speaker-0: Hi everyone, this is Beers and SEO with Barry and Steve, and I'm Steve.
speaker-1: And I'm Barry.
speaker-0: And we're gonna be talking about some of the most recent biggest stories affecting search at the moment, and also we're gonna be talking about that fancy new functionality in Google Search Console where you can remove your content if that's what you want to do from Google's generative AI search features. But as we always do. First the beers. Barry, what have you got?
speaker-1: I went back to the very first beer actually and it's primarily because I've been so busy I didn't actually have time to buy anything new or different and I've just opened it and it's decided to spill all over me. So that's a great start to this pot. Great start to this pot. But yeah, it's it's Rockshore Irish lager. It's lovely to drink. Very easy, very clean, made by the same people who make Guinness, so you sold can count on them. once the foam dies I'll actually take a sip of that.
speaker-0: Nice. Well, I I am on Fierce for Good or Fierce Breweries. This IPA says this IPA plants trees. And I got it just because I don't know any I don't know Fierce actually, so I thought I'd give it a crack and it's is a little bit of a heavy, it's a six well six point two hazy IPA. one of those kind of like fruity Fruity IPA, so you've really spilt it over there, yeah.
speaker-1: it's all over my desk. My entire office is going to be stinking of beer for the next week or so, so
speaker-0: That'll be fun. worry, mate. We've all been there. okay, so I thought we'll take there's i you know, sometimes between podcasts it's been it gets a bit quiet on the Google landscape. Obviously AI search is rolling every single second. but there's been a few sort of like decent updates that's got everyone talking at the moment. and so I'm gonna just I chose one because I thought it was an interesting one, which is There's been a piece of research by D Demand Sphere looking at AI overviews invading branded search. and I'll just go through some of the what they said. So they they said that their data showed that AI overviews rose from roughly twenty twenty-six percent of branded queries from the start of September to more than eighty percent by the end of the month. And it's an interesting one. It's caught up caught a bit of attention because I guess with the land of or the the area of Zero Click, people have looked at branded searches as a way to say, right, we might have seen a decline in in traffic, but actually we are maintaining and growing our core audience, th which you can evidence by using branded searches. And So this one instantly got a lot of viral coverage because it was a a panic because it was like please no, please don't affect the branded searches. so look, I'll I'll give you my view on it is obviously when I saw that I immediately started to check the branded searches for my clients and I haven't seen the drop off that was indicated here. now that's so I think my my view on reports like this is absolutely it's an interest it's a very very much an interesting read. but always go and check your own brands, go and check out your own data reports because you know everybody's industry is different. What do what do you think, Barry?
speaker-1: I have been able to replicate this to some extent. I've done a few test searches myself on branded queries for some of my clients and and other relatively big brands. About half of them do not show an AI overview. The other half, when there is an AI overview, it's below the organic result. with some exceptions. If the brand is particularly big, the AI overview might be the first result. And I see why Google is doing this. Because when a lot of times when people do a branded search, they don't necessarily want to go to the brand websites, they just want to learn about the brand. And the A overview is basically an accelerated version of the knowledge panel we already had for branded searches on the right hand side or desktop, or prominently on mobile, where it basically gives you detailed information about the particular brand you're searching for. Plus it has an added benefit for Google is that brands who really want to claim that branded search traffic now have an extra incentive to advertise on Google for their own brand name. So it's a it's a revenue source for Google as well. And let's be honest, you know, they still need to to make money. So you know being worth what is it, two trillion dollars is obviously not enough for Google. so it that way it's another revenue scene for Google. I suspect this is going to accelerate and enhance even more. this is just the first forays into AO Visual branded searches. I expect it to become more of a norm the moment Google understands a brand well enough that it can show an AI summary, it'll show an AI summary. and that also means that as a website, this is probably a surface you can optimize for to an extent with your about pages, with your mentioned content on other websites about your brand. and that actually makes brand reputation even more important. because without a solid brand reputation you run the risk of negative comments being summarized in the AI summaries that Google will show for your brand search. So yeah, like you said, it's a space you need to keep an eye on to see how it encroaches. You probably want to prepare for it sooner rather than later and make sure that when the AI summaries do come for your brand searches, that you you're already there. You already have I wouldn't necessarily say claimed an AI overview, because I don't think you can claim it, but at least you you're feeding the right information into the AI overview so that you have some measure of control over what The AS only says about you and about your business.
speaker-0: Yeah, I think that's absolutely spot on. I've I I I wrote a piece a couple of weeks ago, investigating how news publishers are how can they how can they give some kind of a control to how their brand appears in in various AI search engines. And I looked with focus at the about page, because I'm a nerd like that. And I I definitely saw absolutely the about page is important. But if we look at those things that might be important to your brand, like is it a trusted well, is it a trusted media organization? what is it value for money? you know, what what kind of impact is it making? there'd be variations on that if you're actually selling a a a a product like a car for instance. But yeah, my my current results of it was really to kind of don't just don't just update your about page. It's about just looking f and establishing an evidence architecture, as I've said, because you're gonna have to look at multiple places across your website where you might be talking about your brand. So the about page, absolutely, but definitely the subscription pages or and other pages which I've written about, which have appeared in AI search. one of the things that I saw which was quite interesting is that we've we've kind of become accustomed to thinking about the check a checkout process to about keeping it super basic and clean for that to not get not get in the way of the audience. But I've seen the kind of page just before checkout. a feature a lot in AI search and that's seems to be an area where you could put a couple of lines of information about your brand. Anyway, I think with all these things there's so many fluctuations in in AI search. It's really about make making sure that, you know, you're just reinforcing your your brand messaging, but also, you know, don't piss off your users. Basically, don't do anything wild, because you've read a LinkedIn po post, right? We're all seeing disasters from taking just AI search tactics with no thought of other con main considerations you might need to take care of. Okay, interesting. So moving on to the next one. So this is obviously a biggie with Google saying and r adjusting its documentation to indicate that you should be doing a human review of your AI content. So Google says it's updated its guidance on generative AI content, bringing its public documentation into line with material from the search quality rate guidelines and its developer presentation. So a couple things from this is that Google now explicitly describes four qualities it wants to see in the main content of a page effort, originality, talent and accuracy. It even gives mass generating text of AI without human oversight or creation as an example of low effort. So it's using this phrase of effort. so what do you think of all this then, Barry? I mean it's obviously very key when Google is starts to our update the documentation, right?
