AI and Discovery: Navigating the New Search Landscape

This article is a transcript and a recording of the "AI and Discovery: Navigating the New Search Landscape" webinar. In this session, Marina Petrova, CEO and Co-Founder of Intentful, shares her expertise on the rapidly changing dynamics of search and discovery, highlighting the impact of AI on content strategies for brands.

Search is transforming as AI technologies redefine the landscape. The concept of Generative Engine Optimization (GEO) is emerging as a critical approach for businesses aiming to stay visible in this evolving environment. With even more innovations like OpenAI’s Operator and Deep Research AI agents, along with the integration of AI Overviews in Google’s search results, traditional methods are quickly changing. The webinar explores how AI-driven search engines are changing how content is interpreted and presented, shifting from static SEO tactics to the fluid and unpredictable realm of GEO, Generative Engine Optimization.

Key webinar insights include:

• The shift from traditional SEO to focusing on user intent, context, and relevance in GEO.
• Strategies to maintain content visibility in the face of advancing AI-driven search technologies.
• Steps needed to stay relevant and discoverable.

Webinar Transcript:

Understanding AI Disruption: Generative Engine Optimization and Beyond

Thank you for finding the time to join the webinar today. So many things are changing. We had a similar session back in November. As I was updating the slides for today's session [in February 2025], it feels like it's a completely different world, and it's just been a couple of months. 

My name is Marina Petrova. I am the CEO and Co-Founder of Intentful, and today, we will talk about so many things. There is a lot of information, and please, if you have any questions, save them for the end of the webinar. We should have enough time for Q&A, and I will answer the questions live at the end of the webinar. Please feel free to reach out afterwards, too. 

I was just saying, several minutes before we started the session, that we had a similar webinar at the end of November. It's just been a couple of months since that event, but so many things have changed in two months just because of how quickly AI and all of the AI-related landscapes evolve. The topic today is understanding generative engine optimization. But it is, in fact, way more than that; it is so much more than just generative engine optimization. 

The title I was thinking about for today's session was “Ranking First is Not Enough.” That may sound confusing at first, but keep that thought as we continue going through what AI agent disruption is and what that means for all of us. 

Just a couple of words of introduction: Intentful started in 2021. We are a GenAI company. I'm not going to do any product promotion or sales presentation because the goal of this webinar is really to share everything we know that is happening with generative search and discovery in general. However, I would like to share a couple of words about us: Our core areas of business are in the DMO and travel space, we also work with global enterprises worldwide as well as mobile network operators, also worldwide. Since we started in 2021, beyond our commercial projects, products, and interests, we also have made it our mission to advance the practical use of AI and [share our] knowledge with the world. You can see us frequently speaking at different conferences and events worldwide.

We also like fun projects, not just commercial projects but fun projects as well. Even before the launch of ChatGPT, we worked on a project with the University of Oxford, where we created an AI model that sounded like Shakespeare, Oscar Wilde, and Jane Austen. The AI versions of them took part in the debate at Oxford Union. Another fun project that educates what AI is capable of is the creation of Ritbot. The Ritbot is an alter ego of a well-known marketing professor Mark Ritson. Ritbot was a panelist during the World Federation of Advertisers Week in Toronto last year.

The only reason I mentioned Intentful’s products today is because they are all powered by AI Search. Our products include Generative Response Ads [which essentially allow] you to have a conversation with an ad and have AI respond in real time right within the ad space. We also have AI Assistant and Data-Driven Content. I promised you that I'm not going to spend too much time on the products, and the only reason I mentioned them is because they are powered by AI Search. This gives us a better idea of how AI search actually works, and we are happy to share that knowledge as you work on your AI discovery strategies, AI-first search engine optimization, and everything else as you consider your content strategies for the new era.

I don't know if you saw this, but probably a couple of weeks ago, I think just ten days ago, Reid Hoffman, who co-founded Linkedin, released a book called Superagency. It's quite an interesting read. There is something that I paid attention to, and that is really the point of what's happening right now. That there is so much information, and we, as humanity, have reached the point where we are producing more information that we can effectively make use of on our own. However, the amount of that information somehow needs to be discovered, and the amount of this information is not as consumed as it can be. 

