Ranking in AI Answers Is Hard Until You Fix This

There's an entire layer of search happening right now that you won't see in any web analytics platform, no matter which one you're using. It's stuff like this where people are going online, they're learning about a topic, they're asking follow-up questions, they're doing all of their research, sometimes without visiting a single website. They're using AI mode, AI overviews, and chat GPT. The first thing you might see as a marketer is a click from Google organic search, which turns into a conversion. All you will see in your analytics platform is maybe a referral visit from chat GBT or you might just see a branded search where someone's typed in your brand name on Google and immediately purchased. But where did they come from? How did they get to that place where they were ready to purchase? And how can we get more of them? Well, increasingly they're coming from AI search. They're doing all of their research in AI platforms, not visiting your website. They're finding out all about you, your products, and your services, and making the decision to buy from you without ever visiting any of your web pages. So, there are two things immediately that all businesses need to do given that the number of these sort of dark conversions is increasing for most businesses. There are really two things that everybody needs to do. The first is work on attribution. Work on understanding how many of your conversions, whether they're sales or leads, are coming from this type of search. Increasingly across Exposure Ninja's clients, we're doing this via selfattribution. I.e., we're adding fields on forms that say, "Where did you hear about us?" Yep, the old school route. If you have a sales team, it's a good idea to encourage them to ask customers this question. And this can lead to some really interesting conversations. Where did you hear about us? Oh, I was just looking into so and so on chat GBT. Huh. The other thing that we all need to be doing is making sure that we're regularly running AI search visibility audits. Two ways to do this. You can either do it manually where you put some queries into Google and Perplexity and Chat GPT and see where you rank or you can use a tool like Profound, who by the way sponsor today's video, which will give you visibility score and track that over time. If your AI visibility is really good, then happy days. You're obviously going to want to maintain and increase that further. If your AI visibility is maybe not so good, you've got a bit of catching up to do, great. That's exactly what we're going to cover in today's video. So today we're going to be building an AI search visibility pyramid. This covers the four main areas that you're going to want to focus on, but there's a secret magic bonus layer at the bottom. That's actually we've observed the main reason why most businesses aren't able to improve their visibility in AI search. Our AI search visibility layer starts at the very top with layer one, your reputation, part one. I've just run a search on Google for Vanguard Roth IRA. Any good? This is a bottom of the funnel search, meaning this is searched for typically by someone who's about to make a purchase. And the value of that customer could be quite high. Now, notice what's happened here. We've got an AI overview right at the top, which is dominating this search result. Remember that 83% of searches with an AI overview don't result in any clicks because people are satisfied with the answer that they find in the AI overview. So, for Vanguard, it's incredibly important that what this AI overview says about their Roth IRA product is positive. But how does the Google AI overview or any other AI tool like Chat GBT decide what to say about the Vanguard Roth IRA or your product or service? It's got the information and the data that the large language model has been trained on, the data that the AI already knows about. And then it's got the information that it finds online when it carries out a web search. Now, the good news, bad news is that when an AI tool like AI overviews or chat GPT is asked about your products and service, it's going to share what it knows and what it finds out, both the good and the bad. In the chat GBT version of this search for example, we can see chat GBT referencing and recommending other providers. So influencing this first layer of reputation, i.e. what do these tools say about your product and service is really important. These tools are looking for consensus about what you sell. What are the consistent messages about your product and service? Are people slating it for an outofdate user experience or a narrow set of tools? or are you known for being low cost or having great customer service? And the reality of this first layer of the AI search pyramid, reputation about your product and service, is that actually it's very difficult to promote a bad product. Just like with any marketing, the more attention you get on a bad product or service, generally speaking, the worse the reputation gets. But on the flip side, if you've got a great product or service, that can be amazing leverage to help you improve your AI search visibility. one of our clients, Zugu case. We're now dominating chat GPT and most of the AI search tools when you ask it about the best iPad cases. Now, part of the reason is because of the AI search optimization that we've done, but we would only be able to get these sorts of results if the product was genuinely really good. In AI search optimization, we call this type of reputation