From SEO to GEO: The New Rules of Search, Discovery, and Strategy in the AI Era

Welcome and Agenda

Tom Newton: Amazing. Hey everyone. Welcome. I'm so glad you all can make the time to join us today. We have an amazing session lined up. I'm very excited to be here with you all. I'm Tom, VP of Revenue Marketing at Jasper, and I'm going to be guiding us through today's conversation. We're covering something very timely. You all know this because you're here. Everything from the future of SEO, Geo, AEO and beyond. And we have an amazing guest joining us, Kyle Byers, who's the Director of Growth Marketing at Semrush, who has amazing firsthand experience and data to share as well. Before we get into the meat of the conversation, starting with, I'm going to start with a little housekeeping for the agenda. We'll start with a quick state of the state in terms of the SEO landscape and search marketing. And then we'll jump into the main event, a fireside chat with Kyle. Then we'll have a quick Jasper demo from Mason on our product marketing team, who walk through some quick wins you can implement in terms of AEO and geo. And finally we'll open it up for your questions. Let's jump to the next slide if we can, and the next one got a little bit of a delay, so yes, please don't be shy. Drop your questions into the Q and A throughout and we'll make sure to get to as many of them as possible at the end. Also, a quick reminder, we will be sending the recording of the session out to everyone who registered and yeah, I think that's it. Thank you all for being here. And with all of that housekeeping out of the way, let's dive in, starting with this chart.

The Evolving SEO Landscape

Tom Newton: Many of you have probably seen this or versions of this floating around LinkedIn and the Internet. And to be clear, it is completely wild. Search is changing faster than at any point since Google launched in 1998. And as someone who grew up with Dogpile and AltaVista and Ask Jeeves, Google was a big deal in 1998. But I'd say things are changing even Faster now with ChatGPT perplexity, Google's own AI overviews in AI mode. All of it is rewriting how we all search, discover, find information, and users are making the switch very quickly. I mean, this chart shows a projection over the next five years, but there's a lot of research that says that this is already happening. It's already in full swing. There was a Pew Research center report that said that when an AI summary appears in a Google result, users only click on an organic result 8% of the time, meaning 92% of those searches when there's the Google AI result are no click. Adobe similarly said that 77% of ChatGPT users rely on it for search and nearly a third trust it more than traditional search engines. Again, like completely wild. So the long story short is that the number one Google ranking we've all been chasing for decades, it's not enough anymore. It's no longer enough to have an excellent and well run SEO program. There's more to the story.

Understanding GEO and AEO

Tom Newton: So next slide. What is that more to the story? There's GEO, AEO, AIO, SEO eieio. We have a full old McDonald farm situation going on in Searchland, but it's true there are new layers to search that are critical to keeping you and your business visible and competitive. The two in particular that I think stand out and we're going to talk about today are AEO Answer Engine Optimization and GEO Generative Engine Optimization. I'm going to spend a couple minutes level setting on definitions there because these acronyms are new and I've even been speaking with peers at other marketing companies and not everyone's on the same page in terms of what all of these acronyms mean. So let's just level set starting with geo. So this one makes sense. It's about making content easy for AI engines like ChatGPT perplexity, Google's AI mode to trust and cite in their answers. This isn't a list of blue links. We all know this. These are the synthesized responses that you get within an AI engine and your content has to be authoritative enough to get pulled in. In practice, GEO favors content that's evidence backed, real data, research and stats, structured clean formats like tables, headings, schema that AI can easily read and transparent with clear authorship sources and credibility signals. One thing you'll notice here is everything I just said is also pretty important for SEO. It might be packaged a little bit differently, but still the next one. AEO Answer Engine Optimization is about making content show up in those quick answer first snippets in tools like Google's AI overviews, being copilot, and I think perplexity instant answers. So instead of just fighting again for blue links, you're formatting content so engines can grab and display it instantly. And AEO content works best when it's direct, clear answers to common questions structured again FAQs, Lists, Tables and accessible simple language that doesn't need to be reinterpreted by the engines. One of the things you'll notice as I was sort of framing up AEO geo. If we go to the next slide, is that content is still critical to search. It's always been the case. It's still the case and at a really high level. Like I've just been saying with AEO and geo, overall content should be readable by AI relevant and contextual, expert and trustworthy. Content is the source material for AI answers for generative results. And it's why we, Jasper are hosting this webinar, because we work with our customers and day in and day out building AI powered content pipelines that set them up for success at scale, including for AEO and geo. And so we'll get to that a little bit later on in our demo with Mason. But cutting to the chase, this is one of the lowest hanging fruits. We think most marketing teams have to move the needle on AEO and geo, which is optimizing content at scale with these AEO GEO best practices in mind. Okay. With all of that foundation being set, let's jump into the conversation.

Fireside Chat with Kyle Byers

Tom Newton: Kyle, I am so incredibly excited you're here. Thank you for making the time to join us to chat through all these topics. Before we actually get in to the meat of the conversation, can you just tell me a little bit about yourself and your role at SEMrush?

Kyle Byers: Sure, absolutely, yeah. Thank you Tom. It's great to be here. So I'm currently the director of growth marketing in SEMrush's own media department where I oversee our acquired assets, organic content, SEO and AI SEO. So geo llmo, eieio, as you said, whatever you'd like to call it. Before SEMrush, I co founded a PPC marketing agency as well as a SaaS startup and I worked in B2C E commerce, so my experience kind of runs the gamut.

