Search is moving from a list of links to an AI powered written answer produced by artificial intelligence. For brands, that changes how people find you, what a click is worth and what it means to be visible.
This briefing sets out how AI search changes SEO and how to stay ahead of it, using only dated evidence from the original studies and the search companies themselves.
The short version: the foundations of search engine optimization still hold, but the prize has moved. Being named inside the answer now matters alongside ranking on the page, and the click is no longer guaranteed to follow.
Read this if you run digital marketing for a brand, or you own the results. You will see what the evidence shows today, what it does not show, and what to change first. Nothing here is a forecast. Every figure names its source and its date.
Search engines now put an AI search answer above the list of search results
Traditional search engines return a ranked list of search results and let the reader choose. AI powered search engines that use AI search read many sources, build one written reply and show it first, with links to supporting pages.
Google calls its versions AI Overviews, which sit inside the page, and AI Mode, which is a separate conversational view.
Google explains that both may use a technique it calls “query fan-out”, issuing several related searches across subtopics and data sources to build a response (Google Search Central, 2026).
Other AI powered tools work in the same spirit, and user expectations are shifting with them. ChatGPT search, Gemini and Perplexity all write direct answers instead of listing ten blue links.
Each uses its own crawler and its own rules, so being visible in one does not guarantee being visible in another. SEO is now a multi-platform job.
Readers click less when AI Overviews and other AI generated summaries appear
The clearest user behavior data comes from Pew Research Center. It analyzed the browsing of 900 US adults who agreed to share their activity, covering March 2025. About six in ten of them, 58%, ran at least one Google search that produced an AI generated summary.
On visits where a summary appeared, users clicked a traditional search result link in 8% of visits. Where no summary appeared, they clicked in 15% of visits, nearly twice as often (Pew Research Center, 2025).
Links inside the summary itself were rarely used. Pew found a click on one in just 1% of visits to pages with a summary. Users were also more likely to end their session after seeing a summary.
This is one country, one month and Google only, so treat it as a signal and not a law. It still explains why zero-click search is now part of every plan for organic traffic.
The click-through rate on the top result has fallen, and personalized search results blur it
Ahrefs, which sells SEO software, studied the same pressure from the publisher side. It compared 300,000 keywords, half with an AI Overview and half informational keywords without one, using Search Console data for December 2023 and December 2025.
Its finding: the presence of an AI Overview now correlates with a 58% lower average click-through rate for the top-ranking page (Ahrefs, February 2026).
Ahrefs frames this as a correlation based on desktop data and a modelled comparison, not proof of cause.
Read that as a warning about measurement, not a verdict on your site. A page can hold its position in the search results and still send fewer visitors. If you only watch search engine rankings and search rankings reports, you will miss the change in SEO performance.
Personalized search results make a single rank check even less reliable.
Visibility now means being named, not only ranked
When the answer is written for the reader, a list position matters less than whether your brand appears inside the text. Modern SEO strategies now weigh citations in AI generated answers and AI generated content alongside traditional rankings. There are two outcomes to separate. A mention is your name in the answer.
A citation is a link to your page. Mentions build recognition. Citations send visitors.
What drives a mention? Ahrefs tested 75,000 brands and found that branded web mentions correlated most strongly with brand visibility in AI Overviews, at 0.664, against 0.218 for backlinks (Ahrefs, 2025).
That is a correlation, not a recipe. It does suggest that what other credible sources say about you carries more weight than raw link counts, which is a shift for anyone whose strategy has leaned on link building. Third party mentions and independent reviews shape how easily a brand is discovered.
Semrush looked at what the visitors are worth. Across more than 500 digital marketing and SEO topics, it found the average AI search visitor to be 4.4 times as valuable as the average organic search visit, judged by conversion rate (Semrush, 2025).
Semrush does not set out how it tracked every conversion, so measure AI referrals on your own analytics.
Google says the fundamentals of search engine optimization still apply
Here is the part of search engine optimization that many headlines skip. Google’s own documentation states that the best practices for SEO remain relevant for AI features such as AI Overviews and AI Mode, and that there are no additional requirements to appear in them (Google Search Central, 2026).
To be shown, a page must be indexed and eligible to appear in Search with a snippet. Google adds that meeting every requirement does not guarantee a page will be crawled, indexed or served.
Google also addresses the new acronyms. It says AEO means answer engine optimization and GEO means generative engine optimization, and that optimizing for generative AI search is still SEO. It adds that structured data is not required and no special schema.org markup is needed (Google Search Central, 2026).
That is Google speaking about its own products. Other AI tools have their own crawlers. OpenAI says sites that opt out of OAI-SearchBot will not be shown in ChatGPT search answers, and that this setting is independent of GPTBot, which relates to model training (OpenAI, 2026).
