Google Tests Web Guide Their AI Organized Search Results - #SEOFOMO TL;DR July 29, 2025

Let's go through the latest news in SEO and AI search of the last week. A lot has happened, a lot has been released, and to go through it with me, I have the amazing Cindy Krumm, who's the CEO of MobileMoxie. Hello, Cindy, how are you?

Hey, I'm good. Thanks for having me.

Thank you very, very much for coming. It's a delight to always talk to you and have your insights. And before starting, I want to thank Similarweb, who's today's episode sponsor. Our first news of the week, and I think that potentially the most impactful, meaningful, if it's actually released in public, is the new experiment test from Google that they have called Web Guide. This is a new experimental AI to organize the SERPs. This is something that they had already announced. At Google I/O that they will be eventually testing and also eventually releasing. From what I have seen, and I have been testing it, they have pretty much released it for any type of query, providing a hybrid experience of AI mode with traditional search results. I was actually comparing what is the overlap for a lot of informational queries, and at least from what I could see, there's a very high overlap with the top organic search results that we see right now in the SERPs for the same queries, but that tend to be a little bit different for more commercially oriented or transactional queries. Then along with this announcement, there are people like Andrea Volpini who has already started to assess how the web guide actually works in the background, right? Analyzing the network traffic, the HAR file, and He had identified that similar to AI mode, yes, it also uses the query found out, but a server side also, there's a very high personalization of what we see. The account profile activity, the location, the device also is taken into account to provide the results and the information that we get. Then also he pointed out something that I also found to be very, very interesting how there is a clever architecture and dynamic UX that pretty much fits to the answers that are provided, that are generated in real time when the results are loaded.

Yeah, I think it's all super fascinating and you're right, it's all coming at us so fast. All of these changes are making it feel like it's hard to keep up, but I think what's happening, what it seems like is happening, is we're seeing like different presentation layers of almost the same technology and Google is just using us as their test group and like multivariate testing of which experience do people like the best and how can they measure it in ways that are going to, you know, quiet down the fears of all the loss of traffic and stuff like that. Like, I think they're trying to, A, probably improve the perception on the SEO side with better statistics. And so they want to find a way to incorporate Gemini and the conversational search in ways that the whole community And regular users are going to like. But yeah, it's, I think it's, it's fascinating. And Andrea does a great job of explaining how it's different and also how it's the same, right? It still uses the query fan-out technique. And he talks about the problem of measuring it, which I think we're all worrying about. How are we going to measure it? How are we going to optimize for it? He talks about, you know, semantic understanding and topic clustering, and that we, we're not going to be looking at individual keyword rankings anymore, which lots of people People haven't actually looked at that too much for years, but yeah, it all feels very experimental and unknown still, I think, at least to me.

If this becomes the main experience of Google, I mean, that changes completely the principles that we use and the methodology that we use to measure that, right? And I wonder where—

I don't think Google's gonna go backwards, do you?

I don't think so, but what actually concerns me is the lack of transparency from all of this activity, all of this engagement. They need to keep up. They need to compete with ChatGPT. While making it so in trying to blend in the current interface to keep showing ads, to keep performing as they perform as an advertisement platform. But we need data. We need to understand what is going on and they seem to not necessarily care about that much. No AIO filters in the Google Search Console, no information at all about AI mode engagement or behavior.

Well, I've been thinking about AI mode and Gemini and some of this stuff as kind of like a whole new Google. And even though it's using a lot of Google's, you know, old search technology, if you think back to the very early days of SEO, Google started without Search Console, without Google Analytics, and without having to be as transparent. And then they were, you know, when they realized that people were really embracing it, then they started being more transparent because they wanted to help people buy in to And help people find out that they could make money so that they should do a better job and stuff like that. But I am also very worried about the transparency because in the years since Google started, when they started without Search Console and Google Analytics, people have built their entire business and lifestyle around being able to rank in Google or being able to rank in search engines or, or things like that. And so I don't think they're going to be able to get away with that for very long. But I do think that there's incentives on Google's side to not share as much as they did when they first started engaging with SEOs in the early days of SEO. Like, we got a lot of really great tips and tricks because Google was new and they were, you know, trying to make the internet a better place. And I'm not sure if that's the same vibe that we're dealing with now. And I also think something that Google is really struggling with is I think there's been an erosion of the separation of church and state. At least with the AI models, right? Like, why would you model something once for paid and once for organic when you could use the same model both places? So things like journey modeling and probably sharing data, and that's going to be hard for Google to talk about in the midst of all of the antitrust and anti-competitive lawsuits.