Tom Newton: Amazing. Yeah. Semrush is a huge player in the search space. Has been for a long time. We use it at Jasper. We have a partnership we launched this past summer. So very excited to have you in particular on this conversation to talk through the new world of search. I think to start with, I shared some stats, shared that chart with the projections of Google search over the next five years. And there have definitely also been conflicting reports about the impact of AI on web traffic in particular. So some studies suggest generative AI and zero click search are driving down traffic from Google and other platforms. Others show that AI powered search is actually creating new discovery opportunities and sort of is additive and even increasing traffic for certain types of content. So in your opinion, based on what you've been seeing, the team has been seeing at Semrush. What? I guess there's two questions. Let's get into the data. Before we even get into the data, where should our general marketing anxiety levels be right now around search?

Kyle Byers: That's a fun way to frame it. I think the excitement level is very, very high. That's the way I would put it. It's very exciting because we are seeing a pretty dramatic shift in the way that you can win in search. And there are so many more touch points now where there are legitimate challengers to Google and yet at the same time the opportunity presented by Google is not necessarily in decline. So as you mentioned, we've got different studies that seem to call out different things that in some ways may contradict each other to some degree. But there's nuance here where it all kind of makes sense when it comes together if you zoom out a bit. So. So for example, at Semrush, one of our studies showed that actually multiple studies have shown and lots of people have presented data showing that as Google rolls out AI overviews, AIOs, the AI summaries at the top of traditional Google search results, as that happens, the click through rates go down. So traffic to websites that are normally listed below, that traffic goes down because people don't click through as much because they're having their answers given to them by the AI overviews instead of. And that has caused this very interesting divide. Previously in SEO, what we always saw in the industry was as your impressions, as your visibility went up, so did your clicks. Basically that's the formula, rank better and you get more impressions and more clicks. That is no longer true because of AI overviews. You can now get lots of impressions and actually have your clicks go down at the same time. And so that's a really interesting dichotomy that we're seeing. And in that case, indeed it's true that the traffic from Google is going down across industries, across players and industries. Perhaps some of the only websites that have seen growth in actual traffic have been Reddit, Quora and some of those others. Wikipedia perhaps in some cases depending on the SERP. The other thing that we see happening is that ChatGPT is not apparently actually taking share from Google, by which I mean even though the growth in users of ChatGPT has skyrocketed, I believe it's up 8x in less than a year in terms of daily active users. That is not actually eating into the number of searches people make on Google. That's also supported by Google. Google has talked about how AI overviews and AI mode and indeed also promote continuous additional search behavior. But even from our own studies, we've seen that users who use ChatGPT on a regular basis 10 times or more per week, doing that actually makes their Google search usage go up by about 20%. So they're researching in ChatGPT, definitely, but that's not replacing what they're doing in Google. Instead, it's. It's supplementing it and even causing additional search behavior to happen. And that supports what's called the expansion hypothesis, that basically there's an additive relationship there, which is really interesting. The third thing that I'd love to call out is conversion rates. So we've seen that a visitor from ChatGPT is actually 4.4 times more likely to convert than a visitor from Google traditional organic search. So that's a really exciting thing. Despite the fact that ChatGPT has still a lot less reach, a lot fewer visitors and users than Google, it's growing so rapidly and you get such a higher conversion rate from those users that it's an incredibly valuable potential channel that more brands should explore. And in the future, that's likely to just go up and up as adoption increases. It also makes some sense. The reason why you would see higher conversion rates there is because people are able to learn so much in ChatGPT. By the time they go through a long conversation where they're researching answers to their questions, solutions to their problems, they're actually seeing within ChatGPT all about your brand and your products and your competitors and, and your pricing and what people say about you. All these things, that whole. It's like an entire customer journey that's happening in a chatgpt window. And then by the time they actually click through ChatGPT to your website, they're already warmed up very well. I tell people ChatGPT can be your best salesperson if you train it properly. And so it makes sense that you can see really high conversion rates coming from that traffic as well.

Tom Newton: Yeah, there's so much good stuff there, Kyle. I'm going to maybe go back to both all three of those points because I think there's a lot to dig into. Maybe we'll work our way backwards here. So the higher intent traffic from ChatGPT does make some intuitive sense. Personally, I've found myself using it more for discovery and some of the traditional buyer's journey, even at Jasper, as we're considering new tools. So it is this interesting shift in buying patterns and it also just makes me laugh A little bit about, you know, like the, the research backed buyer's journey of like 10 or 15 years ago of okay, you need like this many ebooks and this many webinars and this many ad touches and like etc. Etc. Not that those channels don't matter. It just, it is like drastically reforming kind of in front of our eyes. To your point, you can have an entire customer journey or a very large percentage of a customer journey happen in a ChatGPT window, which is just again wild. And it's interesting too because those touch points are still quite valuable. It's just that they in many cases are being subsumed into the ChatGPT conversation. So those very same blog posts, podcasts, YouTube videos that humans used to learn from and still do. ChatGPT has now learned from these LLM models as well as the rag retrieval augmented generation that they do when they actually do a web search from ChatGPT. Those are informed by the same assets that you previously built to nudge humans lower and lower in your funnel. So it's still happening. It's just happening in this interesting way that's modulated by this third party and you don't have tracking on it. Yeah, it's wild. I think that gets to maybe one of the other points I wanted to dig in on with ranking. And so the fact that you can get lots of impressions and clicks can get lots of impressions, but clicks actually go down. And so I guess there's a little bit of a translation, a lot of translation that needs to happen from old world to new world. But maybe specifically what does it mean to rank when LLMs deliver direct answers, does ranking still matter?

