SEO in the Age of AI: What Still Matters

Search Has Split Into Two Parallel Systems

Why does SEO in the age of AI require a different approach than traditional search optimisation?

The most important strategic shift to understand is that search in 2026 does not happen in one place. It happens across at least two meaningfully different systems — traditional Google search and the growing ecosystem of AI-powered answer engines — and these systems evaluate content differently, reward different signals, and serve different user behaviours.

Google still processes over 13 billion searches per day. AI Overviews now appear on a significant portion of those results, but for commercial and transactional queries — the searches that drive business enquiries — the #1 organic position still generates substantial click-through.

Ahrefs research from December 2025 found that AI Overviews reduce organic CTR for position one by up to 58% on informational queries, but the impact on commercial-intent searches is considerably lower. Goodfirms’ 2026 survey found that 86.5% of marketers say ranking first on Google still matters — specifically for the queries that convert to customers.

Alongside Google, ChatGPT referral traffic grew 206% year-on-year from January 2025 to January 2026. More significantly, Semrush data shows that AI-referred visitors convert at 4.4 times the rate of traditional organic search visitors. Ahrefs found that just 0.5% of its total AI-referred visitors drove 12.1% of signups.

The implication is clear: AI-referred traffic is smaller in volume but significantly higher in intent. Businesses that appear consistently in both Google results and AI-generated answers are capturing both the scale of traditional search and the quality of AI referral.

The strategic error is treating these as competing priorities. The content practices, technical foundations, and authority signals that make a site perform well in Google increasingly overlap with what makes it get cited by ChatGPT, Perplexity, and Google AI Overviews. Understanding where they diverge — and where they align — is the practical work of SEO in the age of AI.

What Has Genuinely Changed

Which SEO factors have changed most significantly with the rise of AI search?

Several things have changed meaningfully, and conflating them with the things that haven’t changed is where most SEO strategies in 2026 go wrong. Here is what has genuinely shifted.

Content quality now means information gain, not word count

Google’s Information Gain patent — US20200349181A1 — describes a mechanism that rewards content adding genuinely new information to the web rather than paraphrasing what already exists. The December 2025 Core Update penalised generic content farms heavily, and sites with original data saw a +22% visibility gain after the March 2026 update according to JDM Web Technologies research.

‘Average is the new bad’ has become a phrase used widely in SEO practice — in the age of generative AI where anyone can produce decent content at scale, content that doesn’t offer something a reader couldn’t get from any of the twenty competing articles on the same topic no longer has a clear claim to page-one visibility.

The practical implication is that content strategy has shifted from frequency to depth. One thoroughly researched, perspective-driven article with original data, named frameworks, or first-hand case studies outperforms ten adequately written articles covering the same ground from different angles.

This is not a new principle — Google has been signalling it for years — but the March 2026 update made it a hard enforcement rather than a soft preference.

E-E-A-T has become a structural requirement, not a content checklist

E-E-A-T — Experience, Expertise, Authoritativeness, and Trustworthiness — has existed as a Google quality framework since 2018. In 2026, it functions differently. Google’s algorithms now use AI to check author and entity profiles. Brands with a consistent online presence, verifiable author credentials, and third-party editorial coverage are meaningfully more likely to appear in AI Overview results.

Websites with anonymous content, vague claims, or unverified expertise — regardless of keyword optimisation — struggle to appear in SGE results in competitive niches. E-E-A-T has stopped being a checklist item on a content audit and started being the structural foundation that determines AI citation eligibility.

AI Overviews have changed what ‘ranking’ means for informational queries

For informational queries — ‘how to’, ‘what is’, ‘why does’ — Google AI Overviews frequently answer the question before a user ever reaches the organic results. Ahrefs found that AI Overviews reduce CTR for position one by up to 58% for these query types.

This does not mean SEO for informational content is pointless — it means being cited inside the AI Overview has become more valuable than ranking first below it. Pages appearing inside AI Overview citations see their brand impression volume increase even when click-through volume decreases. Visibility and traffic are no longer the same metric.

Brand signals now feed two systems simultaneously

Third-party brand mentions — press coverage, industry publications, Quora, Reddit, podcasts, review platforms — have always mattered for off-page SEO through the backlink signal they often carry. In 2026, they serve a second function: they feed LLM training data and AI retrieval systems.

Airops research found that brands are 6.5 times more likely to be cited in AI answers through third-party sources than through their own website content alone. SE Ranking research from November 2025 found that domains with brand mentions on platforms like Quora and Reddit have roughly 4x higher chances of being cited by ChatGPT.

The same outreach effort that earns a press mention is now simultaneously building Google authority through backlinks and AI visibility through training data coverage.

What Has Not Changed — And Why That Matters

Which traditional SEO fundamentals still determine rankings in 2026?

In the noise around AI search, several foundational SEO factors have been prematurely declared obsolete. The data does not support that conclusion, and businesses that have abandoned these fundamentals on the basis of speculative commentary have paid for it in declining rankings.

Backlinks still matter — but quality has replaced quantity entirely

Backlinks dropped from 15% to 13% weighting in First Page Sage’s 2025 algorithm analysis — still the third most important ranking factor overall. The shift is not in whether backlinks matter but in what constitutes a valuable backlink. Google’s spam detection has never been sharper, and purchased links from low-authority sources now actively hurt rankings rather than simply failing to help them.