speaker-1: Yeah, I mean that's a clear sign that Google takes this a lot more seriously. And I think also clarifies what Google has been saying in more indirect ways for for many months now. that that AI generated contact content in and of itself isn't a problem, but if it's low effort content, it it becomes a problem. So you can use AI to to aid in the creation of your content. But if your content itself doesn't actually do anything particularly interesting or worthwhile It doesn't demonstrate any of those aspects that Google wants you to demonstrate. Like, you know, you put proper effort into it, which is often why AI actually hinders you because it's a obviously a show of low effort, because it makes writing too easy. If it's not original, and again, AI is a hindrance because by definition, AI cannot be original, because it's just rehashing what is found on the web in in different worms. It's a it's a word predictor. Talent skill, same there, and accuracy. Well, accuracy is probably the part where AI does give you a bit of an advantage because it tends to Be very reasonably accurate in in the information it presents to you. The hallucination problem hasn't gone but is more or less solved, fortunately, thanks to BetrieVlogment degeneration. So it's become less of a problem. but I think again it shows something that that both you and I, Steve, have been saying for well, coming on to three years now, since AI ever started making it to the mainstream, is that you you need to actually add something of value to a particular topic if you have any hope of driving visibility and traffic to to that content. You know, you need to look at your originality. You need to look at are you adding information to it? Are you contributing something valuable? Are you giving it an additional perspective? If you are just rehashing what everybody else is writing about a topic, you should not expect that content to be rewarded in any way, shape, or form. And this predates AI significantly already. It's just with AI, Google has gotten better at this. Google has always said, you know, you need proper original high value content. And if you're especially as a news publisher, if you're just reporting what everybody else is reporting and you don't really do anything original or new or different, then yeah, you're you're probably not gonna succeed in driving a lot of traffic to your content. And with AI, because AI has enabled the scale of content generation to be orders of magnitude larger, which become a bigger problem. And I think that's really driven Google's efforts into weeding out the low effort content even more because writing as a pure task is no problem of the simplest things you can do. But writing quality content is something that, you know, Google still wants to reward. So you need to show that in your content. You need to make sure that you demonstrate your your effort, your expertise, your skill level, your originality, all of that. And that your information of course is accurate and and deserves to be shown in in search results. And that you know you can translate that to journalistic processes as well and and where journalism Can use AI and should not use AI.
speaker-0: Yeah, it's I I feel almost feel most sorry for Google in this area because it's trying to indicate that you're you know you need to that your your content needs to be original. and so it starts to introduce language like effort and I feel sometimes people respond to that by trying to work out what effort means. Well really the overall point is Deliver original content, deliver something. I mean, I've I I mean, you have to be careful about using the word content, right? But deliver brilliant written work that delights your audience, that they ultimately, when they finish reading it, they'll go, Bloody hell, that was amazing. And they'll come back to your brand, right? These are like basic things, but I think sometimes people get too fixated on the the words that are changing as opposed to the main point.
speaker-1: Mediocrity is just not acceptable anymore if you want to be taken seriously as a publisher. And I think there's been too many websites out there who've coasted on mediocrity and have relied on tricks and and tactics to get that mediocrity to be surfaced in Google News and Discovery in Top Stories. That just doesn't work anymore. Mediocrity is is not even table stakes. Mediocrity makes you just part of the masses. Table stakes is excellence. Just just the opportunity to get traffic means you have to be excellent to start with. And then layer on top of that. If you're not excellent, and you have to be honest with yourself here, if you're not meeting that minimum threshold of this is shit hot content, do not expect it to be rewarded by the machine systems in any way, shape, or form. You excellence is table stakes. Excellence is just sitting at the table before you even do anything. And then hopefully you can build on that.
speaker-0: That's I tell my kids every day. Okay, so on that, and well connected, is really this whole question now, right? Are we about to have a massive Google Core update based on this change in documentation and the steps that have been taken around it? And obviously I'm I'm pointing here to an article which Lily Ray published a couple of days now, where in her Substack she is pointing to a number of different signs that are indicating to her and her team that that something may be happening on on horizon. One, if I go through some of the points, we've got those changes in in the guidance. Google have been talking and presenting in Barcelona and other areas, talking about this about these about hu you know you have to be review your AI content. Google now calls fake authors a form of of decep deception. That's something that it's called it's come out quite blatantly on. Google's added new language to its helpful content guidelines, which was about a month ago. And Google has released four spam updates. And it's very interesting about these spam updates, right? Because previously, I think last year there was one, as Liddy points out in 2025. Now there's been four spam updates this year. So And it's not just Lily, it's other people like Glenn Cabe and Cyrus Shepherd, Marie Haynes, et cetera. So lots of different people, lots of experts. what's your take on it, Barry?
