AI Search and GEO Videos

Lightly edited for clarity.

When you use AI in a chat interface, you’ll see that it sometimes searches the web. When it does that it creates a query, retrieves text, and then adds that text into the context window, so that it has richer context to produce a response. Large language models predict language, based on the context they’re given, so connecting a model to the web gives it additional context.

If the retrieved content is clear and factual and complete, then the output from the LLM is usually pretty solid. But if the content has gaps, or it’s outdated or it’s conflicting, then it’ll fill that in with statistically probable tokens, but not necessarily correct ones.

So even when you see fewer people visiting your website, they are still likely consuming that information. But through AI, your content is being extracted and summarized, and recombined and presented to the user. So this means that your website isn’t just a place for people to read, and transact although it will likely continue to be that for years, but your website becomes a source of truth for other systems, that are presenting information to consumers based on your content.

So websites can’t just be brochureware, they become more like an encyclopedia, for the niche that your website covers. If your website’s vague or incomplete, then models are going to guess, and if your site is structured, explicit and authoritative, then models can reference it properly.

And this is also why, a lot of your traditional analytics is starting to feel wrong. Influence in the digital space doesn’t map cleanly to visits, time on site or marketing funnels anymore. AI can shape decisions without a consumer ever visiting your website. And even when they do visit your website, they’re often dropped into the middle or the end of the funnel, and then attribution becomes very difficult.

Lightly edited for clarity.

When you notice fewer visitors coming to your website, your instinctive reaction is that your website matters less. But, part of the problem is that we’re measuring visitors based on clicks, and people are now getting information through AI, and while they may not be visiting your website, your content may still be reaching them, but the process and the way to measure it has changed.

So SEO or search engine optimization, was never really about clicks. It was about people finding your website, and SEO determines whether content is indexed and retrievable, and surfaced when someone goes looking for information. And that interface is changing, but the need has not.

When an AI chat searches the web, it relies on existing search infrastructure to retrieve content, and if your content’s not discoverable, it never enters the context window. So SEO now determines whether your content shows up in AI at all, and all of that happens upstream, ahead of any visits to your site.

So, you see declining traffic and assume your content is losing relevance. But what’s happening potentially, is that instead of a person visiting your site, and reading the content on a page, there’s an agent doing that, extracting parts of that content, combining it with other content, and producing a response. So your content can influence people, without producing visits to your website.

And when people do land on your site, they’re dropped into the middle or the end of the process, and funnels, and attributions start to look really strange in your analytics.

And now this is where GEO, or generative engine optimization starts to come in. It does not replace SEO it extends it. SEO makes content retrievable, and GEO influences how that content is summarized and weighted, and represented once it’s inside a model’s context window.

If SEO fails there’s nothing to optimize for generation. But if generation isn’t considered, the content that is retrieved may have gaps or conflicting information, and the user can receive an answer that is coherent but incorrect.

Ignoring SEO means your content isn’t seen, ignoring GEO means your content may be seen, but not accurately reflected. And this is why the website doesn’t become less important, even though traffic declines. It becomes harder to measure and harder to manage, but also more important.

Lightly edited for clarity.

As AI tools increasingly become the way people ask questions, and make decisions, websites start to play a different role. They’re no longer just destinations for people to visit, they’re reference material for systems that answer on your behalf.

When AI generates a response about a product or your industry, or your expertise it can assemble that response, based on content that it retrieves from the web. And that means that your website is increasingly, being read by machines before it’s being read by people.

And machines don’t infer intent or tone, or credibility the way that people do. They work with what’s explicit. If information is missing or vague or outdated, the system will fill in gaps, using whatever else it can find or make up.

And if definitions conflict across pages, the system will average those. And if details are buried in marketing language, they may not be picked up at all.

So content quality stops being just a branding concern, and websites have to behave more like encyclopedias for their domain. Clear definitions explicit claims, consistent language and up to date facts.

You’re no longer just managing how your brand looks, you’re managing how your brand is represented when someone else, or something else is doing the explaining. And that’s a harder problem than page design, or conversion optimization, because once AI systems become the interface, your website becomes the source, and whatever gaps you leave there won’t stay empty. They will be filled in by the model.

Lightly edited for clarity.

Websites have been primarily for people, so content has been structured around pages and funnels, and search terms. You create pages to rank on specific terms, and to guide people through a journey. And that is still part of the picture, because people will continue to use websites.

But an AI world changes how that content needs to be organized. AI systems don’t read your whole website the way a person might. They pick up snippets and headings and definitions, and question answer pairs, that they can insert into a generated response. Content that’s not structured in a way that makes those pieces obvious, may be overlooked by agents.

And this doesn’t mean that you need to abandon topics and sections, and page hierarchies. It means, you need to rethink how you surface the information, that’s most likely to be used. Pages need clear, question focused headings and direct answer blocks up front, modular sections and bullet lists that have facts or steps, and structured data where appropriate, Schema.org and meta tags and llms.txt.

So machines can interpret the meaning of your content, not just the words. And in practical terms that means that your site’s content, architecture, and on page structure are now part of your visibility strategy.

Traditional SEO encourages you to match a search term. GEO encourages you to match the kinds of questions, AI models are likely to answer. In other words the job’s not just about having good content, it’s about organizing that content, in ways that AI can reliably extract and interpret.

Well organized content makes your expertise machine readable, and increases the chances the AI system will use your material, accurately and frequently. And the content itself still has to be valuable to people. Machines may read first, but humans still decide whether to engage or convert, once they land on your site.

So the organization you adopt has to serve both sides, structured clarity for AI and intuitive navigation for people. And this is the same content that you already have, but it is organized in a way that reflects how it’s being consumed. And that means you need to control the clarity, and the accuracy of your content, at the same time, that you need to be producing much more personalized content, to rank well in GEO.

