How To Dominate AI Search in 2026

Hello everyone and welcome to today's webinar by Exposure Ninja. I'm Charlie, CEO at Exposure Ninja. This is a popular session. We've got loads of people coming into the room right now. Before we get into the topic that everyone wants to hear about, which is how to dominate AI search in 2026, a little bit of housekeeping. So the slides and the recording will be available. They'll be sent to you after. If you need to drop out at any point, don't worry. That's absolutely fine. We'll also be taking Q&A at the end. My colleague Dale is working behind the scenes on the webinar. So, you can drop Q&A into the chat anytime you have questions throughout the presentation. We're going to be going for about an hour and we've got a lot to get through. So, I'm going to go straight into it. Search has changed a lot across the past two years. What we've originally thought of as just Google search when we're thinking about organic SEO has completely changed. We now have new search channels. We have AI overviews and AI mode within the Google ecosystem. We have Google's Gemini and then we have other AI chatbot competitors including Chat GBT, the current market leader, uh Gemini, Perplexity, Grock, Claude, so many different ones. And they are changing people's search habits and they're changing how we think about SEO as well. So today I'm going to be sharing the current situation. We're going to go through a playbook of what works right now. But as these AI search platforms continue to evolve, this continues to change. So here's what's going on. Chat GBT is the orange line and they have been coming up the rankings entering into the top 10 websites. They're actually in the top five just behind Google, YouTube, Facebook, and Instagram. a huge huge change in how people are searching and the websites that they're actually using. Meanwhile, Google's Gemini is trying to gain back some of its market share. Chatbt is the biggest AI chatbot. So, the biggest AI search platform that gets used even compared to Gemini, but some of the latest models have been amongst people's favorites and we know that Google is aggressively trying to get back its share from open AI. And the other big trend we know is that many of Chat GPT's users actually still use Google. That's probably true for you too. So it might be that you start a search in chatbt, you're doing research there and later go onto Google to finish your search. This is also causing a phenomenon of what is known as as dark traffic. Very difficult for us to track exactly what traffic had a touch point with chatbt. If actually later someone types a brand name direct into Google because they've researched in chat GBT that won't show in your Google Analytics even though you will be able to see any clickthroughs, referrals, conversions that have come direct from chat GBT. So we can see this chatbt share trying to balloon into Google's significant audience size as well. This of course has meant lots of changes. So some of our own research, we have a fantastic state of AI search report that I will share the link to at the end but you can find on our website which has many many data points around how people are making purchasing decisions now and some of that data showed that 80% of people make more than half of their purchase decisions within AI tools. So, Chat GBT and other tools, which is absolutely huge. If you haven't begun shifting your search marketing strategy to address this, hopefully the stat is what you need to set that fire beneath you. This probably is something you've seen though. This was a very very common trend that many people see in their uh Google Analytics, in their search console data right now. And this is impression ups clicks down. And this is not just a phenomenon happening to one or two sites. This is pretty much happening across the board. Here's a completely different website in a completely different sector. Same trend. Impressions are up, clicks are down. This has been one of the thorns in the side of many people who are focusing on SEO and organic search because what it means is even though people are seeing us in the Google search results, they're not actually clicking through to our website anymore. We've seen a decrease in clicks, a decrease in organic traffic coming from Google, and this has mainly been caused by these pesky AI overviews that have come onto Google search results. So, these AI overviews take up usually the majority of the above the fold, especially if you're on mobile. They are the AI summaries appearing above regular organic results. Um, they also sometimes appear above and below ads and they are AI answers. answering the question that you've asked Google. And what has this caused? We know from the data that AI overviews correlate with almost a 35% reduction in click-through rates. This is significantly huge. Some industries are affected more than others and it overwhelmingly affects top ofunnel searches and some middle ofunnel searches. So, when we're thinking about what's killing our clicks, it's those AI overviews, but there's also a couple of different AI forces at play here. The second is these AI chat bots. Chat TV, Gemini, Plexity, Claude, all of the others. And now, third, we also have Google's AI mode. And more and more we're seeing in those AI overview summaries Google trying to push us through into AI mode which is their version of an AI chatbot sat within our browsers. So if you've not seen AI mode yet definitely jump onto your Google have a look for it. It will be in the top menu bar along with news shopping and you'll see AI mode there as well. And on mobile Google are pushing this more and more in discover as well. pop-ups trying to bring users through into AI mode to normalize that idea of AI search. So all of these things are affecting how we get organic traffic and ultimately organic conversions, customers and which is what we care about. We care about our bottom line. So the funnel is changing. AI answers are now eating these top offunnel searches and they're taking a little bit of these middle ofunnel searches as well. This is making it more and more difficult for businesses to know what to do with their SEO and organic search strategy now that the game is changing. This has pretty much caused a compressed buyer journey. In traditional SEO when we had just the 10 blue links and even with some AI overviews, we were spending a long time actually putting top ofunnel, middle of funnel, bottom ofunnel queries into Google. We had a such a huge amount of touch points to even get to the decision- making stage in the buyer journey. With AI search, this is changing. This can all happen within a single conversation. A user can go from asking a really big top offunnel term like why is my skin dry even if I'm using moisturizer? right to having a personalized recommendation for specific products for their skin in a conversation with chat GPT in a five minute conversation. This is meaning that buyer journey is getting shorter and shorter and shorter. So what does that actually mean? It means we need to think about our mindset. We need to reframe the way we think about organic search. And our older mindset has been if we publish more articles, we get more sessions, more traffic to our website. we'll get more conversions. It's a volume based game. We're just trying to get more and more more traffic. The new mindset of how we need to think about