speaker-1: The cadence sort of fits, yeah. The signs are all there. I'm just not entirely sure if this is because of the spam updates or if Google is really setting the stage for a major core update. Having said that, end of October, start of November and or start of December tend to be times that Google wills out core optics. the timing sort of fits as well. Plus, as a general rule in my SEO career, I never go against what Lily Ray says because there's a ninety nine point nine percent chance. She's right. I haven't really caught her in any mistakes that I can remember. So yeah, definitely something to keep an eye on. we've been reasonably fortunate as the news industry with the last few core updates that there has been some impact but not dramatic impacts. It's more like a little bit of reshuffling here and there. It's not like we've seen entire sites wiped out, with some exceptions. And those sites that have been wiped out deserved it, let's be honest. so hopefully if there is a core update coming and as of the recording of this podcast on October seventh, we haven't yet see the core update roll out and the spam update is still rolling out. So hopefully we'll we'll we'll get clarity on that soon by, you know, the core update starting, the spam update ending, or you know, Google might overlap the updates as they have done before in the past. Yeah. And yeah, hopefully it's not one that's gonna hit the news and publishing industry particularly heavily. I honestly feel In many ways, if you just look at search specifically, that news in terms of how Google has approached news websites has been fairly stable. That Google hasn't really felt the need to interfere too much in there as they have done with previous core updates. I think Google is fairly comfort comfortable with how news is currently being ranked and shown and and which websites show up and which websites don't. In Discover it's a very different picture. I think in Discover it's still a bit of a free for all. Google really hasn't nailed down the span problem in Discover yet. I think in search they've more or less nailed it. And I don't expect the next core update to to cause a massive shake up of visibility in search results for most news focused websites. I might be wrong on that, but that would be my prediction.
speaker-0: well we'll we'll have to wait and see, right? I think we'll all be eagle eyed on the various sensors that are out there. speaking of like different various simple like changes, have you seen top stories in AI mode or AI overviews yet?
speaker-1: I have not yet seen them myself, no. When I see a top stories box, it's either the main box on the search results page or a supplementary box. I sometimes see an AI overview and a top stories box, but never integrated into one. But there's an AI overview that summarizes it, usually below a top stories box that actually covers the later stories. I have seen screenshots of top stories carousels in AI answers, but I haven't spotted them in the wild yet myself.
speaker-0: And do you think obviously we're being pushed more into AI mode on on Google's on the call it page one if you like, whatever page one means now. Do you think we'll eventually be pushed into AI mode? AI mode will be the default experience or it will be a combination of different elements?
speaker-1: I've said this several years ago and I still stand by it. I don't think AI mode will become a default search experience. I think Google will always need to give at least the illusion of choice in the form of the blue links. I think if Google steps away from that, they are at a very strong risk of undermining what made them big in the first place. And I think they realize that to an extent. Yes, they want people to go to AI mode, but they want that to be a conscious choice that they make, that a user specifically makes, to go into AI mode and ask questions and interact with Google in a different way. I do still think that. they will keep the relatively classic search experience that they have, which yes, includes AI summaries and AI overviews, but also includes the tambour links, includes the top stories boxes, video results, all those sorts of things. Because they realize people trust those results more. The illusion of choice is a real thing, a psychological phenomenon where if you give people the at least the the illusion that there's a choice, that they can source different content from different publishers, that they trust the results more, especially Because people trust the search results as much as they trust the brands that are being mentioned in the search results. If you have a search results page that shows New York Times and The Guardian and The Independent and a few other high value news publishers, people put more faith in that result than if you have a a search result which is full of Joe Bloggs' home website that nobody knows about and never heard of. So I think there's a strong incentive for Google to keep a form of a classic search result page. with which contains, you know, s different elements that people can click on and not go full into AI mode. But yeah, it's obvious Google wants to push more people into AI mode and the data backs it up. More people are using AI mode. But that tends to be complementary to people also using classic search. It's not that AI mode replaces search. It's I think what Google wants to do is position AI mode as a Chat GPT alternative in addition to its classic search results. I think that's what the overall game is that Google wants AI mode to become a free version of ChatGPT where people can do those sorts of basic research that they use ChatGPT very heavily for, but also still have classic search, to what extent we can call it classic search, to deliver that initial search experience that people have gotten so used to.
speaker-0: Hmm. because I I I only mention it because it relates basically to like what one of the well, the main topic that we're talking about today, because you know, we're very keen on knowing what where Google is moving. Now, particularly from a publisher point of view, right? Because now we know and and that now there's a funky new bit of functionality in Google Search Console where you can block Google from showing their content in the generative AI. experiences, right? So AI overviews, AI mode, and also in the AI generated areas in Discover. So the question is, who's going to do this? Because I think, you know, with the with the we I think from the various different sort of like newsrooms that we speak to, Barry, we know that there's some form of experimentation happening. But it's it feels like it's limited at the moment because it feels like it's limited at the moment because it's only at sight level, right?
speaker-1: Yeah, it's an all or nothing kind of thing at the moment. so you have your either your entire site included as a citable source in AI overviews and AI mode or nothing. I d have not yet seen examples of publishers who've opted to opt out of AI overview in AI mode. And I think the business case for that is very clear. Whether you opt out or not, the AI overview is is gonna sit there. It's gonna sit there, it's gonna be there. And yes, we know when there's an A overview, you get significantly less traffic, but you still get traffic. You still get clicks. And I think there is still some value in being a cited source in AI overview unless you are a very hard paywall website and you think, you know what, any engagement with the content, you need to come to a website and pay for it. Which can work if you're a very powerful brand that people will actively seek out. Not a lot of publishers are have that luxurious position. So I think for publishers there is still a case to be made to say, you what, let's stay included in AI overviews and AI mode is a cited answer, because, you know, some traffic is better than no traffic at all. Because whether or not you're included, the A overview is still gonna sit there. And if they cite your competitors overview, you know, they might get that little slice of traffic versus others. It does point out the p systemic issues, of course, that AI has has created for the publishing industry in that they take all our content, summarize it, and then don't send traffic back. and there's just a a deeper issue there that has been playing for years that hasn't been solved yet. Google has started to show signs they might be willing to solve that. by making noises about compensating publishers if they are a a substantial part of an AI overview. That's a beta project they're running at the moment. So we'll see how that pans out. But I wouldn't expect that to become a major revenue source for any publisher in the future. the the opt out mechanism exists at the moment because again, Google is legally forced to do that. I think they don't want to give e publishers that, but competition markets authority in the UK have said now you need to allow Publishers to opt out of AI while still being included in search. So this is what Google has built in Search Console, rather than like a robot.txt rule or some something else. And it's just one of many things Google is now trying to provide publishers with slightly more granular controls for those that want to make use of that. But it's probably not granular enough because it is site level, it's not page level. I think if it were page level, I think with like a meta tag or something, I think more publishers would do it for specific pages. For example, paywalled articles versus free articles.