work sentiment. and we run AI sentiment analysis where we seek to understand what these different AI platforms think of your products and services. If you set up your brand and you go to sentiment, you can see what these AI platforms think of your products and services. So, for example, on the index fund side of things for Vanguard, we can see that the sentiment is broadly positive. And we can see some examples of recent queries that people have made. Uh here's one for Microsoft Copilot talking about Vanguard's um Roth IRA product and we can see the sentiment and we can see how positive it is. Now the next layer of our pyramid is also reputation which is why I call it reputation part two. But rather than being reputation about your products and services, this is reputation about your brand. You'll probably notice that when you ask a tool like chat GPT about a product and service, it will answer about that product and service, but it'll also give you some feedback about the brand itself. For example, here Vanguard is widely respected for low expense ratios on index funds and ETFs. So, yes, that's some feedback about the product, but it's also about the wider brand. Here's some more. Vanguard has decades of experience and billions in assets under management. Huge credibility there, but that's at a brand level, not a product level. And this means being really clear about exactly what your brand stands for and making sure that that is communicated across your different platforms online. We recently worked with a massive multi-billion dollar company that owns lots of personal care brands, brands that you would know about. Now, this business wanted to be known as being innovative. But the problem is when you asked any of these AI search tools about what this brand was known for, innovative was never amongst the answers. Now the fix for this wasn't just tweaking a few words here and there on their website. This actually required a lot of integration between different departments in the business. Yes, the content team needed to be focusing more on the innovative side of the business. The PR and the com side needed to be talking about the innovations that were going on across the organization. The social team needed to talk more about innovation as well. And of course, they needed to feed all of this information through the other layers of the pyramid that we'll talk about in a minute. And whilst this work might sound a bit annoying or like you're just tempted to skip over it because it sounds a bit messy and change managementy, actually you can get a huge amount of leverage from this as we'll see in some later examples in this video. The next step is to make sure that those messages are clearly communicated online. I've just searched for an M&A lawyer for a fintech acquisition and chat GBT has given me a whole bunch of recommendations here. It's really pretty good reputation. So, if this is your space, you might be thinking, well, how do I get my brand recommended here? Do I just need to update my website? But that's where things get a little bit tricky. Look at the number of sources that ChatGBT has pulled from in order to provide this answer. It's scanned through 40 websites plus consulted its own internal training data. There are a few types of websites being cited. Yes, of course we've got company's own websites, but there's also a lot of third-party websites like Legal Business or Legal 500, Leaders League. In this case, you can think of these as kind of directory type sites that combine lots of information about different law firms. But increasingly, these AI tools are choosing to pull their recommendations not from a company's own website, but from third-party websites. Over the last decade at Exposure Ninja, we've been using digital PR to get our clients featured in all sorts of publications, mainly to improve their visibility in organic search by improving their link profile. But digital PR takes on another level of importance in this world of AI search optimization because it's how you get featured in the sort of websites that these AI tools are pulling from to recommend you to their users. And it's important to recognize that the websites being cited by Chat GBT aren't necessarily just the top ranking websites for this sort of phrase on Google. In fact, only 12% of the URLs cited by ChatGBT appear in Google's top 10. So digital PR has moved from just hey we want sites that have really good high authority talking about your brand to we can use third-party websites to shape these LLM's understanding of your business your benefits your positioning and of course if we target the sort of websites that these AI tools are drawing their recommendations from we can increase the likelihood that they're going to recommend you and each of these different AI tools will take recommendation from different websites. Going back to our client Zugu case here, I am in Google's AI mode and I've asked it what's the best iPad Air case. It's recommended the Zugu iPad Air case. Great choice. But look at some of the websites that it has referenced in this answer. For example, this ZDNet website is the one that is cited top. If we go on that, guess what? Zugu case is being recommended here. So helping Zugu get visibility on sites like ZDNet and Wired and loads of coverage on Reddit helps it be recommended in these AI search responses. And by the way, this ZDNet article was being referenced by ChatGpt and Google AI overviews within a couple of days of being published. So results can be really rapid in this