AI search is that visibility in these AI answers is increasingly important over being buried in 10 blue links. Those 10 blue links, even though they still drive some traffic to our websites, and if you can get them for bottom of funnel terms, that's amazing. But often top of funnel, middle of funnel, those blue links are becoming buried. They're getting much less traffic. They're underneath the AI overviews, answers. Some people are doing that top ofunnel research in AI mode. Many, many people are doing that top off ofunnel research in chat GBT. Now, so the top offunnel is existing to influence AI answers, not to generate clicks. This is the new challenge that we have, the new challenge we need to address in our search marketing strategies. So let's take a little look at why that's important and how we do it. Let's first look at AI overviews. So this is an AI overview for a complicated term which is moving towards middle of the funnel. Can I use nioinomide with hyaluronic acid? So this kind of term you already need to know a little bit about skincare to even know what you're asking. The interesting thing about this AI overview is how dominated it is by the ordinary. And it's almost the first thing you realize about this brand. If you've seen their labeling before, you'll recognize it from that image and then two citations on the right side panel with their name, their logo. Even though this may not result in clickthroughs, it might result in a couple, there is a big big brand stickiness factor going on here. And what happens with brand stickiness? Pretty much the searcher sees your brand name repeatedly. They start to see it in all sorts of different places. That's going to be AI overviews, some organic results. Hopefully, you've got some ads running if you're an e-commerce business like this. And you're seeing that visual brand identity. You're learning to trust that brand. And that repetition drives familiarity. Then all it takes is for someone else to say, "Oh, I use retinol from The Ordinary." And you think, "I'm sure I've heard of that brand somewhere. I think I know who they are." That is how we start to become familiar with brands. And ultimately the brand becomes associated with reliable answers. If you're informing AI overviews, if you're informing AI results and your brand is actually being shown visually and by brand name and surfacing answers from your content. Fantastic. Now users believe that they can trust your brand, that you are informed on specific topics. So let's have a little look at how we can get these AI overviews. I'm going to talk you through an exact workflow that we use here at Exposure Ninja that you can also replicate for your business. Um, this is using Semrush who is a fantastic search tracking tool. Many people use it to track their keywords. Others are available like SE Ranking, MOS, Hrefs, etc. So, step one, you can jump into Semrush's organic research tool to find keywords where you own the AI overviews. So, you want to filter by domain ranks. you can see in the yellow boxes here and AI overview. This will show you the ones that you already own, where you're already in that result. You can keep a track of those because you want to maintain that result longterm as well. Step two, check the keywords where you don't own the AI overview. So, this is domain doesn't rank and then filter by AI overview. You're going to see a whole list of keywords where an AI overview appears in the search engine results page for keywords that you care about, that you have tracked for your business, but you don't yet own. So, you can then decide which ones are going to be most influential. Now, I would recommend filtering by the top 10 positions to find the keyword opportunities to get those AI overviews. And the reason I say this is because if you're in the top 10 positions on Google already, it will be much easier for you to own the AI overview by making some tweaks and updates to your content. If you're not on page one of Google for a keyword that has an AI overview, that's the first place to get to. We know that AI overviews are much much more likely to surface content from websites that are in the top 10 for that keyword than something that is on page two or page three. So top 10 is a great place to find the lowhanging fruit. Then you've got a really large list of keywords. It's going to look something like this. You can see the keywords. You can see the position that you currently own. And then you can also look at volume. So the number of people searching every month and the competition. And you can look to see is itformational? Is it quite top of funnel or is it actually becoming commercial? Some of these are both commercial andformational intent. So, I would always prioritize the ones that are commercial andformational intent because you're more likely to actually get some buyers, some people who are further down that buying journey. So, those have got to be the number one priority. You can sort this list, find out which URL, which page on your website is being referenced already for that keyword and start optimizing to try and gain those AI overview positions by making tweaks to your content. And then the final step is of course keeping account. This is an example of one of our clients and their three target regions that we were working on an AI overviews project for and the percent increase in AI overviews owned. So you want to keep a track period on period so that you know the number of AI overviews that you're showing up in and you can correlate this to the traffic and ultimately what we really care about the conversions that are coming through to your website as well. So this is pretty much still what we know as regular search, what we think of as regular Google search. Even though in the old days we were looking mainly at 10 blue links, now we're often looking at AI overviews. This is us directly interacting with the internet. But chat GBT and other AI chat bots have pretty much changed the game. They've become new search interfaces for us. Now that that yellow person, that's us. That's the user. We're having conversations directly with ChatGpt. We don't need to browse through 10 blue links to find the answers to all of our questions. ChatGBT interacts with the internet for us. It does some of that searching for us. And I think we're going to look back on the way that we search, the way regular search works, and think it's really antiquated one day where you think so much more of this process is automated even though we're only just becoming familiar with how AI search is actually working. So let's have a little look at ChatBT's process. And this is very similar across AI chatbots. I'm just using chatbt because they're the biggest. They tend to be the main traffic and conversion driver of the AI chat bots for most businesses. So, how it actually works, a user inputs something, we input a prompt, we input a question, we give lots of context, we actually tend to, I've seen studies showing even put an average of 60 words into a chat GBT prompt. If you compare that, the average Google search is just three or four words. So so much more context is inputed when we're asking chatbt something. Then what happens is chatbt's model analyzes that input and it has a pre-trained model. That pre-trained model then weights the amount of searching that it needs to do. So it's got its own knowledge. It's got its own pre-trained data