Google is secretly changing its search results: Why AI answers are suddenly slipping down – The risky game with search results

New data reveals: This is why Google hides its own AI answers

AI Traffic Shock? What Google's Latest Search Results Mean for Websites

Google is reshaping the internet – and finds itself in a massive strategic dilemma. Current data shows that while so-called "AI Overviews" in search results are rapidly expanding to cover an ever-increasing number of search queries, they are paradoxically losing visibility at the same time. Especially for information-driven searches, AI answers are increasingly being pushed down the list to make room for traditional results and ads. What seems contradictory at first glance turns out to be a careful balancing act: The tech giant must maintain its dominance in the AI ​​sector without cannibalizing its own multi-billion-dollar advertising business or exposing itself to endless copyright lawsuits from publishers. The following article sheds light on these quiet but far-reaching upheavals in Google's search engine, reveals why traditional SEO rules are increasingly being disregarded, and shows how website operators must prepare for the algorithmic volatility of the future.

Google AI Overviews in transition: When market power meets a crisis of confidence

The quiet change in search results is becoming a test for Google's business model

Google is currently transforming one of the most valuable digital assets of all—the layout of its search results page—at a pace and with a consistency that would have been unthinkable just a few years ago. At the heart of this shift are the so-called AI Overviews, AI-generated summaries that, since their widespread introduction, appear directly above the traditional organic search results, providing users with a complete answer without requiring them to click on a single website. Recent analyses from the food sector, conducted using a proprietary SEO tool over the months of April to July 2026, reveal a pattern that extends far beyond a single industry. The coverage of the observed keywords with AI Overviews increased from 4.9 percent in April to 14.0 percent in July, while the average position of the AI ​​block within the results page fell from 1.47 to 2.13 during the same period. In other words, Google now displays the AI ​​summary significantly more often, but simultaneously places it less frequently in the very first position. This contrast between growing reach and declining prominence is no accident, but rather the result of a corporation's strategy, which balances an extremely fine line between technological innovation leadership, regulatory pressure, journalistic conflicts of interest, and the need to keep its advertising business alive.

The figures presented here from the food sector are corroborated by numerous independent studies across various industries and can even be placed in a broader context. Ahrefs data from December 2025 and March 2026, based on the analysis of 863,000 keywords and four million AI overview URLs, show that the share of AI overview citations from the organic top 10 results fell from 76 percent in July 2025 to just 38 percent in March 2026. This decoupling between traditional ranking and AI citation probability fundamentally alters the entire logic of search engine optimization. At the same time, the overall coverage of AI overviews has grown exponentially across industries, from around 13 percent of all search queries in the first quarter of 2026 to estimates of 48 percent in July of the same year, although the methodology and the mix of search queries captured vary considerably between the studies. This discrepancy in the absolute numbers already underlines a key finding of the entire debate: there is no uniform, monolithic AI overview experience, but rather a highly fragmented treatment by Google's algorithms, which varies according to industry, search intent and user device.

Caught between two stools: Why Google is downgrading its own invention

The real key question arising from the observed data is why a company like Google, which has invested heavily in a feature for years and which, according to its own figures, now reaches 2.5 billion monthly active users, is downgrading its prominence instead of further developing it. The most obvious explanation lies in the simple physics of the click economy. When an AI-generated block appears above the traditional results and fully answers the user's question, the likelihood of a click on an organic result drops drastically. The Pew Research study from July 2025, which analyzed nearly 69,000 searches, found that when an AI overview is present, users click on a traditional result in only 8 percent of cases, compared to 15 percent without an AI summary—a relative decrease of 47 percent. Ahrefs reports a decline of up to 58 percent in click-through rates for position one with an available AI overview, while Seer Interactive measures a drop in organic click-through rate from 1.76 percent to 0.61 percent, which corresponds to a reduction of 65 percent, and sees the click-through rate for paid ads even plummet by 68 percent from 19.7 percent to 6.34 percent.

These figures reveal the real dilemma Google faces. The company still earns most of its money from search ads, whose value is directly tied to the visibility and click-through rate of the paid results. An AI block that provides such convincing and comprehensive answers that users don't want to leave the page essentially sabotages its own core business. The observed decline in the average AI overview position from 1.47 to 2.13 in the food segment can be interpreted precisely in this light: Google is apparently systematically testing how far it can push the AI ​​response down without completely devaluing the function that users now expect, while simultaneously creating more space for classic organic results and ads at the top of the screen. Studies on grocery and food keywords confirm a particularly high stability of AI overview presence in this segment, with a retention rate of 56 percent for recipe and ingredient queries, while other product categories such as furniture are treated far more volatilely. This suggests that for information queries with low purchase intent, such as recipe searches, Google is more willing to keep the AI ​​answer prominent, because little advertising revenue would be expected here anyway, while for more commercially valuable search queries, a more cautious weighting is applied.