There is a shift that we're witnessing right now. [This shift] is happening based on what Google introduced many years ago, and we're incredibly grateful to them for doing that. Their mission has always been organizing the world's information. Today, we are seeing how interaction with information is evolving to augmenting human knowledge. There is so much information, and AI makes it possible for us humans to benefit from that knowledge.

What this also means is that this is changing how we, as customers, visitors, users, and consumers – people – interact with that information. [This change] means that we can get faster answers, we have faster access to whatever we want to find, a shorter journey, therefore, and have better experiences.

So, the topic today is not so much about Search but about understanding AI-powered discovery as it stands in February 2025—or should I say February 11—because it is likely that in a couple of days from now, something else will change. We'll be talking about more changes and more interesting things.

So what is being disrupted right now is how people discover information, how people plan what they are going to do, what they're going to buy, where they are going to go, and so on. I will be referencing many travel and destination-related examples because there are many participants today on the webinar from the DMO and travel space. [This includes] planning the trip, planning their itinerary, and also how they are booking and buying — not just in travel, but across industries. So, [it's about] discovery, planning, and buying.

Today, we'll talk about what's still relevant in traditional search. If you see changes in organic traffic, what is zero-click search, what is happening with AI overviews, and what the difference between SEO and GEO is. Of course, we will talk about Agents. This is topic number one, or it has been topic number one for several weeks and months now. But actually, agents are already live. And should technology work as planned today, we'll have a live session and a conversation with an AI agent. Then, we'll translate all that into the GEO framework and what that means for your businesses. And then a couple of interesting things also came up. If you joined the webinar in November, there is an overlap with what we discussed earlier, but there is a lot of new stuff, especially related to AI agents. 

The Evolving Landscape of Search: Understanding Changes in User Behavior and Traffic Dynamics

So, traditional search still exists. It did not go anywhere. Google is not dead, and they won't be. I mean, I don't know, but it is still important to continue looking at Google as one of your top channels for discovery and traffic. Even though the way information is shown is changing, Google remains key. Even if your share of discovery was 80%, now it is 40%, it is still 40%. This is still significant. And yes, this will be changing. But just don't disregard it completely. 

Just a little bit of an overview. When Google launched, almost every interaction with the Internet was happening through Google, mostly because of the desktop. However, when mobile phones were introduced—and that feels like it was ages ago—what happened is that we no longer go to Google for first interaction; we go directly to the app. This was a major shake-up in the search ecosystem and landscape because search started migrating to a channel where the customer is. Google was no longer the first point of entry to the Internet.

[This shift] includes all sorts of channels: social, messengers, voice search, and, of course, ChatGPT, Perplexity, and other AI tools now. Google lost [its position] as the gatekeeper of the Internet. Search is everywhere your customers are.

When we look at the discovery at large, we need to look at three items: where are our users? What do they do? What do they want? What's important for them? As obvious as it sounds. Technology, in terms of how it is evolving, what is possible, and what needs to be taken into account [is crucial]. And, of course, [we must consider] what makes sense from the business point of view—not just for us as organizations, businesses, and companies, but also for the platforms, how they are making money, and how those commercial decisions define their strategies. 

Going back to Google, I’d like to quote their CEO from early 2023, when he shared how search is changing, because the way people search for information is changing. They come to ask mostly for follow-up questions, and that drives the change in the nature of the search itself. 

A side note. A lot of changes also have to do with shopping because all of the shopping-related queries no longer start on Google. Most of them start on Amazon or a retailer in a particular market. This [shift] has influenced and ultimately redefined, going back to what we just talked about, the business side of search. The change in revenue sources for Google, because Google remained the core search engine, has influenced the way organic and paid searches are displayed.

The reason I am giving this context is because it's important to understand the search landscape to know how to structure your content and information to be discoverable. Remember those great times when it was so easy – you just had to be in the top 3 positions to get a 75% percentage of clicks? Yes, there was a lot to be done to be in those 3 positions, but at least you knew more or less what you would be getting. However, things started changing because Google had to evolve to be relevant to its users. 

