Does SEO still matter now that buyers start with AI?
Yes. AI engines ground their answers in crawled web content, so the crawlability, structure and credibility work that drives SEO is the same work that gets you cited in AI answers. What changes is the shape of the pages you publish and how you measure success — not whether the underlying fundamentals still apply.
The honest version
"SEO is dead" is a marketing line, usually from someone selling the replacement. The accurate statement is narrower: the click is at risk, not the content. When an engine answers a question directly, the user may never visit the page it learned from — but it still had to find, read and trust that page. The page did the work; it just got paid in a citation instead of a session.
For an industrial manufacturer this is less alarming than it sounds, because your traffic was never the point. A $2–25M specialty manufacturer does not monetize pageviews. It needs to be on the shortlist when a buyer is choosing who to RFQ. A citation inside an AI answer serves that goal at least as well as a click, and arguably better — it arrives with an implicit recommendation attached.
What still works exactly as before
- Being crawlable. If bots cannot fetch and parse it, nothing downstream matters. This includes not blocking AI crawlers by accident.
- Structured data. Schema markup is arguably more valuable now, not less — it is unambiguous machine-readable meaning, handed over for free.
- Being referenced by credible third parties. Directories, associations, trade publications, distributor listings. The engine reads those pages.
- Technical hygiene. Fast pages, clean HTML, working canonicals, an accurate sitemap, content that is in the HTML and not injected by JavaScript.
- Genuine subject expertise. Specificity — real tolerances, real materials, real failure modes — is what makes a page worth quoting.
What changes
Page shape
The classic SEO instinct was length and comprehensiveness: cover everything, rank for many terms. The GEO instinct is resolution: answer one question cleanly and early, then elaborate. You will publish more pages, each narrower.
Measurement
Rank tracking degrades as a proxy. Position 3 for a term whose result page is now topped by a synthesized answer is worth much less than it was. You need a second instrument alongside it: a fixed set of buyer prompts, re-run on a schedule across several engines, logging who is named and what got cited. Both instruments, not one replacing the other.
Where the work lands
A meaningful share of the effort moves off your own site, onto the third-party pages the engines cite. That is uncomfortable for anyone used to controlling their own domain, and it is where a lot of otherwise-good content programs stall out.
Do not rebuild your SEO. Add the missing layer: find the questions your buyers ask, publish clean answers to them, and check what the engines currently say. Most industrial sites already have the hard part — real technical depth — locked in formats machines cannot read.
Practically, for a manufacturer with no marketing staff
You do not need two programs and two budgets. The overlap is large enough that one content effort, built answer-first and marked up properly, serves both. The additional GEO-specific work is the prompt audit and the third-party placement — a layer on top, not a parallel track.
Find out who gets named in your niche
The $500 audit runs 15 real buyer queries across ChatGPT, Perplexity, Gemini and Google AI Overviews, logs who gets cited query by query, and comes back with a 90-day content map and two ready-to-publish articles. Fixed scope, two weeks. The fee is credited to your first retainer month if you continue.
Start the $500 auditAsk a question