Legal headwinds as an underestimated driver of realignment

Besides the pure click economy, a second, often underestimated factor plays a crucial role in the observed repositioning of AI Overviews: the growing legal and regulatory risk that Google faces simultaneously in several jurisdictions. Publishers, news organizations, and content creators worldwide accuse the company of extracting copyrighted content without adequate compensation and presenting it in a way that systematically cuts off the original traffic to the source. This so-called zero-click problem, in which, according to several studies, between 60 and 80 percent of search queries using AI Overview end without any click on an external page, has already led to antitrust complaints and regulatory attention in several countries. A company that simultaneously acts as the dominant search engine operator, the largest advertising marketer, and now also as the largest automated content aggregator, is caught in a structural conflict of interest that can hardly be completely resolved in the long term, but can at best be mitigated by constantly readjusting the visibility weighting.

Against this backdrop, the assumption expressed in the initial data—that Google is attempting to better balance AI Overviews, organic results, advertising interests, and legal risks—appears not only plausible but almost inevitable. A company of this size doesn't operate from a single motive but must simultaneously manage a complex web of economic, legal, and reputational objectives. The observation that a classic organic result is appearing more frequently before the AI ​​Overview, as also noted by other industry observers, can be interpreted as a cautious retreat from a potentially overly aggressive initial rollout. It is significant that a study published in June 2025 by the platform seoClarity found that AI Overviews no longer occupied the absolute top position in almost all US desktop search results, but ranked below position one in 12.4 percent of cases—a significant increase compared to less than 2 percent just one month earlier. This trend is clearly continuing and even intensifying, as the current food keyword data shows.

The quasi-in-house solution: How Xpert.Digital closes operational gaps in B2B marketing and sales – Smart Content-Driven Business - Image: Xpert.Digital

Xpert.Digital is a data-driven B2B industry hub led by Konrad Wolfenstein . The company acts as an external, quasi-in-house solution for industrial partners, closing operational gaps in marketing, content, and sales – without requiring additional resources on the client side.

More information here:

• The quasi-in-house solution: How Xpert.Digital closes operational gaps in B2B marketing and sales – Smart Content-Driven Business

The citation lottery: When a good ranking is no longer enough

For companies that rely on organic search traffic, this development creates a paradoxical situation that seems contradictory at first glance, but reveals a coherent new logic upon closer inspection. On the one hand, the average position of AI Overviews is decreasing, which should theoretically create more space for traditional results at the top of the page. On the other hand, the reach of AI Overviews—that is, the number of search queries for which an AI summary appears at all—continues to grow, as impressively demonstrated by the increase from 4.9 to 14.0 percent for food keywords. In practice, this means that while slightly more traditional organic space remains available for each individual search query, an increasing number of search queries are affected by this mechanism. Therefore, the net impact on the total organic traffic of an average publisher remains negative, even with the decreasing AI Overview position, because the sheer number of affected search queries continues to expand.

This problem is exacerbated by the aforementioned decoupling of traditional ranking and AI citation probability. While in the summer of 2025, three out of four pages cited in an AI Overview were also among the top ten traditional search results, this ratio had fallen to approximately one to three by March 2026, with some studies even suggesting a ratio of one to six. Google is thus increasingly drawing its AI answers from a much broader pool of sources, extending far beyond the traditional top-10 logic of classic search engine optimization. This phenomenon is often referred to in the field as "query fan-out": Instead of directly answering a single search query, the underlying AI system—now based on Gemini 3 in most AI Overviews, and the standard model since its global rollout on January 27, 2026—breaks down the original question into several sub-questions and searches a significantly wider range of sources before compiling a synthesized answer. For content providers, this means that focusing solely on ranking optimization is no longer sufficient to remain visible in the new AI search ecosystem, even though, according to Ahrefs analyses, position one in the classic results still carries a citation probability of 53 percent, compared to 36.9 percent for position ten.

The food industry as a precedent for special sectoral treatment

The data collected by Chefkoch from the food segment exemplifies a larger picture in which Google evidently treats different industries differently. A BrightEdge study from autumn 2025 on e-commerce keywords around the Christmas season shows how drastically and rapidly Google adjusts the visibility of AI Overviews depending on the product category, with jumps from 9 percent coverage to 26 percent within a week, followed by a drop back to 9 percent. At the same time, food and grocery keywords showed a significantly more stable retention of 56 percent, compared to just 3 percent for furniture queries. This strongly suggests that Google is considerably more cautious with the placement of AI Overviews for product categories with high purchase intent and correspondingly high ad value than for purely informational queries, such as those typically encountered in recipe searches.

