For more than two decades, SEO (Search Engine Optimization) has worked according to very specific logics: keyword research, link building, technical site structure. An SEO specialist's job was to make a piece of content easy for Google to find and interpret, following measurable signals such as keyword density, backlinks or loading speed.
Today, however, we are no longer faced with a list of "blue links."
With the arrival of Google's AI Overviews and the spread of engines such as Perplexity, ChatGPT, Copilot and Gemini, the search experience has been transformed. According to Gartner, search volume on traditional engines will drop by 25 percent by 2026, in favor of AI-based tools, chatbots and digital agents.
A metaphor for understanding the change
Erik Wikander, the CEO of Wilgot.ai, explained in an interview with Forbes the difference between google search and search with AI, and he did so with a simple metaphor.
"Using search engines like Google is like going into a library and getting a list of book titles that might be useful. Then you have to take the books from the shelves, browse through them, and try to find the answer yourself."
"Search based on artificial intelligence completely turns everything upside down."
"It's like having a super-intelligent librarian who reads each book for you and then explains the answer, in your language, adapting it to your context. And if you don't quite understand, you just ask another question and he clarifies. It's fluid, interactive and helpful in a way that traditional research simply is not."
Herein lies the paradigm shift: we no longer optimize our content for a ranking algorithm, but for a language model capable of interpreting meaning, context, intention, and, most importantly, anticipating users' needs and questions.
The impact of AI Overviews on organic results.
If traditional engines rewarded those who could gain the first position, today with the advent of AI Overview the situation is more complex.
AI Overview what is it
AI Overview (or AI Overview) is a feature that shows an AI-generated summary at the top of organic results, with concise answers to even very complex questions.
AI Overview what it is for
Google's stated goal is to speed up search and offer simpler and faster key information, accompanied by links to original sources.
AI Overview how it works
According to Google's official document "How AI Overviews in Search work," this technology is based on a custom Gemini model that works in conjunction with existing ranking and quality systems and the Knowledge Graph.
AI Overviews appear only when Google's systems determine that they can add value: specifically, when the use of generative AI is useful in better clarifying the intent of the user's query.
The consequences of AI Overview on organic results.
The consequences of AI Overview on organic results are significant in both quantitative and qualitative terms.
Ahrefs found that when an AI Overview appears, the CTR of the first organic result drops by 34.5 percent.
According to another analysis done by Indig in collaboration with Eric van Buskirk, organic traffic can drop by as much as 66 percent in some scenarios.
The problem is not only quantitative, but qualitative: the content selected by AI Overviews does not always coincide with the "top 10 "Google results. A research by Semrush showed that the links mentioned in the AI Overview are in 60.9% of cases placed beyond the 21st place.
Another study by Advanced Web Ranking showed that 46.5% of the URLs included in AI Overviews are not in the top 50 organic results.
So Google's AI does not just choose who is higher in SERPs, but who is more useful and more authoritative in answering a specific user question and intention.
AI Search: beyond AI Overviews
If AI Overviews represent Google's internal evolution, the other major transformation comesfrom artificial intelligence-based search engines.
Tools such as ChatGPT (currently the market leader), Perplexity, Copilot, Gemini, Deepseek, Claude, and many others are not just summarizing SERP results, but directly building an answer, leveraging Large Language Models (LLMs) and drawing on a wide range of web sources.
AI Search what it is
AI Search is the common term to describe the search for information, services and products on generative AI engines.
Specifically, it is:
- a search experience mediated by LLMs, or neural networks trained on billions of words and documents,
- capable of understanding, synthesizing, and generating fluent natural language responses
- from a simple prompt or a conversational interaction.
In AI Search, these models act differently than traditional engines: they do not simply list Web pages, but build a unique answer by drawing from both internal databases and the Web in real time, summarizing, comparing, and combining content to offer customized, often link-free answers.
How AI Search Ranking Works.
Ranking on artificial intelligence-based search engines does not work as it does on classic search engines.
There is no longer a SERP made by a ranking of blue links to climb.
