Traditional SEO vs AI Search Optimization (GEO, AEO) - #SEOFOMO TL;DR June 9, 2025

Let's go to the latest in SEO news and AI search in today's SEO FOMO, Too Long, Didn't Read. In today's edition, I have the great company of Chris Green, who's the technical director at Torque Partnerships. Hello, Chris. Thank you very much for—

Thank you for having me. Pleasure.

I want to thank Similarweb, who's our today's sponsor. Let's go through the top news in SEO because there's so much going on. The first one is this post that I did covering Google AI modes adoption versus other verticals via SimilarWeb clickstream data in desktop in the US. They had shared some initial data a few days ago, but then I asked if they could please follow up giving a little bit of context. How does this relate with the other tabs, with the other search verticals that Google has? They They sent that to me. And additionally, they also share the data of the latest days too, which shows that pretty much the growth had peaked. It actually became flat.

I mean, when you initially shared the data that didn't have that context, it still felt a little high. I know, like, 1% isn't large, but relative to all US search, if we extrapolate, that's— that felt high. Like as a general stat. But when you compare it to news and shopping, which is sort of 1 to 2%, depending, maybe 3% on a good day, this feels like a significant change or difference. Perhaps more surprising to me is how few people are clicking on shopping and news. And I think as those 2 streams are relatively small, I know the SERPs are integrating those that might remove the need to click on that tab, but this feels high. It's such a short timeframe. And I'd love to see what the tail of that is and it's stretching off, but I would imagine this will increase. It will have to. I think Google are going to make it a feature that it will increase whether or not we can read it as people liking it or not.

We need to realize that so far this is highly tied to the inclusion of AI overviews. At the end of the day, a lot of AI overviews are very connected right now with informational type of queries, not necessarily that much of transactional commercial because of the implications. Now Google has started to show adverts in AI overviews. So my feeling is that now that they have A way to monetize them, the share of inclusion of AI overviews for these more commercially oriented queries will increase on one hand, and also more people will be, I guess, also incentivized to click more on continue digging on AI mode. And also eventually what you mentioned too, what if Google starts to showcase AI mode first for many scenarios in the future, right? Pushing the user more towards an AI mode experience. Whenever they feel ready that they are monetizing AI mode too, because this is what pretty much restricts them. Let's see how this evolves. In any case, in the next week, I will keep you updated. The second news that I want to cover today is connected to that, is the SEO versus geo or AIO or whatever we want to call it. I want to continue and I will keep this updated. Showing the differences and similarities to it. And to make it even clearer, not only I expanded the areas showcasing the level of difference of each activity of the different areas from user search behavior optimizations areas, I also expanded here to include the results presentation and delivery because these were a lot actually changes and impacts, of course, the behavior and also the goals and metrics. So Each one of them have now a much more specific explanation and much more straightforward, let's say, difference level that hopefully allow us to understand what actually changes or not. And I believe that here it is pretty clear that in the optimizations areas, the concepts, let's say the areas themselves are very, very similar, but the way that they actually change versus each other is many times the way that it is executed or should be executed. Because of the query fan-out technique that LLMs or AI search relies on, for example, the level of personalization, for example. And in the case of the user behavior, of course, it is highly, highly impacted by the very big difference in the results presentation and delivery. So hopefully this will help to clarify many of the doubts or how different it is or not, how much do we need to change or not.

I think for me, the starkest contrast right now is how we track and report on it. I think this highlights, and we kind of look out there, is how a lot of the old paradigms of tracking, reporting, the tools we might have used seem to feel very old and very dated very quickly because the interface is different. How the results are generated is different. The significance in how you track the results Search Console gives you, for example, a lot of that is so fundamentally different that it has caused a really sharp turn or a need for us to To really look at it, I think, and visualizing it in this way has shown how stark that kind of contrast is. It's not different in a way that scares me because frankly, we as an industry have already solved a lot of these problems one, two, or three times since SEO has been kind of a thing. I mean, this is probably one of the most significant changes I've ever seen. The thing that I take comfort from, though, however, is a lot of what the industry has pivoted to over the last, I guess, two years around. what good or quality looks like, a lot of that is still close enough that it doesn't represent a complete change of actually what we do. I think that this shows that the detail and the minutiae is very different. There's a lot of learning. There's a lot of things we need to change. But then when we translate that into something that's actually useful or implementable, it doesn't change that much.

It's not that it's directionally different, isn't it?

I guess I'm extrapolating this piece of content to what everyone needs to do, and it is just Kind of chill a little bit, like write down everything. How does it actually change what you need to do now? And what do you need to act on right now versus the future? Going back to the data you were showing us, the first piece, like it's 1 to 2% right now. I know there's this kind of worry of this existential threat that is around the corner. You know, when AI mode takes over, it's like, it may do, it may do in a similar format, but we're not here yet. This is the perfect time to kind of prepare to To calm yourself, to calm your stakeholders.

