AI has already changed how people search.
Google now places AI-generated answers directly into search results. ChatGPT can search the live web, cite sources and help users compare products, services and information. Buyers are increasingly using generative AI as part of their research process before they ever reach a company website.
The behavioural data shows that this is more than a superficial interface change.
Pew Research Center’s analysis of 68,879 Google searches found that users clicked a traditional search result on 8% of visits when an AI summary appeared, compared with 15% when one did not.
Only 1% of visits containing an AI summary resulted in a click on a source cited within that summary.
Pew has also found that 60% of US adults say they read AI summaries at the top of search results.
For B2B organisations, the shift is equally significant. Gartner’s 2026 B2B buyer research found that 45% of buyers used generative AI during a recent purchase, primarily to gather information about vendors and products. Buyers consulted an average of seven different information sources during the process.
Separate Gartner research has also reported that 60% of B2B buyers use generative AI in some form to influence purchase decisions.
So AI and SEO can no longer be treated as separate future concepts.
AI is already part of search behaviour.
The practical question is now how businesses should adapt their SEO strategy so they remain visible across traditional Google results, AI Overviews, AI Mode, ChatGPT and other large language model-driven search experiences.
The important point is that this does not mean starting again from scratch.
The fundamentals of strong SEO still matter.
In fact, Google’s own guidance confirms that many of the same principles used to improve traditional search performance remain relevant to AI-driven search.
The search experience has changed.
The foundations of discoverability, relevance, authority and trust have not.
What Is AI SEO Optimisation?
AI SEO optimisation is the process of improving visibility across both conventional search engines and AI-powered search experiences.
That includes traditional Google organic results, Google AI Overviews and AI Mode, ChatGPT search and other LLM-powered discovery environments.
You may also see this described as AI optimisation, AIO, GEO, AEO or LLM optimisation.
The terminology is still evolving, but the underlying strategy is more familiar than many discussions make it sound.
Effective AI SEO optimisation still depends on technical accessibility, useful content, strong information architecture, clear search intent, topical authority, trusted external references and a good user experience.
The difference is that businesses now need to consider whether their information can be discovered and understood not only by traditional search engines, but also by AI systems that retrieve, combine and summarise information from multiple sources.
That is an evolution of SEO, not a replacement for it.
AI and SEO: What Has Actually Changed?
One of the most important changes is how users can express what they are looking for.
Traditional search often required people to reduce a complicated need into a short query.
For example:
enterprise CRM
CRM for financial services
CRM Salesforce integration
AI-powered search allows the same user to ask:
“Which enterprise CRM platforms are suitable for a multinational financial services company that needs Salesforce integration, European data residency and SSO?”
That is a much richer expression of intent.
The search system can interpret several requirements at the same time and retrieve information that contributes to a more complete answer.
This has major implications for AI and SEO.
Optimising one page around one exact keyword is becoming less important than building a sufficiently strong body of information around the subject.
Longer Searches Are More Likely to Trigger AI Results
The data supports this shift towards more detailed and conversational search behaviour.
Pew found that Google AI summaries appeared for:
- 8% of searches containing one or two words
- 20% of searches containing three or four words
- 36% of searches containing five to nine words
- 53% of searches containing ten or more words
Queries beginning with question words such as “what”, “why”, “who” and “when” generated AI summaries around 60% of the time.
That matters because longer queries often reveal more commercially useful intent.
Compare:
cybersecurity platform
with:
“What cybersecurity platform should a UK bank use to manage encryption keys across AWS and Azure?”
The second search may have very little measurable keyword volume in an SEO tool.
Commercially, however, it could be far more valuable.
This is why modern AI SEO optimisation needs to consider topics, entities, attributes, requirements and user intent rather than relying exclusively on search volume and exact-match keywords.
AI Search Has Already Changed Click Behaviour
AI-generated answers are also changing what happens after somebody searches.
Pew found that traditional-result clicks dropped from 15% to 8% when an AI summary appeared.