Content is selected and cited by a language model that works to best meet the intent of the query. This means that AI does not just reward those who have optimized keywords better, but those who can provide complete, clear and contextualized answers.
The three crucial factors for ranking on AI engines.
- Contextual understanding → LLMs favor content that not only provides a list of tips or data, but also explains why and how. In this way, the text becomes more useful to the user and more "citable" to the model. For example, when faced with a question such as "how to improve one's nutrition," content that not only lists healthy habits but also explains the rationale and benefits will be more likely to be chosen.
- Content structure → Implementing structured data (schema markup) allows AI to better interpret the context of a page. Elements such as FAQs or step-by-step guides help models extract information into neat blocks that can be easily reused in summaries. In this sense, semantic organization becomes a signal that simplifies the AI's job and increases the likelihood that content will be selected.
- Depth → AI models reward comprehensive content that can explore a topic from multiple angles and answer all the implicit questions in the query. An article that covers a topic comprehensively, integrating explanations, examples, and sources, is much more likely to be considered valid and, therefore, cited as a reference in a generative response. A recent study by Semrush confirmed the importance of this element, showing that content with citations and statistics is up to 40 percent more visible in AI results.
EEAT for AI: the only way to get chosen
Actually, these three factors are not new.
They represent the natural evolution of Google'sE-E-A-T principles (Experience, Expertise, Authoritativeness, Trustworthiness), which are now more central than ever to generative search as well.
Google has for years called for creating content that is useful, trustworthy, and designed for people, and artificial intelligence only amplifies these criteria. In fact, AI models choose sources to draw from not based on keyword density, but on the overall credibility of a piece of content.
EEAT in AI content
- Experience → tell about direct experiences, case studies, and real-world results that demonstrate lived expertise
- Expertise → providing specialized, signed insights that show the author's authority
- Authoritativeness → citing data, studies, and industry-recognized sources, enhancing the reputation of the content
- Trustworthiness → maintaining transparency, up-to-dateness, and consistency, elements that solidify the trust of the reader and the algorithm
Positioning oneself in AI Search does not mean (anymore) "being found," but BEING CHOSEN AS A SOURCE OF TRUST, and the content that embodies the principles of EEAT are precisely those that AI models summarize, cite, and recommend in their responses.
How much does AI Search weigh (today) in numbers?
According to a study by Onelittleweb in the year from April 2024 to March 2025, chatbot traffic accounted for only 2.96 percent of total search engine visits. In other words, chatbots generated 34 times fewer visits than search engines.
As of March 2025, average daily visits for search engines reached 5.5 billion, while chatbots recorded only 233.1 million.
The gap remains wide, but the trend is clear: conversational search is growing rapidly and redefining how content is discovered.
Being ready is the first step to not falling behind.
So what?
It's true: AI Search still accounts for a small portion of global traffic today. But it would be a mistake to read this figure as a signal of calm.
Because while generative chatbots are still growing, AI Overviews are already here and rewriting the rules of ranking.
Every day, millions of Google searches are filtered, synthesized and returned by artificial intelligence, and those who continue to write only for traditional SEO risk being left behind: invisible in the AI summaries that will drive the choices of tomorrow's users.
We need deeper, clearer, more human content. New writing habits and editorial logic are needed, built around context, structure, and credibility.
Subscribe to our newsletter or follow us on LinkedIn to be among the first to read a dedicated insight on how to write for AI engines.
https://onelittleweb.com/data-studies/ai-chatbots-vs-search-engines/
https://www.roberto-serra.com/news/chatbot-ia-vs-motori-di-ricerca-2025/
https:// gs.statcounter.com/search-engine-market-share
https:// www.sap.com/italy/resources/what-is-large-language-model
https:// www.semrush.com/blog/ai-search-optimization/
https:// www.semrush.com/blog/ai-mode-comparison-study/
https:// www.semrush.com/blog/ai-search-visibility-study-findings/
https:// www.semrush.com/blog/how-do-you-optimize-for-ai-search/
https:// www.allaboutai.com/resources/ai-statistics/ai-search-engines