This week in particular, this guide attracted a lot of popularity, a lot of attention, and I can understand why. It's a guide from Wordlift's Dorit Adat explaining AI mode in simple terms, a clear guide to the new era of search results. And he goes with very layman's terms and in a very actionable way explaining what it is, and how it works, the real implications in how we do things. Whereas query fan-out, which might sound very complicated and sophisticated, is like Google asks many questions, not just one, in parallel, right? Building a custom corpus, AI creates a personal bookshelf, or LLMs assemble the answer, AI works like a team of editors, right? So I love how he explained this in very simple terms, giving comparisons, And what it means for SEOs and content creators, very actionable tips, specific tips that we can start incorporating.

I like this piece firstly because the objective of this piece is to help people understand it. And I don't feel that's shared across everything that's been written so far. I think this is, this is one of the first that seems really to, the primary objective is let's bring everyone for the journey, along for the journey. And I'm, I've always got so much time for that because You know, that conduit, that communication is the most important for us to all get on board. I think for me, there's a couple of elements from here that really stuck with me. The first one, and this is kind of a drum I've been banging or a sermon I keep making for years, it's ranking versus visibility like this. The ranking paradigm has been changing beyond a point of usefulness for a while now. If AO mode comes out and everyone says, oh, rank trackers are now useless, or they need to fundamentally change what they do. What they do, I think the SEO industry has needed to really change how we consider ranking for a while because personalization, 10 blue link paradigm has been dead for almost as long as it hasn't. Um, and various like that. But what's really powerful about this here is that you have to be visible in the contexts that are valuable to you. So an arbitrary keyword search at a given point, at a given time, repeated daily, weekly, or monthly is No longer a great proxy for actually how well you perform. And I think that's, if now is when people don't listen to that, like there's almost no hope for them. They're lost, you know, just, just go, just keep going in that direction. So I think that's super useful. I think I know, um, Mike's also issued, um, a challenge to the tool part, tool industry within SEO that basically said you're, you're not rising up to meet the challenge yet. And I would say he's right, but also give him a chance. It's kind of.

Yeah.

still fairly new, but I think you mentioned, we talked before, like the tool providers, there's been an influx over the last couple of days. And I think that the next couple of weeks and months, we'll see, we'll see an explosion and then maybe there'll be too many in a couple of months' time. But so that was really key. And then the other part that's really stuck with me is the whole think in passages, not pages element. And this is, this is for me, this is the really important part. I like custom bookshelf analogy, but I've been thinking more as if like we're all scrapbooking, like we're all taking snippets from things we like and we're all putting them together to create something for us. Yeah. I think Google is now scrapbooking for us. And I think this is recognition that a single page has never been enough for that person at that time. And if you think about the range of different intents that people have, the fact that when people search, they might not even know what the right result for them is, like this kind of gluing together of all of these various passages is so intrinsic to just making this experience quicker. And we have to stop thinking of the page as the document and of the parts of the different journeys. And that's— Again, that's, that's a big mindset shift for most people because again, SEOs, like we have been preoccupied with URLs, with information architecture across URLs, with clusters of content, like—

Make it very obvious for search engines that, oh, this page is about this, it should rank for this.

The one thing that's missing here, and this isn't necessarily criticism of this blog, but this is of the overall conversation, is, is the talk of originality. The SEO FOMO event back in London where we had Gary and One of his sort of more candid moments was around the focus on originality. Like if you have all of the world's data and you've gotta try and pick these passages that are most important, like how do we signify that we have the original take and we are not a summary of a summary? And I think, I don't know if we, if it's just 'cause we don't know how we're gonna address the originality piece or how Google is going to do it properly in this new kind of vector space world, but. I still, we're still missing that a little bit. I think we're still at a little bit of a risk of saying query found means lots of FAQs everywhere in short chunks, and we'll just keep churning that out. And I, I think if I, if I want to see any more, and I need to do some stuff myself on this, is how do we factor originality into this?

That's a really good point, especially because AI generates content out there. So this ties very well with this piece from Amanda Keane about revisiting useful content in the age of AI-dominated search and connects also well with this other article that performed really well, and I'm not surprised why, because it's actually HubSpot sharing how they create quality content in the age of AI. So it's a little bit of 2 guides that cover pretty much the same thing, but in different ways with different angles. The key here is thinking about the customer, not necessarily keyword popularity and so on, and thinking about the business and staying on topic with your business, on brand with your business content, and focusing on creating or building an authority towards what actually matters for your product or service expertise at the end of the day. She provides very specific examples here, goes through actions, taking into account also the Google quality raters guidance that makes total sense. So I love how actionable this advice is, how brand also recognition and authorities is created. And in the case of HubSpot, I love in particular this graph that they created about what is quality content, because this is such a topic that is so ambiguous, right? As top content that helps your audience and that drives business goals. So the intersection of this is what will allow you to create this information that actually makes sense for you rather than thinking, oh, is this something that LLMs will be able to spit out anyway? And the user will be satisfied and won't generate traffic anyway. And it's pointless because of that. No, no, no. different metrics, different goals. And this is content that even with zero clicks will still be valuable for us because it actually communicates what we do and establish the real expertise or unique selling proposition of our company and business, the authority of it. Right.

