeat data reflects it directly. Content publishers monetize through pageviews — when AI answers the question without a click, the publisher’s business model breaks. The 60% traffic decline for small publishers is existential.

What to do: Shift from “ranking for keywords” to “being cited as a source.” Build original research and proprietary data that AI engines must cite. Develop a direct audience through email and social to reduce dependence on search referral traffic.

The bottom line

Your traffic isn’t gone. It moved into AI answers. The question is: when someone asks ChatGPT about your topic, do they get your answer — or your competitor’s?

The Chartbeat data shows the cost of not answering that question. The Matomo data shows the window is open right now: AI is already reading your site. Give it a reason to cite you.

The companies that act on this data in 2026 will build AI citation advantages that compound over time — just as the companies that invested in SEO in 2010 built organic search advantages that still pay dividends today. The window is open. The data is clear. The question is whether you move now or wait until the competitors who did move are entrenched.

Start with a free GEO score to see where you stand. Or explore the full GEO service if you want this done for you.