Users were also more likely to stop browsing after seeing an AI summary. Approximately 26% ended their browsing session after a page containing an AI summary, compared with 16% when one wasn’t present.
That does not mean every website should expect organic traffic to collapse.
The impact is likely to vary significantly by search intent.
A user asking for a simple definition may get everything they need directly in the search result.
Someone researching a complex B2B service, evaluating a medical treatment, comparing software platforms or considering a major purchase usually needs considerably more information.
That makes an important distinction increasingly relevant:
Not every organic click has the same commercial value.
A modern AI and SEO strategy should therefore focus less on maximising traffic at any cost and more on improving visibility around the searches, questions and buying journeys that matter to the business.
Does AI Mean Traditional SEO Is Becoming Less Important?
No.
This is one of the areas where AI SEO is often overcomplicated.
Google’s official guidance for AI search optimisation makes it clear that existing SEO best practices remain relevant to AI Overviews and AI Mode.
That is because Google’s AI experiences still depend on its underlying Search infrastructure and ranking systems to discover and retrieve useful information.
In practical terms, Google still needs to find your pages, crawl them, interpret their content, understand how pages relate to one another and determine whether the information is relevant and useful enough to surface.
If a website is technically inaccessible, badly organised or filled with weak content, an additional layer of “AI optimisation” is not going to fix those problems.
This is why we still consider the fundamentals of SEO the starting point for AI visibility.
AI search has not replaced SEO fundamentals. It has expanded the number of places where those fundamentals can create visibility.
How Google Uses Content in AI Search
Google has provided more transparency around how its generative search systems retrieve information.
Its documentation describes techniques including retrieval-augmented generation, or RAG.
Put simply, this allows the AI system to retrieve current information using Google’s search infrastructure and use those sources to help construct an answer.
Google also describes query fan-out.
Instead of treating a complex question as one search, the system can perform several related searches covering different elements of the original query.
This has an important practical implication for SEO.
You do not necessarily need one page perfectly optimised for every possible conversational question.
Instead, your website needs enough depth around commercially important subjects for search systems to retrieve relevant information from across your content.
That increases the value of strong topic coverage, sensible site architecture, supporting pages and contextual internal links.
None of these are entirely new SEO strategies.
What has changed is the sophistication with which search systems can retrieve and combine the information.
AI SEO Optimisation Is Not About Keyword Stuffing
Keywords remain important because they tell us how people describe their problems, products and requirements.
But AI SEO optimisation makes simplistic keyword density an even less useful strategy.
Consider a company trying to improve visibility around “enterprise identity management”.
A comprehensive resource may naturally need to discuss SSO, identity governance, authentication, access controls, user provisioning, compliance, integrations, security and deployment.
Those concepts help establish what the solution is, what problems it solves and when it may be relevant.
That does not mean throwing every semantically related phrase onto a page.
The information still needs to satisfy a real user need.
Effective SEO content strategy should therefore connect keyword research with search intent, topical coverage and commercial relevance rather than turning every keyword variation into an isolated article.
Make Important Information Explicit
One of the most practical elements of AI SEO optimisation is also one of the simplest: clarity.
Marketing websites are often surprisingly vague.
They talk about “transforming experiences”, “unlocking potential” and “accelerating innovation” without clearly explaining what the company actually provides.
That is frustrating for users, but it can also make an organisation harder for search systems to understand.
If you sell enterprise software, your website should clearly establish what the product does, who it is designed for, the problems it solves, its important capabilities and integrations, the industries it serves and relevant considerations such as deployment, security, compliance and pricing.
The same principle applies in other industries.
A medical website should clearly connect clinicians with specialisms, treatments and clinic locations. A professional services company should clearly communicate its expertise, services and target customers. An ecommerce website should provide complete, accurate and unambiguous product information.
This is not simply “optimising for ChatGPT”.
It is good information architecture, good usability and good SEO.
Optimise for Questions and Real Buying Scenarios
AI search makes it easier for users to combine multiple requirements into one question.
Keyword research therefore needs to be supplemented by a better understanding of the questions people ask during real buying journeys.