Kyle Byers: The short answer is yes, but it's also, it's strange now, right? Google has been personalizing results for ages. But the extent to which Google personalized results previously was typically okay language based on your geography or your browser settings, et cetera, you could see very significant differences based on that geo or language. But other types of personalization, like your browsing history, what they know about your demographics, those things, typically you would see relatively minor shifts in rankings given one person's search engine results page was going to be pretty similar to another person's if they're in the same country especially. And one person's search engine results page for again, the same query was likely to be the same from day to day, more or less, with some small adjustments. That is really not necessarily true with ChatGPT or with Google AI mode for that matter. Or clawed perplexity. For multiple reasons, these systems. Probably the biggest reason is that these systems are non deterministic. Everything that they provide is in a probabilistic basis, which means that there is some roll of the dice and some randomness baked into it every single time. Even for the same user. If you just open a different chat window, right. There's also a lot more variance in the way people are querying these. So a prompt is effectively a very, very long tail keyword. People don't use ChatGPT to say things like, you know, what is SEO? As frequently as they used to, as they type in something like what is SEO and how can I use it to blah, blah, blah, I am a. Et cetera, et cetera. And so there's these very long prompts that they use which naturally, depending on what that prompt is, it's going to be a little bit of a different or a very different result from what a shorter keyword might provide. Then there's the fact that these are multi threaded conversations. Someone might start off the conversation asking, how can I get more customers for my small business xyz? And then it's only like four or five statements deep into the conversation that they actually get around to asking about SEO or AI, content generation or what have you. At that point, the prompt is not just the prompt anymore, it's the prompt plus the whole history. Then there's even additional layers here of you know, whether what tool set they have enabled, what tier of pricing plan they're on, what model they're using, what custom instructions they have set up, whether they're in a project, what the other memory is around that user specifically, and all the things that ChatGPT knows about these people that Google never knew before. So it's a really interesting time from a personalization perspective. And that does indeed make it harder regarding what ranking actually means. Now, on the other hand, ranking still is a thing. Making sure that you show up where you expect to get value from these prompts, making sure that you're showing up in a positive light and ideally showing up above your competitors. If someone asks a question like what are the best AI visibility tools? Then you better believe that we're working on making sure that Semrush shows up very close to the top there in that list of options.

Tom Newton: Yeah, yeah, 100%. It's interesting because a lot of people I've spoken with about this talk about how a lot of the fundamentals haven't necessarily changed for content strategy, search strategy, the types of things that are important in terms of, you know, relevant expert trustworthy, all of these things. And it's interesting what you just mentioned about showing up above competitors in generative results, because at the end of the day, it's a little bit of the same game, like, how do we use this algorithm, in this case, probabilistic model to get our results and our sort of messaging and positioning above our competitors. Before, we were really hyper focused on the blue links on a SERP page and trying to get, you know, one through three, ideally on that first page. And now it's how do we, to your point, show up at the top across almost an infinite variety of prompts, long tail, keywords, context, all of these layers. And so the complexity increases pretty massively. So I think I have some questions, sort of branching off of that, the first one being, okay, you know, we're all marketers. A lot of us have grown up for however many years in the world of SEO and backlink and domain rating and all the things that have mattered for that strategy. What metrics matter in this new world? Right? Like if you're a marketer and you're thinking about AEO geo, where visibility doesn't always lead to a click or a visit, how do you measure success? How do you measure impact as you're leaning into these programs?

Kyle Byers: Sure, definitely. Visibility in general is really crucial for the prompts that you care about for that universe of topics that you care about. How often are you showing up? How often is brand showing up? AI visibility is probably one of the most crucial ones. Share of voice is also very important. So tracking how your competitors are showing up, what is your share of voice percentage against that group and, and how is that going up or down over time? Additionally, the number of citations you're getting. So how often is it not just your brand getting cited, you've got just regular brand mentions and then you've got linked citations. So how often are you actually getting linked to from these responses? Because that's obviously a crucial factor as well. Those would be the most important ones that I would highlight. Certainly ranking can be a good one as well, but it's very difficult to gauge that because of all the personalization. But it's a good one to follow as well. And sentiment, it would be a good one also, although that's a little bit deeper of a metric.

Tom Newton: So is that what your Semrush search dashboard looks like?

Kyle Byers: Yeah, absolutely. So we've got two main tools that are really great for this, the AI SEO toolkit, which is excellent for businesses of any size. And then we have for larger businesses we have SEMrush Enterprise AIO, which is quite powerful. And so these two tools really allow us and allow our customers to get very deep insights in terms of not just how they're performing in those landscapes of the prompts they care about and the industry topics they care about and versus competitors, but also what are those things, what are those leverage points that they can pull to actually make a change and show up more frequently in those places the way that they need to. What specific pages are getting cited there, not just on their own websites, but also on external websites. So if you see that, for example, one of my favorite dashboards is being able to see what Reddit posts are ranking for those prompts that you care about, so you can get in there and make sure that you're present in those posts. Similar for external third party websites where there might be a guest posting opportunity or some outreach you can do to make sure you show up on listicles that are getting cited as sources. Those are all really crucial touch points. And it's interesting because in some ways this hearkens back a little bit to old school SEO back when everyone was doing guest posting. Well, that's back on the menu again.

