Google AI Overviews: Influencing, managing, and generating traffic
Listen to the article:
Imagine you type a question into Google, like: “How to choose a coffee machine for home?”
Instead of the usual list of links, you immediately see a brief, coherent answer where Google explains the main types of coffee machines, what to consider when choosing, and even gives tips on which brands are most popular with users. All of this is an AI Overview – essentially Google’s AI-generated answer.
What is Google AI Overviews and how are AI results generated?
Google AI Overviews is a new search feature based on generative artificial intelligence. It presents the user with a concise, summarized answer that cites multiple sources — most often sites from the top 10 search results.
The technology is built on the Gemini model (formerly Bard) and other AI developments at Google. The system analyzes a multitude of documents, extracts key fragments, and composes a coherent answer using several sources. The user gets a quick answer without clicking any links, but still sees mentions of the websites where the information came from.
In other words, Google itself is increasingly becoming the “answer provider,” and classic SEO is facing a new reality. You now have to think not only about your organic rankings, but also about how to appear in AI Overviews and capture traffic from them.
As of July 2025, according to various analytics platforms (e.g. SparkToro and Similarweb), AI Overviews “eat up” about 18% to 30% of organic traffic for informational and navigational queries.
This effect is especially noticeable in niches where answers can be generated without visiting a site: finance, medicine, food, travel. In commercial topics the impact is lower, but up to 15% of traffic can still be lost due to reduced click-through on regular links.
At the same time, the share of Google searches that trigger AI Overviews already exceeds 20–25% of all queries, and it continues to grow.
In this article, the experts at RegisTeam explain how the AI-driven results work, where they get their data, and what businesses can do so as not to lose positions but, on the contrary, to use the new system to their advantage.
How does Google use generative AI to create short answers?
Since mid-2023, Google has been actively incorporating generative artificial intelligence (AI) into search results. Instead of simply showing a list of links, the AI analyzes dozens of sources and creates a summarized text that answers the user’s query — this is essentially how AI Overviews work:
1. Identifying the query intent. Google first determines what exactly the user wants: find information, buy a product, compare something, etc.
2. Gathering and analyzing data from sites. The algorithms then process relevant pages (primarily from the top 10 results), extracting key points, facts, and frameworks.
3. Generating the answer using AI. Based on the Gemini model (formerly Bard), a concise, coherent text is generated. It typically cites 2–4 sources, though sometimes the links are hidden behind a “Show sources” button.
Our services:
When does AI Overviews appear?
AI answers do not appear for every query, only in cases where the system is confident in the quality and accuracy of its summary:
| Type of Query | Example | Why an AI Overview appears |
| Informational | “What’s the difference between green and black tea?” | AI can quickly summarize knowledge from multiple sources. |
| Educational / How-to | “How to fix a bathroom faucet?” | Step-by-step procedures are easily structured by AI. |
| Overview / Comparative | “Best smartphones under $500 in 2025” | AI aggregates opinions and provides a concise summary. |
| Factual | “How many calories are in a banana?” | AI provides a specific numeric answer. |
| Step-by-step recipes / Guides | “No-bake cheesecake recipe” | The answer format is easily constructed by AI. |
| Navigational + Informational | “Which bank is best for a deposit in Ukraine?” | AI gives a brief analysis of the options. |
Commercial and transactional queries (e.g. “buy a laptop”) are much less likely to trigger AI Overviews, since Google prioritizes ads and regular organic results.
How does AI search work?
To understand how Google AI Overviews impact search traffic and user behavior, it’s important to compare them with the classic organic search results:
| Aspect | Classic Google Search | Google AI Overviews |
| Sources of information | Top 10 organic results (plus ads) | Selected snippets from multiple sites |
| Display format | List of links with titles and descriptions | AI-generated answer with source mentions |
| Click-through rate (CTR) | High for positions 1–3 | Lower – many users get the answer without clicking |
| Content control | Full – everything depends on the site owner | Partial – Google compiles the text itself |
| Brand visibility | Website name is visible in the title | May be hidden or only mentioned in the sources section |
| Speed of getting answer | Only after clicking through to a site | Instant, directly on the SERP |
| Ranking algorithms | PageRank, E-E-A-T, user engagement signals | Emphasis on E-E-A-T, content structure, and completeness |
| Optimization capability | Full – via SEO, content, and user engagement metrics | Partial – via content structure and schema markup |
| Query types | All types (including transactional) | Primarily informational, how-to, comparisons |
Having your site appear in an AI Overview doesn’t guarantee traffic. Even if your page ranks at the top, some clicks will be taken by the generated answer, and the user might not visit your link. To maintain visibility and benefit from AI-driven results, classic SEO alone isn’t enough — you need to adapt your content structure for the AI algorithms: make it more understandable, structured, trustworthy, and ready to be reused in answers.