Some of the most useful information already exists inside the business.
Sales teams know what prospects repeatedly ask before signing a contract. Customer success teams know what customers struggle to understand. Product specialists know which capabilities require explanation. Support tickets, RFPs, Search Console data, site searches, reviews and competitor comparisons can reveal the same patterns.
You may repeatedly encounter questions such as:
- Does the software support SAML?
- Can it be deployed on-premise?
- Does it integrate with Salesforce?
- Is it suitable for businesses with more than 10,000 employees?
A keyword tool may report almost no search volume for those exact questions.
That does not make them commercially unimportant.
If they repeatedly appear during purchasing decisions, there is a strong case for answering them clearly online.
Commercial relevance should inform content strategy alongside search volume.
Build Content Around Complete Topics, Not Isolated Keywords
AI-powered retrieval provides another reason to think in terms of connected information rather than collections of unrelated pages.
For a SaaS company, the journey might move from a problem to a solution, then into features, integrations, industries, comparisons and customer evidence.
For a healthcare organisation, somebody might begin with a condition before researching treatment options, specialists, clinic locations, costs and recovery.
Those relationships should influence website architecture.
Supporting content can answer detailed informational questions while commercial pages explain the actual product or service.
Contextual internal links connect the two and allow users and search engines to move naturally through the subject.
Again, this is not a completely new AI SEO tactic.
It is an established SEO principle that has become increasingly relevant as search engines become better at understanding complex topics and questions.
AI Has Made Original Expertise More Important
Generative AI can produce competent generic content extremely quickly.
That is precisely why generic content is becoming less useful.
If every competitor can generate another introductory article explaining the same subject in seconds, publishing your own slightly rearranged version does not create much additional value.
Google’s current AI search guidance recommends creating unique, valuable and non-commodity content.
For businesses, this creates an opportunity.
Your engineers may know how difficult technical problems are actually solved. Consultants know what commonly goes wrong. Sales teams hear questions that do not appear in keyword tools. Customer service teams see recurring problems. Senior specialists may have decades of expertise, while the business itself may hold proprietary data nobody else can publish.
Good content strategy should bring more of that knowledge onto the website.
The question is not:
“How much content can we create with AI?”
It is:
“What can we contribute that somebody could not get by asking AI to summarise the existing search results?”
That is a much higher bar, but content that clears it is considerably more defensible.
Original Research Is Particularly Valuable for AI and SEO
Original research is one of the clearest ways to create something genuinely distinctive.
Businesses frequently sit on useful information without recognising its content value.
Product usage patterns, anonymised customer data, surveys, benchmarks, sales insights and internal research can potentially reveal something useful about an industry.
Published well, that information can generate organic rankings, editorial links, PR coverage and industry citations.
It can also give journalists, publishers, researchers and potentially AI systems an original source to reference.
That is a very different proposition from producing another generic article targeting a high-volume keyword.
The objective is to become a source worth referencing.
Authority Beyond Your Website Matters More
Your website is not the only place where search engines and AI systems can learn about your business.
Industry publications, customer websites, review platforms, partners, professional bodies, news coverage, research, forums and other independent sources can all contribute to the wider picture.
Pew’s analysis provides an interesting example.
Wikipedia, YouTube and Reddit collectively accounted for 15% of sources cited within the Google AI summaries it analysed.
Government websites were also proportionally more common in AI summaries than in conventional search results, representing approximately 6% of cited sources compared with 2% of traditional results.
This does not mean businesses should manufacture Reddit discussions or create Wikipedia pages purely for AI visibility.
The more useful lesson is that digital authority extends beyond your own domain.
Third-party references can help corroborate who you are, what you sell, what you are known for and whether credible organisations recognise your expertise.
This is where digital PR, brand building, relevant link acquisition and AI optimisation increasingly intersect.
E-E-A-T Still Matters in AI SEO Optimisation
Trust becomes particularly important when AI systems synthesise information from multiple sources.
