Mike King: Open for bands like The Strokes, Yeah Yeah Yeah, Seal, and more. She can break dance and play DDR on expert level and still holds the Marin count country? County? County Middle School. It says country on this slide – Middle school record for girl shot put. Presenting The Vicious cycle of SEO, How We Got Here, Where We’re Going, please welcome Lily Ray.

Lily Ray: Wow. Hello everyone. Thank you all for being here. You’ve officially made it to the last session. Forty speakers. It’s no small feat. We’ve talked about so many things and I’m excited to hopefully wrap it all up for you guys today. So we will be talking about the vicious cycle of SEO, how we got here, and where we are going. Again, my name is Lily Ray. I’m the VP of SEO Strategy and Research at Amsiv. We are based about twenty blocks away from here, so excited to be on the home turf for once. Thank you so much to Mike King and iPullRank. This has been such an incredible experience. Speakers have been amazing. Last night with Busta Rhymes was amazing, legendary. So just so excited to be here with you guys today. And, yeah, this has been a little bit of a milestone for me personally this year because 2025 marks my 15th year of doing SEO professionally. I know a lot of people here have been doing it for even longer, so it’s a we’ve seen a lot of changes and a lot of things happening especially in the last couple of years as it relates to what we’re doing day day in day out, but 15 years felt like a big milestone. And I will say, if there’s one thing you learn after doing SEO for 15 years, it’s that SEO tends to follow somewhat of a predictable cycle, kind of a vicious cycle. So first, SEOs like us here today, we might find some type of cool tactic for basically exploiting search engines or manipulating search engines and getting a lot of organic traffic. And whether it’s at an SEO conference like us, here we are today talking about different tactics, or maybe you post it online or you post it on YouTube or maybe just your competitors reverse engineer the tactics that you’re using to drive traffic. So they kind of like share that publicly and this is what I refer to as an SEO gold rush.

And Google and the other search engines, of course, are watching very carefully. In many cases, they’re literally watching the posts that we make on social media. And they work on systems to kind of counteract and nullify those different tactics that we’re using because they basically treat them as spam in many cases. And once successful SEO strategies are then basically nullified, treated as spam by Google, they stop working as well as they used to. And being the industrious SEOs that we are, we go and look for the next opportunity. But the reality is that the tactics that we use as SEOs to gain visibility often result in a poor search experience for search engine users as much as we don’t like to admit it. So the reality is how SEOs kind of see their own projects is often very different from how Google and search engines feel about those projects. And if you want to know about how Google feels about spam, I talk about this video a lot because I still think it’s very relevant today. If you haven’t seen it, it was published in 2019 on Google’s official YouTube account. It’s called Trillions of Questions, No Easy Answers. And Google specifically talks about spam in this video. These are some quotes from real Googlers in that video. What they say is that spam is one of the biggest problems that they face at Google. They also call fighting spam a cat and mouse game. And they say this is a war that we’re fighting. It’s not something that they think will ever truly be solvable. And when SEOs watch this, I think a lot of us have this kind of default reaction where we’re like, this doesn’t apply to me because my content’s high quality. You know, what I’m doing is great. Every time there’s an algorithm update or a spam update, you hear all these SEOs saying, but my content’s amazing. Why did this impact me? Right?

The the reality is that the rules are continuously changing and the definition of spam also continues to evolve. And as much as we hate to hear it, a lot of the tactics that we have been doing industry are very much the same tactics that Google and the other search engines consider spam. So the Google versus SEO cat and mouse game is a tale as old as time. And we’re going to talk a little bit about those cycles today. So some of the old cycles of the past. And I’ve been kind of working on grouping these together this year and understanding this cat and mouse game between Google and SEOs. So first, way back when, I’m sure many of us remember doing keyword stuffing, doorway pages, cloaking, exact match domains, really manipulative stuff where you can just put white text on a white background and search engines gave you credit for that, which led to updates like the Florida update and the Panda updates. Then we had manipulation of page rank. Right? So we had link buying, blog roll and footer links, spammy anchor text, and private blog networks, which Google then met with, of course, the Penguin updates and link spam updates and nullifying a lot of those tactics over time. For a while, we had the ability to rank pretty easily on Google just by following SEO best practices even if the content that we were sharing and the information that we were sharing kind of like wasn’t really in line with scientific and medical consensus, which led us to the medic update and lots of core updates that really prioritized EEAT for your money or your life content.