speaker-0: Yeah. I think I think the the page level may be coming around March time, although that's always subject to change. And yeah, so at the moment it's it feels more like a a reaction to regulators to offer something. So it's like a false acquiesce in in that it appears to offer publishers more control, but I I'd really genuinely haven't seen evidence of anyone that could use it. because the threat of competition, right? They're they're gonna if you've if remove yourself from the space, someone will fill it. but I guess what we could discuss is really, you know, if publishers were considering to this kind of like talk in advance of let's say March is the time when they get the page level and publishers may be considering removing substantial parts of their content, what kind of considerations they might want to put in place before before making that business decision, right? So I thought we could t talk about a number of different points on that. So obviously I c I think like with any kind of like business decision there, it's really about creating a baseline and working out where where is your visibility to date, right? So it's definitely trying to work out do the publish does the publisher or the brand really understand what they are currently allowing Google and AI platforms to use, right? So you mentioned it before, but you know, doing things like checking your robots TXT file, what information in i i i is in there, many companies may be unaware of what that file actually does. obviously you can put in your own hosted solution, there'd be controls at that level, Cloudflare offer very specific controls in the area. But You know, it's looking at things like no snippet, what instruction instructions you'll send into Google extended. You've obviously got your paywalled content as well. And I know that you've written extensively about things that you have to put in place, Barry, about paywalled content. but yeah, it's definitely around working out what is your visibility and and that's almost like that's always worth doing a stock check, right, Barry, in that kind of regard?
speaker-1: Absolutely. And also understand that the AI chatbots use different mechanisms for different use cases. And so they have different bots for training data versus retrieve logmatic generation versus user initiated actions. And the user initiated actions are very interesting because they ignore robots.txt rules. because they're basically mimicking a browser session. So you can use ChatGPT to go and fetch a web page and then summarize it. And it will not obey robots.txt allow rules because far as chat GPT is concerned, that is a user mimicking a browser session. And the page, regardless of robots.txt, will always be browser accessible. So again, you know, there's different context for different blocking rules. For robots.txt versus CDN level blocking versus metadata on the actual web pages. And there's different mechanisms for different use cases. And you like you said, you start with with understanding what the bots are currently allowed to do in which context and then you have to make business decisions about each of those contexts and decide do we want this to happen or do we not want this to happen. And and that is a tough challenge as well, because there's often not a very clear value exchange for allowing AI access or disallowing access. Because we just don't understand the landscape yet. Yes, the traffic implications are pretty well understood by now, but AI isn't just a traffic generation tool. It's not like a search portal. or a social portal. It's of course a very different context in which users use AI and for what they use it. And what kind of value is there in being a cited source even if you do not get the click? What kind of brand power does that generate for you? What kind of recognition? What kind of complex funnel is that a part of that might eventually lead to a subscription or or whatever sort of conversion you associate with that? And that's the tough challenge and that's something that, you know you can't just pull out of a dashboard and and and slap onto a report. And that's more about understanding where you fit in in user journeys, what the value is of your output in a broader context and how you can monetize that. And yeah, I don't think anybody's really cracked that particular nut just yet.
speaker-0: No, I mean it's it's obviously harder for news publishers where their content or their it is the pages, right, for some of the brands that outside of news that I'm advising, you know, it's more a port important, so brands that might have multiple products. It's it's more important about the relationship between the parent and the products, you know, Do are the products competing with each other as opposed to complementing each other? So it's more around the misinformation or sending clearer signals to AI so that it can g so it can have a clearer idea about the clump the complex nature of your of your business. But you know, ultimately a lot of those brands are actually selling a much more expensive, either a physical product or maybe there might be selling information which is locked down in a tool which is not open to crawlers. so premium data offerings, for instance. So but it's different it's a different ecosystem than than journalism. yeah, I I guess like we I mean I mean I a absolutely this whole process of should we remove our content from AI overviews? Really, it it it does boil down to well, have you done an audit on how original our content is in the in the first place, right? Because it's not worth I don't I don't think it brands should kind of investigate this if their content is already what they call the ugly word commodity content, right? Because What's the point in in doing all that if the bot's gonna go elsewhere anyway?
