There's a conversation happening in marketing teams across the US right now that goes something like this: traffic is softer than it should be, but nothing in the dashboard explains why. Rankings haven't dropped. The site loads fast. Content is current. Spend is steady. And yet, month over month, organic sessions quietly decline, and no one can point to the cause.
Most teams respond by doubling down on what they know — more content, better keywords, another technical audit. That's the wrong diagnosis for the right symptom.
The problem isn't where you rank. It's what happens to most Google searches before anyone decides to click at all.
724 Out of Every 1,000 Google Searches Never Leave Google
A June 2026 study by SparkToro and SimilarWeb — drawing on billions of US search sessions — found that 68.01% of Google searches now end without any click to an external website. Factor in searches that stay inside Google's own properties, and only 27.6% of all searches result in a visit to a site outside Google. (Source: SparkToro, June 10, 2026)
For every 1,000 people searching on Google right now, 724 of them never leave.
That is not a traffic anomaly or a bad month. It is the structural reality of how Google operates in 2026, and it has been moving in this direction for years. The businesses most affected are those optimizing hard for the 276 who click, while giving no thought to what's happening with the 724 who don't.
Where Those 724 People Actually Go
They're not disappearing. Google is answering their questions before the click is needed.
AI Overviews now appear for a growing share of commercial and informational queries. When they do, the search is resolved on the results page itself — the user gets their answer without opening a single external link. Google Maps captures local intent: "IT services in Philadelphia," "managed cloud provider near me" — these queries end inside Google's own interface, directing users to business listings rather than websites.
YouTube handles video-intent searches. Google Shopping intercepts product queries with a visual carousel. Knowledge Panels answer brand searches. People Also Ask captures follow-up questions before they become separate searches. Each of these surfaces resolves intent without sending traffic anywhere.
None of it shows up in your analytics as a session. All of it shapes what someone knows about your business before they ever visit your website.
Why Ranking Higher Won't Fix This
This is where the standard response fails. The instinct to climb from position five to position two is understandable. But if the query triggers an AI Overview that answers the question completely, position two below the fold gets a fraction of the clicks it once did. You can rank higher and still lose ground.
The Google May 2026 Core Update — which completed June 2 and continued producing ranking volatility through mid-June, per ongoing reporting by Search Engine Roundtable — has accelerated this pattern. Well-structured content with clear entity signals is gaining. Individual product and brand pages that haven't adapted their content architecture are losing ground regardless of their historical ranking position.
The Brands Appearing in AI Overviews Aren't Necessarily Outranking You
Many of them aren't. What they've done is structure their content the way Google needs it to be — specific, factual, attributed, and machine-readable — so it can be extracted and presented with confidence inside an AI-generated answer.
That's a different optimization target than keyword ranking. And most SEO strategies were built for the old target.
The question that actually matters now is not "where do I rank for this keyword?" It's "does Google know enough about my business to represent it when a buyer asks a relevant question?" Ranking and representation are two different outcomes, and in 2026, they require different inputs.
What B2B Buyers Are Doing While You're Counting Clicks
This shift in search behavior isn't happening in isolation. It aligns with something B2B marketers are already feeling in their pipelines.
LinkedIn's 2026 Global B2B Marketing Outlook — a survey of 1,299 B2B marketers and CMOs across the US, UK, and Europe, published June 10 — found that 9 in 10 CMOs say buyers now need to know and trust a brand before they're willing to engage with it. 74% said traditional marketing is losing effectiveness with the buyers they're trying to reach. (Source: Marketing Week, June 10, 2026)
That's not a coincidence. As Google keeps more search sessions inside its own surfaces, the content a buyer encounters during their research phase — AI Overviews, Knowledge Panels, reviews, entity data — is increasingly Google-mediated rather than website-mediated. You don't always get to make your case in your own words. Google makes it for you, from whatever signals it has.
The businesses that show up well in that mediated layer are the ones that have given Google reliable, structured, entity-rich signals to work with. The ones who don't appear are not just losing clicks — they're losing consideration at the moment it forms.