Now let’s look at some recent examples of this cat and mouse game. So starting with the great AI flood. I know this is the main topic that we’ve talked about this week. But you know, these generative AI tools and all of us using them and SEO is jumping on the bandwagon and content creators of using these tools has led to a massive influx of AI content in the search results. So starting in about 2022, we see this huge boom of AI content. Lots of companies starting to automate their content creation process by using AI. We also get a huge surge in interest in AI content starting in about 2022. And it’s led to a lot of this kind of very homogenous feeling content in the search results where maybe there’s no author listed or maybe it’s a fake author and the content’s just kind of saying the same thing that everybody else has already said, not really offering anything original or anything unique. So we’re getting this kind of just feeling of this huge influx of content that just is written by AI, doesn’t really have that human element to it. And to add to this, we get a lot of people who are sharing their tactics very publicly. This is again what I refer to as an SEO gold rush. A lot of people are saying, wow, I can become a billionaire with ChatGPT and SEO. SEO can make you rich. I can make $414,000 per month with SEO just using these simple ChatGPT prompts and hacks and everything. So these videos actually have millions and millions of views. Lots of people are jumping on the bandwagon and trying to automate what used to be very complex SEO processes just with generative AI. And this also led us to really kind of a gold rush over the past couple of years, even with bigger brands that found these really high volume keyword and topic opportunities and pretty low effort in terms of generating the content to kind of match those queries. So you might have seen every publisher in the whole world in the last few years start to write about quotes. You know, best quotes for Mother’s Day, best quotes for your birthday, best quotes for Father’s Day, whatever it is. Jokes, you know, funniest jokes for your husband and funniest jokes for this and that. Every publisher has this content right now. There’s riddles, there’s what is your angel number, there’s all types of horoscope content. Every publisher has all kinds of horoscope content now. We have celebrity net worth, we have who are the most beautiful women in the world, and everyone starts to scale all this content because there’s a lot of demand for it. It’s relatively easy to publish, and of course a lot of them are using generative AI to publish this type of content.

We also get parasite SEO, so that’s what we refer to it as, as an industry. But really the idea here is you’re publishing third party content on a high authority website in an effort to kind of capitalize on that website’s ability to rank really well and really quickly with third party content. Which is usually some sort of like sponsored content or it has links back to your website or links to affiliate products. So there’s some type of incentive there. But that content tended to rank really, really quickly up until recently. And we’re still seeing different versions of Parasite SEO. So there’s a few different types. What I would say like one of the first types is user generated content Parasite SEO. So basically publishing on Reddit and Quora or Medium and LinkedIn Pulse or Hub pages is actually still working really well. And you can kind of publish whether it’s an article that has links back to your site or it has links to your affiliate products. And because Google’s prioritizing user generated content so much directly in the search results, this can and does rank extremely well, which we’ll talk about shortly.

There’s also sponsored guest posts. So this was working for a long time up until it didn’t recently. But basically here there’s a lot of publishers who were opening up new methods of monetization where they’re accepting sponsored guest posts. You pay to submit a post to their website. You can include links. You can include affiliate links and links back to your website or whatever it is that your goal is and very quickly rank because a lot of those companies rank really well, really easily on Google. So there’s a lot of that happening as well. And then there was this more kind of like secretive third party partnership thing happening over the last few years, which is another type of parasite SEO, which is essentially a lot of big publishers partnered with different types of companies like coupon companies or content creation companies or product review companies who were providing a lot of that content in the back end and they’re kind of partnering and they have these deals where there’s the external companies that are producing, whether it’s promo codes or product roundups or affiliate articles, partnering with those publishers and the publishers basically host that as like a separate site section or sub domain or sub folder on the site. That’s third party partnerships. So these are different types of parasite SEO. 