Tom Newton: I know what's old is new again, a little bit. It's full circle. Okay, yeah, that's really helpful I think, you know, as, as everyone approaches this, new, new approaches, new strategies, it's helpful I think to ground ourselves in like what success looks like, what metrics we want to focus on. And I think maybe taking that a step further, getting even more tactical. So if you were, you were someone, you know, sitting in, sitting in whatever xyz, mid market or enterprise company, you're just starting to lean in on. AEO Geo, you have a pretty solid SEO program, but you're trying to get your feet wet and you want to run a geo audit, what would you check first of that list, what is your starting point? And maybe some tactical recommendations there.

Kyle Byers: Yeah, sure. So definitely plugging your domain, your brand into a tool tool set like the SEMrush AI SEO toolkit is a great place to start or enterprise AIO understand what that landscape looks like. These tools also will help you come up with a list of prompts if you already don't already have one in mind. It's effectively similar to keyword research, but for prompts instead. And of course those are varied based on what the Persona is. So making sure that you have that set up, checking how you're showing up again, it's AI visibility, share a voice et Cetera, and then looking at who else is showing up better than you, who has more AI visibility for those prompts? If anyone else is leading, what other web pages are you not covered in? What does that gap analysis look like? Is also something our tools can help with. What are those opportunities? So in many ways LLM optimization or GEO is actually quite similar to traditional SEO. You know the basics of content marketing, making sure that you have your brand represented in the topical coverage areas that you care about, both on your own website and externally. As I mentioned, blog posts are a real huge driver of this social media as well. Reddit, Quora, Wikipedia is really, really important as well. All these places are touch points that maybe previously have been considered in some cases more brand marketing, but now they drive real AI visibility as well. So making sure that you're showing up in those different places. There are a couple of very interesting nuances that make AI SEO different from traditional SEO. First, there is actually a bit more uncertainty around it. There are ongoing debates in the community around what really drives value here, what really drives visibility, and how certain can we be of the business impact of that. All those different things. But from what we've seen, traditional SEO tactics, like I said, technical SEO, making sure everything is crawlable. In the case of AI SEO, making sure that you're not relying on client side rendering, server side rendering, or static HTML is key for LLM crawlers, more so than traditional SEO. And making sure also that you are phrasing things in your content in a way that is easily parsable by LLMs and rag. I actually spotted an interesting example of this recently where if you searched on Google for something, the query was something like SEMrush 50% discount for nonprofit organizations. If you made that search at the time, the Google AI overview would say yes, SEMrush offers a 50% discount to nonprofit organizations. And it cites the source. You click through the source and the source is a blog post on a third party website about SaaS, companies that offer discounts to nonprofit organizations. And in this article Semrush has listed like number three or something. I think there are 10 companies or so and it says correct information. In the SEMrush section it says SEMrush offers a discount to nonprofit organizations, but you need to talk to them, their customer success team or their sales team to get more information and the specific discount. That is the truth. But the piece of the article that Google was citing, remember Google said that Semrush offered 50% discount to all nonprofits, which is not true. You scroll down to that part and what it was actually citing was the Asana section. Asana was like number seven way lower in the article. And it said Asana offers blah blah, blah, blah blah. That's one paragraph. Then you get to the next paragraph and it says the product also offers a 50% discount to nonprofit organizations. So basically what Google was doing was using the retrieval augmented generation to grab that piece of information and combining it with what they saw. An article that mentioned SEMrush as well, but in such a way that was not true. And that's just because it didn't have the subject object relationship were not close enough. Just because that one paragraph started with the words the product or the tool, rather than saying semrush or Asana. That led to this completely incorrect paraphrasing. That's another thing to pay attention to in this world of AI SEO. Whereas previously it wasn't as important. We had passage based ranking, but the effects were not nearly as extreme as this. Making sure that the key passages in your content are directly quotable, even out of context is an important thing now.

Tom Newton: Yeah, yeah, I think that's actually an amazing real world example and sort of having the, I mean getting down to the basics of subject object relationship in your content marketing and being very explicit and what is the antecedent, you know, all of those pieces, it's interesting, like the granularity and the level to which the thinking goes, especially as you're thinking about content strategy. Okay, a couple more questions because I think we're coming up surprisingly on our Fireside chat time. First one. So Jasper, we partnered with Semrush as part of launching our optimization agent this summer, which is one of the first autonomous SEO agents built directly into content workflows. But agents are also a hot button topic. Do you see, or I don't know how much you've seen or used. Do you see agents like this becoming the new foundation of modern content operations, content engineering? And if so, what changes do you think that brings to how teams structure, optimize and scale content today? So, more or less, how do you see agents fitting into this new world of search?