So traditional SEO foundations stay: pages that can be crawled, read and trusted, which is what technical SEO and clear content structure deliver. What moves is the target. The Brand Library explains answer engine optimization as the practice built for that target, and our guide to answer engine optimization goes deeper.
| Traditional SEO | SEO for AI answers | |
|---|---|---|
| Where you want to appear | A ranked list of links | Inside the written answer, named or cited |
| What you are judged on | Relevance, authority, page quality | Whether a passage answers the question cleanly and who else vouches for you |
| What you track | Rankings, clicks, organic traffic | Mentions, citations, AI referrals and branded search |
| What stays the same | Crawlable, helpful, trusted content | Crawlable, helpful, trusted content |
What Google’s own guide tells brands to do in the AI era, and what it says to ignore
Google has published an official guide to optimizing for generative AI features, and it is more specific than most commentary. It explains that these features use retrieval-augmented generation: Google’s core ranking systems retrieve relevant, up-to-date pages from its index, and the model then writes a response with links to supporting pages (Google Search Central, 2026).
The guide asks for three things. First, non-commodity content. Google contrasts commodity content, such as “7 Tips for First-Time Homebuyers”, with pages that carry a unique point of view or first-hand experience, and says the second kind will likely influence presence in generative AI search more than any other suggestion in the guide.
Second, a clear technical structure. Pages must be crawlable, indexed and eligible to appear with a snippet. Third, accurate business details. Google says Merchant Center feeds and Google Business Profiles help products and local services appear in AI responses, so local businesses and online stores should keep both current.
For measurement, Google points to the Generative AI performance report in Search Console, and warns that no third-party tool has access to its internal ranking or AI systems.
The guide is just as direct about what to skip. Google says Search does not use llms.txt files or special markup, that content does not need to be broken into tiny chunks, that there is no ideal page length, and that writing differently just for AI systems is not required.
It also says seeking inauthentic mentions is not as helpful as it might seem.
Read that last point beside the Ahrefs finding. Mentions matter, but they are earned by being worth mentioning, not manufactured.
AI crawlers, conversational search queries and website traffic need a fresh look
Technical checks now cover more than one crawler. Google tells site owners to confirm that crawling is allowed in robots.txt and by any CDN or hosting infrastructure, that important content is available in text, and that structured data matches the visible text on the page (Google Search Central, 2025).
That page was last updated in December 2025.
OpenAI runs separate AI crawlers for separate jobs. OAI-SearchBot surfaces sites in ChatGPT search, GPTBot collects content that may be used to train its models, and ChatGPT-User visits a page when a person asks a question. OpenAI says each setting is independent, so a site can allow OAI-SearchBot and disallow GPTBot (OpenAI, 2026).
Review those settings on purpose. A blanket block on AI crawlers that nobody has reviewed can quietly remove your brand from the answers you want to appear in.
Google also describes how people use AI Mode. It is helpful for queries that need exploration, reasoning or complex comparisons, and people can ask nuanced questions that might once have taken several searches.
The search queries worth writing for are therefore fuller questions from your target audience, so lay out comparisons and trade-offs plainly on the page.
On website traffic, Google makes two points. AI Overviews are shown only when its systems judge them additive to classic Search, so they often do not trigger. And clicks from results pages with an AI Overview, it says, are higher quality, meaning people are more likely to spend more time on the site.
That is Google’s own observation, and it sits beside the Pew and Ahrefs click data, so check both in your own analytics.
Local SEO belongs here too. Google lists up-to-date Merchant Center and Business Profile information among the fundamentals for AI features.
Fold all of this into your wider digital marketing strategies. AI visibility across other tools still has to be checked by hand, as step 4 below describes, and human expertise is the one input an AI system cannot generate for you.
Ground Truth, weekly
How AI search changes SEO for brands that want to be found
Four things change in practice. None of them needs a new department, only updated SEO strategies.
SEO strategies that stay ahead treat traditional SEO as the base and leveraging AI as the layer on top. Search engine algorithms, AI powered answers and AI tools all reward the same habits: clear answers, trusted sources and a brand they can describe. Watch search trends, but do not chase every one.
1. Content creation starts from the question
AI systems extract passages to build AI generated responses and AI generated answers, so each section should answer one real question in its first sentences, then give the detail. A page that circles the point for four paragraphs gives a system nothing clean to quote. Optimizing content around direct answers is the cheapest change.
Clear content structure is easier for AI algorithms to parse, and it also serves a hurried human reader.
Match the search intent behind each page, not only the keyword. Search intent is the thing to get right. The Brand Library defines search intent. Search engine algorithms now use machine learning models to read user intent and context, so the conversational phrasing of the question matters more than simple keyword matching.
User intent now shapes what AI search shows. Writing for user intent is the core of SEO now, and it is how you meet user intent where AI search finds it.
