In short: yes. Google answers the question "Is SEO still relevant for AI search?" in its own documentation with "In short, yes!". The AI Overviews in Google Search rely on the same ranking and quality systems as regular search. A page that isn't indexed won't appear there either. What has changed is the way people search and click.
Why are so many people asking this question?
Since March 2025, Google has also shown AI Overviews in Germany. They sit above the classic results and summarise an answer, often with links to sources. Google announced the feature for Germany on 25 March 2025, in German and English, initially for signed-in users aged 18 and over. On top of that, chatbots such as Google Gemini and other AI assistants answer questions directly.
For website owners, this means that some searchers get their answer without clicking on a result. That raises the worry that working on rankings no longer pays off. How much clicks are falling overall is disputed and depends heavily on the topic.
What does Google itself say?
Google has set out its position in several official documents. Three statements matter in practice:
- No special rules: The documentation "AI features and your website" states that there are no additional requirements for appearing in AI Overviews or AI Mode. You don't need special files, separate markup or particular structured data either.
- Same foundation: According to the guide to optimising for generative AI features (as of July 2026), these features are based on the core ranking and quality systems of Search. From Google's point of view, optimising for AI search is therefore still SEO.
- Indexing remains a must: A page has to be indexed and eligible to be shown with a snippet before it can appear as a source in AI features. Even then, inclusion is not assured.
In May 2025, Google also wrote on the Search Central blog that clicks from results pages with an AI Overview were of higher quality, with visitors staying on the website longer. Keep in mind that this is Google's own statement and no independent measurement.
What does AI change for SEO?
Simple answers lose value
Pages that only repeat knowledge found everywhere have a harder time. In its May 2025 blog post, Google explicitly recommends unique content that can't easily be swapped for something else. Personal experience, your own data, practical examples and clear recommendations are hard to replace.
Being cited as a source
An AI Overview often lists several sources side by side. Besides your position in the list, it matters whether your page shows up as a reference. Content with clear statements, good structure and verifiable facts is better suited to this than long texts without a core message.
Brand and trust gain weight
When people search, they often see provider names in the AI answer. If your brand is known and mentioned on other websites, users are more likely to recognise you and search for you directly. That is classic brand building, and it complements SEO.
Other search systems join in
Other AI services search the web with their own systems. Unlike Google, they offer no official guidance. The basics still help: reachable pages, a clean technical setup, clear content. Under the label GEO, short for Generative Engine Optimization, people discuss how to appear more often in such answers. You can read what is actually backed by evidence in our article on GEO and AEO.
What stays the same?
Most of the basic rules still apply. Google has to be able to find, crawl and index your pages. Titles, headings and text should match the search intent. Your site needs to load quickly and work well on a phone. Links from other websites remain a signal of trust.
Structured data also stays useful, for example for products, events or recipes. Google points out, however, that it must match the visible content. There is no special AI code.
On the other hand, Google explicitly calls some things unnecessary: artificially chopping content into small bites, creating extra AI text files such as llms.txt, or collecting inauthentic mentions. Generating pages in bulk just to cover variations of search queries falls under the spam policy on scaled content abuse.
Technical basics: crawling and indexing
Before a page can appear in search results or in an AI Overview, Google has to discover it, fetch it and add it to the index. The technical terms for this are crawling, rendering and indexing. If any one of these steps fails, even the best content won't help. Technical SEO is therefore the foundation for everything else.
Internal links and sitemap
Google discovers new pages mainly through links. According to Google, every page you want in the index should receive at least one internal link from another page on your site. Google can only read links that are set as a normal HTML element with an href attribute. An XML sitemap adds to this: it lists your important URLs and helps Google find them faster.
Using robots.txt correctly
The robots.txt file controls which areas crawlers may access. Google describes it as a tool for managing crawler traffic. A common misconception: blocking a page in robots.txt does not reliably keep it out of Google. The URL can still appear in the results, just without a description. If a page should not be in the index at all, Google recommends noindex or password protection. After every relaunch, also check whether the robots.txt from the test phase is still blocking the whole site. A leftover like that is easy to miss.
Canonical tags against duplicate content
Many websites serve the same content under several URLs, for example with and without parameters or as a print version. With a canonical tag in the page head, you tell Google which URL you prefer. According to Google, redirects and rel="canonical" are strong signals, while a sitemap entry is a weak one. Google treats these signals as hints and makes the final decision itself. For canonicalisation, Google advises against using robots.txt.
Page speed and mobile display
A slow page, or one that is hard to use on a phone, loses visitors wherever they come from. Our article on improving website speed shows how to tackle this.

Which pages are losing traffic? How to check
General claims about falling clicks won't help you much. What matters is what happens on your own website. Google Search Console is free for this and shows data straight from Google.
Performance report with a date comparison
- In Search Console, open the "Performance" report for Google Search. It shows four metrics: clicks, impressions, click-through rate (CTR) and average position.
- Click the date filter at the top and choose "Compare". This puts two periods side by side.
- In the table, switch to the "Pages" tab and sort by the difference in clicks. The pages with the biggest loss will then be at the top.
- Click one of these pages and switch to "Queries". Now you can see which search terms are missing clicks.