I'm sure you all know about the feature snippets and all the sections, but this has evolved even further. 

It has evolved into what is now called zero-click search. If you've never heard about this before, I encourage you to look at a company called SparkToro. I have a QR code on the next slide, so you can scan it and see it or follow up, and I can send you the link. 

Basically, what happens is that 58% [of users] in the U.S. and 59.7% in Europe do not click on any links. A user would come to Google, ask a question, get the answer through a featured snippet that we just looked at, and then move on. This means that you are not getting that click and not seeing that traffic in your Google Analytics. It's a significant number, it’s 60%. Then, if we look at what happens with other clicks, out of this 40%, 75% go to unpaid, organic results, 24% go to Google's properties like YouTube and Maps, and only 1.5% go to paid ads.

This requires some time to let it sink in because when you look at Google Analytics and try to understand what's happening and why you may be losing organic traffic, this is one of the reasons. This trend is not new; it started before COVID, it started in 2018, but it was progressing and changing. As Google introduced AI Overviews, if you haven't seen a traffic decline yet, you might start seeing it soon because AI overview is zero-click in most cases. Sometimes, they do include a link to sources; sometimes, they don't, but you are not getting those clicks to your website. The role of search is evolving, but your website does remains the source of information; you are just not registering those clicks.

This has been changing fast. We are still talking about Google, and we’ll come to ChatGPT and Perplexity in several minutes. Back to Google. Back in August 2024, some six months ago Google AI overviews were less than 8%. That was clear because they had to experiment and see how it happened. In just a short three months, they announced that AI Overviews and searches are coming to more places around the world. And that means—well, I believe there is a short video here. Let me play it. Yes, so that means this AI Overview appears for almost everyone, every user, billions of users worldwide. Yes, the responses in AI Overviews include links that are somewhat visible. Again, I don't have the exact percentage of how many people click on those, but I assume that zero-click is what is happening. You are not seeing these users in Google Analytics. So if your traffic starts dropping, don't panic; it's not just you. It's a change in the ecosystem.

This was October 2024. Now, it is pretty much everywhere. And, of course, Google will continue experimenting with how this works and how to continuously improve it. But AI Overviews are the norm now. What it means from the discovery point of view is that the core SEO principles remain the same, and this is not about SEO 101. This is not a technical presentation. This is more about the approach in general. Google will continue updating its algorithm for the response relevance. However, what changes is how that information is presented. So, instead of the blue links we all know, Generative AI is now used to form a response.

The Role of Long-Form Content and Contextual Understanding in the AI-Driven Search

Now, a part where I'm applying a little bit of our knowledge from our products: how does AI know actually what content to choose and what content to serve? It splits large information, or rather it splits the information that it collects, into pieces called chunks. Then, it understands the context from that chunk and also the surrounding information so that it can see the broader meaning and the patterns of that context.

So, if your website is mostly marketing copy, it will be very difficult for AI to understand what you are talking about or selling just from a beautiful slogan or a few sentences. It really needs to be information-rich. It scans the content, splits it into chunks, and then tries to understand what it is about.

Then, we have the way LLMs work. As you know, it predicts the best next word based on what it was trained on, which also defines context-driven predictions. So that is another layer into what goes into chunks. 

I understand that's a lot of information. But I still want to give this to you so that when you start thinking about how you define your content for discoverability, you can go back, reread, rethink, and analyze what that means for your content. 

Then, to generate the response, it combines those multiple chunks it created. So, it retrieves, merges, and generates responses from multiple sources within the right context because it is generative AI. What this means is that your information can be combined with information from a competitor or another relevant website. It may not be relevant from your point of view but it is from the point of view of the crawler. 

Something that I really would like to emphasize is the importance of long-form content. Honestly, I couldn't understand for many years why, in the world of mobiles, in the world where decisions are being made in the blink of an eye, who reads these long-form 800 or 1,000-word articles? No one needs that. However, once we started working with AI, it became clear that long-form content and context are important for AI to understand what your website, page, business, product, or destination is about, because a short piece is simply not enough. 

