SEO isn't dead, it's becoming three jobs instead of one

Every six months a new essay declares SEO dead. The argument keeps mutating. In 2023 it was ChatGPT. In 2024 it was AI Overviews. In 2026 it is the disappearance of the informational click. The pattern is the same: someone notices that one slice of search behavior has shifted, then writes the obituary for the whole channel.

What has actually changed is narrower than the headlines. Informational queries that used to send people to a blog post now resolve inside an AI Overview, a Perplexity answer, or a ChatGPT search response. Commercial queries, the ones with money behind them, still resolve the way they always have. Someone searching 'plumber near me' or 'best CRM for agencies' is still clicking a link, comparing options, and converting on a website.

So the obituary is too broad. The skill that needs to evolve is the part of SEO that targeted informational keywords as a top-of-funnel play. That work used to be one discipline. It is now three.

Classical SEO still owns the commercial and navigational queries that drive direct revenue. The fundamentals have not moved. Title tags, internal linking, page experience, backlink quality, content depth, semantic HTML, schema for rich results, crawl budget management on big sites. The Lighthouse score still matters. Indexing still matters. The Search Console performance report still tells you most of what you need to know about traffic at the URL level.

If anything, classical SEO is easier to win at now because the field is thinning. Plenty of teams pulled budget out of SEO when AI Overviews arrived and have not been replaced. The competitors who stayed and adapted are picking up rankings on the queries that still send clicks. Median time-to-first-page-one ranking on tracked client sites in 2026 is around eleven weeks for medium-competition commercial terms. That is faster than 2023.

AEO is the new top-of-funnel. If the user is asking an informational question, the answer is happening inside an AI surface. Your job is to be the source the model cites. That changes what 'good content' looks like.

Long, narrative blog posts that ranked well in 2021 are not what AI engines want to extract. They want self-contained answer blocks of forty to eighty words placed under question-form headings, with the entity named explicitly, the concept defined cleanly, and the answer standing alone without surrounding context. We engineer every long-form piece to lead with a TL;DR block sized exactly for citation.

We track AEO performance with monthly snapshots from the DataForSEO ChatGPT scraper, manual prompt panels across fifty target queries per client, and citation diffs that surface what is new month over month. Median lift on tracked client engagements at the six-month mark is plus thirty-eight citations per month from a baseline of near zero.

Generative Engine Optimization sits next to AEO but is not the same discipline. AEO is about being extracted into an answer. GEO is about being recommended inside a generative experience: a ChatGPT response that names you as a vendor, a Perplexity answer that includes you in a shortlist, a Claude response that suggests your product as a fit for the user's stated need.

The mechanics are different. GEO leans heavily on entity strength, sameAs links to LinkedIn, GitHub, Crunchbase, and Wikidata, third-party reviews that are themselves crawlable, and brand mentions in places models trust as authority. A page can be perfectly optimized for AEO and still not show up in a 'best tools for X' GEO answer if the surrounding entity graph is weak.

Most agencies still treat AEO and GEO as the same thing. They are not. AEO answers questions. GEO answers recommendations. The content that wins one does not automatically win the other.

The practical move is to stop calling it 'SEO budget' and start tracking three sub-budgets. Classical SEO still gets the largest share for most B2B and local businesses, because that is where the commercial intent sits. AEO and GEO together get the rest, weighted toward AEO if the business sells to people who research before buying.

The shared infrastructure across all three is real and worth pricing in once. Schema work, fast and crawlable HTML, semantic structure, internal linking, entity definition. Build that foundation once and all three disciplines benefit. The discipline-specific work, classical content depth versus AEO passage citability versus GEO entity strengthening, gets layered on top.

We treat the three jobs as one engagement with three measurable tracks. Every monthly report at FPWS now has three sections: classical organic performance from Search Console, AEO citation performance from our DataForSEO and manual prompt tracking, and GEO recommendation performance from a smaller panel of vendor-shortlist queries we run by hand each month.

Clients who used to ask 'why is traffic flat' are now asking 'where is the conversion coming from' and getting answers. AI-referred traffic is harder to attribute because the referrer headers are inconsistent, but UTM hygiene, direct-detection heuristics, and self-reported attribution in form fields close most of the gap.

SEO is not dead. It just stopped being a single job. Staff and measure each track separately or you will keep wondering why the line on the chart is not moving.