speaker-1: Yeah, I mean, i if as a publisher you're still relying on commodity content and I'm sort of starting to get a bit tired of that phrase myself, to be honest. But why are you in business in the first place? You know? you have to have something unique. And I feel like a a broken record. I say this almost every podcast, don't I, Stephen? You and I are so sick of talking about it. But if you're not doing anything different and unique, why are you in business in the first place as a publisher? and I think some publishers overestimate their own uniqueness to an extent. They're not really aware of of what their added value is. or aren't honest with themselves about it. but others are very clear about what their brand is. And it doesn't necessarily mean that you have to report something different than everybody else, but it can mean that you report the same news with a different angle. The different slant, a different perspective. And that too is valuable. And this is what you see with a lot of the UK press where you have the same news being reported on the Guardian versus the telegraph, just they have different slants and different angles on them. And they cater to different audiences. And that's fine, that's great. That's how you build a brand and that's why both the Guardian and the telegraph were reasonably successful with the Their their own respective forms of subscription models, because they know their audience and know what the audience wants, and they feed that into that audience. And yeah, I I work with too many publishers, especially in specific niches, for example, Formula One or Football News, where the site just doesn't really take that unique box where the site feels that they do something different, but when you actually start reading the content, it it's just the same reports that everybody else is reporting. and that the the site sort of overestimates its own importance and value. yeah, but we have paddock access to Formula One. Yeah, everybody in Formula One has paddock access if you write about it. That's not new. you know, w w what can you do that is new, that is different? And and that is challenging for some websites. it's not easy to distill that and find out what makes you a a unique proposition. And sometimes you need to go outside of your own constraints as a written word news publisher and try it something different. Okay, you might not be able to write unique articles about what's happening in Formula One, but you can start a podcast, like we are, or you can do a video series or or you can do something else that that makes you stand out from the masses of of news reporters in your p particular niche. That just adds that layer of differentiation to it. and, you know, you have to really put effort into that now and you know, the journalism itself is is just one part of the the overall picture that you're presenting to users and in the end you're still, you know, wanting people to read and engage with your content. And that means you have to have something that that is worth engaging with.
speaker-0: As we keep saying, yes. I guess the the other step absolutely when considering pulling your brand out of these generative AI experiences in Google is just working out how dependent your brand is on Google anyway, right? Because I'm just thinking about a a scenario where let's say somebody outside of an audience specialist team might say on principle We should definitely opt out of Google's generative AI experiences for these for these reasons, without actually knowing how dependent that your brand is on there. So that's definitely worth sort of like measuring where you are in that space before you pull the plug.
speaker-1: Yeah. I'd also say that you have to be careful in how you measure that and what your analytics tool tells you. I've had discussions with clients recently and also saw a LinkedIn post, based on chart beat data. Now I have a big beef with chart beat data, because chart beat data has a traffic source they call internal. It means every internal click is is listed as an internal traffic source and gets a separate graph in that bar. In that in that report. And I've had discussions with clients where they look at the chartbeat report and say, Google has only five percent of our total traffic, and therefore we're not the dependent, we can just kill off Google entirely. And that is such a narrow and naive view because it means you don't understand what that data is telling you. Give you an example. You get a user who clicks from Google result to your website, that is one page view associated with Google. Then that same user from that initial page view clicks on three more articles. For most web analytics programs, that would mean four page views associated with Google. But not for Chartbeat. For Chartbeat, that is one page view associated with Google and three associated with internal, because they're internal clicks, not Google clicks. Which means if you kill Google off, you don't lose the one page view, you lose all four page views. Which is what's happened with some of my publishers who didn't take Google or SEO seriously because it's only five percent, and then they end losing 20, 30, 40% of that traffic because they just didn't understand what the data was actually telling them. Yeah. Which is one of the reasons I don't like Chartbeat as a as a tool that the default reports has this internal traffic as the biggest source, but because it strips attribution from that internal traffic entirely. And that is a very, very dangerous place to get in because internal traffic isn't just people clicking. They arrive to your website through a mechanism and Chartbeat doesn't tell you what that mechanism is. It's just an internal click. That internal click i comes from a source. And that source is usually Google. So you have to be very, very careful with that data and don't make rash decisions on a you have understood graph that somebody shows you. You have to actually understand what the data means, what the implications are, before you decide to opt out of Google entirely.
speaker-0: Yeah, I mean I mean that's just around I mean that's a that's a absolutely key case study there. But yes, it's it's a reminder because we are definitely in the land of multiple pretty dashboards at the moment, right? And some of them are very expensive and it they seem to promise the world. So particularly in our roles as consultants, Barry, it's Quite often it's our role to really get to grips with the data because sometimes internal teams don't have the time to go through the ins and outs of of these tools. So if they were to hire us, we can help in that in that area, right?
speaker-1: If I get a penny for every time I get approached on LinkedIn by another prompt tracking tool wanting me to test them out, I would retire because I wouldn't need to work anymore. It's just ridiculous. I mean, stop people reinventing the wheel. prompt tracking has its own issues and I'm not a particularly huge fan of it unless you do it in a meaningful way based on actual user prompts rather than stuff you pull out of your backside yourself. so yeah, don't buy into the latest hype without understanding what it actually means and what the value of it might be to your business.
speaker-0: Yes, absolutely. I advise clients. I think the worst there's nothing there's nothing wrong, I think, with using synthetic prompts if you want a quick snapshot of something. So you might want a quick snapshot of a of how a product sits alongside in jet very broad terms, the landscape or of where it sits in alongside its competitors. But if you want to then apply resource and get internal teams focused on it. Absolutely get serious about it. I mean I I I use I use open source prompt tracking using methods where is extremely cheap for me. but that's only when I want to do some very brush broad brush strokes preliminary analysis. But as soon as it gets serious then absolutely like you say use proper tools, but also you know, get have a governance model around prompt tracking. try to use real questions that are gathered from what your customers are saying and there'd be multiple inputs for that. But yes, this is that is a separate podcast, I think, Barry. but yeah I guess another thing with the this ability of being able to switch off Google gener generative AI features. you know really yes you could do that but it's just a reminder as you say unique quality content but it's also absolutely convert trying to convert your user into you know try try to forge that direct relationship right so for instance okay you could take that app action but elsewhere for instance what are you doing by way of news l newsletters you know I have recently sort of transferred clients from one email database or service over to another and that's interesting in itself to look at w who are actual users and well cleaning up the the the the email database basically and working out who who are the real real engaged audience here and how can we not only appeal to that core audience but also build the awareness of that newsletter. So I think this time's like really, really interesting because, you know, we're all matching the clients' needs because it's not just search that or that they want they want clients want to know all the different sort of like channels that they could switch on and experiment with. So I think that's where we come to the fore on that. So I think you know we'll we'll kind of wrap wrap up there. there's lots more that we that all publishers might want to consider. before they turn off the tap in Google Search console. Hopefully we've gone through some of those points there. I think we would you know obviously, you know, encourage journalists always to invest in original journalism to really challenge themselves to really look at the brand messaging to make sure that it's really, really clear and you know, test themselves against it because You know, it's such a competitive landscape up there. There's so many AI startups. Google is obviously responding to that as fast as it can, but it's a bit whack a mole. so you know, it's our roles really to advise what to prioritize and where to best spend resource. do you have anything to add to that, Barry?