This has always been the case for traditional search and now for AI as well. AI can now detect so much more, the models are smarter, and so on. But we see it over and over in different results that long-form content wins because it is easier for AI to make a decision. It just chooses a chunk of information that it ‘thinks’  is right. It won't make an effort to search for something else; therefore, it ‘reads’ the long-form content, splits it into chunks, finds the relevant pieces, and serves it as a result.

I am not saying that every single web page of your website needs to be in long form; absolutely not. But make sure that you do have a blog section or event descriptions where you have a mix of something that is for the user, [allowing] a human to understand what it is about and make a quick decision. And you have a piece of content for AI. We will also see this when we go to the demo of an AI agent.

Why is this important? Because an AI agent, while super intelligent, can still skip pieces of information. When you watch the AI agent in action, you just want to say, "No, no, don't go there. You need to choose this." Sometimes, it just does not have enough context, and that's where long-form content helps. 

Adapting to Generative Engine Optimization (GEO): Key Strategies for Enhancing SEO in the AI Era

Okay, let’s look at what is changing in the transition from  SEO to GEO, and as a reminder, GEO stands for generative engine optimization. Accessibility remains key. Make sure that the search engine robot can read your website. I can't stop emphasizing this enough, and I do this in every presentation I make. It is an issue for almost every single website. You need to ensure it is open and available to the crawler.

What's happening right now, because of the competition between different LLM providers, is that they scan everything. Whatever piece of information they can find, they grab it, and they will add it to their LLM because they need more data, more text, and more context. So, if that can help with the search at some point, why not? So make sure it is open to the crawlers, be it search or other crawlers.

Relevant content. We just talked about this. It is still important to make sure that whatever the user is searching for can be answered with the information on the website, not what the marketing team thinks it means. So, it really does need to match the response to the user's query.

An interesting thing is that keywords are going away. Keywords are still important; however, there is now a bigger opportunity because you no longer need to look at the keywords. You can look at topics because AI can understand context and get these chunks of information. It's not just going and trying to identify a single keyword in the text; it can serve more relevant contextual content to the user. When you think about your content plans, don't just think about injecting a keyword here and there. Think about the topic you want to communicate, and yes, of course, include the keywords.

Mentions are something that is also relatively new. Backlinks have always been important, and they continue to be important. Don’t get rid of your backlinks, but because of how AI is advancing and how the crawlers are advancing, they can read when your company, brand, or destination is mentioned elsewhere without needing a URL. Mentions would work as an additional signal that helps your content rank. This is fantastic for PR, and if you have a PR agency or are doing PR on your own, the more content you can distribute, and the more mentions you can have, the better it would work for you.

Clicks and traffic, including paid, also work. Crawlers would ‘think’, “Okay, you’re popular, people click,” so that’s why you have good traffic. Therefore, if people read [your content], you can be included in the search results. The 7 to 10 are still relevant as they were in traditional search. This number 7 is similar to Number 1. It needs to load fast. It needs to be mobile-first because crawlers now go mobile-first. They don't view it as the desktop experience. So ensure that your website loads within seconds because if it doesn't, the bot will move on to another website. They ignore you; they’re not going to wait because their priority is to provide the user with the experience they need.

We talked about links and, yes, optimized title URLs and meta descriptions. This is also important for the agents because it's new for everyone, right? We can make assumptions, but we see what it sees on the search page, and we’ll do the demo in a minute. It makes decisions about relevance based on the initial signals it gets. So, numbers 9 and 10 are still very important in terms of how you should be optimizing your content. I talked about the user experience. It needs to be super fast. It needs to be perfect. It needs to be logged in quickly. 

Navigating AI Search: ChatGPT and Perplexity's Impact on Content Discoverability

And ChatGPT search — I know many people used ChatGPT as a search even before OpenAI introduced the search function. But it was not a search before. So they ‘connected it to the open Internet,’ as some people would call it, at the end of October, and this has changed everything again. 

OpenAI is really good at publishing all of the information on their website. So, if you want to dig a little bit deeper to understand how it works, you can find a lot of it on the official OpenAI website.

