SEO basics in the AI search era: what changed?

Even in the AI search era, the skeleton of SEO stays the same. What changed is the shape of search results and how citable content is; understanding search intent, quality signals, and technical basics still drive both rankings and citations. You often hear that "AI changed everything, so SEO is now an outdated concept," but when you break down how AI engines pick content and build answers, it isn't very different from the problems SEO has dealt with for a long time.

SEO (Search Engine Optimization) was never really "a technique for raising rankings"; it was "the process of creating content that matches what searchers intend." AI search didn't overturn that definition. It only changed how the results get consumed. So in this post, we'll separate what actually changed from what's still the same.

3 things that changed in the AI era

1. The shape of the search results page (SERP)

As Google AI Overviews take the top of search results, a growing share of users get their answer without clicking a link ^[2]. The old formula of "rank 1 = traffic" used to hold to some degree, but now a paradox can occur: when the content of the top-ranked page is summarized and shown in an AI Overview, clicks can actually go down.

What this change means isn't that rankings have become meaningless, but that the link between rankings and traffic isn't as simple as it used to be. Rankings are still an important input in deciding which sources AI refers to.

2. Content "citability" emerges as a new metric

When AI engines build answers, they prefer content with clear definitional sentences, specific numbers, and trustworthy sources. The old approach of simply repeating keywords makes it hard to get cited, and a 2023 joint study by Princeton and IIT Delhi showed empirically that content with added citations, statistics, and sources is chosen more often in generative engine answers ^[1]. That study is where the concept of GEO (Generative Engine Optimization) began.

3. More search channels

Unlike the days when Google was all you watched, AI search platforms like ChatGPT, Perplexity, and Gemini have now become real search channels. That's why SEO practitioners need to widen their view from "optimizing for Google" to "optimizing for AI search as a whole."

3 things that haven't changed in the AI era

1. Understanding search intent is the starting point

Understanding what users want when they search is still the first step of SEO. Search intent is usually split into four types (informational, navigational, transactional, and commercial investigation), and that classification still holds in AI search. Questions people ask AI ultimately start from one of the same four intents.

2. The content quality standard: E-E-A-T

Google points to Experience, Expertise, Authoritativeness, and Trustworthiness as the core criteria for evaluating content ^[3]. These criteria work the same way in AI search. AI engines also tend to cite authoritative sources ahead of content with unclear origins.

3. Technical basics

A crawlable site structure, fast loading, mobile optimization, and correct structured data are a foundation that holds no matter how search changes. Google's "How Search works" shows that the bottom layer, crawling and indexing, works the same way in the AI search era ^[4]. AI engines also end up referring to this index or re-reading the same pages with their own crawlers. A page where crawling is blocked never makes it into AI's candidate pool in the first place.

A map of the core SEO elements

SEO can be split into three broad areas.

The three areas aren't independent. No matter how good your content (on-page) is, slow loading (technical) makes visitors, AI or human, drop off before they get there, and without trust signals (off-page), AI is less likely to pick you as a source.

How SEO and GEO connect

Some see SEO and GEO as competitors, but in practice it's more accurate to see SEO as the foundation and GEO as an extension built on top of it. That's because AI re-reads and cites content that search engines have crawled and indexed. If you build a GEO strategy without SEO basics, you never get into the pool of sources AI refers to, so there's nothing left to optimize.

Conversely, if you do SEO without paying attention to GEO, you can rank high and still be left out of AI answers. So rather than treating the two as separate projects, it's more efficient to manage them together on top of one content quality standard.

Common misconceptions in practice

• "Once AI Overviews appear, SEO is over": In reality, AI Overviews also use content indexed by search engines as sources. A page that isn't even indexed can't appear in an AI Overview.
• "Just place your keywords well": Sentence structures that answer the question directly, meaning answer formats that match search intent, now matter more than keyword density.
• "Technical SEO is the dev team's job and has nothing to do with marketing": Structured data and crawl optimization directly affect whether you get cited by AI, so they've become inseparable from content strategy.

An SEO self-check for the AI era

Once you've separated what changed from what didn't, it's time to check where your site stands now. Go through the items below one by one.

If two or three or more of these five items are missing, shore up the basics of your existing pages before building a new GEO strategy. Crawling and loading speed in particular knock you out of AI's source candidates regardless of content quality, so they're the top priority.

You don't need to rebuild your team structure

When people hear "responding to AI search," it's easy to assume you need a separate team or new tools, but in practice it's often enough for your existing SEO staff to add a few checklist items when they create content. Once habits like putting the definition up front, attaching sources to claims, and never skipping structured data take hold, your chances of being cited by AI rise naturally without a dedicated "GEO specialist." In fact, splitting SEO and GEO into separate jobs easily leads to the inefficiency of creating the same content twice.

Wrapping up

The arrival of AI search didn't throw out the concept of SEO. The shape of search results changed and a new criterion, citability, was added, but the principles of understanding search intent, creating trustworthy content, and building a solid technical foundation remain. What's needed now isn't to abandon SEO, but to add one more layer on top of it: a structure that AI finds easy to cite.

If it's hard to judge by gut feel how often your site is being cited in AI search right now, diagnosing your visibility at the prompt level in TRAIL Search (search.traillabs.ai) can be a starting point.

For related reading, see What is GEO (generative engine optimization)? Which pages to start with, and in what order, What you don't need to do for GEO: checking llms.txt, AI search schema, and AI-only writing styles, and GEO vs. SEO vs. AEO: how they differ and what to do first.

You can continue with Brand entity SEO: keep AI from confusing you with others, What Are AEO and GEO? Preparing Content for AI Search, and ChatGPT optimization: get your brand mentioned in AI chats.