And if you want to see what that actually looks like in practice, some people have made their entire life mission exposing who’s behind the scenes of the different types of articles that you might read on some of the major publishers. Many different publishers were playing in this space where they were working with third parties to produce that type of affiliate content. And there’s an ex account named Googanuff who has been almost daily revealing who these partnerships are. I believe Google’s web spam team is looking very carefully at the post that this person does because they’re piecing together who it all is. So a lot of this was kind of being done in secrecy for a while. But in the last year or so, a lot of people in the SEO industry have actually taken it upon themselves to share what’s happening behind the scenes, make it very transparent, which is very helpful for Google, which we’ll talk about in a second.

And another big gold rush that I would say had a pretty big dramatic impact on the type of content that we’re seeing on the internet. I know a lot of us here have dabbled in this space. But basically programmatic SEO, which is using templates and databases to automatically generate very hyper optimized keyword pages that are hyper optimized, targeting hyper optimized keywords at scale and creating many, many pages at once. So you might see for like Zoom info, you might see a certain page template where there’s tens of thousands or hundreds of thousands of pages that kind of use these databases to generate. They have some unique insights but in many cases they’re kind of doing the same thing over and over and over. But they kind of match the same pattern, same template. And in many cases Google and search engines feel that this is not extremely helpful content. There are some exceptions to that. So some big examples of programmatic SEO that are done effectively are companies like Wise who has currency converters. And they have 30,000 pages that do currency conversion that generate two point seven million monthly organic clicks per month. And then Adobe with just 26 URLs, they do PDF tools. And that’s a different programmatic SEO strategy on a much smaller scale but it’s generating 4.6 million clicks for them per month. So when executed well, I’m not standing up here saying programmatic SEO is not great, but these initiatives here are really tied to core products and services offered by those companies and they really do provide helpful tools for customers.

But that being said, there’s been a lot of work being done in the SEO space and in the basically spamming space to try to mass auto generate lots of pages using some type of template, in many cases using very, very repetitive title tags. So you can often crawl websites that are hit by algorithm updates and just sort by the H1 or the title alphabetically. And you’ll see that they tend to follow these templates, kind of like cookie cutter templates for creating lots and lots of pages at scale. This was working really, really well up until, I would say, two years ago, or even last year during the March core update.  This is what Google means when they talk about writing content for search engines. So many sites that we’ve worked with at Amsiv on our team, because we work with a lot of companies hit by algorithm updates, We’re doing a version of this where they’re like, well, have such great content. It’s unique content. But do you really have expertise to write about all the best things to do in every city in the entire world? Right? A lot of companies were playing in that space. So in the past couple of years, it started to feel like this SEO content was everywhere. But the reality is real searchers, like real users of Google’s products and other search engines, were not so excited about this proliferation of SEO content. And we’ve talked about this before. I know Don Anderson mentioned this yesterday or a couple days ago. People are not loving this, like what SEOs have done to the internet. In fact, they’re kind of blaming us for a lot of the shortcomings of Google’s search quality lately. I know a lot of us have read and been interviewed in the people that ruined the internet by The Verge a couple of years ago. But they’re basically blaming the SEO industry for a lack of good quality results on Google and the inability to find anything because the content is so overly optimised. So this is the feedback that Google was hearing from its users and they were taking it very seriously. 