They announced initially that the key focus for their search would be particularly in areas like shopping and travel, and it's clear why. But again, because we have a lot of travel companies or travel-related companies on the webinar today, please continue paying attention to what OpenAI does in the travel space. It is a fantastic area for experimentation for them. And that means that we also need to make sure that the industry content is discoverable and optimized for that.

In the traditional old search, even though so much had to be done to rank in those first three results to ensure you do get 75% of clicks through positions one to three, you knew exactly what a user would see. You knew what you were putting into that content, and if you were successful and your link came up in the search results, then you knew what was there.

With the new search, we don't know what's going [to appear in the system response]. I hear it quite frequently now that companies are trying to reverse engineer what the results are going to be. I'm not sure if that’s effective because the search results are generated every time differently, and they’re different for every user. In the new search reality, we have less control over what appears in the AI-generated responses, and this is also because it comes from multiple sources.

In the old search, there was owned media, paid media, and earned media, and the user journey was to click a link, visit the website, go back, ask a follow-up question, go somewhere else, and then repeat. Now, it is conversational and aggregates real-time data from multiple sources. It understands context and user intent( and I have to say that’s the reason we are called Intentful). The follow-up conversations happen right there within that search moment, where the user is, redefining the user journey.

I understand that's a lot of information, and I'm happy to have a follow-up conversation if you'd like. This is important in understanding how your information gets in front of the user. 

I used this example before and want to use it today as well. It's from my personal experience. I was looking to rent a car and went to Google, and I initially received the traditional results we all get, along with links. The team at Hertz Car Rental clearly worked on what would be there. They know the type of content and descriptions that have to be here, and it was quite good because everything there was branded, with no competitors and not even ads.

Then I asked a related question, and I got an AI Overview. It took this information and presented it differently. I did this on the day when the ChatGPT search was launched. I asked the same question to ChatGPT, without asking for anything specific, "Tell me about Hertz, Norfolk." then I got this: it started with, “It offers generally mixed experiences. While some customers report smooth drop-offs and helpful service, others have encountered significant challenges," and then it goes about pros and cons and what to expect. It included links because this was already a search and not just something ChatGPT made up. Then it said, "It can be a viable option if you are prepared to encounter some variability in service quality." I was surprised because I didn't ask for this [detailed analysis], but what it did was scan information, and the links were from reviews from all over the Internet.

Before I summarize, I did the same search with Perplexity, which was very similar to what ChatGPT said. It said customer experiences have been mixed, with long wait times, inefficiencies, and so on. In the case of Google, it summarized what it had from the traditional search results. AI engines scanned a far greater range of sources and then summarized them by going through that chunking process, providing me with what I think. With ChatGPT, I considered that perhaps it has a memory and knows me already, and that’s why maybe it thought that was what I was looking for, and I never mentioned it to it. However, with Perplexity, I was not a user; I was not even logged in, yet it provided with similar information. 

So, basically, the AI search becomes the truth serum. This means that in everything you do for your content, brand, or destination, you must keep in mind that a) you will not see exactly what comes up in the search results, and b) it will not just go to your website, it will go to multiple sources to collect that information. The shift that happens is from controlling the owned channels to ensuring that your organic content is present in the broader data ecosystem. 

Exploring OpenAI's Operator: The Rise of Autonomous AI Agents in Search and Discovery

You can no longer control what users see, which becomes even more interesting. About three weeks ago, in mid-January, OpenAI launched Operator, an AI agent. It is still not yet available in the European Union. But for those of you joining from Europe, you'll be able to preview and start getting ready because even if it's not going to be OpenAI, it will be someone else; the experience is going to be the same.

Before we go to the demo, I'd like to clarify the terminology. A lot of people use the term "agent," and sometimes it is used in an incorrect context. An agent is an autonomous assistant. This is not an AI assistant in the traditional sense; it can do things on its own. You can tell it what you want it to do, and it will surf the internet, book tickets for you, and do actions for you.

 I'm doing lots of experimentation with it. It can do a lot just by following a couple of simple prompts, and I'll show you now. 

