SEO and AI Search — What Actually Changes and What Does Not

Few topics in digital marketing have generated more heat and less light in the past two years than AI search. Generative Engine Optimisation, AI Overview optimisation, GEO, AEO — the acronyms are multiplying faster than the evidence. This guide cuts through the noise and offers an honest, evidence-based assessment of what AI search actually means for SEO practice, what has genuinely changed, and what has stayed the same.

A disclosure upfront: this is a fast-moving area and this page will date. Where I am making an assessment based on current evidence, I will say so. Where the honest answer is that we do not know yet, I will say that too.

What AI search actually is

AI search is not a single thing. It covers several distinct developments that are often conflated:

Google AI Overviews

Previously called Search Generative Experience (SGE) during its testing phase, Google AI Overviews are the AI-generated summaries that appear at the top of some search results pages. They pull content from indexed web pages, summarise it, and display it with citations. They are most common on informational queries and least common on transactional or navigational ones.

The key point that is often missed: AI Overviews are generated from indexed web pages. To appear in them, your pages need to be indexed, authoritative on the topic, and well-structured enough for the model to extract and attribute your content. This is not a new set of signals — it is the same signals that determine organic rankings, applied to a different presentation layer.

AI-native search engines

Perplexity, ChatGPT Search, Google Gemini, and Microsoft Copilot are AI-native search interfaces that generate responses drawing on both indexed web content and their own training data. They represent a genuinely different search paradigm — users interact conversationally rather than through keyword queries, and the results are synthesised rather than listed.

These platforms are growing in usage, particularly for research and information-gathering tasks. They are not yet displacing traditional search for transactional or commercial queries at scale, but the trend is clear. Their share of information-seeking behaviour is increasing.

AI in traditional search

Beyond AI Overviews, Google has been incorporating AI into its core ranking systems for years — RankBrain (2015), BERT (2019), MUM (2021) are the publicly confirmed examples. The quality signals these systems assess have consistently rewarded the same things: clear, accurate, well-structured content from identifiable, authoritative sources.

What has genuinely changed

Three things have materially changed for SEO practitioners in the AI era, and they are worth being specific about.

Zero-click behaviour is increasing

AI Overviews answer questions directly on the results page, reducing the need for users to click through to source websites. This is most significant for purely informational queries — definition queries, simple factual questions, how-to content with short answers. If a significant share of your traffic comes from informational queries where the answer can be summarised in two sentences, AI Overviews are a real threat to click volume.

The practical implication is that informational content needs to either be genuinely deep enough that AI Overviews prompt users to read the full source, or strategically positioned to earn citations that drive brand visibility even when clicks do not follow.

SERP real estate has changed

The visual structure of search results pages has changed significantly. AI Overviews push traditional organic results further down the page on affected queries. This means that position 1 on a query with an AI Overview generates less traffic than position 1 on the same query without one. Average click-through rates on informational queries have declined in markets where AI Overviews are deployed.

This makes transactional and commercial intent queries — where AI Overviews are less prevalent — relatively more valuable. It reinforces the case for keyword strategies that prioritise commercial intent alongside informational content.

Brand and entity visibility matters more

AI systems — both Google’s and the AI-native platforms — appear to favour sources they can identify as authoritative entities. This means that entity establishment (being clearly identifiable as a known expert or organisation in your field) is becoming more important. Structured data, consistent author attribution, About pages with clear credentials, and citations from credible external sources all contribute to entity signals.

This is essentially E-E-A-T (Experience, Expertise, Authoritativeness, Trustworthiness) applied more broadly and more systematically. It is not a new concept — it is an existing concept becoming more consequential.

What has not changed

Despite the volume of content suggesting that AI search requires a completely different approach, the foundational requirements for organic visibility have remained remarkably stable. These things have not changed:

Technical accessibility

AI systems cannot cite content they cannot access. Pages that are blocked, not indexed, or rendered in a way that crawlers cannot process will not appear in AI Overviews or be cited by AI-native search engines. The technical SEO fundamentals — crawlability, indexation, structured data, page speed — remain as important as they have always been, and arguably more so as the systems that assess them become more sophisticated.