And all throughout 2023 and 2024, there were a lot of complaints that Google felt like it’s just too full of AI content, too full of spammy content, parasite SEO, all these things that people were complaining about. And Google said, buckle up. We’re working on it. I would say for a year or so they were working very seriously on making significant changes to Google’s algorithms to counter all these different kind of like spammy SEO tactics that were so prolific at that time. And that led us to the great SEO crackdown of basically starting in 2023 leading up to right now. I would say the helpful content update was a big turning point in 2023, September, for a lot of smaller sites, a lot of niche websites like travel bloggers and recipe bloggers, where they were doing a lot to drive a lot of SEO visibility and a lot of SEO traffic up until literally September 17, 2023. I remember exactly where I was because my inbox absolutely exploded. Thousands of companies were affected by this. And many of them were literally generating most of their revenue with SEO and lost 85% or 95% or more of traffic and visibility overnight. And there’s been little to no recoveries among many of these affected companies. So this is a really, really strong reaction by Google against that type of content. There was also a lot of collateral damage with this update because Google said that they’re using machine learning to basically understand what type of website it is and apply classifiers to websites. And lo and behold, there’s not a lot of precision involved in that process. There’s a lot of people that were really trying to do things right, were really trying to write with EEAT in mind and provide real experience and real expertise who were also destroyed by the helpful content update. And again, these sites that were classified as helpful content update sites have been mostly just wiped off the map, and very few of them have seen any type of meaningful recovery on those domains.

Which leads us to the biggest update in Google’s history, was very similar in nature. During the March 2024 core and spam update, which lasted 45 days, Again, it was the biggest update in Google’s history. They came out saying that they had a war room at Google where every day they were making sure they weren’t breaking the search results as they rolled this thing out. And according to Google, it reduced unhelpful content in Google search results by 45%. When you look at the types of queries that were impacted by this, remember before I mentioned every publisher is writing about quotes and jokes and riddles and all this stuff? A lot of it was that. A lot of it was companies who found these kind of opportunities and exploits where I can publish a lot of traffic really quickly or I could do these programmatic SEO strategies and capture all the possible ways that people are searching for this with AI and generate as many pages as possible, and it was working really, really well for them. In March, a lot of them lost you know, 70% or more of that visibility overnight, and very few of them have seen recoveries as well. So you don’t want to get in hot water like that with Google because it’s very hard to recover. And they didn’t stop there. Google continued to launch more and more updates throughout the year, Not just core updates, but spam updates as well. We saw some pretty big brands affected by Google’s spam update last year in December. And these were brands that were actually known for having pretty good SEO strategies. And a lot of content that a lot of us here emulated as like, this is ranking really well. It must be working. Google started to treat that as spam. And a lot of the times, it’s because these companies were, again, venturing into these very high volume keyword opportunities that didn’t have a lot to do with the business’s core products and services. And again, that’s what Google means by you’re writing content for search engines. You’re not writing content for humans. And if you go too far in that direction, you can expect to be impacted by Google’s core algorithm updates and spam updates.

So something that we like to do on our team is use tools like Ahrefs to kind of understand in certain site sections that were impacted, were they producing a lot of content very quickly? So you can see the organic pages report in Ahrefs and compare that to organic traffic for the same subfolder or subdomain or whatever specific site section you’re looking at. You can often see that there’s a big rise or exponential rise in the production of content and the traffic drops off at a certain point. That’s because it’s highly visible content that’s being produced really, really rapidly in many cases. And sometimes it takes Google algorithms a minute to catch up to the fact that this is not providing a good user experience. So even though it’s ranking well for a while, after a certain point you start to drop off dramatically because they’re saying this is content written for search, not really written for people. And there was another interesting example that was a full site pure spam manual action. So this is a website that literally was de-indexed from Google because they received this manual action. They were treated as pure spam. This is Cyrus Shepard posting about it on LinkedIn, but what he says is things were going really well until this site started to publish thousands of off topic articles on every topic under the sun. And guess what they wrote about? Happy birthday wishes, the most beautiful women in the world, Instagram usernames for girls and boys. It’s always the same thing. It’s always these very high volume topics that everyone’s trying to get a piece. Right? Because there’s so much opportunity there, except there’s not anymore. As far as parasite SEO, user generated content, like posting parasite content on Reddit and Quora and LinkedIn and everything, somehow is still okay. If you want to leave the conference today and you want to go spam Google, I would say that’s the best way of doing it.