But again – it is autonomous. It is not just an AI assistant or a chatbot. When you hear this definition from anyone trying to tell you it is an AI agent, double-check because it needs to be truly autonomous. This also means a lot of privacy-related things, so please be careful once you start considering rolling out AI agents.

OpenAI was not the first company to introduce agents; Anthropic and Microsoft launched those in preview at the end of 2024. OpenAI now released two agents. One is Operator, which we'll look at right now, and the other is Deep Research. Deep Research is mind-blowing. I think this is the best search and AI that I have used so far because you can find absolutely any information, and it won't be just a short summary. It will provide you with a lot of information and all with the relevant references.

We'll go into the demo right now. Before we do, I want to say that while there is so much conversation about this, we have time to get ready for this agentic shift, because the adoption by humans will not be immediate. OpenAI Operator is still in preview. It is still in research. It is not perfect or ideal. I compare it now to GPT-3. When we started Intentful in 2021 and worked with GPT-3, it was already mind-blowing but still required a lot of work. Comparing it with the models that we have right now, oh my God, there's a huge, huge difference. This will not take five years, but we probably have at least a year to get ready for agentic AI advances, or maybe even a little more. 

But now let's go and see what we have with the Operator. Okay. I'm going to share my screen. There we go. Okay, so this is Operator. I want to tell it what to do without going into too many details. So I would say, "I'm interested in a beach vacation, end of February. Ideally, the flight should be no longer than 4-5 hours. Let's look at adult-only options. I also need good internet”, because, you know, I will work.  And that's it. I'm not giving it anything specific. I am a user in the, let's say, discovery phase. I still have not decided where exactly I'm going. Oh, I need to say from where, right? From EWR or JFK, okay.

So now it says, "I look for suitable beach vacations," and then it opens its own browser. See what it does. It goes to search using the summarized query from me. It quickly looks at the search results. The first thing it does—and I did not plan this; I had no idea what it would come back with — is read a blog. It reads long-form content with descriptions and facts that would answer my query. I'm sure the reason it chose it is also because it has adult-only travel. 

It uses the Command+ F search function because it is trying to find information, and here (pointing at Operator action log) I can see everything that it is doing: ‘searching beach vacation, exploring the Caribbean, scrolling through an article, searching for adult-only resorts, and retrying back for navigation.’ 

It's my ‘favorite’ when it gets stuck because of navigation issues. This is happening everywhere. A lot of companies would have to update their interfaces because if an AI agent cannot click on some navigational element, that most likely means that the traditional search crawler cannot do that either.

It looks at 17 direct flights, and you see that at the same time, it continues to open multiple browser windows. Now, it goes to search again because it already chose something it thinks would work for me. Then it went to Microsoft Bing Travel. This is the first time I've seen this happen. In my previous experiences, it tried getting me to Expedia and Booking.com, but Expedia had a CAPTCHA, and it couldn't go through it, so it decided to go to a different website.

Let's give it a couple more minutes because sometimes it just gets excited and continues on and on, but the result at the end of the day is worth it. Theoretically, it should ask me what kind of hotel I'm interested in, like the number of stars. Still, for some reason, it decided not to, so it probably makes assumptions, or maybe it already knows based on my browsing history.

See, it continues going now to the news website. It went to the U.S.A. travel. Yep, and it's again to confirm that long-form content [is important]. Right now, we are looking at what the agent is doing. -. But you can take control of the screen, and you can tell it to stop or click something if it gets stuck. But normally, and this is what I already do in my day-to-day work, I will just give it a task and move on and do some other things.

I think this is the adoption we will be seeing. It's not going to be overnight or immediate that people start trusting AI agents, but it does save a lot of time, so you get the idea. It continues searching. For some reason, it does not think what it sees is good enough, so it continues. It is now looking at luxury-only options, closing image previews, and accessing article links. So it reads.

But what I also wanted to say is what we're looking at now is as an agent, right? The reality is that search crawlers are similar — well, not exactly the same because the technology is different — but this is a very good visual representation of how an AI crawler, a bot, gets access to your content and how it makes decisions about whether it wants to include it in the search results or move on.