set. And if it can, it uses less resource just to answer the question it's been asked with that pre-trained knowledge. If it needs to, however, particularly if you ask something more recent, if you ask a very complicated query, something newsworthy as well, then it will search externally. And that means it's going to go and look on the web to find the information it needs. And it will combine that with pre-trained knowledge to give an answer. So that external searches is influenced by your website, what traditionally we've thought of as SEO, how we have actually optimized our website to try and show up and also other websites. And the waiting of this, which we'll come to later, is significant. More than 85% of what is surfaced in chatbt tends to be from other websites talking about your business rather than your own website's content. So this is huge, huge, hugely important. If you're a business who's been doing digital PR for years, you're probably in a much stronger position in AI as a starting point. If it's new for you, you're probably seeing your visibility tends to be a lot lower. So this is why SEO and digital PR are so important. Then what happens is the model generates a response using all of the information and it outputs it in a format that uses natural language very similar to the type of language in our original question. And one of the important things when we think about SEO is just because we're optimizing for chatbt Gemini perplexity AI mode. SEO is still important. This is because they have an underlying search function. So if you check column three here, you can see chat GBT, Gemini, and AI mode all use Google search. Even though Chat GBT sometimes apparently uses Bing index, we more and more see Google search being the significant one. So, Google is still an incredibly important player and Perplexity uses its own proprietary tech sometimes also uses Bing search index as well. However, none of these platforms just give you the same results as Google. If you Google a question and you chatb a question, you're going to see two very different answers. You're not going to see the same top 10 in Google as you would do in chatbt. They're going to be different. There's going to be a little bit of overlap. And this is because all of these AI chatbots have their own training data and they have their own crawler bots. They're not using Google's algorithm even if they are sometimes using Google's underlying search index. So let's talk a little bit about how these are working. This is Chach Gemini Perplexity Claude. They all take in information and surface information from websites across the internet. Here's just a couple of really influential ones in how they search and the types of information they bring up. But they are looking everywhere. When you ask a question, when you put a query into ChachiBT, it's scouring across loads of websites to give you an answer. And this is because LLMs are not ranking websites like Google. They're using information from all of these different sites instead to give you an answer, not to just give you a list of 10 websites. And this means that we need to redesign our search strategies. We need to be designing for the way that LLMs actually answer the questions. So, here's the problem. These LLMs, they don't rank pages by keywords in the way that Google does. They instead recommend brands they understand as authorities within a topical area. What does this mean? Let's take a look. So, here's a keyword that you might be going after if you sold organic veg boxes. Something like best organic veg box delivery near me. Now, let's compare the topical area is organic veg boxes. It's much broader. And what chat GBBT and other AIS are looking for is to understand that a business is an authority in this particular topic. And that's going to be through lots of different types of content both on the business's own website and on third party websites. So that could be things like recipes, customer stories, and reviews. Reviews are incredibly influential in AI. It might be even talking about tangentally related topics like farming practices and seasonality and things like supply chain transparency and environmental impact because AI is searching so much more broadly to understand that a website has authority within a specific area. And this means that we need to think differently about how we actually design our search strategies. We need to be redesigning for what search looks like now with these AI platforms influencing so many of searches. Not what search looked like two or three years ago when everyone was purely using Google, maybe a little bit of Bing and Duck.Go, but mainly Google. Now, we have so many other platforms at play. So, I'm going to take you into the three pillars of AI search, and these are incredibly important because they influence those answers that AI gives about our business. They they influence the way AI speaks about us, the sentiment that it has. They influence whether the information AI gives about our business is correct and actually accurate and they influence how often we show up in the AI answers we care about, how visible our business is. So the first is technical foundations. Hopefully no surprise to everyone here that technical foundations for your website are incredibly important. things like having a great structure, actually making sure your website is crawable to AI bots. If it's not, they won't be surfacing content from your website. They'll just be using third parties, no errors, and extremely up-to-date schema. The second is clear positioning and content. Very, very well-written, well structured content that's easy to site and addresses query fan outs, which I'll come on to later. And then digital PR, citations, and reviews. pretty much your online reputation. And this is probably the juiciest and the most influential part of the three pillars we need to look at for AI search. So let's have a little look a deep dive into technical foundations. I'm hoping some of this is familiar from just general best practice even when you're optimizing only for Google. The first is a great website architecture. If your website is a mess, then humans will struggle to navigate through it and so does AI. So AI can look across your whole website architecture to find what it needs to find when a query is asked of it. I like to think of this a little bit like a library. If you walk in and you're looking for a book, you want to be able to go and find it alphabetized by author on the shelf where you think it should be. If you walk into a library and all of the books are just thrown on the floor, there's no categories, there's no alphabetization, there's no authors, it would be a nightmare. The easier it is to actually navigate through your website, the cleaner the architecture, the more well organized it your filing cabinet is, the more likely AI is going to be to trust your content and be able to find what it needs to surface it. Second, your schema. This is incredibly important. There are many, many different types of schema. And because AI will read the schema, the backend code of your website pages to understand what the page is about and what the business does, it is more likely to reference this. So, you must must must have it in place. Many websites have some version of schema in place from their original SEO, but not usually very thorough versions. So, 100% worth checking what schema you've got