Google’s quality framework considers Experience, Expertise, Authoritativeness and Trustworthiness, commonly referred to as E-E-A-T.
These considerations are particularly significant for health, finance, safety and other areas where inaccurate information could cause real harm, but the underlying principles apply much more widely.
It should be easy to establish who is responsible for important content, whether they have relevant expertise, where factual claims originate and whether the information remains current.
Company ownership and contact information should be transparent, and important claims should ideally be supported by evidence rather than simply asserted.
An author biography alone does not create E-E-A-T.
Authority is the cumulative result of the content, people, organisation and wider evidence supporting what is being said.
And in a web increasingly populated by synthetic content, demonstrable expertise may become more valuable rather than less.
Technical SEO Is Still Critical for AI Visibility
For all the sophistication of generative AI, these systems still need access to information.
Google explicitly recommends ensuring that important content can be crawled, indexed and accessed in text form.
That keeps familiar technical SEO priorities such as crawlability, indexation, internal linking, canonicalisation, JavaScript rendering, structured data, site architecture and page performance firmly relevant.
This becomes particularly important on large websites.
A technical problem affecting one page may be insignificant. The same issue built into a template used across 100,000 URLs can materially affect organic performance.
No amount of AI-focused content optimisation compensates for commercially important pages that search engines struggle to discover or understand.
This is why our managed SEO services combine technical SEO, content and authority rather than treating AI optimisation as an isolated bolt-on.
Do You Need Special Schema or an llms.txt File?
Probably not in the way many AI SEO articles suggest.
Google has confirmed that there is no special schema required for AI Overviews or AI Mode.
Existing structured data remains useful when it accurately describes information users can see on the page, but adding schema does not make weak content authoritative.
Google has also said that llms.txt currently neither helps nor harms visibility within Google Search, and businesses do not need to artificially break content into tiny sections simply because somebody claims AI systems prefer them.
Structured data is best viewed as a way of reducing ambiguity, not as a shortcut into AI-generated answers.
That is a useful principle for AI SEO more generally: make genuinely useful information easier to discover and understand rather than searching for technical tricks that supposedly manufacture relevance.
AI SEO Optimisation for ChatGPT
ChatGPT introduces another discovery environment because users can search the live web, receive cited answers and continue refining their requirements conversationally.
A user researching an agency, for example, might begin with:
“What are the best enterprise SEO agencies for a multinational SaaS company?”
They could then narrow that search by asking which agencies specialise in technical SEO, which have international experience and how the remaining options compare.
Your ability to appear somewhere within that research journey is not simply determined by whether one page contains the phrase “enterprise SEO agency”.
The wider information environment needs to make it possible to establish what the organisation does, where its expertise lies and whether those claims can be supported.
For businesses, practical AI SEO optimisation for ChatGPT therefore comes back to many of the same fundamentals: clear product and service information, strong topical coverage, accessible expert content, relevant external authority, accurate information and a technically sound website.
There is no individual optimisation that guarantees ChatGPT will recommend a company.
The objective is to make the evidence supporting your relevance as strong and unambiguous as possible.
AI SEO Should Be Commercial, Not Just Informational
This is one of the most important parts of the strategy.
AI SEO discussions often focus heavily on blog articles because informational queries are where AI answers are particularly visible.
But businesses ultimately need search visibility to produce commercial value.
Product, service, solution, industry, integration and location pages therefore deserve just as much attention.
If an AI system is helping somebody compare potential suppliers, it may need to establish which services you offer, who you work with, whether you provide a particular capability, which industries you understand, where you operate and how your proposition differs.
Those answers frequently live on commercial pages rather than blogs.
Optimising informational content while leaving the pages closest to revenue vague creates a disconnect between AI visibility and business growth.
Give People a Reason to Visit Your Website
This becomes particularly important if AI answers continue reducing informational clicks.
If somebody’s entire question can be satisfactorily answered in three sentences, there may be little reason for them to visit the original source.
Businesses therefore need to create more things that are difficult to replace with a summary.