Posting on Hub pages and LinkedIn posts still somehow works because Google loves user generated content. But as far as sponsored posts and as far as a lot of the third party partnerships that we talked about, Google’s penalized those. A lot of those companies have received site reputation abuse manual actions. It’s very hard to come back from those. You have to no index that content. And Google also changed the rules a bit. So in November of last year, they actually changed the site reputation abuse guidelines one day before they penalized dozens of companies. So what they said is that freelancers, and when you use freelancers in a way that abuses the host site’s ranking signals, those two things have to be true, that’s a violation of site reputation abuse, And it doesn’t matter if the freelancers or the third party content creation company worked with your first party website, the people at your website, for your first party staff to review the content, to edit the content, to co-author it. That doesn’t matter. Google said if there’s any level of that freelancer or third party content creation happening in a way that we believe abuses the host site’s ranking signals, that’s site reputation abuse. And you basically get a penalty and you cannot come back from that unless you no index the content. And if you want to come back from that after that, you have to publish that content with first party staff exclusively. And this is the Google web spam team doing this. So we have the web spam team now making judgment calls about the nature of these types of relationships. Google’s been very careful to say it’s not all freelancers. Freelancers are not the problem. They just have to be writing freelance content in a way that violates the host site’s ranking signals. So pretty subjective, but that’s where we’re at and this has affected dozens and dozens of major publishers over the last few months.

It’s been a bloodbath. This is not a secret. There’s a lot of very big companies and big publishers who have lost tens of millions of dollars as a result of Google’s change of new policies as it relates to site reputation abuse. And a lot of these companies are now moving that type of content production in house. Some of them are hiring the freelancers to write the content for them. But Google’s changing the rules here. And this is something that sites got away with for a long time that they’re no longer able to. So where does this leave us? It feels like for every SEO opportunity that we’ve had over the years, like directory submissions and exact match domains and PBNs and niche sites and product reviews and parasite SEO and AI content, it feels like Google has basically patched each and every one of them up as an exploit. You know, they see what we’re doing and they say, nope, not anymore. We’re going cut that off. We’ve had the helpful content update, reviews update, scaled content abuse, spam updates, everything’s just been patched up. And at the same time, we’ve talked about this a lot over the last few days, but Google’s revealed its shiny new AI toys. So we get AI overviews and we get AI mode. So not only is it harder and harder to use SEO to drive traffic, but the traffic is obviously dropping off dramatically as a result of these new products. I think we’ve seen this table a few times already this week. This is from Wil Reynolds and Sierra Interactive. But basically they did an early study and now we’re starting to see some new studies about the fact that AI overviews obviously cut into organic click through rates in a pretty dramatic way.

So it leaves a lot of SEOs wondering, is the opportunity all gone? Where do we go from here? And I think that this requires a shift in mindset. I think that you can do great SEO without being stuck in this perpetual cat and mouse game. So the history of SEO reveals that the rules are always changing and adaptation is key. So this leads us to what I’ve been calling kind of like a new frontier of SEO. I know like Mike and a lot of us have talked about this is a really exciting moment in SEO actually, even with all these challenges. So we’re going to talk a little bit about what I call authentic SEO, as well as some tactics for optimizing large language model visibility in what I believe is an ethical and authentic way. So first and foremost, we’ve had some of these tools here with us today, like Profound did a talk. But there’s brand new tools on the market. And if you work at an agency or you work in house, you should be considering adding these tools to your repertoire.

Our team is working with Profound and ZipTie and a few others to kind of understand who’s appearing in the large language models and how can we kind of give that data and those insights to our clients. ZipTie is fantastic for understanding granularity as it relates to AI Overviews on Google specifically. There’s more here, but these are some of the big ones that have been developing these products over the last year or two.  And of course, we’ve talked about this a lot already at the conference, but you have to be tracking large language model traffic refers in Google Analytics four or whatever analytics tool you’re using so you can understand user behavior and which tools they’re coming from, how they’re interacting with your website, are there conversions, are we seeing good user experience on our site from users that are coming in from large language models. Very easy to do in GA4. And a lot of people that are doing this type of tracking, including some of the clients on our team, were seeing the conversion rates and the user experience is actually higher from traffic that comes from large language models, which isn’t too surprising if you think about asking ChatGPT what’s the best pair of running shoes for women that like to do whatever. And it tells you this is the exact shoe you should buy. It sends you to the product page. Of course, you’re going to see higher conversion rates from that traffic.