set up on which pages. particularly your highest priority pages, your highest priority service and or product pages depending on the type of business you run. Then making sure there's no errors. An example of a 404 error, but it works exactly the same way as humans navigating a site. If we're looking around a site and we go to a page and it's broken, most likely we're not going to try and go back and navigate around the site. We're going to get frustrated. We'll lose patience. We'll just exit that website. What happens in AI and especially if AI is using an agent. So if anyone's tried Chat GPT's agent mode where the agent just goes and can browse the website for you in its own browser without you needing to control anything. It will exit websites that are broken. If links are broken, if they are not following good redirects. So make sure you've cleaned up everything around that as well. And also make sure that website crawers can access your website. If you're using Cloudflare, then please, please, please check that you have enabled AI crawling on your website via Cloudflare. They updated their default settings about eight months ago. So, if you didn't make a change or double check it, then you might find that AI crawlers aren't even accessing your website. Okay, part two of our three pillars is our positioning and our content. So this is your brand and the way you describe your brand is incredibly important. You should be able to say what your business does in one very clear sentence. And that same sentence about what you do should be everywhere. It should be on your website, your about page, your product pages, your socials, your PR mentions, your reviews. It should be really clear. And you want your brand consistent. The same way that your brand visuals, your brand logo, your brand imagery would be consistent across all of these different areas. You want the same in the text that you write about your brand because you want AI to associate you with those topical areas. You want it to be very clear on what it is that your business does. So, make sure you always describe your brand accurately. What you don't want is five or 10 different ways or descriptions of what your business does because then what happens is AI becomes confused. It doesn't know which is the real truth and it will look across the internet both your website, third party sites, review sites, socials to try and amalgamate the information to find the most relevant source of truth from all of those different sources. So you must be extremely clear on your positioning. Then we come to our prompts and keywords. Before we're doing anything with our content, we actually need to understand what people are going to be searching in AI. And there's no easy way to do this. The best way I would recommend doing this is combining your traditional keyword research for Google, which actually then has search volumes and intent ideas behind it, with customer persona data and understanding their natural language. So if you have sales calls transcripts for example, particularly if you work in a lead generation based business, fantastic. Use that information from customers. If you have customer feedback from feedback calls, fantastic. Use those to amalgamate with your keyword research so that you can find types of language that your customers use and create prompts that are likely to be searched within AI platforms like chat GBT. There's a fantastic tool called PKI which I highly recommend. If you're thinking, I have no idea where to start, fortunately, you can go and put your website into PKI and they offer a free trial, so you can just have a play with it for a week and it will recommend you some prompts to get started with. But these will tend to be longer tail versions of things like best X, whatever it is that you actually sell, whatever sector you're in. So, start there, then move into content planning. And I like to think of this as a kill, a keep, and a create. A three-step content plan. And when we're talking about killing content, sometimes people can get a little nervous, but pruning content is incredibly important. And this is because you don't want AI confused about what you do. If you have a very, very neat site with not much content and it's all very clearly delineated, relevant to your business, you'll have almost nothing to kill. If you have a website that is decades old and people have been adding content to it and you've got thousands and thousands, tens of thousands of blogs andformational pages, you will find that there is content that you need to kill. And you can find that actually by going into your Google Analytics to understand what doesn't even show up, what's not even ranking, that gets zero traffic, and you can prune it back. Get rid of all the stuff you don't need. This is most likely going to be generic explainers. So, what is X that isn't capturing any demand for you, that isn't that closely related to what you do? Content that AI can summarize very, very easily and isn't using your content for. Um, things like rehash definitions or content that doesn't have any brand perspective because it's probably not performing very well. Then you've got your keep and some updates as well. They're going to be higher performing pieces of content that you've got or mid-performers that you think could be doing well, maybe a few low performers that you think if they had a total rewrite they could also perform well. And AI really, really loves to reference content, things like comparisons and alternatives, which X should I buy, which shampoo should I buy, which is the best car insurance, all of those types of contents. pricing and use case pages. That might be different for different types of businesses. It might be that you're not able to put your pricing on your website, but if you are, then do. Reviews, awards, accolades. Make sure those are really, really clear on your website. They can be on your homepage. They could also be on your about page. And then high intent FAQs as well. the questions people are likely to be asking close to the bottom of the funnel when they have an idea of what it is they need to purchase or how they're going to solve their problem. And then create new content. We're going to be looking at things like entity rich reference pages. So that might be about a product, it might be about an ingredient, a concept, a brand. You also want to publish as much original data and tests as you can. Some of that might be from your own case studies that you have if you're a service-based business. Some of it might be including reviews. Some of it might be doing customer data studies to find both PR angles and create original unique content for your website. Looking at opinionated pieces and best of articles and listicles that often get surfaced at the moment in AI as well. The other thing to add, judge your top off ofunnel content by its potential to be cited in AI, not only the traffic potential. You want to be thinking about both how likely it is to get traffic from Google, but also how likely it is to be visible in AI. So let's talk a little bit about what's going on on under the hood with content. So these AI platforms, I'm talking about chatbt, Gemini, Plexity, and AI mode. This is Google's example from AI mode. They use something called the query fan out technique. And when you write a query in, so in this example, best things to do in Nashville this weekend with