Another thing I want to mention, and then we can continue with the slides( and then we'll come back to Operator to see the results it produced), is that sometimes we assume that the search crawler is incredibly intelligent and perfect and knows what to find. It's not. And so we make sure that we give it more chances to be discoverable, more content, and more information. What AI agents are teaching us is that those first three top links that used to generate 75% of organic clicks, are no longer enough because agents do not always click on the first link in the search results. They click on what they consider relevant. Yes, this is still early. Yes, we are still just several weeks in after the agent was released, and it is still in review. But agents don't just pick the first result. They choose what they think is solving the problem and the task they are given. The AI will go to multiple results, and it all happens incredibly fast, in seconds.

So it's important that the content on your website is not just informative but also viewed as something that can be trusted and relevant. Structured data is absolutely a must. If you've never looked at whether your website has structured data, you can Google it or ChatGPT it; you can look it up and then talk to your development team to ensure that you have structured data. We work with a number of DMOs, and we know that most, unfortunately, do not have structured data. It is very important for discoverability in this new, amazing world.

Adapting to AI Content Discovery: Ensuring Relevance and Visibility

Yes, there is more. This is one of the experiments we did when the Operator was launched. We had Operator speak with an AI assistant created by Intentful on one of our customers' websites, a DMO website. OpenAI's Operator came to the DMO website and, instead of clicking all the links, started asking questions to the AI assistant because that provided a much faster and better experience. Then, the Operator AI would come back to the user with a response that was created in several seconds.

I know that's a lot of information, but it all comes back to making sure that what you have on the website is relevant and that you are the source of truth. What you provide on your website or in other channel partners needs to come from you so that you are the source of truth and control your story. 

Why organic? OpenAI has said that they are not launching, or at least not yet, any ads in ChatGPT. Perplexity is launching an ad program, but it is still early. Google ads will, of course, evolve, but the ad ecosystem will also be changing and will align a lot with these new different knowledge and intelligence layers of discovery. 

All of this means that there is a lot of content work ahead if you want your destination, organization, or brand to be discovered. Content planning and content-specific programs need to become a constant exercise. It’s no longer just "why don’t we write this blog" or "why don’t we write about this in an email." It needs to be a structured approach around user intent, the context in which the content would appear, and how relevant it is. I’ll quickly walk through this. 

The intent is based on user behavior, questions, and multiple signals. When someone is searching for a beach vacation versus when someone says, "Book me a five-night hotel stay," these are different types of intent. So, the type of content the AI would tap into would also be different. If you want to be discoverable across all of this, you need to make sure you do have content that is clearly discoverable and where you explain what it is that you offer so that you can be relevant to the user at every stage of their journey. My recommendation is to create a spreadsheet (because I love spreadsheets), but use whatever type of content plan format works for you. Have intent as your column two. Column one is the message you want to deliver and what you are selling, promoting, or communicating, and then split all of that by type of intent: informational, navigational, transactional, problem-solving, or inspirational. For each type, have content so that the AI can discover you.

When we talk about relevance, make sure you talk about 2025 and not some topics from 2017. Avoid fluffy, unnecessary stuff. Be direct because that's what AI likes; it needs information. And, of course, be consistent.

In terms of execution, and again, this recording will be available, but I’m happy to share this list as well. Remember that it needs to be facts; it is information and not just marketing copy. Make sure that it is accessible, easy for the bot to understand, and everything else we discussed. Ensure that your content is relevant, consistent, and easy for AI, which would mean it is also easy for humans to interact with as well.

Speaking of formats, text remains very important because text is the first thing that crawlers look at, but images and videos work, too, and they're important. So when you upload images and videos, make sure that at least they also have a text description of what they are and not just a file name like ".final" and a set of characters that only you would understand.

So the zero-click search we discussed earlier today gets to a completely new level. It is going to be even more confusing before it starts getting into some sort of structure. But it is absolutely fine to start exploring this new world of content discovery because the future is already here. It's not an assumption; it is already here. 

Remember to be the source of truth and ensure you have the content distribution channels.