speaker-1: No, except to say that bad data is worse than no data. And don't make important business decisions on bad data. get multiple sources, multiple perspectives. And properly understand your own business. I'm still surprised, shocked sometimes at how badly understood publishers understand their own business. and and not really sure what what their business model is. and how often commercial people are left making important decisions Which generally ends up in bad decisions 'cause commercial people have quarterly targets to hit and don't really care about the long term viability of a of a publishing business. So don't fall into that trap, people. build on longevity, build on long term value, build on reader value and the commercials will follow. And there's been so many examples of recent times of websites who have succeeded by doing exactly that.
speaker-0: And invest in good beer as well.
speaker-1: Absolutely. Probably the most important thing.
speaker-0: Okay, well, please reach out if you've got any questions for me and Barry and we'll feature those in further pods. But for now, is it goodbye from me?
speaker-1: And goodbye for me.
speaker-1: And I'm Barry.
speaker-0: And we're gonna be talking about some of the most recent biggest stories affecting search at the moment, and also we're gonna be talking about that fancy new functionality in Google Search Console where you can remove your content if that's what you want to do from Google's generative AI search features. But as we always do. First the beers. Barry, what have you got?
speaker-1: I went back to the very first beer actually and it's primarily because I've been so busy I didn't actually have time to buy anything new or different and I've just opened it and it's decided to spill all over me. So that's a great start to this pot. Great start to this pot. But yeah, it's it's Rockshore Irish lager. It's lovely to drink. Very easy, very clean, made by the same people who make Guinness, so you sold can count on them. once the foam dies I'll actually take a sip of that.
speaker-0: Nice. Well, I I am on Fierce for Good or Fierce Breweries. This IPA says this IPA plants trees. And I got it just because I don't know any I don't know Fierce actually, so I thought I'd give it a crack and it's is a little bit of a heavy, it's a six well six point two hazy IPA. one of those kind of like fruity Fruity IPA, so you've really spilt it over there, yeah.
speaker-1: it's all over my desk. My entire office is going to be stinking of beer for the next week or so, so
speaker-0: That'll be fun. worry, mate. We've all been there. okay, so I thought we'll take there's i you know, sometimes between podcasts it's been it gets a bit quiet on the Google landscape. Obviously AI search is rolling every single second. but there's been a few sort of like decent updates that's got everyone talking at the moment. and so I'm gonna just I chose one because I thought it was an interesting one, which is There's been a piece of research by D Demand Sphere looking at AI overviews invading branded search. and I'll just go through some of the what they said. So they they said that their data showed that AI overviews rose from roughly twenty twenty-six percent of branded queries from the start of September to more than eighty percent by the end of the month. And it's an interesting one. It's caught up caught a bit of attention because I guess with the land of or the the area of Zero Click, people have looked at branded searches as a way to say, right, we might have seen a decline in in traffic, but actually we are maintaining and growing our core audience, th which you can evidence by using branded searches. And So this one instantly got a lot of viral coverage because it was a a panic because it was like please no, please don't affect the branded searches. so look, I'll I'll give you my view on it is obviously when I saw that I immediately started to check the branded searches for my clients and I haven't seen the drop off that was indicated here. now that's so I think my my view on reports like this is absolutely it's an interest it's a very very much an interesting read. but always go and check your own brands, go and check out your own data reports because you know everybody's industry is different. What do what do you think, Barry?
speaker-1: I have been able to replicate this to some extent. I've done a few test searches myself on branded queries for some of my clients and and other relatively big brands. About half of them do not show an AI overview. The other half, when there is an AI overview, it's below the organic result. with some exceptions. If the brand is particularly big, the AI overview might be the first result. And I see why Google is doing this. Because when a lot of times when people do a branded search, they don't necessarily want to go to the brand websites, they just want to learn about the brand. And the A overview is basically an accelerated version of the knowledge panel we already had for branded searches on the right hand side or desktop, or prominently on mobile, where it basically gives you detailed information about the particular brand you're searching for. Plus it has an added benefit for Google is that brands who really want to claim that branded search traffic now have an extra incentive to advertise on Google for their own brand name. So it's a it's a revenue source for Google as well. And let's be honest, you know, they still need to to make money. So you know being worth what is it, two trillion dollars is obviously not enough for Google. so it that way it's another revenue scene for Google. I suspect this is going to accelerate and enhance even more. this is just the first forays into AO Visual branded searches. I expect it to become more of a norm the moment Google understands a brand well enough that it can show an AI summary, it'll show an AI summary. and that also means that as a website, this is probably a surface you can optimize for to an extent with your about pages, with your mentioned content on other websites about your brand. and that actually makes brand reputation even more important. because without a solid brand reputation you run the risk of negative comments being summarized in the AI summaries that Google will show for your brand search. So yeah, like you said, it's a space you need to keep an eye on to see how it encroaches. You probably want to prepare for it sooner rather than later and make sure that when the AI summaries do come for your brand searches, that you you're already there. You already have I wouldn't necessarily say claimed an AI overview, because I don't think you can claim it, but at least you you're feeding the right information into the AI overview so that you have some measure of control over what The AS only says about you and about your business.