Why does quality content from the top 10 get priority?
Google isn’t using a completely separate ranking system for AI Overviews versus regular organic results. The AI looks for information in places where the search engine already “trusts” the sources.
If a site made it into the top 3 for a competitive query, it means:
- It passed the quality filters.
- The content is deemed expert-level.
- Users interact positively with the site.
Therefore, the AI prefers to use snippets from those pages, rather than looking for material down on the 2nd or 3rd page of results.
For example:
For the query “How to choose a robot vacuum for an apartment,” the AI Overview provides a brief answer:
If your site lacks a clear structure, an original approach, or Google’s trust, the chance of being included in an AI answer drops.
What sources does ChatGPT use?
The basic version of ChatGPT (without internet access) relies on static data extracted from open sources, including high-authority sites like Wikipedia, Stack Overflow, government portals, public knowledge bases, technical blogs, and major media. This data was pulled during the model’s training up to a certain cutoff date (for example, GPT-4’s knowledge goes up to the end of 2023).
Extended versions of ChatGPT with internet access (via Bing or plugins) leverage live web search. In such cases, the model looks at search engine results and can provide links to actual webpages, including your site.
When and how can AI link to your website?
ChatGPT with internet access can add a link to your site in two cases:
- if your page directly answers the user’s specific question;
- if your site appears among the visible search results for the query’s topic.
In addition, some plugins (for booking, news, shopping, etc.) can use your content directly if it’s connected via an API or indexed through special feeds.
However, even without a direct link, the AI might form an answer based on snippets of text from your site. This boosts your brand authority if the user later searches for the source of that information.
Brief retellings from AI in search results: Official Source
How to get into Google AI Overviews?
| Content Type | Example Query | Why it Gets Picked for AI Overviews |
| Answers to questions | “What to do for a burn?” | Provides a precise, concise answer to a specific user query. |
| Step-by-step instructions | “How to set up a router?” | The step-by-step format makes it easier to generate an answer. |
| Overview articles | “Best smartphones under $500” | Covers the topic in full, including comparisons, pros and cons. |
| Comparative materials | “Sole proprietorship or LLC — which to choose?” | Offers point-by-point arguments and helps in decision-making. |
| Frequently asked questions (FAQ) | “How to choose a bicycle: common questions” | Provides short answers to sub-questions all in one place. |
| Tables and lists | “Comparison of mobile carrier plans” | Make it easier to extract data and structure information. |
| Short paragraphs | “What is E-E-A-T?” | Improve readability and reduce “noise” in the text. |
| How-to guides with subheadings | “How to clear cache in Chrome” | Break the material into logical blocks, which is convenient for parsing. |
| Pages with schema markup (schema.org) | “Instructions for installing an air conditioner” | Enable algorithms to clearly interpret the text’s structure and meaning. |
How does credibility and expertise affect getting into AI Overviews?
Google gives preference to sites it trusts. The AI Overviews algorithms use E-E-A-T signals (Experience, Expertise, Authority, Trustworthiness) to judge how suitable a source is for generating an answer:
1. Clearly identify the author and their expertise. AI ranks pages higher when an author is listed along with their credentials in the topic — education, work experience, professional status.
2. Include external links to primary sources and research. Content without evidence is seen as weak. AI more often chooses text that includes references to authoritative sources like academic papers, government portals, or international standards.
3. Boost your brand presence and media mentions. The more often your site or brand is mentioned on other reputable platforms, the higher its authority in the algorithm’s eyes. This is important for both organic search and AI selection.
4. Maintain a detailed “About Us” page and contacts. Providing transparent information about your business increases trust: legal details, addresses, real phone numbers, links to social media, and public profiles of your team.