Kyle Byers: That's a big question. Yes, I would love to. As best you can with your crystal ball today, you know, sure, yeah, yeah, yeah. I'd love to hear your answer to this. Also, my perspective on agents is over time it seems clear that over time they're going to become more and more crucial and the design that the way we design our accessibility and our data and our user interfaces need to be more and more geared toward what is usable for an agent. Anyone who has used ChatGPT in agent mode has probably noticed that there are certain things that trip it up sometimes, and those kinds of patterns can and should be optimized for in the future. As it stands right now, it's probably a rounding error in terms of the business value to a given business, but 12 months from now it could be very substantial. So making sure that you are thinking about it from that perspective, obviously accessibility from a technical perspective is something I already mentioned as well, client side rendering being contraindicated in this case. But I also think that agents, given their design to fulfill human needs in a human world that's designed for human interaction, I think these entities are largely being built already to take advantage of what's being optimized for humans. I think that in all likelihood, at least for the next six to 12 months, the immediate future, the same optimizations that help humans get their jobs done faster will also help agents get their jobs done faster. And in the future, I'm sure there's going to be a point where it makes sense to split two different so depending on the user agent, your website provides a much different interface for agents versus humans, and a larger and larger chunk of that will be AI agents. But as it stands right now, for a company that is not spilling with resources that they can spend on additional versions of their website, most likely just making sure that it's also easier for humans is likely to help with the agent aspect as well, I would think. But I would love to hear, like I said, your thoughts on that as well.

Tom Newton: Tom yeah, personally I find the crystal ball of marketing just so interesting and agree with everything you said. I think there's actually an interesting parallel that we can all look back to, which I feel like it was relatively common Part of the discourse in we're going way back to 90s 1998. Websites in the early 90s were like billboards in the desert. There was no way to get to them unless you had the exact URL or maybe it was listed up in some chat or forum, the AOL homepage. Whatever it was, there's no real way to get to them. Search engines change that. And all of a sudden the content on the website wasn't just for the user you were trying to speak to, it was for the user in the search engine. And that's how we've gone through decade of now evolution of keyword stuffing and backlink Building and all of these things that led us to where we are today and our websites are today. And I feel like there's this feeling maybe a little bit of oh agents, like how am I going to optimize my content for an agent? And now I have to create a website that's accessible and agent optimized. But we've been here before. We had to redo all of our websites for mobile in the late 2000s, early 2010s. We just have a new audience in the mix and that audience is agents. So now it's actually the end user, it's still search, it's agents. And how that actually plays out we will see over the ensuing months and years. I also think it's interesting because there's a meta element to this. Like Jasper is an AI marketing platform. We're focused a lot on building content pipelines for marketing teams, especially in mid market enterprise organizations. And we have agents that are trained on your brand, your product, your audiences and can create content at scale. It gets into this interesting agents building for end users, agents building for SEO, agents building for agents and at least for now, AEO and geo. So yeah, we're in this. It's meta being a marketer at an AI marketing company and it's going to get even more meta being a marketer using agents to optimize for agents. But I think that is generally the direction we're going and we're already there in some ways.

Kyle Byers: Yeah, completely agree. Well said.

Tom Newton: Cool. Okay, last one, promise. Then we're going to get to quick demo and Q and A and I've seen a bunch of really good questions filtering in so get to as many as we can. Last question back to maybe Geo aeo. What does a mature or sort of a well functioning Geo AEO function and team look like inside of a marketing org in 2026? Is Geo AEO? Should it be a sub team? Is it a mindset? Is it a center of excellence? And I guess there could be two answers. How you have it at Semrush and maybe like your ideal state or maybe those are the same but like what? What does it look like?

Kyle Byers: Sure. So I think that the future of this and marketing in general must be much more cross functional, much less siloed. Some organizations already do a very good job of this but in my experience most don't. The because all those touch points that we discussed earlier are really important for driving AI SEO. So for anyone tuning in late, you know, blog posts, both on your own websites as well as externally Reddit, other social media, Quora, Wikipedia, PR and digital pr. Anything having to do with press, even emails that get shared on, on dark social communities, all these touch points, webinars, podcasts, videos, all these things now feed into the AI corpus of learning. And because of that, it is so much a team sport to make this work. SEO has always been something that could and ideally should be informed by and supported by other channels and other teams as well. But it was typically in most organizations, a little bit siloed. That cannot be true anymore when it comes to AI search. AI search really demands this very cross functional cooperation between teams. I think a natural thing that's happening for many organizations is that the existing SEO team may be the one stepping up to lead on this because they're the ones that are already doing the thing that's closest to it. As we talked about, traditional SEO is in many ways the foundation of AI search as well. But having to work with these other, with your teammates, with other department heads to make sure that these things get done across all those other touch points as well is just so crucial. I would even go one step further than that and say, just zooming out a bit, I think that the organizations, the marketing organizations that win in 2026 will be the ones more than ever before, will be the ones that really deeply understand their customers. So it's not even just about cooperating amongst all of marketing. It's cooperating with product, with customer success, customer insights, with sales, all these touch points, every single touch point with customers, as well as the data behind that user data, et cetera, understanding all those things. And that's especially crucial because attribution is becoming a bigger and bigger challenge when your investments in PR in social media pay off instead through AI search. And even when it pays off through AI search, it's not always tracked back to AI search. You lose a lot of visibility of what that flywheel really looks like. And so you're less able to make decisions that are purely data based. And you actually have to bring in some of that human understanding of what your customers really resonate with and make some judgment calls there in a way that perhaps we didn't need to as much just a year or two ago.