speaker-0: Yeah, I think that's absolutely spot on. I've I I I wrote a piece a couple of weeks ago, investigating how news publishers are how can they how can they give some kind of a control to how their brand appears in in various AI search engines. And I looked with focus at the about page, because I'm a nerd like that. And I I definitely saw absolutely the about page is important. But if we look at those things that might be important to your brand, like is it a trusted well, is it a trusted media organization? what is it value for money? you know, what what kind of impact is it making? there'd be variations on that if you're actually selling a a a a product like a car for instance. But yeah, my my current results of it was really to kind of don't just don't just update your about page. It's about just looking f and establishing an evidence architecture, as I've said, because you're gonna have to look at multiple places across your website where you might be talking about your brand. So the about page, absolutely, but definitely the subscription pages or and other pages which I've written about, which have appeared in AI search. one of the things that I saw which was quite interesting is that we've we've kind of become accustomed to thinking about the check a checkout process to about keeping it super basic and clean for that to not get not get in the way of the audience. But I've seen the kind of page just before checkout. a feature a lot in AI search and that's seems to be an area where you could put a couple of lines of information about your brand. Anyway, I think with all these things there's so many fluctuations in in AI search. It's really about make making sure that, you know, you're just reinforcing your your brand messaging, but also, you know, don't piss off your users. Basically, don't do anything wild, because you've read a LinkedIn po post, right? We're all seeing disasters from taking just AI search tactics with no thought of other con main considerations you might need to take care of. Okay, interesting. So moving on to the next one. So this is obviously a biggie with Google saying and r adjusting its documentation to indicate that you should be doing a human review of your AI content. So Google says it's updated its guidance on generative AI content, bringing its public documentation into line with material from the search quality rate guidelines and its developer presentation. So a couple things from this is that Google now explicitly describes four qualities it wants to see in the main content of a page effort, originality, talent and accuracy. It even gives mass generating text of AI without human oversight or creation as an example of low effort. So it's using this phrase of effort. so what do you think of all this then, Barry? I mean it's obviously very key when Google is starts to our update the documentation, right?
speaker-1: Yeah, I mean that's a clear sign that Google takes this a lot more seriously. And I think also clarifies what Google has been saying in more indirect ways for for many months now. that that AI generated contact content in and of itself isn't a problem, but if it's low effort content, it it becomes a problem. So you can use AI to to aid in the creation of your content. But if your content itself doesn't actually do anything particularly interesting or worthwhile It doesn't demonstrate any of those aspects that Google wants you to demonstrate. Like, you know, you put proper effort into it, which is often why AI actually hinders you because it's a obviously a show of low effort, because it makes writing too easy. If it's not original, and again, AI is a hindrance because by definition, AI cannot be original, because it's just rehashing what is found on the web in in different worms. It's a it's a word predictor. Talent skill, same there, and accuracy. Well, accuracy is probably the part where AI does give you a bit of an advantage because it tends to Be very reasonably accurate in in the information it presents to you. The hallucination problem hasn't gone but is more or less solved, fortunately, thanks to BetrieVlogment degeneration. So it's become less of a problem. but I think again it shows something that that both you and I, Steve, have been saying for well, coming on to three years now, since AI ever started making it to the mainstream, is that you you need to actually add something of value to a particular topic if you have any hope of driving visibility and traffic to to that content. You know, you need to look at your originality. You need to look at are you adding information to it? Are you contributing something valuable? Are you giving it an additional perspective? If you are just rehashing what everybody else is writing about a topic, you should not expect that content to be rewarded in any way, shape, or form. And this predates AI significantly already. It's just with AI, Google has gotten better at this. Google has always said, you know, you need proper original high value content. And if you're especially as a news publisher, if you're just reporting what everybody else is reporting and you don't really do anything original or new or different, then yeah, you're you're probably not gonna succeed in driving a lot of traffic to your content. And with AI, because AI has enabled the scale of content generation to be orders of magnitude larger, which become a bigger problem. And I think that's really driven Google's efforts into weeding out the low effort content even more because writing as a pure task is no problem of the simplest things you can do. But writing quality content is something that, you know, Google still wants to reward. So you need to show that in your content. You need to make sure that you demonstrate your your effort, your expertise, your skill level, your originality, all of that. And that your information of course is accurate and and deserves to be shown in in search results. And that you know you can translate that to journalistic processes as well and and where journalism Can use AI and should not use AI.
speaker-0: Yeah, it's I I feel almost feel most sorry for Google in this area because it's trying to indicate that you're you know you need to that your your content needs to be original. and so it starts to introduce language like effort and I feel sometimes people respond to that by trying to work out what effort means. Well really the overall point is Deliver original content, deliver something. I mean, I've I I mean, you have to be careful about using the word content, right? But deliver brilliant written work that delights your audience, that they ultimately, when they finish reading it, they'll go, Bloody hell, that was amazing. And they'll come back to your brand, right? These are like basic things, but I think sometimes people get too fixated on the the words that are changing as opposed to the main point.