area, although of course it takes absolutely ages to get those features arranged in the first place. As we said earlier, a lot of these tools tend to prioritize what other websites are saying about your products rather than what you're saying about your products. Take a look at this one. Best hyaluronic acid serum. I'm not looking to spend loads. And you'll see our client, The Ordinary, showing up here as the number one recommendation, which we are very happy about. But you'll notice the information being provided here is not from The Ordinaries website. It's from British Vogue. It doesn't matter if we edit the content on the ordinaries website talking about their hyaluronic acid. This product is being recommended cuz it's being shown in an authority publication to be the best. Same here in Google AI overviews. Now, Google AI overviews is generally happier to pull information from your own website. But for this search, what's the best value hyaluronic acid serum? The ordinary is still being recommended, but the link is not The Ordinary's own website. It's actually Cult Beauty and the Independent third-party websites. Now, there's no shortcut to doing this. This involves an immense amount of work, typically with either a digital PR team like the one at Exposure Ninja or someone else who is perfectly okay with hitchhiker levels of rejection. And that's why the DPR team at Exposure Ninja have to mentally prepare themselves every morning for another day at the rejection coldface. But when it comes off, the results can be extraordinary. Now, inside profound, you can head over to the citation section, which shows you where each different AI tool is pulling its responses from. So, you can break this down by different platforms. If you wanted to just, for example, see the types of websites that chat GPT is citing, you could choose chat GPT. Or if you wanted to go for all of the platforms to see where you get your the most bang for your buck across all of them, then you obviously go select all. In this case, we can see if we're promoting Vanguard, we can see the number one domain for citations, i.e. across all of the platforms, across all of the prompts is Nerd Wallet, then Bankrate, then Fool, then Reddit. So, we know which domains to put our work into getting citations on. Now, if you want, you can filter this by topic. So, let's say that we just wanted to improve visibility for the Robo Advisor terms. We could choose Robo Advisor. And let's say we wanted to filter it. So we're just focusing on chat GBT. We could put that filter in as well. This shows us then the domains that we want to go after if we want to improve our visibility for these areas. And by the way, if you're interested in getting some coverage for your brand, comment down below with what you're selling. We'll comment back on the first few to give you some ideas for how to get some coverage. And by the way, if you want help improving your business's visibility in AI search, that's exactly what the team here at Exposure Ninja can do. You can get started by requesting a free website and digital marketing review. We'll take a look at your current visibility and suggest some priority actions that you can implement to improve that a bit. Now, we will often ask you for a call first. We're not trying to sell you anything or pitch you anything at all on this call. We just want to understand a bit more about your current situation so we can tailor our recommendations to be as useful as possible. This free review is typically delivered two to three working days via video since your email. It's completely free of charge, but not everybody is eligible, so you do need to apply for it over at exposurinja.com/re. Okay, onto the fourth layer of our pyramid then. All this visibility is for nothing if we don't at some point get people onto your website in order to make a purchase. In order to do that, we have to have some visibility for your own brand, your own domain, your own website pages. And actually, if you have a look in profound, you'll see that typically your own website is one of the top cited sources unless there's something going wrong. Of course, the holy grail is to get your brand recommended in the answer and your own website being cited as the source. That way you get both the mention and you also potentially get the click from the user. We found that AI tools like Chat GBT, like Google's AI overviews prefer content that has been structured in a certain way. Things like content chunking, clear headings, FAQ sections, statistical data points. All of these seem to increase the chances that your content is going to be referenced as a source. And this is also work that we did for the ordinary. For example, if you go on to Google and search for how do I build my skincare regimen, you'll see that the AI overview gives you some recommendations here and then it links through to the skincare regimen builder on the ordinary website. We wanted to help this content get featured here and ranking organically because yes, you can get some information about how to build a skincare regimen, but actually it' be so much better if you just went onto the website and used the tool. Not least because if you use the tool, you're going to be end up buying a bunch of the ordinary products. A lot of this is about the structure of the page and making sure it's produced in a way that these AI tools like to draw from. And there can be a lot of manual work involved in analyzing these pages and fine-tuning them. Actually, one of the areas that you can use as a shortcut