Tom Newton: Yeah. As someone who's been sort of deeply involved in attribution questions, arguments, thought exercises for, you know, the better part of my career, I am also personally extremely interested and frankly curious about how we solve this even more complex world of attribution, but also potentially how some of these tools can help us solve Right and I think that's the interesting piece of being at this crossroad, at this sort of inflection point with AI from a marketing perspective is that things are getting more. A lot of the foundational elements of good, excellent marketing are staying the same, but a lot of things are shifting under our feet and the tools that are available to us are also shifting. So it's very much this sort of like ballet back and forth of like challenge shifts. Tools evolve mostly with AI. How do we apply those tools? Challenge. You know what I mean? Like at some point we'll reach a new baseline, but it's going to be an exciting ride to get there. Okay, Kyle, this was amazing. So much, really thoughtful, so many thoughtful answers. I honestly feel like we could keep going for another hour, but I am going to cut us off there. Like I said, we have, I think, see right now, 88 questions that have come in which we love to see. We're not going to be able to get to all of them, but we'll get to as many as we can. And before we do that, I do want to ask Mason from our product marketing team at Jasper to join the call to give a quick demo of what SEO, GEO and AEO optimization looks like within Jasper. Because I think that can help make all of this a bit more real. A lot of what you've been saying, Kyle, the foundational pieces are the same. It requires being cross functional and cooperating across teams. It is a mindset shift, but at the end of the day it's also a content challenge. How are you making sure that the content you are publishing across all of these channels that we know and love are optimized for every audience, which includes end user LLMs, maybe eventually agents. It's just this ongoing challenge. So yes, with all that said, we'd love to invite or hand it over to Mason.