A single editorial mention from a high-authority domain in the relevant niche is worth more than dozens of directory links. Semrush research found that AI Overviews show approximately 86% domain overlap with traditional Google search results — meaning the sites that rank on Google largely are the sites appearing in AI Overviews. Building genuine backlink authority remains foundational to visibility in both systems.

Technical SEO is the prerequisite that nothing else overrides

No amount of content quality or backlink authority rescues a page that Google cannot crawl, cannot index, or cannot render correctly. Technical SEO has not changed in importance — it has never been more important, because the cost of a technical failure now extends beyond organic rankings to AI citation eligibility.

Pages that AI crawlers cannot access — either because of robots.txt misconfigurations or JavaScript rendering failures — cannot appear in AI-generated answers regardless of content quality.

Core Web Vitals continue to matter as a ranking differentiator. INP (Interaction to Next Paint) fully replaced FID as the primary interactivity metric — teams that optimised for FID and assumed the work was done need to re-audit their INP scores separately, as the metrics measure different interactions.

Pages that combine solid technical foundations with clean, low-friction user experience consistently outperform technically compromised pages of equivalent content quality.

Search intent alignment remains the primary relevance filter

Google’s AI systems now interpret search intent at a granular level — evaluating context, problem depth, decision stage, and expected outcome. Two pages covering the same keywords can rank wildly differently based on how accurately each matches the intent behind the query.

A blog post optimised for an informational keyword will lose its ranking if the SERP has evolved to show primarily transactional pages — because the intent has shifted even if the keyword has not. Monitoring SERP composition alongside keyword position remains essential for catching these shifts before they produce unexplained ranking declines.

Internal linking structure continues to compound authority

Internal linking — the deliberate architectural connection of pages within a site — continues to be one of the highest-leverage, lowest-cost SEO actions available to any business. Well-structured internal linking distributes link equity toward commercial pages, builds topical authority clusters that Google’s quality systems reward, and ensures every published page has a clear path through which crawlers can discover and re-evaluate it.

The work is unglamorous and rarely headline-grabbing, but across client work the SEO team at DIGITALOPS has consistently found that internal linking restructuring produces among the fastest ranking improvements of any intervention — because it improves performance using authority that already exists on the site.

SEO in the Age of AI: What Has Changed vs. What Remains Foundational

How to Win Visibility Across Both Google and AI Systems

What is the practical SEO strategy for ranking in Google and getting cited by AI in 2026?

The businesses that are building durable visibility in 2026 are not choosing between Google optimisation and AI optimisation. They are building content that performs well across both — and the overlap between what those systems reward is larger than the differences. Here is what that looks like in practice.

Write for Information Gain, not keyword density

Every piece of content published should answer one question before going live: does this add something to the web that readers could not get from the twenty other articles on this topic? If the answer is no, the article is competing for a position it cannot justify holding.

Original data, proprietary frameworks, first-hand case studies, expert analysis that goes beyond surface synthesis — these are the content characteristics that both Google’s quality systems and AI citation mechanisms reward consistently.

Between 65% and 85% of ChatGPT prompts do not match any traditional search keyword, meaning content written in natural conversational language that addresses the questions people actually ask — rather than the phrases they type — reaches audiences that keyword-optimised content cannot.

Structure content explicitly for AI extraction

AI systems cite content that is easy to extract and reproduce as a direct answer. The practical requirements are: opening paragraphs that give the answer first before elaborating, subheadings framed as natural questions rather than topic labels, clear self-contained sections that can be quoted without requiring the entire article for context, and FAQ sections structured with explicit question-and-answer pairs.

This structure simultaneously improves Google Featured Snippet and People Also Ask box performance — it is one of the rare content optimisations that serves every discovery channel simultaneously.

Deploy schema markup across all key page types

Schema markup does what prose content cannot: it gives AI systems and search engines unambiguous, structured information about what a page contains. FAQPage schema makes Q&A content explicitly machine-readable. Article schema with author credentials feeds E-E-A-T signals.

Speakable schema marks the passages a voice assistant or AI should extract. Organisation and LocalBusiness schema establishes entity signals that build brand recognition across both Google’s Knowledge Graph and AI training data. These are not large technical investments — implemented once, they continue paying dividends across every system that uses structured data for content evaluation.

Build topical authority clusters, not isolated articles

A single well-written article about SEO builds limited authority. A network of twenty deeply interconnected articles covering SEO from multiple angles — each linking back to a central service page, each cross-linking to related cluster content, all demonstrating depth of expertise across the topic — builds the kind of topical authority that Google’s Niche Expertise signal rewards and that AI systems recognise as domain ownership.

The pillar-cluster architecture is not a content format preference — it is an authority structure that compounds in value every time a new cluster page is published and properly connected.

Earn third-party citations — they feed both systems

The same digital PR and outreach effort that earns editorial backlinks for Google authority is simultaneously building the third-party brand presence that feeds AI training data and retrieval systems. Domains with profiles on review platforms like G2, Clutch, or Google Business have 3x higher chances of being selected by ChatGPT as a source according to SE Ranking research.

Press coverage in recognised industry publications, expert commentary in roundup articles, and consistent presence in online communities relevant to the business all build the off-site brand signal that AI systems use as a proxy for trustworthiness.

Priority Framework: Where to Focus SEO Effort in the Age of AI