Indeed. Speaking of Google releases and integrating everything in a single experience, What actually was released to the public is this try-on style with AI feature that they had been testing for a while, focused on the shopping vertical, one of the most easy to monetize and profitable vertical as well. The try-ons feature allows you to take a photo of you, upload it, and check how whatever you want to buy, any shirts, any pants and shorts, whatever looks on you. And Google has been building all of their shopping graph, adding the product packs that pretty much look like a POP, like a product listing page. And now this is like they are adding all of this sort of features that you will tend to expect from an e-commerce, a proper retail website. And this only allows them to let engage the user more and more and keep their journey within Google rather than referring them to their e-commerce website right away and referring the potential customer at very end of the journey. This feels more and more like, yeah, they want to be real competitor from Amazon. They are not hiding that out at this point. In any case, and what is interesting is that they are not even waiting to release AI mode or further integration for shopping with an LLM experience, but releasing this right away.

I think Google has 2 main goals here. And this relates exactly to what we were just talking about with the lack of break between church and state in terms of paid and organic. Because I think if you think about like, why would Google do this? Why would they invest so heavily in this? Why have they been trying to do it for years? They want to fight back against Amazon, like you said, but how can they do that? Well, if I've uploaded my picture and put in all my sizes and my preferences and my price points so that Google can start to show me stuff that it knows I'll like, because I've already put that information in, they can do that organically, but they can also do that in paid. So like if Google has an understanding of what one store has in inventory and they don't have my size, they can know to show a different ad. Potentially. And that's what I'm worried about with the lack of transparency. Like, should they be showing me ads for something I would like but that is not available in my size? Probably not, especially when they can make as much money showing me an ad for something that I'll like that is available in my size. And that's going to help them make more money and do better on the advertising side, providing an advertising platform that actually results in conversion. And maybe they'll do that on the organic side too. I think that's probably part of the plan, right? Like if they have access or know what's in inventory or what's not, or can see that through Merchant Center, whatever, they probably want to do that in organic too. But they really want to do it in paid first. And that's kind of what I'm concerned about. So that's one thing. But then the other thing is, for years, Google has been trying to undercut the affiliate marketing that they created, right? They created the world of affiliate marketing. And now they don't like it as much anymore. And fashion try-ons are a huge part of affiliate marketing. But a lot of it sends the traffic straight straight to Amazon, right? And so like lots of people who are really into fashion will have a group of people that they follow that are roughly their size. And those people will try on the outfits and give them feedback and be like, oh, it's, you know, too tight here, too loose here, whatever, whatever. Or it's great, buy it. And those links all go to Amazon. And so Google's idea here, I think, is if they can cut out that kind of affiliate marketing, then less traffic is going to go to Amazon. And you're going to know exactly what it looks like, not just on someone who looks like you, but actually you. And so they, they're thinking like maybe they can hijack that traffic and get it back. That's my thought.

I completely agree. I hadn't thought on that particular angle of the affiliate relationship that a lot of publications end up having with Amazon. So at the end of the day, this is revenue going to Amazon, right? So rather than them keeping that traffic and provide an experience for this sort of needs themselves, 100%. Sadly, at the end of the day, you know who are going to be the ones who end up getting hurt the most, right? Like the middlemen, the, the publications, and especially, and, and feel for the ones who are indie, the independent ones, not necessarily the major networks that will end up diversifying and finding many more revenue streams in other type of partnerships and any other ways, but the smaller ones that had a heavy reliance on Google. All of this. Connects very well with the next 2 articles, which are different sides of the same coin. On one hand, Google in their earnings call of Q2 disclosing that AI Overviews drive 10% of more queries, so more page views in their SERPs on one hand, but at the same time, shockingly or not, the Google click-through rate drops 32%. for top results after AI Overview rollout. So more traffic, more engagement in Google, but less traffic refer to websites, which are at the end of the day, the ones that keep Google alive with their content and the one that Google use for their SERPs and to provide a good experience to their users. So I believe that unfortunately we have reached that point with Google that the balance is on the extremes at this point, and there's only so much that it will subsist like this. If the traffic continues to drop more and more, the incentives will be gone for original, expert-driven, authoritative, unique, original content that is human-written. But these 2 insights, the 2 sides of the coin, shows very clearly that Something has changed and Google is not losing, will never lose. They will continue releasing things to keep users engaged. The ones that end up losing, unfortunately, are ultimately the website and business owners.