A recent Semrush study from July 2026 on commercial search intent also confirms the special role of the food and beverage sector, where AI Overviews surprisingly appear in 16 percent of transactional search queries and 14 percent of commercial search queries. This is an unusual pattern compared to other sectors and contradicts the general observation that Google is more reserved when it comes to search queries with immediate purchase intent. This anomaly in the food segment can likely be explained by the fact that recipe and cooking queries, even when they contain a certain transactional component—such as searching for a specific dish to recreate—remain primarily informational in nature and are ideally suited for structured AI summaries, for example, in the form of ingredient lists or step-by-step instructions. This very structural suitability of the food segment for AI-generated answers could also explain why coverage there has grown so rapidly and significantly, from 4.9 to 14.0 percent in just four months, while other sectors are growing considerably slower or more erratically.

The value of having your own database in an era of algorithmic volatility

One aspect of this analysis deserves special emphasis because it extends far beyond the purely technical discussion of SEO and has fundamental strategic implications for any data-driven company. The ability to capture daily, granular SERP data with a custom-developed tool, rather than relying entirely on the pre-configured metrics of commercial SEO platforms, proves to be a crucial competitive advantage in such a volatile environment. While large commercial providers like Ahrefs, Semrush, or BrightEdge rely on broad, aggregated studies of hundreds of thousands of keywords, which inevitably reflect general trends but rarely capture the specific dynamics of a single niche industry like food and recipes in their full depth, a proprietary database allows for a much more precise and rapid response to industry-specific shifts.

This methodological consideration is by no means merely a technical detail, but a central strategic point. If, as the available data shows, the algorithmic handling of AI overviews can fundamentally change within a few months, and apparently even within individual industries, the ability to independently collect data frequently and flexibly becomes a crucial factor for corporate resilience. Companies that rely solely on standard metrics provided by third-party providers risk either recognizing important shifts too late or not being able to identify them at all, because the relevant questions are simply not available in the interface of the respective tool. The ability to retrospectively formulate new analytical questions on the complete raw data set at any time represents a structural advantage that can hardly be overestimated in an era where search engine algorithms evolve virtually in real time.

A new, more fragile equilibrium logic instead of an endpoint

The initial assessment, that the observed shifts do not represent the end of AI Overviews but rather a rebalancing within the search results page, can be clearly confirmed and even refined based on the available data. Google is in a continuous calibration process driven by several, sometimes conflicting, forces simultaneously. The need to avoid cannibalizing its own advertising business with overly aggressive AI responses is juxtaposed with the technological and competitive necessity of keeping pace with emerging competitors in the field of AI-powered response systems. At the same time, legal and reputational pressure from publishers and content creators is growing, while in parallel new search interfaces such as the so-called AI Mode, which according to Google at the I/O conference in May 2026 already reached one billion monthly active users, are establishing a completely separate, even more AI-centric logic that differs significantly from the classic AI Overviews in its hit citation, with an overlap of the cited URLs of only around 14 percent between the two systems.

Therefore, neither a complete withdrawal of AI Overviews nor a return to its undisputed top position is to be expected in the future. A more realistic scenario is a continuous, fine-grained adjustment that will further differentiate itself in individual markets depending on search intent, industry, device, and regulatory environment. The combination of increasing coverage and declining position observed in the food segment could be repeated in other industries in the coming months, but this is not necessarily guaranteed, as the underlying incentives vary considerably depending on the competitive intensity of the advertising markets and the proportion of informational versus transactional search queries. For content creators and companies dependent on organic search traffic, the key recommendation remains to pursue both traditional ranking optimization and targeted optimization for structured, AI-citable content formats in parallel, relying on the most granular, independent data collection possible to identify industry-specific shifts early on, before they materialize as noticeable traffic losses.

☑️ Our business language is English or German

☑️ NEW: Correspondence in your native language!

Konrad Wolfenstein

I and my team are happy to be available to you as your personal advisor.

You can contact me by filling out the contact form here [email protected]:or simply call me at +49 7348 4088 965. My email address is

I'm looking forward to our joint project.

☑️ SME support in strategy, consulting, planning and implementation

☑️ Creation or realignment of the digital strategy and digitization

☑️ Expansion and optimization of international sales processes

☑️ Global & Digital B2B trading platforms

☑️ Pioneer Business Development / Marketing / PR / Trade Fairs

From visibility to trust: Your scalable path with Xpert.Digital - Image: Xpert.Digital

In industrial B2B, sustainable business relationships rarely emerge overnight. They develop step by step – through visibility, professional relevance, recurring touchpoints, and growing trust. Xpert.Digital's 4-stage model addresses precisely this: It offers a structured path that begins with a manageable entry point and can evolve into deeper collaboration in business development if needed.

Instead of relying on loud marketing promises, this model puts the relationship at the forefront. Companies start with clearly defined, easily calculable measures and then decide, based on their own experience, how far they want to expand the collaboration. A key factor for this undisturbed trust-building process: The platform completely avoids annoying advertising ads, so the editorial focus remains solely on the companies' expertise.

More information here:

• From visibility to trust: Your scalable path with Xpert.Digital