friends, we're big foodies, we like music, chill vibes, exploring off the beaten path. That's the query that we put in. What happens underneath the hood is that instead of just directly searching, AI is able to break this down and it fans it out into multiple different keywords and queries. So in this example, you can see that even though this search query didn't ask about independent bookstores or breweries, it's been included in the query fan out because what AI is trying to do is break your search down to understand the intent and it's using different queries in the fan out to do its best job to answer what it thinks the person is looking to understand. So this means that now that we have AI searches, they are much much more sophisticated in the number of queries that they can actually look for. And on AI mode, this can be hundreds of different queries, but you'll often see a smaller version. Plexity is my favorite one to look in because they actually show you the queries, the keywords that they break it down into and that they search under the hood. This won't happen if they just answer from their own training data, but anything where they're using a search index, you will see this. I 100% recommend going in, putting a couple of queries into Complexity, and seeing what the breakdown looks like. And the reason this is important is because it means that in order for you to have your business surfacing in the AI answers you care about, you need to be doing well in these underlying queries. And your content on your website needs to mimic the kinds of content that would be answering these queries. So if you're writing something, you need the depth and the detail reflected in the query fan out. If you're thinking, "Oh my goodness, this seems like a lot. I have no idea where to start." My fantastic colleague Abby, who is our head of digital um PR and content at Exposure Ninja, has written a fantastic blog post breaking down a template for exactly what it needs to look like to be optimized for AI and have your best chance of actually showing in AI answers. Okay, our final pillar, the biggest and the juiciest one, digital PR citations and reviews. This is all about having a great reputation online across the internet, which is incredibly important. And here's why. When AI sites sources, it favors third party websites and not just your own website. I know it's not the information that we all wanted to hear, but it's true. So AI is much much more likely to source to show information from third party websites that is written about your business rather than just your own website. So 80% of the time it's using third party websites. 9% third party websites that have your brand name in the URL and then 9% of the time your website. So your website's content is only being referenced a small portion of the time. So, you're probably thinking, okay, well, how does AI decide what sources to site? Why is it not thinking my website is the source of truth? So, remember, AI is looking across the internet. It's looking across Reddit, Quora, Trust Pilot, Google, Wikipedia, your website, and then relevant articles. So, for example, if you work in the tech sector, it's going to be looking at Nerd Wallet. It's going to be looking at tech reviews, anything that it considers relevant to the topic being searched. So, we don't know 100% why certain sources are favored by AI systems, but there are lots of commonalities and trends we can draw from. The first is that they're authoritative. They have they're a well-known website and they have authority in that topic, just like the tech sector example I just used. And LLMs tend to favor firsterson accounts as well. So, things like someone writing a product review if it's the tech sector. They also love well ststructured websites just like we spoke about earlier with the library. So they want content to be easy to pass. They expect good schema, page title, metad descriptions, no broken links, all of that on the third party and any website that they reference. They expect it to be relevant to the query. So AI uses natural language processing and it pattern matches to find content it believes is most relevant to what's being asked. This is one of the reasons it's incredibly important to use your customers language in the types of content and FAQs that you write on your website because the tone of voice, the way of phrasing things that your customers ask, you want to be reflecting that. You want AI to be able to match your content to the type of language they use and fresh content. We know that AI generally prefers recently updated content. Um, and it if you can change publish dates and update timestamps, that can be a signal for accuracy. And now what I mean by that is please please don't go and put the 25th of February on all of your blog posts immediately. If you are updating timestamps and publish dates, then you need to add updates to make your blog posts or yourformational content new, fresher. And that could be adding an additional paragraph. It could be adding an update specific to 2026. It could be creating a new version. If you've got a 2025 best of list, then updating that to be 2026. So, you do need to make some changes to the content as well. Don't just go and update all of the publish dates. It won't work and it will have taken a lot of time for someone. So, here's an example of that actually in progress. So, one of our clients, Zugu Case, they sell fantastic iPad Air cases. You can look them up. We did a PR campaign for them, published an article in ZDNet. Very interesting to see Perplexity referencing that exact article in ZDNet in the sources that it cited the very next day. Here's another example. This was an article we published on Wired, also featuring Zugu case, talking about their folio case. The very next day, Gemini is referencing Zugu case, sourcing this published article on Wired. But we also see the same just recently a couple of weeks ago I did the same search even though we published these articles back in Jul around July August last year and we're still seeing Google's AI mode referencing both that ZDNet article and that wired article and you can see Zuku iPad Air cases dominating the search here. So just a couple of wellplaced publications if done right can mean that your business is surfaced in the answers because the sources that it's using where AI is pulling the information is from those third-party websites. Um if you want to read the case study in more detail you can find it on our website significant AI traffic increases and revenue driven for that business. So let's have a look then at how these citations actually work. So if you go on to a chatbt answer and it works the same in Gemini perplexity and others as well. I'm just using chat as an example. You will be able to find a toggle for the sources. This is in different places in different AI chatbots. In chatbt it's at the end of the answer and what do is summarize for you all of the websites all of the articles where they sourced information. So a little bit like writing an essay at university. At the end you had to write down everything that you referenced, who said it, where it was published. AI does exactly the same for us. So you can see here on the right hand side the citation panel all of the articles where this answer was informed by. This is fantastic news, right? Because if we know this, then we can start to reverse engineer the types of