I think you're totally right that these statistics are both in, in Google's favor. And I think one of these articles talked even about how when AI overviews are present, The one of the kind of positive things for SEOs is like when AIOs are present, people do actually sometimes search and things in lower rankings like 6 through 10 are getting more clicks. But to me, what that says is like AI overviews aren't answering the question and people are looking lower and they're not finding the answer. And I think this goes back to some of the stuff that's been published about Google actually actively from the DOJ trials, Google actually actively testing degrading the results and seeing no negative impact to their business. And so number one, like seeing an increase in the number of searches and a decrease in the CTR, that's necessarily going to happen when there are more searches. People aren't clicking, not necessarily clicking more. Maybe it's not necessarily going to happen, but it's expected, right? And these are both statistics that you would also expect to see with conversational search increasing. Because when you're doing a conversation, it's going to be an iterative process to get to the final question and the final answer. And so that's going to have more queries and fewer clicks. The hope, you know, what we hope is that Google is putting the right things into the AI overviews and the right things into all of the different variations of Gemini that they're testing everywhere, and that people will eventually click But right now, you're right, websites are the ones that are suffering. And Google, I think, is going to continue to harp on this. Oh no, queries are increasing. That means people love search. Like, no, that means people aren't finding what they want, or they're engaging in search differently. But a successful search means that you get it right on the first query and you're done. So unless the population of searchers is going way up, I wouldn't say like that necessarily the number of queries going up is a positive statistic for Google. They're trying to spin it that way. And if, if what they're going for is conversational search, then yeah. But if they're going for just answering questions, not necessarily.

Not necessarily. This also ties well with our next article here that was the outcome of the tests that I did, that funnily enough, ironically enough or not, ChatGPT is using Google SERP snippets for its answers in those circumstances, in those scenarios in which you press for ChatGPT to find an answer from the real web. Of course, they don't have that in their trained data from their own internal data. And rather than going and trying to look for this information with Bing, which is what I would have guessed, that they will do that, they end up getting the information from the Google SERPs. And a lot of people came back to me after I released this test saying, oh, but how do you know that this was actually from the SERPs from Google and not from the ones of Bing? First, because it wasn't indexed in, in Bing until much later after I did the initial round of tests. And then this snippet, the description snippet here is not taken from the meta description of the page. It's the one that has been rewritten by Google because there was another person telling me, oh, but what if this is from the Common Crawl or whatever? No, no, no, no. This is, this is not the meta description here. This is the rewritten snippet from Google. So yes, I get that Cindy is speaking about monopolies, et cetera, right? Like Google has this massive advantage that we thought, oh, ChatGPT will be able to be the leader, has all of this new technology, is able to fulfill better the users, but they end up even going back to the SERPs. And I guess also another point regarding this, this shows once again why SEO is not dead. Hello. And this shows quite clearly that ChatGPT is behind Google.

So I think this is fascinating. It was a really great test and I was so glad to see that you did it and, and all of your findings, but Think about this test in the context of the stats that we just looked at on the 2 previous articles. The queries going up and click-through going down. ChatGPT might be causing part of that, right? If every time someone does a ChatGPT search query, they're hitting Google, then that's going to drive Google queries up and CTR down. If ChatGPT isn't actually clicking on anything, which it doesn't appear they are, they're just scraping the description that Google has put in, not even the meta description. They're not even requesting the page. So I think what What that says is, number one, like ChatGPT may be making those statistics like look like they're more— be more warped, and it might be double counting because if someone is doing a query in ChatGPT, it's also hitting Google as a query. And so like all of these statistics are kind of messed up if that's the case. And so, you know, ChatGPT is doing queries to respond to Users' queries, and maybe they're doing queries preemptively to collect information. We don't know. So that's, that's all weird, right? It's messing up the data when one entity who's trying to become a competitor is using the data from the existing monopoly. I guess that's how it's gotta be. That, that's part of them trying to overcome the barrier to entry because the barrier to entry is all of the data that Google has collected in the topic layer and understands. Over the years, and ChatGPT is trying to make up the distance there. But you're right that it also means that the answers that ChatGPT gives when search is on is, are already pre-screened by Google. So Google, whatever Google thinks is good is also being fed directly to ChatGPT. So it means that we're still not getting, you know, the, the original marketplace of ideas that the internet was supposed to be.