And the reality is we’ve talked about this a lot, but SEO is no longer just about rankings and traffic. So a lot of this is just communicating with your bosses and your clients about the fact that the KPIs that we used to be focused on, hyper focused on, are really changing. So it’s about visibility now. It’s about how much our brands and our products and services and our people are being mentioned within large language models and also the sentiment around them. I know that there’s been a little bit of controversy around whether or not it makes sense to actually do sentiment analysis within large language models. Tools like Profound are helping us understand this. I personally think this is a really, really cool opportunity because what you can see here with this is a view from Profound. What it’s basically saying is that for the different products and services that, in this case Intuit QuickBooks offers, what is the general consumer sentiment around those products? And if you see that there’s a lot of negative sentiment, you can kind of take a look at like what are those citations? What are people saying about it? And how can we then as a brand go back and take a lot of that consumer feedback that wasn’t positive and reframe our products in a different way. This is also Profound, you can kind of emulate this yourself with whatever tools you’re using. But basically, what is the accuracy of the answers within large language models? So if you’re Samsung, for example, when people are asking about Samsung products and services within large language models, how often is it getting the information right? Profound has some new visualizations to give you that gap of where the information’s wrong. And I think there’s a lot that brands can do to get ahead of that. So for example, if you ask, what are the accessories that come with a new iPhone, you know, and it gets it wrong, then maybe apple dot com should think about like, let’s make sure that we are very clear and we’re writing a very clear language that they’re citing us as the source as opposed to some third party or some Reddit thread or something that might be inaccurate.

And a lot of this is we talked today, I think it was Dejaune that mentioned this, but it’s kind of like basics, like fundamentals of SEO. At one point, my title recently changed in the past couple of years. And unfortunately, my old title was all over the internet. And in order for me to really start to see my new title appearing consistently in large language models, It’s just a matter of going across all those different places that I’ve been cited and asking for them to change the title or changing the title in my bio. It’s a lot like NAP consistency in the local SEO space. Name, address, and phone number, just making sure that it’s consistent across the whole ecosystem. Believe it or not, to this day, I’m still going out there and finding new examples of my old titles. So it’s a lot of work. But of course, the more consistent that you are and the more times they’re seeing that same consistent title or tagline or whatever, you’re much more likely to see that appearing correctly in the large language models. And this is also some data from Procound. I’m sorry it’s a little bit hard to read, but I will read it out loud to you guys. These are basically the sites that are most highly cited within the different large language models. So for OpenAI, we have Wikipedia and G2. For Perplexity, we have Reddit and YouTube. For AI overviews on Google, unsurprisingly, have YouTube and LinkedIn. And for Copilot, have Forbes and Gartner.

I think with AI Overviews, it’s very interesting because it’s relatively easy to influence if you’re publishing content on YouTube and publishing content on LinkedIn. Whereas with Perplexity, they’re much more reliant on those Reddit conversations. But if you start to get an understanding of which large language models are sending the most traffic to your site, then you can develop a strategy around like, well we know that those really draw upon Reddit threads. Or we know that those really draw upon YouTube conversations. So that can kind of frame how you want to target or produce content across those different platforms. And laser focus on highly cited articles in your niche. This is from ProFound. You can get a good understanding of how frequently articles have been mentioned or domains have been mentioned as part of the citations. They also give you a breakdown of how much of that is social traffic and direct traffic. So if it’s social or direct, you have much more ability to influence the conversations that are happening across those platforms and get ahead of those conversations. You can also do something where you can basically take the most frequently cited articles within a certain category in a large language model. And you can then run that article through a large language model like ChatGPT and compare it to your own products and services and say, hey, like this is an article that’s being really frequently cited in large language models about best running shoes or whatever. And my client is Nike and I want to understand what they’re saying about all the best running shoes in this article. Compare it to what we’re saying on our website or what people know about us and what are some of the And it can give you some examples of like, well, if you want to consider being more heavily cited and have a better experience, a better sentiment across the answers in those large language models, these are some areas where the site can improve how it describes those products.