2. The brand has to be one consistent entity
If your website, LinkedIn page and directory listings each describe you differently, AI systems have no clear picture to repeat. Settle one name, one category and one set of facts, and use them everywhere. AI systems map semantic relationships between a brand, its category and its claims.
This is positioning work before it is technical work.
3. Keyword research widens to questions and mentions
Good keyword research starts from these questions. Keyword research still finds demand, but the unit of work shifts from a phrase to a set of buyer questions. Write the questions your customers ask, put them to the AI tools, and note who is named.
Then work out which outside sources shape those answers, and pick relevant keywords and long tail keywords from the questions people actually phrase.
4. Measurement adds the AI generated answer itself
Rank tracking does not show whether an AI tool names you. Add a monthly check: 15 to 25 real buyer questions, put to ChatGPT, Gemini, Perplexity and Google AI Mode, with each result recorded as named, cited or absent.
Answers change between runs, so read the trend, not a single result. The check gives valuable insights into your website’s visibility. Update your SEO strategies to include one more metric: whether AI powered answers name you.
Search engine optimization teams and SEO professionals are adding new metrics, such as how often they are cited and their share of voice across AI answers. Our piece on why brands go missing from AI answers sets out that audit in more detail.
What SEO professionals should adjust in their SEO strategies this quarter
AI SEO is not a separate discipline, but it does change the weekly checklist of SEO professionals. Treat what follows as an update to existing SEO strategies and digital marketing, not a new program.
Start with keyword research. Search engines and AI systems use natural language processing to read whole questions instead of strings of words, so list the full questions your target audience asks, and the phrases around them.
Automate keyword research where a tool saves time, then check the output against real user behavior in your own analytics.
Look next at user behavior after the click. If search results send fewer visitors, each visitor who arrives matters more. Track what they read, which pages they start from and where they convert, because that shows which questions your content already answers.
Give voice search a short test. People speak to assistants, including Google Assistant, in full sentences, which looks more like a conversational question than a typed keyword. Read a top page aloud. If the opening sentence does not answer the question in plain speech, rewrite it. Each blog post should answer one question first.
Keep the traditional SEO practices that still earn their place: titles that describe the page, internal links, fast pages and technical SEO that keeps important content crawlable. Traditional SEO methods are the base for AI SEO, not a legacy to retire.
Do not chase keyword density. Google’s guide says you do not need to write in a specific way just for generative AI search, so there is no density number to hit.
Finally, report in terms a leader can act on: organic traffic, search engine rankings, conversions from search, AI referrals from your analytics and the named, cited or absent audit. Effective SEO strategies read these together, because search engine results pages now hold both links and written answers.
Across AI platforms, reshaping search will keep changing the mix of the two.
Machine learning, voice search and data analysis shape the future of SEO in the AI era
The future of SEO is shaped by machine learning inside search engines and by how people speak to them. Voice search is one visible edge of that change. Spoken questions arrive as whole sentences, and voice search assistants rely on natural language processing to match them to answers.
Voice search will not replace typing, but it rewards pages written in plain language.
Machine learning also sits under the ranking systems themselves. Search engine algorithms and AI algorithms learn patterns from large volumes of content and user behavior, which is why a single tactic ages quickly.
Data analysis is the practical response: compare organic traffic, conversions and AI referrals month by month, and let the trend steer your search engine optimization.
SEO professionals who pair data analysis with human expertise read these signals best. Ask where users interact with your pages, where they leave, and which search queries bring them. Then refine on page optimization for content optimization: headings, opening answers, internal links and a title that matches search intent.
Optimizing content is also a content creation habit. Draft the answer first, add the detail, then check keyword optimization against the questions you want to answer. On link building, favor reviews and earned coverage over volume.
None of this makes the future of SEO mysterious. In the AI era, AI SEO is mostly traditional SEO done with more care about questions, entities and measurement. Google’s AI Overviews and the other AI tools simply make the quality of that work easier to see in search engine results.
What changes for search engines, AI algorithms and AI driven SEO practices
Search engines have always changed their algorithms, and SEO professionals have always adapted. What is different now is that AI algorithms sit between search engines and the reader.
AI driven features such as AI Overviews read pages, weigh sources and write a response, so search engines are no longer only ranking pages. They are assembling answers.
For SEO professionals, that means reading AI driven results the way you read any search engines update: look at what changed, which queries are affected and what your own data shows. The same SEO practices, including clear structure, helpful content and crawlable pages, still decide whether AI algorithms can use your work.
Search engines also differ from each other. Google, Bing and the AI assistants each run their own AI algorithms, so a digital marketing team should test the same question across several search engines before deciding that a change is real.
Link building follows the same logic. A link from a credible site still helps search engines understand your authority, and accurate local SEO listings help AI driven answers describe a business correctly.