I think for me, these 2 pieces are super, super useful because this is the next question in our kind of our journey, isn't it? It's kind of like, okay, we're starting to fundamentally get get to the grips of what the implications of AI mode are. But then there's the, but what do we do differently? How do we deal with it? I think the HubSpot one sticks out mostly because of where they actually start giving their own examples of how they're doing it. And some of the examples you can do because you're HubSpot, you know, there's a degree of scale and authority that comes behind it, but some of it I think isn't just that. And I think there are some real useful nuggets in here that you can kind of take away. I think a lot of people need to invest time into working out, well, What are the implications? How do we actually make this relevant to different brands? The writing content in chunks advice is useful, but a lot of brands might not write specifically in that way, or their brand voice might not be entirely like currently working alongside some of these new demands and requirements. So there's just a degree of how do we actually do it, which I really enjoy. And I find this stuff kind of interesting. The other thing that really how we might need to go back to personas again for our content journeys. And I know like user personas is something that the marketing world either loves or hates depending on what month or year or how long it is. But the idea that the personalization element that mixes this up is say, well, you actually need to start working out what possible things might your specific customers want in much more depth because that might be rewarded more within the search experience. I know we've been saying that for years, like personalize your content or make your content like useful for customers, but. Search results haven't been that personalized, really, not significantly between people.

Minor, minor personalization.

We haven't had to worry about it. Like, I still think, speaking more towards SEOs than wider marketers, is we're still a little bit ignorant of, and again, speaking broadly, but we're still ignorant of what our specific customers are trying to achieve. I think a lot of people haven't gone that extra mile. And I think there is the need for that here. So again, I found that equally, equally interesting. The, the final point is this also addresses another anxiety I think is here in the industry, but maybe we're not worried enough about it is I don't really think a lot of SEOs truly understand what great and useful content is. A lot of trying and a lot of people are doing it. And again, I'm not speaking ill, but it's when we get into this scale, it's a 2,000-word blog, or for this client, we're just going to employ the glossary technique. So we're going to do definitions of everything. It's like. But you're doing that because it's a trope or a content genre that's in your playbook, but actually, is it the most useful thing to that given customer? Is that truly useful?

I think that this pretty much boils down to how strategic you are, how much you actually understand the business. Unfortunately, in SEO, especially for a long, long while, we have gotten away too much on being checklist-oriented, very tactical. And I understand, and you make a really good point about not everybody has the resources of HubSpot to scale stuff. So for example, speaking about originality, conduct original research to set yourself apart from AI and competitors. But who has the capacity to create the state of marketing report of their own industry, right? Like very few companies at the end of the day. But yes, because of resource constraints, because we have had this incentive to create content at scale, I think that we should more about quality rather than volume as a whole. And thinking more strategically about what type of content is actually really needed. To start wrapping up, the last article here, 12 New KPIs for the Generative AI Search Era, Duane Forrester. It's great to see how he goes through how these all metrics we have been taking so far a lot into consideration needs to be refreshed. So speaking about chunks, chunks retrieval frequency, embeddings relevance score, attribution rate in AI outputs, AI citation count, vector index presence. Retrieval confidence score. I believe that many of this overlap. LLM answer coverage, AI mode call success rate, semantic density score. So again, many of this more or less overlap with each other, but I believe that 3 or 4 of this, along with what ultimately ends up paying off being the goal of our work, which is conversion revenue, is what we need to start focusing more on. I also like that he provides useful tools and resources and data sources that we can start already using. We have mentioned we still need more insights, more tools, better features, but a lot of it is already here in maybe in a little bit of a raw way, but it is here already.

This is the start of the new conversation for me. I don't think this is the end where we end up, but I think this is absolutely fantastic and provocative piece to make people see it. I kind of think back to The SEO metric chain that I did where it's kind of like goes from technical to financial, it's like those key business metrics, as you said, what money are we making? That's not going to change. We still need better attribution beyond the last click for SEO there. But I think from a technical point of view, for technical marketers that really want to understand how successful our content will be as kind of a leading indicator, a lot of these feel like they have a huge amount of promise. I'm Certainly going to go and test creating some of these. I don't imagine a lot of these will feature on any exec level or C-suite reports anytime soon, but for, for technical teams, I think Duane's done some of the most significant work I've seen in the SEO space towards these metrics.

Completely agree. So thank you. Thank you very much for your time, Chris. It has been amazing to have your inputs, your insights, giving a refreshed views and different types of angles to this week's Top news, and you don't need to wait until the next week in order to keep up with SEO. Go to the SEO FOMO News. Remember to like today's video if you have liked it. Remember to check out SimilarWeb. They have one of the very few data sources that actually provide prompts in their AI traffic report. So definitely go and check it out. And thanks again for their sponsorship today. And thank you for watching our today's episode. Like it, subscribe it, don't miss it, and stay tuned for next week. Thank you very much again, and thanks, Chris. Bye-bye.