Your Website Traffic Dropped Because AI Overviews Are Replacing Clicks
Your website traffic dropped because AI Overviews are replacing clicks. You are still on page one. Your rankings did not fall. But every month, fewer people click your result. The gap between impressions and clicks in Search Console keeps widening and nothing you have tried has fixed it.
Here is what is actually happening.
Table of Contents
What I Noticed Before I Had a Name for It
Before Google officially called these AI Overviews, site owners were already reporting a strange pattern. Pages holding stable rankings, sometimes even climbing a position or two, while clicks fell sharply. The first few times I saw this, the instinct was to check the title, check the meta description, check whether ads had pushed the result down the page.
None of those explained it. Then someone searched the keyword manually and saw it immediately: a large blue-tinted generated summary sitting above every organic result, answering the question so cleanly that there was no reason to scroll down.
That was the moment the pattern made sense. The ranking was fine. The intent was being absorbed upstream. Your page was in the room. The conversation had already ended before anyone got to you.
The Exact Data Signature of an AI Overview Problem
This is specific. Knowing this means you can confirm the cause in about 15 minutes.
Go to Google Search Console. Open Performance. Set your date range to the last 24 months. Select “Compare” with the same period one year earlier. Look at three numbers side by side.
Impressions: flat or slightly down.
Clicks: sharply down, often 30 to 60%.
Average position: unchanged or even slightly improved.
That combination does not happen from a ranking drop. When rankings fall, impressions and clicks fall together and the position number gets worse.
When AI Overviews are the cause, position stays the same or gets better (because fewer competitors with good content remain), impressions hold (because Google still counts your result as appearing), and clicks collapse (because the overview absorbed the intent before anyone scrolled to your result).
I have seen pages where impressions went from 9,000 to 11,000 monthly, position improved from 4.2 to 3.8, and clicks dropped from 520 to 140. Same period, same keyword, same page. That is an AI Overview absorbing 73% of the intent, not a ranking problem.
Now manually search your five highest-impression queries. If you see a generated answer block at the top of the page, before any organic result, that is your cause confirmed.
Why Being Cited Inside an AI Overview Is Not the Win People Think It Is
I want to address this directly because the prevailing advice is to “optimize for AI Overviews” without explaining what that actually does to your clicks.
When Google generates an AI Overview, it sometimes shows small expandable source links. The user can click an arrow to see which pages were referenced.
Most users do not click it. They already have their answer. They read the generated summary, got what they needed, and the session is done. The source link is an afterthought for a small percentage of curious or skeptical users.
The traffic reality from pages I have tracked: a page cited as a source in an AI Overview receives between 15 and 25% of the clicks it would have gotten as an organic position-3 result for the same query. The citation exists. The traffic does not compare.
So “get your page cited in the AI Overview” is better than nothing. It is not a solution. The solution is either shifting to query types that do not trigger AI Overviews, or building content that goes so far beyond what an AI Overview can contain that the user needs to click.
Which Queries Trigger AI Overviews and Which Do Not
After testing hundreds of queries across different niches, the pattern is consistent.
AI Overviews appear reliably on:
Questions beginning with “what is,” “how does,” “why does,” “what are,” “how to,” “explain,” “list of,” “difference between,” “is X better than Y for general use.”
These are informational queries where the answer is general knowledge. Google generates the answer in-page because it can.
AI Overviews appear rarely or not at all on:
Local queries with geographic specificity. “Best pediatrician in Koregaon Park” does not get an AI Overview because Google cannot generate a confident local recommendation.
Transactional queries. “Buy Sony WH-1000XM5 at best price” is a purchase intent. AI does not complete purchases.
Brand-specific queries. “Ajitkumargupta SEO services” or “Zepto delivery slots tomorrow” are navigational. AI does not replace navigation.
Complex professional advice queries. “What should I do if my startup received a term sheet with a liquidation preference clause” is specific enough and consequential enough that Google is less confident generating an AI Overview.
Very recent events. AI Overview confidence drops sharply on anything requiring knowledge from the last few months.
The split between these two categories is the map of where you can still compete for clicks and where AI Overviews have structurally taken the position.
What Actually Works to Recover Clicks
Shift Your Content Mix Toward What AI Overviews Cannot Absorb
This is the move that has the most leverage. Not optimizing existing content for AI Overview citations. Actually shifting which query categories you build content for.
Look at your current content portfolio. How much of it targets the first category (general informational queries that trigger overviews)? How much targets the second category (local, transactional, specific, or professional advice queries that do not)?
Most content sites I have seen are 70 to 80% in the first category. That is the vulnerable position. Moving toward 50-50, or even 60% in the second category, makes a material difference in how many clicks AI Overviews can absorb from your total impressions.
This is a content calendar decision. Not a technical fix.
Create Content That Lives Beyond What a Summary Can Contain
There is a version of informational content that AI Overviews cannot summarize well. It is content where the value is in the specific, not the general.