100%. And to start wrapping up, I'd love to cover the last news of the week. A lot of us actually were able to keep up with this with the shares from Dan Taylor, also Natalia Wittek, also Kenichi Suzuki. A few SEOs were at the latest Google Search Central event, but this time was a little bit different. It was a 3-day event. Much more in-depth event. Actually, one of the things that I want to highlight was that from some of the photos that were shared, actually Google went much more in-depth and were much more actionable than in any other event that they, they have done. They actually share geo-targeting factors, for example, or what affects the crawl budget of a website or not. If, if What HTTP status affected crawl budgets, or if, if something was an indexing signal or not, for example. So I, I can definitely see how we have been talking about lack of transparency, et cetera. But all of a sudden there is this new type of event that they also took or leveraged, let's say, to announce new releases. Funnily enough, they didn't announce nothing about the Web Guide or these new features, but they announced that. New API for Google Trends, which is great, but nothing related to AI. And also here we have how Kenichi actually highlighted all of these insights regarding AI and search in particular, and how the bigger point here from Gary is that AIO and other AI features on Google Search use the exact same infrastructure, signals, and systems as traditional blue links. And he explicitly stated there is no need for a new acronym. Or a separate discipline, and the core principles of creating helpful, reliable, people-first content remain as the foundation for visibility in all of these new search formats. So yes, very interesting moment for this new type of event. I am the first one who embraced more insights from Google from one hand. So great to see the confirmation of all of this indexing signals, geolocation signals, crawl budget. factors and so on. But yes, it's interesting to see the, the take, let's say, from Google regarding AI search and the need or not of something new to optimize for it.

This is the first time that Google really does seem like they've gone out of their way to be actionable, especially in relation to all the changes in AI. And, and so I think that's great. And the coverage from Kenichi and Dan both was like Super great, super helpful. I was really glad that they were there. I hope that Google continues to do stuff like this and actually ups their representation at industry events so that we can get really good, clear guidance about what they want us to do. Because I do think that they've gone back and forth on how they think publishers should be using AI. Because they, they definitely think that advertisers should be using AI and they've created a bunch of tools. For advertisers to use AI in their managing their campaigns, creating their campaigns, creating their creative. And then they seem like they want us to be using AI to some degree, but then they, they'll double down and say, but it has to have human oversight. But I think there's an interesting conundrum that they've created where, like, everything is now AI-able, like, and that makes everything kind of infinitely Spammable. So you, you, your first take is that the Google Trends API has nothing to do with AI. My concern is that it absolutely does. Like, why wouldn't every news publisher latch into the Google Trends API and start pre-writing every article on everything that's trending in every country everywhere around the world? And that's a problem for Google though if they do that. Like, they've created this thing where if they give us cool things We're going to use AI to exploit them, and that means there's more that they have to index. I mean, I just don't know if it's how they're going to manage all of these things and how they're going to manage the messaging around how and when to use the tools that they've given us. And I worry that it'll just be another excuse for them to give us less data, less tool.

Yeah, it needs to be consistent, 100%. I guess that at the beginning when they released certain best practices, they never thought that it was going to be manipulated as they did. And All of these AI-generated spam ended up making their lives much, much harder in terms of indexing. Well, we see how much more difficult it is to get indexed now than before. So I guess that they have had to fine-tune the way that they communicated that. As usual, we need to use our own critical thinking, I guess, and the best benefit of the companies that we work with and represent rather than naively just follow Google's best practices, right? It's like, is this actually applicable and what are the potential backlash that we can suffer, especially for each website's own characteristics and context. I would love to finish on one hand, first, thanks Similarweb again for their sponsorship today. They actually have very useful AI traffic related features and reports, as you can see here. They're able also to provide insights like the ones that Google don't yet provide themselves or even ChatGPT regarding top prompts, for example. So if you haven't taken a look at it yet, go and check it out. is definitely useful. On the other, Cindy, I would love to hear your last tips. Any thoughts to finish today's episode?

Yeah. So I think now is a time to be paying a lot of attention, but not overinvesting in any one decision that Google makes, because I think Google is just testing. And so they may do this, they may do that. They may try this for 2 weeks and it doesn't work. So don't, don't live and die by the things that Google does or says right now. Like you said, make good decisions for yourself. Think about your customers. Think about your business as a whole. And then when you're trying to anticipate what Google is doing or why they're doing it, think about all of these things that Google is doing and relate it back to your business. You have to break it down as if you were Google as a business and think about how is Google going to leverage this? How is Google going to make money on this? How does this benefit Google? And that will give you kind of the insight into what you need to do to be a part of it.

Amazing. Amazing. That is definitely a really, really good Last recommendation for sure. I really appreciate that you joined today. And hopefully this was eye-opening for a lot of the people who are watching, not only to keep up with the latest news, but also to identify the trends and the opportunities for sure. Thank you again for joining, Cindy. Thank you again for watching. Remember to like the video if you like it, to subscribe to the channel to avoid missing out, and to visit the SEO FOMO News in case you don't want to wait until the next edition of the SEO FOMO Newsletter. Until the next one, bye-bye.