Skip to content
Digital & performance marketing agency
SEO · AI series · Updated 12 October 2026

SEO in the Age of AI Search: How to Stay Visible in AI Overviews, AI Mode and ChatGPT

Ranking first no longer guarantees the click. This guide explains how generative search engines choose what to cite, what has changed in the data, and the specific page and site changes that earn visibility in both the blue links and the AI answer.

··6 min read

For twenty years SEO had one goal: the top organic position. In 2026 that position often sits beneath an AI-generated answer that has already told the searcher what they wanted to know. Google’s AI Overviews appear on a large share of informational queries, AI Mode offers a full conversational search experience, Bing’s results are wrapped in Copilot, and a meaningful number of buyers now start research in ChatGPT, Perplexity or Gemini rather than a search box. The discipline that has grown up around this is called generative engine optimisation (GEO) or answer engine optimisation (AEO). Most of it is SEO done properly, with a few genuinely new requirements. This guide covers both.

What has changed in search behaviour

Three shifts matter. Queries are longer and more conversational, because people type to AI the way they would ask a person. Informational clicks have fallen, because the answer is on the results page; the Search Console AI reports Google released in 2026 let you see impressions in AI Overviews and AI Mode separately, and for most sites those impressions are rising while clicks fall. And the funnel has compressed: a buyer who once visited six sites now reads one synthesised answer, then visits two or three of the businesses it named. Being one of those names is the new first page.

The good news is that transactional and local intent still sends clicks. “Dentist in Hisar”, “affiliate management agency”, “Google Ads pricing” and “book a demo” are searches where the user needs a business, not an answer, and those pages still get visited. The traffic you lose is mostly the top-of-funnel traffic that converted least anyway.

How AI search chooses its sources

Google’s AI features retrieve candidate pages from the normal index, so classic ranking factors are the entry ticket: if you are not in the top results for the sub-queries the system generates, you will not be cited. From that candidate set, the model favours passages that answer a specific question directly, in self-contained sentences, with a clear subject and verifiable facts. It also prefers sources whose identity it understands: who wrote this, which organisation stands behind it, and whether other trusted sources agree. ChatGPT and Perplexity work similarly, blending live web retrieval (often via Bing’s index) with the model’s training, which means Bing visibility and brand mentions across the web both matter more than they did.

Step 1: Protect the searches that still click

Pull a year of Search Console data and sort queries by clicks, not impressions. The pages behind commercial and local queries are where to invest: faster load times, clearer offers, prices or price ranges, reviews, FAQs that match People Also Ask, and strong internal links from the rest of the site. For local businesses this is inseparable from the Google Business Profile, which AI answers use heavily for “near me” questions; our local SEO service covers that side.

Step 2: Make pages quotable

A quotable page gives a direct answer in the first paragraph under each heading, uses headings that are actual questions or clear topics, keeps one idea per paragraph, and includes concrete facts: numbers, dates, names, steps. Tables and ordered lists are extracted particularly well. Add a short summary at the top of long pages, a named author with a bio, a visible published and updated date, and sources for any statistic. Then check the result with our free on-page SEO checker, which flags missing structure. Avoid the opposite failure, which is pages so stripped down for machines that a human finds nothing worth reading; the models are trained to reward the same qualities people reward.

Step 3: Build a consistent entity

AI systems reason about entities, not keywords. Your business should be described identically everywhere it appears: the same name, address, phone, description and services on your website, Google Business Profile, LinkedIn, directories and social profiles. On the site, implement Organization, Person (for authors), Service, Article and FAQ schema, link them with sameAs to your profiles, and keep them accurate; our free schema generator produces valid JSON-LD for each type. Every inconsistency is a reason for a model to be unsure who you are, and uncertain entities do not get cited.

Step 4: Earn authority signals AI engines trust

Links still matter, and mentions now matter too. Being named in articles, directories, podcasts, industry lists and community discussions, even without a link, increases the chance that a model associates your brand with a topic. Original research, benchmark data and tools are the most citable assets a business can publish because they are the things an AI cannot generate itself. Reviews on Google and industry platforms feed directly into local AI answers. And authorship is a real signal: a consistent named expert with a profile, a bio and a history of content on the topic is weighted differently from an anonymous “admin”.

