SEO is not dying; it is splitting. The single Google results page is becoming three surfaces: classic blue links, AI Overviews and answer engines, and LLM chat answers. Each rewards different work. The sites losing ground treat them as one channel. The sites winning produce content each surface can rank, cite, or quote.
This matters because most SEO advice still optimizes for a results page that fewer buyers see whole. A founder searching "how much does a fractional CMO cost" might read an AI Overview, ask ChatGPT a follow-up, and click one blue link across the entire journey. Your content either shows up on all three surfaces of that journey or it competes for a shrinking slice of one.
We run this playbook on our own site: the page you are reading is built the way it recommends. What follows is the surface-by-surface breakdown, what has not changed underneath, and a this-quarter checklist.
Three surfaces, three different games
Each surface has its own selection logic. Blue links reward depth and authority accumulated over time. AI Overviews reward passages a machine can lift cleanly. LLM chat answers reward being a recognized entity with citable material. Optimizing one does not automatically earn the others, but the work overlaps more than most teams assume.
| Surface | What ranks (or gets cited) there | What to change |
|---|---|---|
| Classic blue links | Intent-matched pages with real depth, site-level topical authority, earned links, clean technical foundations | Less than you think. Keep building intent-matched pages, internal link structure, and page speed — this surface still carries most commercial-intent queries |
| AI Overviews & answer engines | Self-contained quotable passages, answer-first structure, FAQ and Article schema, clear headings a machine can parse into an answer | Restructure pages so the first paragraph answers the query in 40–60 words and every section stands alone; add FAQPage schema; write for extraction as much as for reading |
| LLM chat answers (ChatGPT, Perplexity, Claude) | Recognized entities with consistent identity signals, crawlable content, llms.txt, original definitions and data worth citing by name | Build entity clarity: consistent org and author markup, an llms.txt file, published frameworks and terms the models can attribute to you |
Note the asymmetry: the first column is the same content strategy wearing three outputs. A genuinely authoritative page, structured answer-first, with clean schema and clear entity signals, is eligible on all three surfaces at once. You do not need three content programs; you need one program built to a higher structural standard.
The zero-click reality, briefly
A growing share of searches now end without a click, because the answer appears on the results page itself. We have made the full argument, with the demand-gen consequences, on the zero-click page, so this section will not re-litigate it. The operational summary: visibility and traffic have decoupled, and a page can win the query while losing the session.
What that means for SEO planning is a measurement change, not a surrender. Track impressions and citations alongside sessions, watch branded search as the downstream signal of answer-surface presence, and judge the program on pipeline contribution rather than on traffic alone. A buyer who read your answer in an AI Overview and searched your brand two weeks later is an SEO win no sessions report will credit.
What stays true
Underneath the split, three fundamentals have not moved: search intent, real authority, and technical hygiene. Every surface (link, Overview, or chat answer) is still trying to hand a person the best available response to what they meant. The selection mechanics changed; the selection criteria mostly did not.
Intent still decides what kind of page can win. A commercial-intent query wants a pricing breakdown, not a think piece, whether the response arrives as a link or a generated paragraph. Authority still decides who is trusted: models and ranking systems both lean toward sources with a track record on the topic, which is why thin sites do not get rescued by schema. Technical hygiene still gates eligibility: a page that cannot be crawled, or takes seconds to render its content, is invisible to every surface equally.
If your current SEO program is built on these three, the split is an extension of your work, not a demolition of it. If it is built on volume, pumping out keyword-stuffed pages a machine can now out-generate for free, the split is the end of the road, and no amount of AI-surface tactics will compensate.
What to do this quarter
Four moves, in priority order. None require new tooling; all require editorial discipline.
Rebuild key pages answer-first
Take your ten most important pages and restructure each so the opening paragraph answers the primary query in 40–60 words, directly and completely. Then make every H2 section open with a sentence that could stand alone as a quote. This single change serves all three surfaces: readers get the answer faster, Overviews get a liftable passage, and chat models get something attributable.
Ship schema that matches the page
Article, FAQPage, and BreadcrumbList on editorial pages; Organization and Person markup with stable @id references across the site. Schema does not create authority, but it removes ambiguity about what a page is and who stands behind it. Ambiguity is what keeps eligible content out of generated answers. The tactical detail lives on how to rank in AI Overviews.
Establish the entity: llms.txt and consistency
Publish an llms.txt file describing who you are, what you do, and which pages state it canonically. Make your organization name, founder, and service descriptions consistent everywhere they appear: site, profiles, directories. Chat models assemble answers from entity understanding; a business that describes itself five different ways fragments its own citations.
Write passages worth citing
Machines cite what is quotable: a definition sharper than the consensus, a framework with a name, a number from your own operation with the context attached. Generic summaries get paraphrased without credit; original material gets attributed. One genuinely citable passage per page beats a thousand words of restated common knowledge — this is the discipline answer engine optimization formalizes.
One production note. AI now drafts much of the SEO work above (the pages, the schema, the audits), which compresses what used to be a content team into a supervised pipeline. The judgment does not compress: deciding which pages deserve to exist, and standing behind every claim on them, stays with a person. That trade is how one operator runs a program of this scope, and it is the same shift reshaping every discipline covered on the future of marketing hub. What a program like that costs, run as part of a full growth engagement, is on services.
Frequently asked questions
Is SEO dead?
No, but the version of SEO that assumed every search ends in a click is dead. Search demand has not shrunk; the results page has split. Classic rankings still exist, AI Overviews now answer many queries above them, and LLM chat tools answer some questions with no results page at all. Visibility and traffic have decoupled: you can be more visible than ever and get fewer sessions. SEO now means earning presence on three surfaces (links, AI answers, and chat citations) and measuring branded demand and pipeline rather than sessions alone.
What is the future of SEO?
The future of SEO is one query stream splitting into three surfaces: classic blue links, AI Overviews and answer engines, and LLM chat answers. Each rewards different work: depth and authority for links, quotable self-contained passages and schema for AI answers, entity clarity and citable original material for chat. The fundamentals carry over: search intent, real authority, and technical hygiene still decide who is eligible on every surface. What changes is the deliverable (content built to be quoted and cited on top of being ranked) and the measurement, which shifts from sessions to presence and pipeline.
Will AI replace SEO?
AI is replacing parts of the SEO workflow, not the discipline. Draft production, technical audits, brief writing, and metadata passes are being automated, which means SEO programs need fewer production hands. But AI search increases the need for SEO judgment: someone has to decide which pages deserve to exist, make content citable, structure entities and schema, and verify that what gets published is accurate. The discipline is shifting from producing pages to earning citations, and the practitioners who direct AI production while owning that judgment layer are the ones AI makes more valuable, not less.
How is SEO changing with AI Overviews?
AI Overviews answer the query above the classic results and cite a handful of sources, which changes SEO in three ways. First, clicks concentrate: many informational queries resolve inside the answer, so ranking first no longer guarantees traffic. Second, being cited becomes a distinct goal from ranking: it favors self-contained, quotable passages, clean schema, and clear entity signals over long unstructured pages. Third, measurement shifts: impressions, citations, and branded demand matter alongside sessions. Pages built answer-first, with each section quotable on its own, are the format AI Overviews reward.