Content quality and depth

AI Overviews draw from pages that rank well. Ranking well still requires content that comprehensively addresses search intent, is accurate, is well-structured, and comes from an identifiable source with relevant expertise. There is no shortcut here and no prompt engineering that replaces this. The content that surfaces in AI Overviews is, in the overwhelming majority of cases, content that was already ranking organically.

Authority signals

External links remain the primary signal of authority for both traditional and AI-influenced search. The sites that appear most frequently in AI Overviews and AI-native search citations tend to be the same sites that have the strongest link profiles and the most established topical authority. Building genuine authority through quality content and earned links is not an old-fashioned strategy — it is the strategy that the evidence supports.

Search intent alignment

AI search engines are, if anything, better at assessing whether content genuinely matches the intent behind a query. Thin content that includes keywords without answering the underlying question is less likely to surface in AI-generated responses than in traditional ranked results. The intent alignment bar has not been lowered — it has been raised.

What about Generative Engine Optimisation?

GEO has emerged as a term for the practice of optimising content specifically to appear in AI-generated responses. It is worth being direct about the current state of this discipline: most of what is described as GEO is either repackaged SEO best practice, speculation based on thin evidence, or both.

There is no reliable evidence that specific GEO tactics — beyond the fundamentals described above — produce consistent, measurable improvements in AI citation frequency. The systems generating AI responses are not fully transparent, they change frequently, and the relationship between specific content choices and citation outcomes is not yet well understood. Anyone claiming to have a GEO system that reliably drives AI citation volume is overstating what the evidence supports.

That said, there are reasonable, evidence-consistent practices worth adopting:

• Clear author attribution and credential signalling — AI systems appear to favour attributable expertise over anonymous content
• Structured content with clear answers to specific questions — well-organised content is easier for AI systems to extract and cite accurately
• Factual accuracy and citing sources — AI systems are sensitive to misinformation signals; content that references credible sources and makes accurate, verifiable claims performs better
• Comprehensive topic coverage — thin pages that cover a topic partially are less likely to be cited than pages that address a topic thoroughly from multiple angles

These are not new recommendations. They are what good content has always looked like. The AI era has made the penalty for ignoring them larger, not introduced a new set of requirements.

What this means practically for SEO strategy

The strategic implications of AI search, taken together, point in a consistent direction:

• Prioritise transactional and commercial intent keywords — these are less affected by AI Overviews and continue to drive high-value traffic
• Build genuine topical authority through cluster-based content architecture — AI systems reward depth of coverage on a topic, not just the presence of individual optimised pages
• Invest in entity establishment — clear authorship, structured data, consistent brand signals, and credentials visible on the site
• Keep technical foundations robust — if AI systems cannot access your content, nothing else matters
• Monitor AI citation presence alongside traditional rankings — tools like Ahrefs now track AI Overview and AI-native search citations, which provides useful signal about topical authority
• Do not abandon informational content — being cited in AI Overviews has brand value even when clicks do not follow, and informational content continues to drive organic rankings that support commercial pages

A note on this site

AlphaSERP has 120 citations in Google AI Overviews, 56 in ChatGPT, and 217 in Perplexity — data from a client engagement that produced these results as a secondary outcome of content architecture and schema work. I mention this not as a claim that there is a reliable formula for AI citations, but as evidence that the same content quality signals that drive traditional organic rankings also drive AI citation presence. The two are not separate problems requiring separate solutions.

If you want to discuss how AI search affects your specific SEO situation, get in touch. The honest answer is usually that the fundamentals still apply — and that getting those right is the most effective thing you can do regardless of where AI search goes next.

For related reading, see the technical SEO guide and the guide on why your website is not ranking.