Another thing that we’ve been doing is using tools like Profound or ZipTie to extract all the scroll to text highlighting within the URLs. Because basically when it scrapes AI Overview answers, it also collects that scroll to text highlighting. If it jumps to basically the paragraph of text that it’s pulling from, you can extract that. You can take all that data, clean up that data, and then you have all the different paragraphs of text that it’s highlighted for AI Overviews. You can put that into a large language model and say, help me summarize all the different answers that AI Overviews have jumped to. And what have they said on their page? And how can we kind of produce that content on our page as well? Jess Scholes talks about question queries in Search Console. You can basically come up with some type of regex rule like this for all the different question queries somebody might be asking. This is a directional way of maybe understanding which keywords are generating reviews for you because we don’t actually have that data in Console. But then you can understand which questions are people asking and how can we surface answers to those questions directly on our website. Also, we talked about this a few times in this conference, but podcasts and audio and video tend to get picked up in large language models as well. So if you’re doing an interview, you can kind of guide the questions. A lot of times a podcast interviewer will say, is there anything you want me to ask you? You can tell them what you want to talk about because lo and behold, that gets picked up very quickly in large language models.

Always have a Q & A on your brand website or your personal website. Get ahead of any conversation or any question that you think people might be asking about you. I have an FAQ on my personal website where I talk about what are my favorite cities or whatever. If you ask any large language model, it’s always referencing directly what I say on that page. So it’s a really good way to control information about you. Watch out for something called pronoun resolution failure. Basically, if you I had a I had a page on my site that talked about my dog. She’s a female dog. I said, I own a dog named Marcy, and she is nine years old. Large language model didn’t know that the she was about Marcy, they thought it was about me. So for a while it was calling me nine years old on Google. So there’s some really quick work that you can just do to make sure that you’re using the right pronouns so that large language models don’t get confused. 

I have a few more minutes. SEO is not all doom and gloom. So this is some data from Citrix over the past couple of years. You might want to take a picture of this because these are all the brands, not all, but many of the brands that have been seeing significant visibility increases in Google over the last two years. So there’s a lot of really great examples of what’s happening in the SEO space. And some common traits among them, they have a lot of brand awareness, a lot of navigational searches, really strong EEAT for your money, your life topics. They’re often the direct providers of products and services, and they often house a lot of user generated content, real world experience, and authentic reviews. You might be wondering about affiliate sites. What are they doing? It’s very hard, but basically the ones that are succeeding are using first first person opinions and insights. They have robust review methodologies that they include on every page. They have real videos of them actually doing the product reviews, original photography, and in this case it’s a fitness website. They have real personal trainers that are reviewing all the content.

There’s still a long list of approaches that search engines can never take away from us. This is a website called WalletHub that’s doing a lot of original research that’s driving a lot of links and monthly organic traffic and referring domains to their website. So this is an example of some of the research studies and surveys that they’ve done, how many links that’s generated for them. Turns out that it’s not just SEO traffic and links that they’re driving. They’re being cited in large language models very frequently. In many cases, if you type the name of the study like property taxes by state in 2025, can see WalletHub being mentioned sometimes ten times in a ChatGPT response. And if you have original content, you need to repurpose it into many different formats. So if you have an article, you want to make it into a video, splice up the video into short videos, make it into podcasts, talk about it on Reddit, about it on forums, use visual references, make it into a conference presentation, upload your deck to SlideShare, talk about it on social media, so you can take one piece of content and repurpose it in many different ways. YouTube and TikTok, we’ve talked about that a lot. I’m running out of time.

