GEO and SEO: How Search is Changing with AI

Artificial intelligence has quietly reshaped our habits: how we search for information, how we choose which products to buy, and how we form our opinions.

For anyone with a digital product, this changes the rules of the game. In the past, the goal was to improve ranking in the SERP (Search Engine Results Page). But what about today?

In 2026, taking care of SEO (Search Engine Optimization, meaning optimizing visibility on search engines) also means addressing a new concept: GEO (Generative Engine Optimization), which refers to how results are ranked by generative models such as ChatGPT, Gemini, or Perplexity.

How artificial intelligence has changed online search

A large share of online users has already started using AI-powered search engines instead of Google. Why scroll through dozens of results—without even knowing if you’ll find what you’re looking for—when you can simply ask ChatGPT and get a single, complete answer, possibly followed by more specific questions if the result isn’t satisfactory?

The numbers behind generative AI are constantly growing: ChatGPT is approaching one billion active users, Perplexity has grown by 800% in just one year, and Gemini is also performing strongly, with over 100 million new users between the third and fourth quarters of 2025.

Moreover, when users perform a standard search on Google, in most cases the first result they see is an AI Overview—a response generated by combining the best available sources into a single answer.

Why GEO is becoming important

If you have a product you want to promote, today it’s no longer enough to aim for a high ranking in search results—you also need to be part of the context created by AI and be cited as a source.

In fact, if a large portion of users no longer browse through all results but instead rely on AI agents or read generated summaries, more and more searches will become “zero-click”—meaning users find the answer without opening any links.

It is estimated that when an AI-generated summary appears, the CTR (Click-Through Rate—the percentage of users who click on a link out of total impressions) for results not cited by the AI drops significantly, by approximately 35–60%.

For companies with a digital presence, the focus—and the way to maintain visibility—is to appear in AI-generated results by optimizing content through GEO.

Why GEO matters for businesses

The goal of GEO is to increase the likelihood of being selected in responses from systems like Google AI Overviews or ChatGPT, much like SEO does for traditional search results.

For a company, being cited primarily means greater brand exposure—being visible to an audience in a context that inherently conveys authority. This directly impacts customer acquisition, brand reputation, and opportunities for partnerships and even fundraising.

Additionally, when your content is used as a knowledge source, you can shape your own narrative. This means being able to communicate your perspective on a product instead of leaving that space to competitors, indirectly influencing public perception.

Finally, if users want to explore an AI-generated answer further, they are very likely to click on the cited sources. For this reason, click-based metrics still remain relevant.

How Generative Engine Optimization works

How can you be selected as an ideal source by artificial intelligence? There is, of course, no method that guarantees selection 100% of the time. However, there are several guidelines you can follow to increase the likelihood of being chosen as a source.

To gather useful content, LLMs (Large Language Models) use crawlers that visit web pages, read their content (including embedded links), and ultimately build a structured map of the web.

Traditional crawlers are used to rank and index pages. LLM crawlers, on the other hand, extract information that can be used to answer common user questions.

As a result, the content that is preferred typically has:

• Clear semantic structure
• Explicit definitions
• Direct answers to questions
• Single-topic sections

But how do LLMs interpret content? They don’t just read text—they rely on tools specifically designed to provide explicit information to crawlers, such as llms.txt files and JSON-LD.

What is the llms.txt file and what is it used for?

The llms.txt file is a text file designed to provide crawlers with information about the most important parts of a website. Its format is markdown, readable by both humans and machines, similar to the traditional robots.txt file used by search engine crawlers.

This file summarizes the content of a digital product by listing:

• Project name
• A short description
• The main sections of the site
• Links within each section

You can find an example of an llms.txt structure provided directly by Google at this link.

Thanks to this file, crawlers can access the right information—especially the content you want to highlight, such as pricing pages or up-to-date documentation.

From another perspective, it also allows you to filter information, determining what is relevant and what is not, with implications for the product’s reputation.

What is JSON-LD and how to use it

JSON-LD is a data format that translates a website from human-readable text into machine-readable content. This helps AI understand the relationships between elements, rather than relying solely on keywords.

In practice, it is used to make answers about a digital product explicit. A single website typically includes multiple files of this type—called JSON-LD snippets—which structure different kinds of information.

For example, one snippet might refer to a single product, a blog article, or even the entire organization, linking together its name, website, contact details, social profiles, and more. There are also FAQ snippets, which are particularly valuable for GEO because their question-and-answer format matches how AI systems operate.

Another noteworthy snippet is LocalBusiness/Place, which is especially relevant for companies because it communicates geolocated information to AI in a clear way, such as:

• Business location
• Opening hours
• Price range
• Links to social profiles

This means that if it’s important for your product to appear for potential customers in a specific geographic area, AI systems are more likely to prioritize your site for queries like “SEO experts in Padua.”

Traditional SEO principles still apply

Not everything we’ve learned about traditional SEO should be discarded—quite the opposite. Many principles are still valid and are also used by AI systems. One of the most important is the E-E-A-T principle: Experience, Expertise, Authoritativeness, Trustworthiness.

These criteria are considered important in both traditional SEO and GEO: those who demonstrate these qualities provide more value and therefore rank better in search results.

If your digital product’s growth has stalled and you’re looking for advice on how to optimize it for GEO and start scaling again, get in touch with us. We can discuss the best solutions for your product together.