5. Show reviews, case studies, proof of experience. AI trusts pages that display real signs of hands-on experience: examples of work, client testimonials, photos, demonstrations.
This approach shows Google that your content is created by experienced people, for people, and can be safely used in AI-generated results.
Useful articles:
AI trends in automotive website promotion
How AI Overviews are changing organic traffic and what website owners need to know
Technical SEO Optimization for AI Overviews
Google’s AI doesn’t just read your text — it analyzes the page structure, code, markup, and technical factors:
1. HTML structure and content logic
| Before | After |
| A long, unstructured text with no clear sections, headings, or logical flow | A clear structure with H1–H3 headings, paragraphs, logical sections with labels |
| Non-semantic HTML markup (using <div> and <span> everywhere) | Using semantic HTML tags like <article>, <section>, <header>, <main> |
2. Schema.org markup and FAQ/HowTo blocks
| Before | After |
| Plain text with no additional structured markup | Pages marked up with structured data (e.g. FAQPage, HowTo, Article schema) |
| Answers to questions presented without clear formatting | Q&A sections wrapped in proper FAQ schema markup |
3. Loading speed, responsiveness, and Core Web Vitals
| Before | After |
| Slow loading times with heavy images | Optimized images (WebP), lazy loading, and resource compression |
| A desktop-only design not adapted for smartphones | Mobile-first responsive design |
| The page fails Core Web Vitals | Core Web Vitals in the green zone (e.g. CLS < 0.1, LCP < 2.5s, etc.) |
Google’s AI algorithms consider your site’s technical performance when selecting sources. Even good content can be ignored if a page takes 5 seconds to load.
Adapting commercial pages for AI responses
Commercial pages rarely appear in AI Overviews, since they’re often focused only on selling. But if you properly adapt their structure and content, such a page can become a useful source even for generative search results. Here’s how:
Step 1: Add a useful information block at the top of the page
What to do: At the beginning of the page, insert a short, structured block that explains what the product or service is, who it’s for, and its key features.
This should be a neutral, descriptive paragraph – not a promotional one.
Step 2: Implement microformats and structured markup
What to do: Add structured data markup such as Product, Service, Review, AggregateRating, and Offer. This gives Google’s AI clear signals about the content type, its properties, and user ratings.
Step 3: Add real testimonials and case studies
What to do: Include a section with real customer testimonials, ideally with photos, names, and usage details. In B2B contexts, describe case studies with concrete numbers and results.
Example:
“Thanks to the ad campaign setup by RegisTeam, we increased inquiries by 64% in 3 months (testimonial from Dmitry, CEO of AutoParts).”
Step 4. Create a unique description rather than copying the characteristics
What to do: Avoid using the supplier’s boilerplate descriptions. Write a first-person explanation of why the product/service is valuable and what problems it solves.
Step 5: Find a balance between a salesy and informative tone
What to do: Begin with the informational content at the top, and put persuasive selling points further down the page. Don’t blend everything into one block. The AI is looking for clear answers, not emotional sales copy.
These adaptations help turn a commercial page into a source of useful information, rather than just an ad landing. This increases the chance of being included in AI Overviews — especially for queries like “which one to choose,” “how it works,” or “is it suitable for…”.
Recommendations for creating content that goes into AI Overviews
Here is a list of effective techniques that significantly improve your content’s chances of appearing in AI Overview blocks:
| Technique / Method | What to do | Why it works |
| Q&A format | Formulate subheadings as questions and provide a direct, concise answer. | AI can quickly locate this snippet and use it as a ready-made block for its answer. |
| Lists (numbered or bulleted) | List out benefits, steps, features, or arguments in list form. | Lists are easy to parse and are often carried over into AI Overviews as structured content. |
| Step-by-step instructions | Break processes down into individual steps, describing each one separately. | AI prefers step-by-step formats — they simplify generating how-to answers. |
| Comparison tables | Compare characteristics, prices, pros and cons in a table format. | Tables structure information in a way that’s convenient for generating comparisons. |
| H2 and H3 subheadings | Divide the text into logical sections using second- and third-level headings. | Clear headings help the AI navigate your content and identify the needed sections. |
| One idea per paragraph | Ensure each paragraph delivers one main idea, without unnecessary fluff. | AI processes compact, logically formatted fragments more effectively. |
| Logical content structure | Start with a general conclusion, then provide details and examples. | This structure aligns with how the AI builds an answer: from the main point to the specifics. |
These approaches not only make your text easier to read, but also adapt it to the principles by which Google selects information for its AI summaries.
