What Google AI Overviews Actually Mean for Your Organic Traffic
Google AI Overviews do not kill organic traffic equally across every page. Some query types see click rates fall sharply. Others are barely touched. The difference comes down to what kind of answer the user actually needs, and whether your content is structured in a way the system can read and cite. Getting that distinction wrong means optimising for a problem you do not have, while ignoring the one you do.
What AI Overviews Pull From and What They Leave Out
AI Overviews draw on pages that demonstrate clear topical authority, carry well-structured content, and give direct, quotable answers to specific questions. Thin pages, pages with duplicate meta descriptions, and pages that bury the core point under filler paragraphs tend to get skipped. The system appears to favour content that a human editor would also reach for first.
What gets left out is often predictable. Pages set to noindex by accident, pages with no internal links pointing to them, pages with headings that contradict the search title. These are not abstract SEO warnings; they are concrete exclusion signals. A page that Google's crawler cannot confidently interpret is a page an AI Overview cannot confidently cite.
The selection logic is not purely about domain authority. Smaller, focused sites with structured, specific content regularly appear in overviews alongside much larger publishers. The signal is relevance and readability, not just reach.
Where Traffic Drops and Where It Holds
Simple factual queries and short how-to questions are where zero-click behaviour is most pronounced. If someone asks what the boiling point of water is, or how many centimetres are in a foot, the Overview answers it completely. No click needed. Traffic to those pages was never high-intent anyway, but if your content strategy leaned on informational micro-queries to generate volume, that volume is now softer.
Commercial, local, and high-consideration queries tell a different story. When someone is comparing suppliers, looking for a local service, or weighing up a significant purchase, the Overview often prompts more research rather than ending it. CTR on these queries has held up more reliably because the user still needs to evaluate, contact, or book. That is a click no overview replaces.
How query type shapes click-through exposure after an AI Overview appears
Schema and Structured Data as an AI Signal
Structured data is not a ranking factor in the traditional sense, but it is a readability factor. When a page emits clean, consistent schema, AI systems can extract meaning without having to infer it. That distinction matters more now than it did when the only audience was a human reader clicking through search results.
What works is a unified schema graph per page, built from a single source of truth rather than competing markup blocks added by different plugins at different times. Overlapping or contradictory schema confuses both crawlers and answer engines. A page with three conflicting LocalBusiness blocks, each showing a slightly different phone number, is a page a system cannot cite with any confidence. Our AI search optimisation approach builds schema this way by default, drawing every property from one central record so the markup stays consistent across every page automatically.
Rank Tracking When Position One Only Tells Half the Story
Why position data is no longer enough
A site ranking first for a query is not the same thing as capturing the traffic that query generates. Once an AI Overview sits above the organic results, position one behaves more like position four used to. The impression is there. The click often is not.
That means rank tracking needs to expand beyond position data. The signals worth monitoring now include impression share by query type, CTR broken down by query category, and whether a page is being cited within overviews at all. A site that tracks position only is missing the layer that explains the actual traffic number sitting in Search Console.
Building a monitoring setup that accounts for it
What rank tracking metrics actually need to show you covers this in more detail if you want to build a monitoring setup that accounts for the shift.
Practical Steps for WordPress Site Owners
Start with thin and low-authority content
Pages with fewer than 300 words, no internal links pointing to them, and no clear topical focus are the first to be excluded from AI Overview citations. They are also the easiest to fix or consolidate, so this is where an audit should begin.
Tighten your structured data
Check that your schema is consistent and draws from one source rather than several competing plugins. The properties Google uses most, business name, address, opening hours, service type, need to be populated accurately. Contradictory data across plugin-generated blocks is one of the more common issues we see on sites that should otherwise be well-placed to appear in overviews.
Build topically, not just prolifically
A cluster of pages covering one subject from several angles gives a site far more surface area to be cited than a collection of isolated articles with nothing connecting them. That depth is what Google's own quality rater guidelines describe as demonstrating expertise. Publishing more pages without any structural relationship between them rarely moves the needle here.
Fix the site-health issues that exclude pages from consideration
Broken internal links, accidental noindex tags, missing H1s, images without alt text, pages buried more than three clicks from the home page. These are all concrete exclusion risks, and addressing them is not glamorous work, but it is where the actual gains tend to come from.