"Everyone searches with AI now, so why bother with SEO?" It sounds reasonable. It gets the causality exactly backward.
Before AI answers anything, it uses the search engine's own ranking system to pull a pool of candidate pages, then extracts an answer from them. If your site can't even get crawled and indexed by Google, AI has no idea you exist — let alone cites you. SEO wasn't replaced. It became the gate that comes before everything else.
In other words: the battlefield moved from "what rank am I" to "am I written into that answer" — but the ticket to even get on the field hasn't changed.
01What died is the playbook, not SEO
Customers used to type keywords into Google, scroll page after page, and compare vendors one by one. Now they ask ChatGPT, Perplexity, or Google's AI summary directly: "which company's reliable for this kind of service?" AI gives them a paragraph, drops a couple of names, and the conversation is over. No page two. No scrolling.
The entry point really has changed: it used to be a row of blue links, now it's a paragraph AI already wrote for you. So the old playbook — chase rankings, stuff keywords, buy a pile of backlinks — has been paying off less and less.
But that doesn't mean SEO is obsolete. It means SEO's target moved up a level: you're no longer just optimizing for rank position — you're optimizing for whether AI treats you as a credible source worth taking. The foundation is still the same foundation. The bar for passing just got higher.
02It didn't disappear. It stacked into four layers.
The industry consensus now is that SEO evolved into four stacked layers — each one building on the last, not replacing it:
| Layer | The question it asks | What you're managing |
|---|---|---|
| Traditional SEO | Can you be found? | Crawling, indexing, Core Web Vitals, keywords |
| AEO | Are you treated as the answer? | Structured content, citation mechanics, clear entities |
| GEO | Are you recommended by generative AI? | Question maps, evidence chains, cross-platform presence |
| AIO | Are you there when AI makes the decision for a customer? | The combined result once the three layers above are in place |
The key word is "stacked." Every layer above is built on the one below it: skip traditional SEO, and the three layers above it are moot. This is exactly why Google has said, plainly, on the record: the substance of AI search optimization is still part of SEO. Traditional SEO is still the entry ticket.
The industry can keep arguing over whether GEO is just SEO in a new coat. Argue all they want — both sides land on the same bottom line: content that's genuinely useful, backed by verifiable evidence, is the only real long-term answer. The terminology changes. The foundation doesn't.
03The ticket got harder to earn, not easier
Plenty of people assume technical SEO gets to relax in the AI era. It's exactly the opposite. AI crawlers are pickier than Google's, and the bar only goes up, never down. There are three technical checkpoints, and every one of them just got stricter:
| # | Checkpoint | The new AI-era requirement |
|---|---|---|
| 1 | Crawlable | Allowing Google isn't enough anymore. If robots.txt blocks AI retrieval crawlers like OAI-SearchBot or PerplexityBot, you disappear from AI results entirely |
| 2 | Readable | AI crawlers almost never execute JavaScript. If your site relies on client-side rendering, it can't read a single word of your key content — SSR or pre-rendering went from "nice to have" to "the price of entry" |
| 3 | Fast-loading | INP (interaction responsiveness) is the new checkpoint, and it needs to stay under 200ms. The #1 killer is third-party scripts: analytics tools, ad SDKs, chat widgets |
Speed doesn't just affect the experience — it converts directly into business outcomes: three-quarters of users leave if a page hasn't finished loading in 4 seconds; cutting Largest Contentful Paint from 4 seconds to 2 can lift organic ranking by roughly 40 percent; every additional second of page load costs roughly 7 percent in conversion.
These numbers used to fall under "nice to fix eventually." Now they're "skip this and you're out."
04Having the ticket doesn't guarantee you make the cut
Passing the technical checkpoints only gets you onto the candidate list. Whether AI actually writes you into that answer depends on something further up the chain.
First is E-E-A-T: experience, expertise, authoritativeness, trust. AI favors citing content with first-hand experience, clear authorship, mentions from peers in the field, and a trustworthy site overall. For small businesses, this is exactly your strength: your real cases, your pricing logic, the mistakes you've actually made and learned from — that's precisely the material that isn't sitting in public data anywhere else, and exactly what AI most wants to cite.
Second is structure. Organize your site around topic pillars: one deep pillar page per core topic, with a cluster of focused articles underneath it, all internally linked, so both search engines and AI can clearly see "this site is the authority on this topic." Scattered, unrelated posts never build that kind of recognition.
Technical work gets AI to find you. Content and structure get AI to trust you. Miss either piece, and you're stuck at the door of the candidate list.
05Don't spend your effort in the wrong place
SEO evolved, but some old habits are just as wasteful as the new snake oil. When we audit for clients, these are the three things we tell them to hold off on first:
- Don't chase big industry keywords, blast out blog posts, or rush to buy backlinks first. Confirm the underlying pipeline works: are your core pages actually indexed, does every page own the keyword it's supposed to, can your product pages actually answer a buyer's real question.
- Internal links before external ones. Internal linking is something you can finish today, controllable and safe. Earn backlinks deliberately, over your second and third months. Remember: five high-quality backlinks from real sites in your industry beat 500 junk links.
- Don't buy the snake oil. Google has already called out a list of useless tactics: forcing an llms.txt file that does nothing, chopping long content into random fragments, spinning up a separate page for every long-tail keyword string, faking brand mentions, inventing "AI-exclusive schema" that doesn't actually exist. Save that effort and put it into your fundamentals instead.
SEO was a race where ranking higher meant more traffic.
SEO is the ticket to whether you make it into the AI's answer at all.
Where this lands
SEO isn't dead. What died is the one-shot playbook where hitting page one meant you won. It evolved into a ticket that's harder to earn, and worth more: technical work gets AI to find you, content earns AI's trust, structure tells AI exactly who you are.
We handle brand, web, SEO, and AI workflows — from strategy through launch, no subcontracting, no dropped handoffs. Get that ticket earned properly, and the moment a customer asks AI, you're actually eligible to show up in the answer.
Rooted in craft. Built for the new wild.