Artificial intelligence changed SEO on two fronts. In the result: Google now answers many searches with an AI-generated summary at the top, before the links. And in behavior: a share of people stopped searching on Google and ask ChatGPT, Perplexity, or Gemini directly. The foundation of SEO still holds, but the goal has widened. Ranking is no longer enough; now you also have to get cited.
What Actually Changed in Google’s Results?
The most visible change is AI Overviews, the summaries Google places above the traditional results, powered by its Gemini model. Instead of showing you a plain list of links, Google synthesizes an answer from several sources and presents it up top, with citations to the pages it used.
For your site, this has a concrete consequence. If your page is one of the cited sources, you gain visibility even without being the number one result. If the answer resolves the question completely, a portion of users do not click anything at all. That is why content that only restates the obvious loses traffic, while content that adds something the summary cannot cover keeps it.
Did Traditional SEO Stop Working?
No. That is the most common misunderstanding. The foundations have not moved:
- A fast, crawlable, well-structured site is still the basis for everything.
- Content that answers the real intent behind a search still wins.
- Authority, meaning that other trustworthy sites link to you and mention you, is still decisive.
AI did not repeal these rules. It made them stricter. When a model decides whom to cite, it rewards exactly the clarity, precision, and reputation that good SEO has always pursued.
What Did Change: From Ranking to Being Cited
Success used to be measured by your position in a list. Today it is also measured by whether an AI uses you as a source when it builds its answer. The goals are similar, but not identical.
A page can sit at position three on Google and still not get cited in the AI Overview, because its information is hard to extract. And the reverse holds too: a clear page, with concrete and well-bounded facts, can get cited even when it does not top the list. The model is looking for the snippet that answers, not the biggest domain.
How Do You Write So AI Understands You?
The underlying question has changed. You used to optimize for an algorithm that ranked links. Now you also write for a model that reads, summarizes, and attributes. That favors a different style:
Answer the question right away, with no preamble. Use headings that match what people actually ask. Give numbers, conditions, and steps instead of generalities. Structure the information so each part stands on its own. A verifiable, well-bounded fact is material an AI can lift with confidence.
So What Does GEO Add?
This is where GEO comes in, generative engine optimization. Classic SEO prepares you for traditional search engines. GEO prepares you for the engines that generate answers: Google’s AI Overviews, ChatGPT, Perplexity, and Gemini.
GEO is not a replacement. It is a layer that sits on top of solid SEO. It shares most of the foundations (a healthy site, useful content, real authority) and adds what is specific to the AI era: extractable content, structured data that reduces ambiguity, and brand presence in the sources the models consult. We cover it in depth in our guide to GEO.
Where Should You Start?
If your base SEO is weak, start there: AI rewards the same fundamentals, so that work is never wasted. If you already have a solid base, the next step is to adapt your content so it is citable and to build presence where the models look.
The important thing is not to panic and not to throw out what works. AI changed the rules of the game, not the game itself. If you want a read on your situation and an orderly plan, that is how we approach each case in SEO consulting.