Jasper Demo: Optimizing for GEO and AEO

Mason: Thank you so much Tom. And thank you to everyone who's here on the webinar today. In today's demo, I'm going to be playing the role of Alex, who's a content marketer at Taskforge. Think of it as like a B2B SaaS company that builds project management tools for modern teams. Let's imagine a scenario where our organic traffic is dipping. If we don't turn this around quickly, we risk losing market share to competitors and I actually risk losing my job. The stakes can be higher. Historically, the Task Forge blog has been one of the biggest drivers of our organic traffic and filling the top of the funnel with thousands of inbound prospects each month. But traffic is slipping. To get this back on track, the first step is going to be creating net new blog pages that capture trending opportunities while also optimizing our existing library of content so it doesn't fall invisible in the era of generative search. Today I'm going to show you how I use Jasper and Semrush to power this workflow end to end all inside Jasper. I'll show you how I research trends and opportunities, instantly identify and apply the right keywords while generating optimized blog posts at scale, and finally reformat existing content so it's GEO and AEO ready. So let's dive in. Before I create any content, I need to ensure that Jasper's AI is trained to understand my audience, which helps to ensure that the content my team creates resonates with our target brand. That's where Jasper IQ comes into play. One key component of Jasper IQ is audiences, which is how we capture engagement patterns, power, words, objections, value, props and more to ensure all content is Persona aligned from the start. As you can see, we also provide things like brand voice and style guide configuration, but audiences goes even deeper, which ultimately helps us get better content at scale. In today's demo, I'm going to skip showing the actual creation of audience's brand voice and style guide, but you'll see exactly how it comes into play in just a moment. Now that we've ensured Jasper is set up to create high quality content, the next step is actually creating that content. But before I can begin, I really need to understand what's trending in my space and where there are opportunities so I know what to create content about. Instead of manually combing through spreadsheets or analyst reports, I can actually use Jasper's research agent to instantly surface opportunities. And it helps me save countless hours on analysis so I can focus on generating results. In this example, I've already prompted our research agent to create a report into Task Forge's space. If we actually look at the results, we can see a highly detailed research report that surfaces trends gaining momentum like remote collaboration tools, AI assisted workflows, and also flags oversaturated areas that I might want to avoid. It also pinpoints some gaps that highlight where our brand has a real chance to differentiate against bigger competitors. And that's the foundation that I need to decide what new blogs to create. While it's great that I now have a research report with trends and opportunities, we all know research alone doesn't drive traffic. I need to turn this research into actionable content and that's where Jasper's optimization agent comes into play. Powered by Semrush, it helps take all these insights and translates it into a prioritized keyword strategy. In the chat I've asked the optimization agent to run a keyword research and analysis based off the insights surface in the taskforge report and I got this highly detailed Let me close this real fast. I get this highly detailed report as well as a keyword table that helps give me at glance view of where I can realistically compete and win. For example, it highlights some high volume but lower competition phrases and really helps me turn strategy into execution. No longer am I staring at a list of trends. I know exactly what keywords I should be prioritizing for my net new content and how to position it today to capture both today's search traffic as well as tomorrow's generative search traffic. Because up until now I've just gathered insights to create the right keywords around. But as any marketer knows, we actually have to take this insights and turn it into content which can take weeks across multiple stakeholders. But with Jasper I can compress that entire workflow to minutes. I'll come into our chat and select our optimization agent. I'll highlight the research reports that's been done. For context, enter a simple prompt to generate a series of blog posts based on these reports and hit enter in a single click. You're going to see the optimization agent generate a series of blog posts. What's great is it actually is going to tell me what it's going to create and the steps it's going to take in order to do that. You can see it's already identified five potential blog posts and it's going to get started on creating that first one right away. I really want to call out that it's not just going to give me an outline or a starting point for a blog post, it's going to write the entire piece. I mean right here you can see the first article being generated. I really want to call out like how it's organized, right? We have a compelling introduction that hooks the reader. We have body paragraphs are clearly organized and flows logically with subheads aligned to SEO best practices. And the conclusion is going to align position. Task force is the natural solution and you can already see it's getting started on the second blog post. Throughout each post the keywords are seamlessly woven in throughout the so the post feels is optimized without feeling forced. In other words, this is content that's truly published, ready from the start. And you can see how the integration with Semrush helps ensure that the optimization insights live inside the content. I'm able to confirm things like keyword usage, readability and search intent alignment without having to leave Jasper. Which means I can go from nothing to a polished, optimized blog in the time that it normally would have taken me to just draft a creative brief. So instead of staring at a blank page, I can start with a fully formed, high quality piece of content that's already strategically aligned to my strategy and ready to publish. And this is the kind of acceleration that doesn't just save time, but creates a competitive advantage. So that's for creating net new content. Let's talk about how we can optimize existing content for this new era of GEO and aeo. If you remember, I said that the Task Force blog has historically been a huge driver of organic traffic, but a lot of our top performing posts were written for this pre GEO era. And if I don't modernize them, they're going to continue slipping down the rankings and potentially vanish from generative engines entirely. Look at these examples of old blog posts, right? This kind of style probably reminds you of what used to work for traditional SEO in 2019, but generative engines don't reward it anymore. Generative platforms want content that is clear, structured, conversational and aligned to user intent, which is why we created a quick Content Strategy checklist so you always know the exact areas you should be optimizing based on what you're trying to do. As a bonus, everyone on today's webinar is going to get exclusive access to this checklist, plus dozen more actionable tips in our latest ebook. And we're going to follow up after the call to make sure you guys get it right away. Now, in just a simple click, let's focus on this future Proofing your project management With Task Forge, I'm going to open up a custom app that I've built to rewrite content for this new era. You can see that I can easily apply a audience, so let's do Agile Project Managers hit next. I can do more keyword research directly with Semrush as part of the workflow, but we'll just skip that for now and hit Generate. What you're going to see is that by running this outdated article through the GEO and AEO rewriter, you're going to get an instantly revitalized version. And if we compare this back to the checklist, you'll see that the rewriter automatically applies nearly every single Best practice it inserts H2s and H3s that align with common queries, replaces walls of text with scannable bullet points and suggests schema markups so AI systems know how to actually interpret the content. It even rewrites in a more natural conversational tone so the content feels more approachable and even more importantly, it understands it anticipates intent. Instead of just having a paragraph about what AI transformation and project management is, it phrases it as a natural question that somebody is going to be searching for on these search engines. What really makes Jasper special though, is the ability to do all of this at scale. So you just saw me take one blog post and optimize it one time for AEO and geo. But what's really exciting about Jasper is that I if I can figure out how to use this, I can select multiple pieces of content at once, Hit generate and you're going to see Jasper automatically start rewriting multiple pieces of content. This is what we really mean about taking content and transforming it at scales. You can take multiple posts that were decaying in search rankings and give them new life and instead of losing traffic year over year, you can re optimize and actually expand visibility on GEO platforms. The time savings are huge because it would have taken me hours of rewriting and Jasper gets it done in second. And it helps breathe new life into my entire content library. So I know we only have a bit of time to show you the demo today, but if we actually look back at what we just accomplished in less than five minutes, we went from a place of vulnerability because organic traffic was dipping, we were scared of losing market share and there's even personal stakes online. And now we've created a clear, repeatable strategy for turning things around. We launched new SEO pages by grounding them in a research report that surfaced trends and opportunities that I would have never found on my own. Next, we transformed those insights into action with the optimization agent by creating a prioritized keyword roadmap and turning it into publish ready content. And then we finally breathe new life into older blog posts by running them through the GEO and AEO rewriter to ensure that they're structured, conversational and aligned to how Generative Engine Surface results today. What this means is that Jasper and Semrush together give me an end to end workflow that is faster, smarter and more future proof than anything I could piece together with legacy tools. I've not only protected my existing traffic, I've opened new doors to capture demand and visibility in a generative first search world. And I think that's the big takeaway from today's demo. In an era where search is evolving faster than ever, standing still is the same as falling behind. But with Jasper and Semrush, you don't just keep up, you get ahead, you safeguard your market share and you ensure your content engine is always working in your favor. So that's it for me today. I'll hand it back over to Tom and Kyle to run an excellent Q and A. Amazing.