publications, the types of articles we want our business referenced in on third party websites because we know they inform the type of answers that we want to be in. So this example you're seeing on screen is for best CRM. If you were selling a CRM or a similar type of SAS product, it's exactly the kind of list you want to be in because you know that people searching for CRM are going to be comparing different CRM before they even come to your website. They're having that comparison conversation in ChatBT or in their favorite AI platform, which means you can look at building citations. Now, if you're thinking, "Oh my goodness, Charlie, I would have to go into multiple different chatbt searches. I'd have to be doing like 20, 30, 40, 50, etc. searches for all of the types of queries that I want to show up on and then pulling all these lists. That is one way to do it. That's extremely manual and I don't recommend it. Again, there's a great tool called Peak AI. There's other options available. One called Scrunch AI. There's one called Profound, Clear Scope, various different tools that all do the same thing. And once you have your prompts in there, they will pull for you a list that just looks like this of all of the sources across all of those prompts. So they will automatically run it at scale for you and they do that daily so you can get an updated list. The great thing about this is you're able to see which articles you're already mentioned in and which ones you're not. So in this example, this is HubSpot various different queries here, including best CRM, best CRM for midsize enterprises, etc. And you can see the green yeses, they're already in those articles. They're already cited. The red nos, they're not cited there. So that gives them an understanding of the publications to go after. And so for example, you can see there the third one down, techraar.com. They are not in that CRM review list for 2026, which is kind of crazy. You can also see that their competitors are monday.com is mentioned in there. So building citations, there's a couple of different ways to do this. One of them would be getting in touch with some of these publishers to try and get updated into the old article. However, that's going to be difficult most of the time. The other way is to predict the trends, the kinds of topics that are coming up and how likely those articles are to be republished. So, for example, anything in here like top AI powered CRM for 2025, we know that publisher is eventually going to write a 2026 list and we know that if a 2026 list existed, it's quite likely that article will be gazumped as a source in AI. So, building those citations, there's a couple of different ways. One is replying to journalist requests that are relevant to your business. You can do that through a great tool called quoted. Um something called Harrow, help a reporter out also exists or you could contact editors directly. You can do press releases, PR stories using unique data that we spoke about earlier. If you have interesting studies, uh interesting data, then definitely create PR stories around that. There's also PR work around referrals and affiliate schemes if relevant to you, partnerships, influencers, you know, those first party accounts that AI absolutely loves. So, getting people actually speaking about your brand directly. And then, of course, things like awards directories and customer reviews. And remember, you're trying to build topical authority. So, you want all of your citation and PR work to be based around the topics that you want to own, where you want your business and website to be seen as an authority. So, the last section, you're probably thinking, well, this is all well and good. If we're employing all of these AI search tactics and we're trying to show up, how do we actually measure the results though? How do we know if it's having any impact on our website? Because ultimately, we care about ROI. We care about our bottom line. We're not just doing this for the sake of vanity. So, first up, there is a myth that you can't measure anything coming from TACBT, Gemini Plexity, etc. That is untrue. In your Google Analytics 4, you can filter for generative AI referral traffic and conversions. Here's an example with all of the AI platforms in one. However, Google Analytics doesn't automatically do this for you. This will show you the traffic that is coming directly referred from Chat GBT, Gemini XT, and other AI platforms you choose to track and if they convert. Remember, not everyone clicks through from an answer in AI. So just because someone searches in Chat GBT for the best CRM, it doesn't mean they click through to go to the HubSpot website. If they didn't, it won't show up here. If they did, it will show up here. If they didn't, more likely it will show up in direct traffic if they actually just typed hubspot.com in and went to it that way. If you want to set this up, you can do it by going into channel groups and you can set up a custom channel. you need to put in all of the AI platform conditions to make sure that you capture all traffic coming across through those different platforms. Then if you prefer, you can use a looker studio report. This is an example of one of our client reports from last year. And look studio can pull through this data automatically. So if you're already using Looker Studio, then brilliant. It's much easier to surface the data in here. You can look at traffic over over time. This type of trend is really really common. the uptick in AI search traffic, the number of sessions, and if you're e-commerce and you're tracking in your analytics, then also the revenue, the number of transactions driven by AI. And you can also split it down by platform. You can see in this example, chatbt is absolutely dominating both the traffic and the revenue splits. This is different for different businesses. ChatBT tends to be the most dominant for the majority. However, particularly if you're in B2B, you might see other players like co-pilot taking up a larger portion of the pie. This is incredibly important because just meaning you're optimizing for touchbt doesn't mean you're also going to do well in perplexity or co-pilot or Gemini. So if you know the platform that most of your customers and your audience use, you can focus in on that particular AI search platform and trying to be visible in the relevant answers there. Then there's also many great platforms. PKI is one. Try profound which is the example on screen is another that allows you to track visibility in prompts. I would always take this with a grain of salt, but what you roughly want to know is for the prompts you care about that your business is tracking, very similar to the keywords you care about, how often you are visible, and you want that to be quite high. So, in this example, this is Sandals Resorts. They are all-inclusive beach resorts for honeymoons and that kind of things in Jamaica. This is just an example. And it can break it down even further into multiple different queries and give you an overall visibility score. You just want to know for the prompts you care about, you are visible. And also in one of these tools, this is also profound. You can check things like sentiment. Now, sentiment is really important in AI because we want our brands to be spoken about positively, right? This is also sandals in the example and you can see overwhelming majority 72% feel really