Make sure you compare periods of the same length. For seasonal topics, comparing with the previous year makes more sense than comparing with the previous month. According to Google, the most recent days are partly preliminary and may still change.
Reading the numbers correctly
Three patterns come up often. If impressions are stable but clicks have dropped, users still see your result yet click it less often. One possible reason is an answer already shown on the results page, for example in an AI Overview. When impressions and position fall together, you have lost rankings. In that case, look at your content, technical setup and competitors. If a page is missing entirely, it may have dropped out of the index.
Keep in mind: in Search Console, Google counts traffic from AI features as normal search traffic. Its documentation does not mention a separate filter for it. That means you can only estimate the impact indirectly.
Checking indexing and individual URLs
The "Page indexing" report shows which pages are in the index and why others are missing. Typical reasons are a noindex tag, a block in robots.txt, or a duplicate for which Google has chosen a different URL as canonical. According to Google, not every URL on a site needs to be indexed. What counts is that the canonical version of each important page is there.
With the URL Inspection tool, you can see for a single URL whether crawling is allowed, whether indexing is allowed and which canonical URL you have declared. Next to it, the tool shows the URL that Google itself has selected as canonical. After a fix, you can request indexing again from there.

Content based on your own experience: how to go about it
In its guide to helpful content, Google describes the criteria experience, expertise, authoritativeness and trustworthiness, E-E-A-T for short. Of these, Google considers trust the most important. As an example of experience, it mentions knowledge that comes from actually using a product or service or visiting a place. That kind of content is exactly what an AI cannot assemble from other texts.
Making experience visible
- Tell stories from everyday work: Describe how you actually solve a task, with the steps that work for you and the mistakes customers often make.
- Collect customer questions: Note down questions from emails, phone calls and consultations. They show what people really want to know.
- Your own photos and evidence: Pictures from your business, screenshots from real projects or your own measurements come across as more credible than stock photos.
- Name the limits: Also write about when a solution is not a good fit. That builds trust.
Who writes, and how is the text created?
Google advises making it clear who created a piece of content and recommends accurate author information. A short author box with role and experience is often enough. If you use AI for writing, a note about it can make sense where readers might wonder how the text was produced. Google also writes that large amounts of AI text without human review take little effort. So check every draft yourself and add what only you know.
Our article on content marketing strategy explains how to plan this kind of content.
Who benefits most from SEO?
- Local businesses: Someone searching for "hairdresser near me" wants an address and opening hours. Here, the Google Business Profile and local SEO matter. Find out more on our Local SEO topic page.
- Service providers whose clients need advice: With questions like "How much does … cost?" or "How do I find …?", people are looking for someone they can trust. Detailed, honest content convinces here.
- Online shops: Product searches rarely end with an AI answer alone. Clean product and category pages remain important.
- Specialist sites and blogs: If you show real knowledge and first-hand experience, you are more likely to be cited as a source.
SEO pays off less for pages that only repeat general facts without adding anything of their own. For those, the AI answer is often already enough.
What should you actually do in 2026?
- Check indexing: Google Search Console shows which pages are in the index. Clicks from AI features count as normal search traffic there.
- Sharpen your content: Answer your customers' questions directly at the start, then in detail. Add your own examples, figures and experience.
- Keep the technical side clean: page speed, mobile display, clear canonical tags, no robots.txt blocks on important pages.
- Strengthen your brand: Profiles, mentions in trade media and genuine reviews help people and machines place you.
- Measure differently: Besides clicks, look at enquiries, revenue and branded searches.
Checklist: SEO basics in the age of AI
- robots.txt does not block any important pages or sections.
- Every important page has at least one internal link and is listed in the XML sitemap.
- The "Page indexing" report shows no unexpected exclusions.
- Duplicate URLs point to one main version via a canonical tag or redirect.
- The website loads quickly and works on mobile.
- Each page answers the main question right in the first paragraph.
- Texts contain your own experience, examples or data that can't be found elsewhere.
- Authors are named, along with their role.
- Structured data matches the visible content.
- Every few months, you use the Performance report to compare which pages are losing clicks.
- Links from other websites are earned in a clean way. Our article on link building explains what is allowed.
SEO or GEO: do you need both?
Many providers sell GEO as a new discipline. For Google Search, Google itself says that no separate approach is needed. Other AI services have no official rules, yet well-reachable, clearly structured and credible pages benefit there too. In practice, GEO therefore widens the scope of SEO to include other search systems.
Summary
SEO is not obsolete in 2026. Search is changing, and so is the way visitors reach your website. If you want to be found, you still need a technically clean website and content with its own value. What's new is the competition to appear as a source in AI answers. You'll find the basics on our SEO topic page.
Sources: Google Search Central, "AI features and your website" (as of 10.12.2025); "Optimizing your website for generative AI features on Google Search" (as of 10.07.2026); Search Central Blog, 21.05.2025; Google blog post on the launch of AI Overviews in Europe, 25.03.2025; Google spam policies; Google Search Central, "Creating helpful, reliable, people-first content" (as of 05.10.2026); "Introduction to robots.txt" (as of 10.12.2025); "How to specify a canonical URL" (as of 10.07.2026); "Make your links crawlable" (as of 10.12.2025); Search Console Help: Performance report, Page indexing report, URL Inspection tool.