Which niches are more likely to benefit from AI output?
AI Overviews don’t cover every topic evenly. The AI is more active where information can be quickly summarized, a direct answer given, or a comparison offered. Some niches gain an advantage, while others, on the contrary, lose clicks.
| Niche | Why it benefits | How to leverage in your strategy |
| Medicine | Many queries have a clear intent (symptoms, first aid, medications). | Create answers to specific questions like “what to do if…”, “how to treat…”. |
| Finance | AI effectively aggregates numerical data, step-by-step instructions, and comparisons of conditions. | Format your content as tables, how-to guides, and explanations of terms and percentages. |
| Education | Frequent FAQs, how-tos, terminology, and reference queries. | Write educational articles: definitions, step-by-step processes, guides. |
| Technology | Queries like “how does it work,” “which to choose,” or “what’s the difference” are easily structured. | Produce comparisons, reviews, instructions, and FAQ sections with technical explanations. |
| E-commerce | Search queries include “which to choose,” “best products,” “top rated…”. | Publish overview articles, product round-ups, top-10 lists, and expert opinions. |
| Travel | Many how-to queries: routes, visas, prices, tips. | Provide step-by-step itineraries, interactive guides, and visual content blocks. |
| Home & Living | AI can easily generate answers from everyday “how-to” instructions and advice. | Structure your tips into clear steps, and create content like how-to articles, video tutorials, and checklists. |
Above, in the table, an analysis of the areas that benefit from AI output and strategy recommendations.
Integrating AI Overviews into content strategy
Start by identifying the topics that satisfy the informational demand in your niche — these could be common questions, practical instructions, or frequently encountered problems. When designing your article structure, consider the search intent and categorize content by type: explanation, step-by-step guide, comparison, or help with choosing. In each piece, use the formats that AI prefers: lists, tables, structured how-to guides, as well as Q&A blocks with short answers. If you’re working on commercial pages, adapt them to include not only a sales section but also useful information relevant to the user’s informational queries. This approach makes your content simultaneously useful, well-ranked, and suitable for generative search results.
How to track traffic from AI Overviews and AI tools?
Google has already started highlighting AI Overviews traffic in GA4 — it appears labeled as source: google and medium: ai_overview. This lets you track visits from AI answers directly, without having to rely on guesswork.
Additionally, you can use Google Search Console to watch for an increase in impressions without clicks and a drop in CTR while your positions remain stable. These shifts often indicate that an AI answer block has appeared for a query.
Several SEO tools, including Ahrefs, have also added features to identify queries where your site appears in AI Overviews. This helps capture not only traffic, but also overall search visibility.
Why does AI sometimes use sites outside of Google's top 10?
Although Google’s AI Overviews are predominantly built from top-ranked sites, the AI algorithms can include material from resources outside the top 10. This isn’t a bug in the system, but a deliberate choice based on additional quality signals.
Firstly, content freshness plays a crucial role. If your page was updated recently and contains up-to-date data on a fast-changing topic (for example, new laws, emerging technologies, trends), it might be chosen by the AI as more relevant — even if it ranks lower than competitors.
Secondly, a unique user experience and original delivery of information are becoming more important than plain SEO optimization. For instance, an author might share a personal case study, experiment results, or conduct original tests. Such signals indicate real experience and strengthen the “Experience” element in E-E-A-T.
If you operate in a niche field, focus not on volume and broad keywords, but on depth, accuracy, a personal touch, and covering topics that competitors haven’t addressed. The AI algorithm is looking for usefulness, even if it lies outside the top 10.
In addition, the AI takes into account user reviews, comments, video tutorials, galleries — anything that helps it feel that the content is alive and not templated. Non-standard content formats (interactive elements, visualizations, a mix of text and video) increase the algorithm’s trust.
This is especially noticeable in narrow topics where competition among authoritative sites is lower. In such cases, the AI might prefer a blog post, expert column, forum thread, or small-business page if its content is unique, clear, and relevant.