One example of this that I did recently was I recently spoke at Brighton SEO. I did a survey where I just quickly put a survey together. I talked to 160 people that have been working in the SEO space for a long time about historical SEO tactics. I then took all my survey responses, mapped all the historical SEO tactics to different historical periods. So we get the SEO Ice Age and the SEO Stone Age and the Golden Age and the Renaissance and we made this kind of like fun and cutesy explanation of SEO history and all the different tactics that have worked over time. My team at Ampsiv have helped put together some infographics about the different eras of SEO history. They had me talk on video which we then spliced into a short video talking about these fun historical SEO periods. And lo and behold, two days later, if you ask ChatGPT what are the seven eras of SEO history, it’s still to this day. If you ask it, it’ll answer with all of our historical periods that we made up. So that was a pretty fun and easy process of kind of creating something new and original and using AI to kind of help you with the creative flow that you’re using. And I asked it, is this actually an authoritative article? And because I published it on the Ampsiv blog and it said it talked about my accolades and said, given her extensive experience and active involvement in the SEO community, Lily Ray’s analyses and writings are considered authoritative. So this leads us to having a strong personal brand and authentic social media presence. Always consider having your experts do everything you can to have them be trusted experts in Google’s knowledge graph because think about the fact that some of the biggest SEO winners in the last couple of years are social media sites, Instagram, Threads, X, LinkedIn posts. These these individual articles and posts and reels on social media are being indexed and ranking much better than they ever have.

And this is what Rand talked about earlier. And this is why you want to appear everywhere and post often across all the different walled gardens. Me every day, you guys probably know this, I post on X, Blue Sky, LinkedIn, Instagram daily, all day, every day, SEO stuff, all day, every day, because it’s being picked up all over the place. And Google loves now what people are saying. It’s a new SERP feature that literally highlights contributions from individual creators, their social media posts, their videos. So there’s lots of different ways that Google uses this user generated content from different platforms. Social posts are also heavily cited in large language models. So I’ve been playing around a bit with just asking different large language models questions about my personal life, which are only answered by my social media posts. But lo and behold, I ask, where does Lily Ray spend January? I’ve spent my last four Januarys in Argentina. Because I talk about that on Instagram and because I talked about that on Facebook and Blue Sky and a few other places, each large language model knew that that’s true.

So becoming a strong personal brand, posting frequently on social media is something that’s really easy to influence what appears in large language models. Dejon mentioned this earlier in the day, but LinkedIn Pulse is super powerful. If you have a good robust following on LinkedIn and you’re considered authoritative in a certain space, basically you can earn top rankings really quickly. So I published an article on LinkedIn Pulse about Google Discover. Within six hours, I was ranking number one on Google for Google Discover Optimization 2025, something like that. Within five hours or six hours, it was number one on Google. It was in the featured snippet. It was also within it was number one on AI Overviews as well as AI mode on Google. So if you have experts that you can work with on LinkedIn or your own personal LinkedIn profile, LinkedIn Pulse is currently very powerful. But proceed with caution because like looking at the history of SEO and these cycles of SEO and this cat and mouse game, my crystal ball tells me that there’s going to be a lot of let’s say unethical behaviors as people start to learn how to optimize large language models. I believe there’s going to be a lot of bribes. I believe that we’re going to start to see, okay, well they’re the most cited article, you know, in ChatGPT in this category, so let’s all just bribe them, send them money to try to get my products listed or my brand listed.

And like with site reputation abuse, you want to think about, is this really the most sustainable strategy for us as especially consumers start to catch on to the fact that they can’t really trust what they’re reading? Or should we be thinking about doing something in a more organic and ethical way? And I also believe that we’re going to be due for some serious anti-spam measures in searching LLMs. We’ve talked about a lot of different topics this week and I’m seeing a lot of people post online about different ways of manipulating LLMs and everything. And I do think that we’re going to be met with some level of resistance from, of course, Google, but also OpenAI as they kind of start to like demote and nullify a lot of the tactics that we’re using to try to influence the answers. So just be careful and consider which side you want to be on as we kind of start to dive into this new era of manipulation.So takeaways. Don’t get stuck in the cat and mouse game. You’ll eventually get caught. Consider investing in original research, community building, authenticity, and thought leadership. Build a strong personal brand for the experts that work at your company because social media is currently generating significant visibility within large language models. Focus on the things that search engines can never take away from you. Thank you so much. Enjoy. See you guys.