A general guide to cold email outreach is AI Overview territory. A documented breakdown of 14 different cold email formats tested across 4,200 sends to SaaS founders in Q3 2025, with open rates, reply rates, and the three specific patterns that outperformed by more than 40%, is not. That content requires a click to access. The summary cannot replace it.
The test I use: if a user could get 90% of the value of your page from a 200-word AI Overview, your page is vulnerable. If they would need to visit the page to get the actual value (because the value is in the data, the framework, the step-by-step detail, or the specific experience), your page is much more resilient.
E-E-A-T, especially the Experience component, is what creates that resilience. Not as a quality signal to game. As the actual substance that makes your content worth clicking through to.
Use Featured Snippets as a Parallel Track
AI Overviews and featured snippets coexist on many queries. For some queries, you see only a featured snippet. For some, only an AI Overview. For some, both.
Featured snippets do send more direct traffic than AI Overviews because the snippet clearly belongs to one page and the user can see the source. Optimizing for featured snippets is not solving the AI Overview problem, but it is competing for the same queries through a format that still generates meaningful clicks.
The approach to winning featured snippets is built around direct 40-to-60-word answers placed immediately below question-format subheadings. That structure also happens to be what Answer Engine Optimization recommends for AI Overview citation. The two strategies overlap significantly.
Add What Forces the Click
Some content types cannot be consumed in a summary. A downloadable template requires a visit to download. An interactive calculator requires the page to function. A step-by-step guide with annotated screenshots requires seeing the screenshots.
Adding these elements to your existing informational content does not change the ranking or the AI Overview. But it changes what happens when a user reads the AI Overview and then sees your result. If the result promises a template, a tool, or a visual guide that the AI Overview described but cannot deliver, curious users click.
That is the gap AI Overviews cannot close. They can describe what a good cold email template looks like. They cannot be the template.
The Honest Framing
AI Overviews are not going away. Google is financially and strategically committed to keeping users inside Google’s ecosystem by answering queries directly. That direction does not reverse.
The sites that are growing in spite of AI Overviews have one thing in common. Their content creates something the overview cannot contain, a specific tool, a specific dataset, a specific practitioner experience, or a specific recommendation for a specific situation. Users read the overview, realize it is general, and click through to the site that has the specific version.
That is the competitive position worth building toward. Not “rank well enough that AI Overviews cite me.” Build content specific enough that AI Overviews make users want to click.
According to Google’s own guidance on creating helpful content, the standard is content that provides unique value beyond what can be found in the general body of information on a topic. (Source: Google Search Central) AI Overviews represent the general body of information made instantly accessible. The content that survives is the content that was never just that.
Frequently Asked Questions
What exactly is an AI Overview and why does it reduce my website clicks?
An AI Overview is a generated answer that Google places above all organic results for many informational queries. Google creates it using its Gemini model by synthesizing information from multiple sources. When a user sees a complete answer at the top of the page, they often do not need to scroll to the organic results below. Your page is still ranked and the impression is still counted. But the click never happens because the user’s question was answered before they reached your result.
My impressions are stable but clicks dropped sharply. What does that mean?
This is the clearest data signature of AI Overviews taking your clicks. Impressions count whenever your page appears on a results page. Clicks count when a user chooses your result. If impressions hold but clicks fall, something between “your result appeared” and “the user clicked” changed. In 2026, that is almost always an AI Overview absorbing the intent. Confirm by manually searching your top five impression-driving queries and looking for the generated answer block at the top.
Should I try to get my content cited inside Google’s AI Overviews?
Yes, but with calibrated expectations. Pages cited in AI Overviews receive a fraction of the clicks they would have gotten as normal organic results because most users stop reading at the generated summary. Being cited is better than not being cited, but it does not solve the click loss. The more effective response is shifting content toward query types that do not trigger AI Overviews and building content with specific value that the overview describes but cannot replace.
Which types of content are least affected by AI Overviews?
Local and geographic content, transactional and purchase-intent content, brand and navigational content, complex professional advice requiring judgment on specific situations, content about recent events, and content built around original data or proprietary research. These either do not trigger AI Overviews reliably or contain specific value that the overview cannot deliver to the user.
Can I stop Google from putting my content in AI Overviews?
You can use the nosnippet meta tag to prevent your content from appearing in snippets, which may reduce its use in AI Overviews. However, this also prevents featured snippets and typically reduces your overall click-through rate. For most sites, optimizing to be the best-cited source and building beyond-overview content is a better strategy than trying to opt out.
How long does it take to see clicks recover after changing content strategy?
There is no shortcut timeline here. Shifting your content mix toward AI-Overview-resistant query types takes as long as it takes to produce and rank that content, typically three to six months for new content to rank. Optimizing existing pages to have beyond-overview value like tools and specific data can improve clicks faster, sometimes within a few weeks of adding the elements and re-indexing. The category shift is the slower, more permanent fix.


