to do this is profound. And it's rolling out these new AI agents which allow you to do this in a really nifty way. Let me just give you a quick demo. So, if we go over to the AI agent section, it's still in beta at the moment, but you can try it out at exposure.ninja/profound. You can start from a bunch of templates or you can click to see all. There's a ton of templates already built here, like you can generate blogs from YouTube videos. So, you just put the URL of the YouTube video in and it can then generate a blog post for your website from that URL. You might want to tweak this a little bit, but this does a whole bunch of that work. But the area that's most useful for optimizing onpage content is the content optimization suggestions. What this tool does is it scrapes your web page. It analyzes it against Profound's answer engine optimization content scorecard to see what you might need to change. It gives you recommendations and then it presents you with the finished article. So let's just stick in a URL here. We're going to use the Vanguard ETF page and we're going to say we want to optimize this for best ETF. If we just click run, it's going to go through its process. And here's what we get from the tool as the final optimization suggestions. You can see these are all really clear. And yes, there's a bit of work to be done to actually get these things built in. But if you do this, the quality of your content will be much higher for users as well as AI tools. Now, also in this fourth layer, the citations layer of the pyramid is all of the technical on-site SEO optimization that is absolutely crucial and required to make your website as crawable and indexable as possible. This is fairly technical stuff. We have an entire team at Exposure Ninja that works on this for our clients and our own brands. I'm not going to get into loads of detail on this here. We've got other videos talking about it and I'll be honest, it's not the most exciting stuff for a video. But without this layer, you can really limit your visibility in these AI tools. And now we come to the real reason why your AI search strategy might not be working. And this is from our observation talking to loads of businesses. In fact, I've just been on the phone with a 200 million pound business that's trying to get their AI search optimization strategy up and running and they've got stuck in exactly this area. This is the friction layer aka getting the support that you need to get this strategy rolled out because the reality is that AI search optimization actually involves lots of different people or teams across the organization. It needs buyin at the seauite level to understand the value of doing all of this. You might have leaders of different brands that also all need to be bought in. There might be a development team that needs to understand the importance of optimizing the website for these AI tools to be able to crawl them. You might need to get buyin from your web development team to make some changes to the site. You might need buyin from your content teams to optimize some of the things that they're doing and change their approach slightly or target some content to different topics that you give them. You might need buyin from your digital PR team or your PR agency to focus their attention in certain publications which you know is going to help your visibility in these AI tools. You might need brand teams, social teams, customer service teams all bought in. And of course, everyone has their own agenda, their own mental baggage. They may have tried some things before. They may have worked with an agency on parasite SEO and that didn't really go well and now they feel burnt. We know that this can be really difficult and getting buyin across the organization to roll something out like this is a massive challenge for marketing leaders. The best approach that we've found is running sessions with as many people across the organization as possible. We start right at the top what's going on in the world of AI search, explaining to people the basics of zeroclick search and why traffic might be dropping and why that's not the end of the world. if you're visible in this new era of AI search tool. It might sound corny, but it really is about taking people on that journey to see how they can contribute to what really is clearly the future of digital marketing. And one of the biggest shifts that we've seen since the advent of search without this full buying across the organization, the danger is that you're only ever able to poke at the edges and you never really get the traction that you need. whilst smaller, more nimble companies are able to roll things out much more quickly cuz they don't have to get sign off from anyone. So there you have the five layers of the AI search optimization pyramid. You've got reputation part one, that's about your product and service. Reputation part two, that's about your brand. You've got the promotion, are you getting your message across the net? You've got the citation layer, this is about optimizing your content and your website to be featured in AI answers. And then you've got the final hidden layer, removing that friction layer to make sure this stuff actually gets rolled out. Now, if you want to work on improving your content for AI search, watch this video. And by the way, if you haven't already subscribed, please do so. We recently hit over 100,000 subs. I've got the plaque up there, and it's the only thing I've ever done in my career that my children actually understand. Until next time.