What the Strategy Looks Like Now
None of this argues for abandoning SEO. Traditional foundations — technical health, domain authority, crawlability, keyword relevance — are still the floor. Without them, nothing else works. But they are no longer sufficient to capture the full range of surfaces where buyers now make decisions.
Entity Clarity: Does Google Know Who You Are?
The starting point is entity recognition. An entity, in Google's terms, is a named, structured thing it can reference with confidence — a business with a known location, category, service area, and online footprint. This is distinct from keyword matching. Google cannot represent you well in AI-generated answers if it doesn't have a clear, consistent entity record to draw from.
In practice: Is your Google Business Profile accurate and fully populated? Does your website carry Organization schema markup? Are you listed consistently on the third-party platforms Google treats as authoritative — Clutch, G2, LinkedIn, and industry directories? These signals build the entity record that makes AI-mediated representation possible.
Content Structure: Writing for Extraction, Not Just Ranking
AI Overviews pull from pages that answer questions directly, in plain, structured text. Marketing copy — "we deliver end-to-end solutions for modern enterprises" — is invisible to AI extraction. Factual, specific, question-responsive content — "our cloud migration service covers infrastructure assessment, phased migration planning, and post-migration optimization with defined SLAs" — is what gets cited.
The audit question that matters is not "does this page rank for my keyword?" It's "can an AI system extract a specific, confident answer to a buyer's question from this page?" Most pages fail the second test. That gap is where visibility is being lost.
| Signal that earns AI Overview Presence | Signal that earns Keyword Ranking |
|---|
| Direct question-answering structure | Keyword density and placement |
| FAQPage, HowTo, Speakable schema | Backlink profile and anchor text |
| Named entity associations | Domain authority score |
| Third-party citation quality | Title tag and meta description |
| Topical depth — pillar + cluster content | Page-level keyword optimization |
The FAQPage schema deserves specific mention. It is one of the most direct mechanisms for earning structured-result presence in AI Overviews and featured snippets, and the majority of business service pages haven't implemented it. It doesn't require rebuilding anything — it requires adding structured markup to pages that already answer questions.
Third-Party Presence: Where Google Goes When Your Website Isn't Enough
AI Overviews don't rely solely on your own website's signals. They pull from the broader web — review platforms, industry publications, forums, and directories that Google treats as trusted references.
A business with a thin presence outside its own website gives Google limited material to work with. Consistent, specific brand mentions on platforms Google retrieves from — with service-level and category-level detail — build the citation base AI surfaces draw from. Gartner's June 2026 consumer research found that 49% of US consumers now say AI-generated content has made overall content quality worse, with that figure rising to 57% among Gen Z and millennials. (Source: Gartner, June 9, 2026) Specific, attributed, experience-based brand mentions stand out in that environment precisely because generic content has proliferated.
Conclusion
The traffic you're not seeing in your analytics isn't going to your competitors. Most of it is staying inside Google, resolved by AI Overviews and Google's own surfaces, before anyone has a reason to click anywhere.
That changes what's actually at stake. The brands getting selected by those surfaces are capturing buyer consideration without needing a click. The brands absent from those surfaces aren't just missing traffic. They're absent from the moment buying intent forms — and that gap doesn't show up clearly in a rankings report until the revenue consequence is already compounding.
Fixing this isn't a rebrand or a platform rebuild. It is a focused shift in what you're optimizing for: entity clarity, structured content that answers questions AI systems can extract, and brand presence on the third-party platforms Google actually trusts. The businesses that make that shift now are the ones that will be represented in AI-mediated search when their buyers are deciding — before the click, before the visit, and increasingly, before any competitor gets the chance.
Intelegencia's Business Findability & SEO team works with US businesses to build search presence that reaches beyond keyword rankings — covering entity optimization, structured content for AI extraction, schema implementation, and topical authority architecture. If your rankings look stable but organic traffic is quietly contracting, talk to us about where the gaps are.
Related reading: Content Strategy & Marketing → | E-commerce Content & Listing Optimization →