Step 5: Technical access for AI crawlers

Check robots.txt for the AI crawlers you want to allow: Google-Extended controls training use but not AI Overviews, which use the normal Googlebot; OpenAI’s and Perplexity’s crawlers have their own user agents. Decide deliberately rather than by default, because blocking them removes you from those answers. Keep content in server-rendered HTML rather than behind JavaScript, make sure canonical tags and hreflang are correct, and keep Core Web Vitals healthy; AI retrieval still runs on the ordinary index, so technical SEO remains the foundation.

Step 6: Measure AI visibility

Use the Search Console AI reports to track impressions in AI Overviews and AI Mode by page. Query the major assistants monthly with the questions your customers ask and record whether and how you are cited. Watch referral traffic from chatgpt.com, perplexity.ai and copilot in GA4. And check Bing positions with our Bing rank checker, since several assistants retrieve through Bing. Report AI visibility separately from rankings; they are now different outcomes.

GEO myths to ignore

“Add an llms.txt file and you will rank in AI.”No major engine has confirmed using it for ranking. It is harmless, but it is not a strategy.
“Write for the AI, not for people.”The models are trained to prefer what people prefer. Robotic, keyword-dense text performs worse in both.
“Block AI crawlers to protect content.”You may have good reasons, but understand that it removes you from the answers your competitors appear in.
“Classic SEO is dead.”AI answers are built from the ordinary index. If you do not rank, you do not get cited.

This article is part of our series on AI’s impact on digital marketing. For the fundamentals, start with the complete SEO guide.

Not sure where your traffic went?

Request a free audit. We will separate what AI answers have taken from what you can win back, and show you which pages to fix first. Or read about our SEO service.

Mayank Rajput

Mayank Rajput is the founder of Axoria Marketing and an SEO and affiliate marketing professional based in Hisar, Haryana. He has run performance campaigns as a CPS and CPA specialist and media buyer, led SEO and digital marketing for international clients, and works hands-on across technical SEO, paid media, affiliate program management and conversion optimisation. He is a regular at Affiliate World and international SEO conferences, and writes about what actually moves qualified traffic and revenue for growing businesses.

Connect on LinkedIn
FAQ

Frequently asked questions

Straight answers to the questions we hear most. Anything else, ask us directly.

What is generative engine optimisation (GEO)?

GEO, also called answer engine optimisation, is the practice of making a website and brand more likely to be retrieved and cited by AI search systems such as Google AI Overviews and AI Mode, Bing Copilot, ChatGPT and Perplexity. Most of it overlaps with strong SEO; the additions are quotable structure, entity consistency, authorship and brand mentions.

Do AI Overviews reduce website traffic?

For informational queries, yes: the answer is completed on the results page. Transactional and local queries still send clicks. The new Search Console AI reports let you see AI Overview and AI Mode impressions separately so you can measure the effect on your own pages.

How do I get my business cited in ChatGPT?

ChatGPT with search retrieves live pages, often through Bing, so rank well in Bing, keep your entity information consistent across the web, publish clearly structured pages with direct answers, and earn mentions on sites the model already trusts. There is no submission process; citations follow authority and clarity.

Should I block AI crawlers in robots.txt?

Only if you have a specific reason. Google AI Overviews use the standard Googlebot, so blocking Google-Extended does not remove you from them. Blocking OpenAI or Perplexity crawlers removes you from those assistants' answers, which usually costs more visibility than it protects.

Is schema markup important for AI search?

Yes. Organization, Person, Service, Article and FAQ schema help systems understand who you are and what each page covers, and sameAs links tie your site to your profiles. It is not a ranking shortcut, but it removes ambiguity, and ambiguous entities are cited less.

Keep reading

Related articles

Ready to turn acquisition into a measurable growth system?

Tell us where you are and where you want to be. We will come back with a candid view of what will move the numbers and what will not.

Call Book a Strategy Call