positively about the brand. AI speaks in a very positive way about them. And that's because of people's experiences. It's because of the types of reviews, the types of content available online about Sandals Resorts. So, it like people love their built-in honeymoon perks, great experience, legal planning and support, all gets done for you nice and smooth. But 27% feel a little negatively, and I think this is probably the guys because they get to those Sandals resorts, they're on their honeymoon, and you can't rent a tux. So, there's venue restrictions, there's extra costs, there's no tuxedos. Then what this allows you to do if you're sandals in this situation is address that negative sentiment, address those issues within your business so the AI is showing more and more positively about you. Okay. So that takes us to the future. The future of SEO is about owning these answers. It's about being visible in those AI overview. It's not about chasing clicks and the amount of traffic that we get. If you are visible in AI overviews, if you are visible in chat GPT and AI mode and relevant AI answers, people will learn about your brand at the top of funnel. The top of funnel is happening inside AI answers. Now, it's not happening in your website's blog section like it used to. And we need to think about how we bring those people to our website when they're closer to purchasing rather than when they're just at the very top of funnel research stage. So I know I mentioned in this presentation a couple of times a fantastic report that we put together that has lots of interesting data around how you can reshape your SEO strategy and some benchmarks for AI search. If you would like to get that report, you can get it on the screen or you can Google state of AI search report exposure ninja and you will be able to download a copy of this loads and loads of fantastic data and benchmarking to give you a place to start. Where do you actually go from here? Where are you going to leave this webinar? What are you going to do? There are three very clear steps to take to sort your SEO strategy and make it relevant for AI set. First is diagnosing those competitive gaps. Looking at your technical SEO, your content and your AI search visibility. You need to know now how visible your website is in AI search so that you can understand what needs to be done to address it. Then you want to work on establishing your authority in AI search. This is going to be fixing those technical blockers that have come up, building entity credibility, building that topical credibility in your website content and on PR content. And it's going to be owning high intent answers. You want to be visible in the answers where someone is likely to reach your business when they move down the funnel, when they're having those conversations in chat GPT. And then the third part of course scaling and defending your market position. And you do this through ongoing optimization, AI monitoring of those results and expanding your authority into the topical areas that you care about. If all of this seems a little overwhelming and you would like a hand, you may book a consultation with one of our fantastic marketing experts. We have an incredible team working behind the scenes here at Exposure Ninja and we are more than happy to talk about your search marketing and have a look at your AI visibility, give you a bit of an overview of where you are now and the opportunities for where we think you can get to. Thank you everyone so much for listening. We have a little bit of time for questions. So my colleague Dale who is beside behind the scenes is about to jump out. If you have any additional questions that you'd like to ask, now is the time. We've got about five minutes, please feel free to throw those into the chat and Dale and I will answer. Hey Dale. Hey Charlie. Just dealing with a couple of issues with the QR code at the moment. Apologize to anyone who hasn't been able to get the QR code to work for the report. I've added a clickable button at the bottom there. It should take you immediately through to the report so you can get it from there. Uh, and thank you to Anna who shared it in the chat as well. Um, thanks everyone also for engaging with me in the chat as well. I love all the questions that you sent through. I'm going to try and get through as many as I possibly can. There are a lot here. Let's just see what we can do. So, um, there are a few that are a bit too long. Uh, so I shared them with our head of search, head of AI search, Andy, and he'll be in touch with you by email to answer that one. But the first question I'd like to share, Charlie, is um, how can you view the fan out? Oh, I love this. How can you view the fan out? This is a great thing to be thinking about. So, the first thing for viewing the fan out, you can't 100% know the fan out for absolutely everything, but there's a couple of things you can do. The first is decide what prompt you actually want to understand the fan out for. So, what your search is, I would use perplexity because it it does the best job of showing what's going on behind the fan out. You can put your prompt into and then it will break down the query. I showed an example screenshot of that earlier. So, if you missed that slide, definitely go back to it and you'll see a breakdown of what it looks like. If you're thinking about AI mode, unfortunately, you can't view everything that's going on, but there's lot of content websites out there that will break down topical ideas. Um, there's a great one called Answer the Public, which is my favorite, that gives you a very good idea of what the query fan out is likely to look like. So, that's where I would start with those. Superb. Uh, next question is from Jinsen who's asking, um, what should the ratio of onpage versus off- page or digital PR and SEO be uh to stay on top of AI search? Oh, it's a brilliant question and it's a difficult one because you need to balance all of these areas. My general thinking is to think of this like a pyramid and at the base of the pyramid is your technical SEO. If your technical SEO isn't working well, everything else is pointless. If your site breaks, if you've got AI crawlers blocked, then it doesn't really matter what else you do because it's going to be really difficult to influence AI answers. So, your technical SEO has to be good. If you're a sort of smaller, medium-sized business, quarterly reviews will be enough for your technical SEO. If you're very large enterprise significant scale website, you'll be looking at least monthly if not weekly at your technical SEO. Onpage work has to come next. You want to show up in specific answers. So, you need to build that base. And then the tip of the pyramid should be your digital PR, your off-site work. Given that over 80% of answers site third party sources, digital PR needs to be a significant priority for AI search compared to traditional Google rankings. Dale, did you want to add something? No, I've just got some related questions. I'm trying to see how I can bring them in. Like one one question we have is um is it fair to say that keyword strong content is still king? Uh you know, do we keep amending the content that we have? Do we, you know, do we go about things differently? If you've got great ranking content for specific keywords and it's doing really well on Google, fantastic. Definitely keep it. You should be able to make updates to content