Audience Q&A

Tom Newton: Thanks, Mason. Yeah, I think um, giving a real world example is really helpful. Um, I think it helps contextualize what it means to take maybe stale content and optimize it for geo. The other thing I'll say really quickly is if you want to learn more about everything Mason just demoed, check out Jasper and we'd be happy to chat with you, um, and go through all the details of what it means to create a content pipeline that supports Geo AEO at scale. We did have. Okay, we're coming up at time and we did pass the hundred question mark. So I'm going to try to pick out maybe some of the top ones if you're ready for it, Kyle, and we'll get through at least a handful and we also will try to follow up with folks after the call for any questions that the majority of questions that we won't be able to answer. So the first one was a clarifying question from Neil. Um, he was trying to connect some dots. ChatGPT and Google currently have an additive relationship in that ChatGPT usage is leading to increased Google searches, which is what you had mentioned, Kyle. But LLMs are set to overtake traditional search engines by 2028. That first chart we were looking at, how is that possible? What is that relationship? And you are muted, Kyle, if you just want to come off quickly.

Kyle Byers: Thanks. Yeah, great question. So a lot of this is simply what Google's roadmap looks like, right? So they have been rolling out their own AI mode to more and more users. They just launched it across I think an additional 180 countries just a couple of weeks ago. It is not the default search setup yet, it's not the default experience, but personally I think that's pretty likely to happen. And if you just also look at the growth of ChatGPT over time, that 8x over the last less than a year, Google responding to that very frequently, with rolling out more AI overviews as well as AI mode potentially becoming the default, it kind of makes sense that there's going to be this crossfade over time where those AI interfaces are just going to be more and more prevalent and really become the default. So that's where that is mostly coming from. It's true that right now there's, there's still kind of an additive. Also, I believe that chart is from a either it's a traffic perspective or a value tracked value perspective. And that's something that we're already starting to see right when we talk about how impressions and visibility can go way up in Google at the same time that your actually clicks go down. Well, that actually is part of this trend line as well. Even if ChatGPT is driving 20% more Google searches for the people who use ChatGPT, that's not enough to combat the vast loss in clicks that's happening as Google cannibalizes its own search results with AI overviews.

Tom Newton: Yeah, it is. I mean this is probably the 10th time I've used the word wild on this webinar. But it is wild that like the search we all have known and loved for 27 years May no longer exist in like the near future. At least as it's constituted, has been constituted for the past almost three decades, which is pretty wild. The next one we had a lot of questions about sites like Reddit and Quora. Do you have recommendations on how a business can increase its presence on Reddit and Quora or any other third party sites?

Kyle Byers: Yeah, one very important way of course is to claim your business username and get, get active on there with your brand account. You can create your own subreddit as well that's based on your brand. If you have enough of a following, a large enough brand, we have one for Semrush. Trying to keep those active as a community, building up that kind of store of content on that third party website is really valuable. Also, you know, partnering with influencers who already really care about your product quite a bit. Brand ambassadors, affiliates, employee advocacy programs. We have a very successful employee advocacy program at Semrush where internal employees team members are encouraged to share their thoughts. And we have certain internal prompts where we say this week why don't you talk about XYZ topic? Then people post on LinkedIn and Reddit and they naturally end up talking about Semrush depending on what the topic is. So those kinds of things are really valuable ways to get more visibility in those places.

Tom Newton: That's amazing. I know we're right at time, but I'm going to try to sneak in two more. So from C Burch, what types of content do you think lend itself to Geo and AO Best without undermining user experience or brand integrity. So for example, homepages versus blog versus discoverable social content like YouTube or Instagram.

Kyle Byers: They're all really crucial depending on the authority of your website. We've seen that Reddit is by far Reddit and Quora are by far two of the most cited sources across almost any industry. So those are very high leverage points that many businesses are not already doing. But certainly you have direct control over your own website and blog posts and homepage and product pages. So making sure that you're framing your product and your brand in the right ways that are in relationship to those prompts, people are querying basically the AI keywords people search for. That's also really important.

Tom Newton: Amazing. Okay, last one From Benoit. Can ChatGPT or other AI tools detect AI assisted content and does this impact your site's visibility or ranking in AI driven search results like Geo Ado?

Kyle Byers: You may want to talk to that as well. Tom, as the Jasper guy, from what I've seen, the answer is no, that there's not a lot of concern here in terms of penalizations.

Tom Newton: Yeah, I would say the same. I think two or three years ago this was a really hot topic issue. But I would also say the quality of outputs has improved massively. And so even with the example that Mason was giving earlier, it's high quality output that you can get from tools like Jasper that are on brand specific for your audience understanding your product. So in a lot of ways it is still net new like original content. Whereas I think two or three years ago a lot of ChatGPT outputs were very formulaic to the point that you could read it and say oh that is AI. But I think we've, if we haven't fully moved away, we're very much moving away from that era of AI content.

Kyle Byers: Yeah. I think the key question is not is it AI? Instead it's is it valuable? And that can go both ways, whether it's human or AI.

Tom Newton: Yes. And I think that's where like the human in the loop is still so critical because there is an element of, you know, a tool like Jasper can understand your audience, can understand your brand, can give you really compelling content that's optimized for Geo aeo. But you person who understands that company the best should still be reading it to say and your audience to say yes, this would be valuable. No, it's not. So that QA check I think still exists for sure.

Closing Remarks

Tom Newton: Cool. All right, 1202, I think it was a. I really appreciate you having. I really appreciate having you, Kyle. Amazing conversation. So many great takeaways and questions. And we should do more of these soon.

Kyle Byers: This was so much fun. Thanks for having me.

Tom Newton: All right, cool. See ya.

Kyle Byers: Bye.