Lightly edited for clarity.

As people increasingly use AI, traditional website analytics becomes more and more incomplete. The process has changed and the measurement has to change as well. Traditional analytics assumes a human visitor, but when AI tools sit between your content and your audience, that model kind of falls apart.

People can research and compare options and form opinions using AI, without ever visiting your website, and your content can shape that decision, even though you never see the session. And when they do visit your website, it’s often in the middle or the end of the process. From an analytics perspective, that looks strange and attribution feels unreliable, because funnels assume linear journeys, and attribution models assume visibility.

But AI mediated interactions hide large parts of that process, and you start seeing symptoms. Traffic declining but maybe your impact’s not, or conversions that don’t line up with campaigns, or content that doesn’t perform but keeps showing up in AI answers. And nothing’s necessarily wrong with the content, but the measurement model is wrong. It is a different way of thinking, when your content is being used outside your site.

And this is one of the things we do for our clients. We help them understand how their content shows up in AI responses. We help to connect SEO, GEO, and analytics into a coherent picture and help them measure success, in this new environment.

You need to know where your content’s being used, how it’s being represented, and what signals still matter when clicks aren’t the whole story, because you are still trying to be visible and accurate, and influence decisions but the interface and the process, and the ways to measure and optimize have all changed.

Lightly edited for clarity.

Search is no longer just ranking links, it’s composing answers. Google and Bing are increasingly synthesizing responses directly into the search result.

And the real question isn’t do we rank. It’s are we included in that composed answer.

Because if AI cites you in the answer, you’re visible, even if you’re not in position one on that list of links. If you’re not included, you don’t exist in that interaction.

So optimization is no longer just about ranking, it’s about being retrievable and citeable. And that changes content strategy entirely.

Lightly edited for clarity.

AI now sits between brands and customers. Research is becoming conversational, and evaluation is becoming synthesized.

And that means your content might influence a buying decision without ever generating a website visit, and traditional attribution models aren’t built for that.

If AI summarizes your content in an answer, you shape the decision even if no one clicked. And that creates an invisible influence layer, and most companies are not measuring it.

Lightly edited for clarity.

For a long time broad positioning worked. If you have a product, you could say things like, “We have enterprise-grade solutions for modern businesses.” And that was enough to get into the conversation.

But that’s starting to break because the way people research is changing. They’re not typing two-word queries anymore. They’re prompting with constraints. They’re saying things instead like, “I need a cloud ERP for a midsize healthcare system that’s under HIPAA constraints and has to integrate with legacy finance tools.”

And so when AI mediates discovery, it tries to answer the exact question being asked. If your content is vague, it doesn’t get retrieved.

So the change isn’t just that AI is replacing search. It’s that specificity now determines visibility. And most companies are still writing their content like it’s 2015.

Lightly edited for clarity.

When people talk about AI search, they often focus on the language model, but that’s not really where visibility is decided.

Modern AI systems retrieve content first. They pull relevant passages into the context window and then generate an answer grounded in those sources, which means the real competition happens before the answer is written.

If your content is not retrievable, technically accessible, clearly structured, semantically precise, it never enters the conversation.

So traditional seo still matters, indexing still matters, but there’s now an additional layer. You’re not just optimizing to rank, you’re optimizing to be retrieved and trusted and cited.

And that’s not just an seo tweak, that’s a content architecture shift.

Lightly edited for clarity.

When people think about AI reading their website, they usually expect that the first thing that changes is traffic or rankings, but it’s not really. What usually breaks, um, what starts to break earlier than that is accuracy. And, like, not on the webpage itself, because the page is still the same, right? But the way that AI pulls that content, the way that it’s understood, and the way it’s reused once it’s combined with other sources.

So an AI tool doesn’t read your website the same way a person does, and it’s not moving from top to bottom and thinking about intent or filling in gaps based on experience. Well, actually, LLMs do fill in gaps, but not based on experience. The LLMs extract fragments from your content, and then they use those fragments as new material to generate a response. And if those fragments are not complete, or they’re loosely defined, or they’re inconsistent across different parts of the site, then the system doesn’t stop and ask which one of those you meant. It keeps going, and it’ll fill in whatever is missing with what’s statistically probable, based on the rest of what it has in its context.

And so that’s how marketing language turns into very ambiguous input. Assumptions become these implied facts, and small inconsistencies that are on different pages, so they were never meant to be read together, start to get merged into a single explanation that no one on your team would ever consciously write.

And the difficult part is that none of this looks like a clean failure, and, in fact, most companies don’t even realize that it’s happening. You don’t see a broken page. There’s not an obvious drop in performance. What disappears instead is something that’s a lot harder to notice unless you’re actively measuring for it. It’s control over how your business is described when someone else, or something else in this case, is doing the explaining.

Lightly edited for clarity.

Most teams haven’t changed the way they build their website, even though the way that content is being consumed has already changed and will likely continue to change.

Websites are still organized around pages, product pages, landing pages, about pages, funnels, because that model works really well when a reader is going to your website and moving through the site.

But AI systems don’t experience your website in the same way. They pull information in pieces, often from multiple URLs at once, and then try to reconcile those pieces into a single explanation.

So when the definition lives on one page and the positioning language lives on another, and constraints or edge cases are buried somewhere else, the system doesn’t know which one is supposed to anchor the others. It just tries to make them all fit together.

And that’s where problems begin. Not because the pages are inherently bad, and not because funnels don’t matter, but because content is now being consumed in fragments while it’s still being written as if someone’s gonna read the entire page.

So when you’re organizing your content on your site, your concepts need to be clearly defined. You need to consistently describe them and scope them explicitly wherever they appear, because that’s what allows those fragments to be used and recombined without losing their meaning.