Original research, calculators, benchmarks, templates, interactive tools, detailed technical resources and genuinely useful comparison material can provide value beyond the generated answer.
They can also attract links and strengthen authority.
Our guide to creating compelling lead magnets explores how useful resources can turn informational interest into meaningful commercial engagement.
The objective is not to put every useful resource behind a form.
It is to give the right audience a genuine reason to continue the journey with you.
Measure AI SEO Beyond Organic Traffic
Traffic remains important, but it can no longer tell the whole story.
Someone might encounter your business in an AI Overview, see it mentioned again in ChatGPT, search for the brand several days later and eventually contact sales.
Traditional last-click attribution may give almost none of that earlier search visibility any credit.
That does not mean we should invent precision where it does not exist.
AI visibility measurement is still developing, and claims of perfect LLM attribution should be treated cautiously.
Instead, businesses can combine conventional SEO metrics such as rankings, organic traffic and conversions with emerging signals including AI Overview appearances, LLM citations, AI referral traffic, branded search demand and broader share of search.
Most importantly, those metrics should eventually connect back to leads, pipeline, revenue or whatever constitutes meaningful commercial performance for the business.
Our guide to SEO ROI explores why traffic and rankings alone have never been sufficient measures of SEO success.
A Practical AI SEO Optimisation Framework
For most organisations, AI SEO optimisation should follow a logical order rather than beginning with experimental AI-specific tactics.
1. Fix the SEO Foundations
Make sure important pages can be discovered, crawled, rendered and indexed before investing heavily in AI-specific experimentation.
2. Identify Commercial Search Topics
Determine which problems, services, products and questions matter to revenue rather than allowing search volume to dictate the entire strategy.
3. Improve Information Clarity
Make important facts about your organisation, expertise, products and services explicit.
4. Build Complete Topic Coverage
Connect commercial landing pages with supporting content around the subjects you genuinely need to be known for.
5. Demonstrate Real Expertise
Bring internal specialists, proprietary data, original research and first-hand experience into the content.
6. Strengthen External Authority
Develop relevant links, editorial mentions, citations, partnerships and other credible third-party signals.
7. Optimise the Commercial Journey
Make sure users who do reach the website have a clear reason and route to take the next step.
8. Measure SEO and AI Visibility Together
Continue tracking conventional organic performance while incorporating useful AI visibility and referral signals as the data matures.
This is a considerably more sustainable approach than chasing whichever new “GEO hack” happens to be circulating this month.
AI and SEO: What Businesses Should Prioritise Now
The practical conclusion is relatively straightforward.
AI has already changed search behaviour.
More people are encountering generated answers. AI summaries are affecting click behaviour. Complex queries are disproportionately triggering generative results. B2B buyers are using AI to research vendors and products. ChatGPT has become another meaningful search and discovery environment.
Those trends could become substantially more significant as the technology improves.
But the evidence does not suggest abandoning conventional SEO.
It points towards expanding it.
Technical accessibility still matters because information needs to be retrieved.
Content quality matters because search systems need something worth surfacing.
Topical authority helps establish relevance.
E-E-A-T and external authority help establish trust.
Usability and conversion become even more important if overall informational traffic declines.
The businesses best positioned for AI search are therefore unlikely to be those chasing every new optimisation trick.
They will be the organisations with technically strong websites, genuinely useful information, demonstrable expertise, clear commercial propositions and credible authority across the wider web.
That is how we approach AI and SEO at Lemon Pulse.
Rather than separating traditional SEO, technical SEO, content, digital authority and AI SEO optimisation into competing disciplines, we look at how they work together to create visibility and commercial opportunity across Google, AI Overviews, ChatGPT and other emerging search experiences.
The objective is not simply to get your company mentioned by an AI.
It is to make sure that when potential customers research the problems you solve, your organisation has a credible opportunity to be discovered, considered and ultimately chosen.
If your SEO strategy is still built primarily around rankings and traffic, or you are unsure how your visibility is changing as search becomes increasingly AI-driven, talk to Lemon Pulse about building an SEO strategy for conventional and AI search.