that is performing really well to get it surfacing in AI, too. The only difference is sometimes content that ranks well on Google isn't necessarily the same topics that AI would be surfacing in their answers. So, you need to do some analysis around the content that you're doing. uh the content that you want to create to make sure you've got the right types of content. What you don't want to do is create the same blog post twice because you're trying to rank one in Google and you're trying to rank one in chat GBT. You should be able to use the same piece of content. A wellperforming piece of content in Google isn't a good place for that. You should also be thinking beyond keywords though. So a keyword is just is just part of it. We're used to writing content just to show up on one keyword. AI doesn't work the same. AI answers it's thinking about a whole topical area. So you want to have lots of relevant content around a specific content nicely organized in your blog section or in whatever yourformational section is. It might be a knowledge base on your website. Uh similar kind of question uh from Lawrence is asking when you said earlier on that blog post and not the future uh maybe paraphrasing there. What would you recommend to replace this with? I think that is a paraphrase. Um, blog posts definitely are still influential. I would say you don't want to be creating lots of top offunnel blog posts. Things like what is X? Like what is a mortgage? Those types of blog posts are very, very top ofunnel. AI is answering a lot of those questions. And it's unlikely someone searching that is going to then be looking for a mortgage lender. They're more likely 18 year olds and trying to find out what a mortgage even means. what that what what the definition of that is. So when you're thinking about blogs, you need to be thinking actually closer to middle of the funnel and bottom of the funnel content. What's actually going to drive people to your website and what is going to be surfacing AI answers that are actually relevant to the business that are actually going to bring people down the buyer journey hopefully again also to your brand also to your business. So that but then also you need to think very very strongly about your third party articles, your PR work, your digital PR, your directories, your reviews, everything that is being surfaced in the AI answers relating to your business and the topics you care about. I got a great question here from Charlie who's asking, "Are you noticing any preferences among audience demographics for particular AI tools?" Yeah, this is a great question. One of the biggest trends is that younger generations are significantly more likely to use AI. This kind of makes sense. They're very keen to adopt. So anyone who is Gen Z, millennials are likely to be using Chatbt, Gemini, Plexity, those types of tools. Other demographics, we also know of course B2B may use co-pilot significantly more. Not always, but because many B2B businesses have to work on Microsoft suite and so they're required to use co-pilot. The thing I would say around demographics for specific AI tools is there's lots of data out there, lots of interesting studies. What really matters is what your customers are using. And the best way to do that is to dig into your Google Analytics to see where the most referral traffic and conversions come from from the specific AI platforms. For the majority of businesses, it will be chat GBT, but given that it would take about 10 minutes to check, I think it's well worth the time before you start optimizing. rather than relying on third party data for de demographics of AI tools. This isn't a question, but it's just uh thank you ever so much. That's absolutely wonderful of you. We love to hear that. Absolutely. I had another question here uh from Philip who's wanted to know looking to the future, is it possible to determine whether the traffic that we're earning is from a human or an agent? Love this. Right now, it's not easy for us to know if the traffic is a human or an agent. My general feeling, if I was looking to the future, is not to get too preoccupied on this. Most likely, the direction of travel with search is that we will have more and more agents doing searches on our behalf. At the moment, search is very very manual for us humans. We have to go and search everything. Now that we have chatbt and other AI, we search a bit less because AI does more of the searching. Eventually there will be lots of tasks things like booking tickets or flight research for example that agents are able to carry out on our behalf. That is the direction of travel. The timeline for which anyone's best guess. I think it's hard to tell. But even if agents are searching rather than humans, you still want them to find your business ultimately at the end of the day. I've got one last question before we finish and it's from Kenna who wants to know um so love uh the line about top of funnel existing now to influence AI answers versus clicks but how can major publishers think about this regarding their bottom line so you revenue generation leads and so on. Oh, can fantastic question. Um, this is really important because what we're also thinking about when we work in marketing is how we report on this information and we're very very used to reporting on clicks and traffic. My recommendation is always to be thinking about the bottom line, the actual conversion. So, if you if you generate leads from your website, the number of leads that come through and if you're e-commerce, the number of sales that you're getting. Everything else is just a progress marker towards that revenue. What we're seeing more and more is even though traffic is lower. If you're getting really qualified good traffic, the conversion should still be good. You should still be getting the same number of conversions at the end, but you're getting less traffic at that top of funnel. You're getting less people just researching on your website. People should be coming to your website when they're middle bottom of funnel and they're actually much closer to making a purchasing decision. They're browsing your product pages or your service pages. They're maybe then hopefully sending you an inquiry. They might be getting a demo, downloading a white paper, doing product comparisons, something like this on your website because they've already done this top off of funnel research in AI. So my recommendation is always to be thinking actually about those bottom line metrics, thinking about your conversions, thinking about your e-commerce sales if you sell online. Superb. Well, thank you everyone for joining us. I imagine many of you got to hop off other meetings and possibly other webinars. Who knows? Or maybe to implement everything that Charlie shared today. If you do have any more inquiries or questions you'd like to have answered, if you head over to exposioninja.com, you should be able to find many of the answers there. But if you'd prefer to speak to the team, just head over to the contacts page. You'll be able to book a call and get some answers today. But if you prefer a review of like your marketing and website more generally, head to exposure.ninja review. Sorry, explosioninja.comreview. Uh, and you'll be able to request uh more deeper advice from our team and they'll be able to have a look over everything more generally for you. Amazing. Thank you everyone for joining us today. So,