SEO is not dead. Google still controls around 90% of all searches worldwide and sends 345 times more traffic than ChatGPT, Gemini, and Perplexity combined, based on analysis of approximately 82,000 sites. What has changed is the shape of the game: AI Overviews now appear in roughly 16% of US searches (more than doubling since early 2025), average search traffic dropped 21% year-over-year across the web, and 45% of Gen Z users now search on TikTok and Instagram alongside Google. These shifts are real, but they do not make SEO irrelevant. They make it harder to coast on mediocre content and easier to win for brands that build genuine authority.
The businesses still seeing strong organic growth are not ignoring AI. They are adapting. Good SEO today still drives rankings on Google, but the same work, done well, also earns citations in AI Overviews, ChatGPT, Perplexity, and Gemini. The signals overlap more than most practitioners realise: 76% of pages cited in AI Overviews already rank in Google’s top 10. Ranking well on Google remains the most reliable path to appearing in AI answers.
The real question is not “Is SEO dead?” but “What does SEO have to become?” The answer is a discipline focused on topical authority, structured information, and verifiable brand presence across the web, not just keyword density and backlink counts.
Does SEO Still Drive Traffic?
Yes, but the volume per query has compressed. AI Overviews reduce clicks by an estimated 34.5% for the queries they appear on, and average search traffic across sites tracked by Ahrefs fell 21% in the past year. But Google’s total search volume is still enormous and growing. Sites with strong organic visibility also receive more mentions in AI chatbot answers, because AI systems ground their responses in the same high-authority, frequently-cited pages that rank well organically. The traffic erosion is real; abandoning SEO in response to it would accelerate the loss.
A useful comparison: 78% of businesses in a Semrush survey of 700+ marketers described SEO as extremely or very valuable, and 62.9% reported SEO improvements in the second half of 2024. Only 9.5% saw traffic declines. These are not the numbers of a discipline that has stopped working.
What Has Actually Changed
Several shifts are genuine and practitioners should not downplay them.
AI Overviews compress informational clicks. When Google surfaces a direct answer in the search results page, fewer users click through to source sites. AI Overviews now appear in 16% of US searches, a 116% increase since early 2025. For purely definitional or factual queries (“what is compound interest”, “how do you treat a sprained ankle”), click-through rates are lower than they were three years ago.
Content saturation has raised the bar. Approximately 74% of all new web content is now created with generative AI. Only 13.5% of pages ranking in the top 20 for competitive keywords are purely human-written, according to Ahrefs analysis of 100,000 keywords. This does not mean AI content cannot rank; it means the volume of passable content has exploded while the volume of genuinely useful, original content has not kept pace. Differentiation through depth and real-world experience matters more than ever.
Platform diversification is real, not hypothetical. About 45% of Gen Z users search on TikTok and Instagram as part of their discovery behaviour, with 35% of millennials having made similar shifts. These users are not abandoning Google for product research; they are using social search for discovery and inspiration, then converting through other channels. SEO strategies that ignore this miss a meaningful part of the funnel.
AI traffic converts unusually well. Ahrefs found that AI Overview traffic generated 12% of signups while accounting for only 0.5% of total traffic, and that AI-referred traffic converts 23x higher than organic search. The implication: a user arriving via an AI answer is much further along in their decision process than a typical organic visitor. This makes AI-citation work commercially valuable even when the raw traffic numbers look small.
The Google-AI Citation Connection
The most important structural fact in this debate: the same work that earns Google rankings also earns AI citations.
Perplexity, ChatGPT’s web search mode, and Google’s own AI Overviews all pull from indexed, crawlable, high-authority web content. Ninety percent of pages retrieved by SearchGPT appear in Google’s first-page results. The correlation between brand mentions across the web and AI Overview citations sits at 0.664, a strong relationship. Brands in the top 25% for web mentions get roughly 10 times more AI visibility than brands in the bottom 75%.
The practical conclusion: building topical authority so Google trusts your site creates the same content corpus that AI engines draw from. There is no separate “AI SEO” track that bypasses traditional optimisation. There is, however, additional work specific to AI retrieval: structured data that names your entity clearly, content that directly answers questions with verifiable specificity, and brand-level signals (citations, mentions, reviews) that establish trustworthiness across multiple platforms.
You can track whether AI engines are citing your brand with Fokal and see your AI search visibility across ChatGPT, Perplexity, and Google AI Overviews in one place.
What “Modern SEO” Actually Looks Like
The practices that have always separated good SEO from bad SEO are now the only SEO that works. There is less room for technical tricks and keyword stuffing because AI-generated content has flooded the low-effort tier.
Topical authority over individual rankings. Google and AI engines both favour sites that cover a topic comprehensively rather than publishing one opportunistic page. A site with 30 well-linked articles on a single topic will consistently outrank a site with one page on that topic, regardless of backlinks. See topical authority for how this clustering works.
Entity-based signals. Google’s Knowledge Graph and AI engines both use named entities to understand what an organisation is, what it offers, and whether it is credible. Claiming your Google Knowledge Panel, structuring your organisation schema, and maintaining consistent NAP (name, address, phone) data all signal entity clarity. A site with a clear, verifiable entity profile appears more often in AI answers.
Structured, directly answerable content. AI Overviews tend to pull from pages that provide a clear, concise answer to a question, then expand. Structuring content with descriptive headings, concise lead sentences, and specific data points rather than winding introductions improves both featured snippet selection and AI citation rates. Explore AI Overview optimisation and how to get cited by AI for tactical guidance.
Brand mentions beyond your own site. The 0.664 correlation between brand mentions and AI citation rates points to a clear lever: getting named in reviews, directories, press coverage, and community discussions builds the signal network that AI engines use to validate a brand. Reddit alone saw a 287% growth in AI citations over three months, with 12.8 million Reddit mentions captured in AI outputs. Being present and cited in credible third-party contexts matters.
The Jobs Argument
One frequently cited piece of evidence for “SEO is dead”: SEO job listings dropped 37% year-over-year in Q1 2024. This is accurate, sourced from Semrush data, but it reflects the consolidation of AI-assisted workflows rather than the elimination of SEO as a discipline. The number of people needed to publish mediocre content has fallen. The value of an expert who can direct AI output, build topical clusters, and interpret data has not.
Google’s market capitalisation grew from $1.76 trillion to $2.14 trillion during 2024, a 22% increase that coincided with peak AI adoption. A company whose core product was dying would not behave that way. Alphabet’s advertising revenue, which depends on search clicks, continued growing through 2024. These are not vanity metrics; they are the direct financial expression of search still driving commercial intent at scale.
What to Actually Do
The “SEO is dead” framing is a distraction from the real strategic question, which is whether a brand is building the kind of authority that works across both traditional search results and AI-generated answers.
The overlapping checklist:
- Publish substantive, original content on topics your audience cares about, structured for direct answers, not just keyword targets
- Build internal linking between related articles to signal topic depth to both Google and AI crawlers
- Maintain schema markup that names your organisation, products, and services precisely
- Earn citations on third-party platforms: reviews, directories, trade publications, and community discussions
- Monitor AI Overview appearances and track which of your pages are being cited
For brands that operate in competitive categories, the difference between appearing in AI answers and being invisible to them comes down to the same fundamentals that determine organic rankings. The work is not new. The urgency is.
Read the AI search research hub for the underlying research, or explore how AI search works to understand why Google and AI engines reach for the same sources.