For a target audience that asks fuller questions, user intent and keyword research remain the starting point. A single blog post can serve content creation, SEO strategies and measurement if it answers one question well.
AI powered tools and AI driven SEO speed up content optimization but do not supply content quality
AI also changes how the work gets done, including content creation workflows and AI powered drafting. Search engines use machine learning algorithms to rank search results, and AI driven SEO tools use the same kind of AI powered technology.
AI powered tools now automate repetitive tasks in keyword research, on page optimization, technical audits and reporting.
McKinsey’s 2023 analysis of generative AI estimated that it could raise the productivity of the marketing function by an amount worth 5 to 15 percent of total marketing spending. It also noted that generative AI can help marketers lower cost and raise conversion through SEO work on page titles, image tags and URLs (McKinsey, 2023).
AI tools also help SEO professionals and content marketers spot search trends, predict search trends from large datasets and identify trends worth a page.
AI tools can analyze large datasets to uncover keyword opportunities, including long tail keywords that convert better than generic keywords. They also suggest relevant keywords and page titles, improve readability, draft metadata from user behavior data and flag mobile experience problems. That is useful search optimization, and it is enhancing SEO without replacing SEO.
AI is revolutionizing SEO workflows, but the goal of relevant traffic from people who need you is unchanged. Leveraging AI frees time for what a tool cannot do.
AI driven SEO tools support SEO content creation and content optimization, but a draft is only as good as the brief behind it. AI generated content that repeats what already exists gives an answer engine no reason to pick it.
In a crowded digital marketplace, content quality is what stands out as AI generated content multiplies. E-E-A-T principles, meaning experience, expertise, authoritativeness and trust, are crucial for credibility. Optimizing content for AI search means offering original data, a clear point of view and named experience, because those decide whether a page is quoted or skipped.
Use AI algorithms and AI models for routine steps, and keep a person responsible for accuracy, so your content remains relevant.
What artificial intelligence does not change in AI SEO, and what the evidence cannot tell you
Content quality still decides whether a page deserves to be quoted. Pew found the same three sites, Wikipedia, YouTube and Reddit, were the most commonly linked in both AI summaries and standard results, at 15% and 17% of sources. Established, credible sources feed both.
Search algorithms keep changing, so judge your SEO efforts by trend, and keep core SEO practices steady. Enhancing SEO still means the basics: SEO practices like crawlable pages, clear answers and trusted sources, and content creation that answers real questions.
Review your SEO strategies each quarter against AI generated answers, because AI powered features and search engine algorithms keep changing how those AI generated summaries are built. Be honest about the gaps.
No platform publishes a ranking formula for AI answers, so nobody can give a verified weighting. Most studies above are correlations. Several are one country or one month.
And the share of your own visitors who arrive from AI tools is something only your analytics can show.
Be wary of anyone who guarantees citations. Models, retrieval methods and AI driven features change without notice, and Google itself says that meeting its requirements does not guarantee a page will be served.
How traditional SEO, local SEO and link building fit the website and the distribution plan
AI search sits inside two things a brand already does.
The first is the website. A site built to rank and to be quoted by AI answer engines has clear page structure, content that answers questions directly and analytics wired to what matters. Our website design service treats search and AI foundations as a named part of the build.
The second is distribution. Outside mentions are earned, not published, and earning them is distribution work across owned, earned and paid reach. Our content distribution explainer covers the three routes, and the Distribution Strategy page shows how we plan them.
If you want a weekly view of what is changing in search and AI answers, with the dated evidence behind it, join the Ground Truth newsletter. It goes out on Wednesdays.
For the fuller explanation of how generative engines choose what to cite, read what generative engine optimization is.
FAQ
How does AI search affect SEO?
AI search adds a written answer above the results. Pew found users clicked a traditional result in 8% of visits with an AI summary against 15% without. The SEO basics still apply, but visibility now also means being named or cited inside the answer.
How is AI going to change SEO?
Measurement and targets change first. Rankings alone no longer show whether you are found, so brands add mentions, citations and AI referrals. Google says SEO best practices remain relevant for its AI features, so the change is in emphasis, not a new rulebook.
Is SEO dead or evolving?
Evolving, not replaced. AI is not replacing SEO. Google describes optimizing for generative AI search as still SEO, and AI answers depend on pages that can be found, read and trusted. What is dying is the habit of measuring success by rank and clicks alone.
What does the 30% rule for AI mean?
The phrase is not a search requirement. We found no 30% rule in Google’s documentation on AI features and generative AI search. Treat any fixed percentage as one author’s guideline, and judge a page by whether it is accurate, helpful and trusted.
Do I need schema markup for AI search?
Not necessarily. Google states that structured data is not required for generative AI search and that no special markup is needed. It can still help describe what a page already shows, so use it for clarity and not as a shortcut. Search optimization for AI needs no special markup.