speaker-1: Mediocrity is just not acceptable anymore if you want to be taken seriously as a publisher. And I think there's been too many websites out there who've coasted on mediocrity and have relied on tricks and and tactics to get that mediocrity to be surfaced in Google News and Discovery in Top Stories. That just doesn't work anymore. Mediocrity is is not even table stakes. Mediocrity makes you just part of the masses. Table stakes is excellence. Just just the opportunity to get traffic means you have to be excellent to start with. And then layer on top of that. If you're not excellent, and you have to be honest with yourself here, if you're not meeting that minimum threshold of this is shit hot content, do not expect it to be rewarded by the machine systems in any way, shape, or form. You excellence is table stakes. Excellence is just sitting at the table before you even do anything. And then hopefully you can build on that.
speaker-0: That's I tell my kids every day. Okay, so on that, and well connected, is really this whole question now, right? Are we about to have a massive Google Core update based on this change in documentation and the steps that have been taken around it? And obviously I'm I'm pointing here to an article which Lily Ray published a couple of days now, where in her Substack she is pointing to a number of different signs that are indicating to her and her team that that something may be happening on on horizon. One, if I go through some of the points, we've got those changes in in the guidance. Google have been talking and presenting in Barcelona and other areas, talking about this about these about hu you know you have to be review your AI content. Google now calls fake authors a form of of decep deception. That's something that it's called it's come out quite blatantly on. Google's added new language to its helpful content guidelines, which was about a month ago. And Google has released four spam updates. And it's very interesting about these spam updates, right? Because previously, I think last year there was one, as Liddy points out in 2025. Now there's been four spam updates this year. So And it's not just Lily, it's other people like Glenn Cabe and Cyrus Shepherd, Marie Haynes, et cetera. So lots of different people, lots of experts. what's your take on it, Barry?
speaker-1: The cadence sort of fits, yeah. The signs are all there. I'm just not entirely sure if this is because of the spam updates or if Google is really setting the stage for a major core update. Having said that, end of October, start of November and or start of December tend to be times that Google wills out core optics. the timing sort of fits as well. Plus, as a general rule in my SEO career, I never go against what Lily Ray says because there's a ninety nine point nine percent chance. She's right. I haven't really caught her in any mistakes that I can remember. So yeah, definitely something to keep an eye on. we've been reasonably fortunate as the news industry with the last few core updates that there has been some impact but not dramatic impacts. It's more like a little bit of reshuffling here and there. It's not like we've seen entire sites wiped out, with some exceptions. And those sites that have been wiped out deserved it, let's be honest. so hopefully if there is a core update coming and as of the recording of this podcast on October seventh, we haven't yet see the core update roll out and the spam update is still rolling out. So hopefully we'll we'll we'll get clarity on that soon by, you know, the core update starting, the spam update ending, or you know, Google might overlap the updates as they have done before in the past. Yeah. And yeah, hopefully it's not one that's gonna hit the news and publishing industry particularly heavily. I honestly feel In many ways, if you just look at search specifically, that news in terms of how Google has approached news websites has been fairly stable. That Google hasn't really felt the need to interfere too much in there as they have done with previous core updates. I think Google is fairly comfort comfortable with how news is currently being ranked and shown and and which websites show up and which websites don't. In Discover it's a very different picture. I think in Discover it's still a bit of a free for all. Google really hasn't nailed down the span problem in Discover yet. I think in search they've more or less nailed it. And I don't expect the next core update to to cause a massive shake up of visibility in search results for most news focused websites. I might be wrong on that, but that would be my prediction.
speaker-0: well we'll we'll have to wait and see, right? I think we'll all be eagle eyed on the various sensors that are out there. speaking of like different various simple like changes, have you seen top stories in AI mode or AI overviews yet?
speaker-1: I have not yet seen them myself, no. When I see a top stories box, it's either the main box on the search results page or a supplementary box. I sometimes see an AI overview and a top stories box, but never integrated into one. But there's an AI overview that summarizes it, usually below a top stories box that actually covers the later stories. I have seen screenshots of top stories carousels in AI answers, but I haven't spotted them in the wild yet myself.
speaker-0: And do you think obviously we're being pushed more into AI mode on on Google's on the call it page one if you like, whatever page one means now. Do you think we'll eventually be pushed into AI mode? AI mode will be the default experience or it will be a combination of different elements?
speaker-1: I've said this several years ago and I still stand by it. I don't think AI mode will become a default search experience. I think Google will always need to give at least the illusion of choice in the form of the blue links. I think if Google steps away from that, they are at a very strong risk of undermining what made them big in the first place. And I think they realize that to an extent. Yes, they want people to go to AI mode, but they want that to be a conscious choice that they make, that a user specifically makes, to go into AI mode and ask questions and interact with Google in a different way. I do still think that. they will keep the relatively classic search experience that they have, which yes, includes AI summaries and AI overviews, but also includes the tambour links, includes the top stories boxes, video results, all those sorts of things. Because they realize people trust those results more. The illusion of choice is a real thing, a psychological phenomenon where if you give people the at least the the illusion that there's a choice, that they can source different content from different publishers, that they trust the results more, especially Because people trust the search results as much as they trust the brands that are being mentioned in the search results. If you have a search results page that shows New York Times and The Guardian and The Independent and a few other high value news publishers, people put more faith in that result than if you have a a search result which is full of Joe Bloggs' home website that nobody knows about and never heard of. So I think there's a strong incentive for Google to keep a form of a classic search result page. with which contains, you know, s different elements that people can click on and not go full into AI mode. But yeah, it's obvious Google wants to push more people into AI mode and the data backs it up. More people are using AI mode. But that tends to be complementary to people also using classic search. It's not that AI mode replaces search. It's I think what Google wants to do is position AI mode as a Chat GPT alternative in addition to its classic search results. I think that's what the overall game is that Google wants AI mode to become a free version of ChatGPT where people can do those sorts of basic research that they use ChatGPT very heavily for, but also still have classic search, to what extent we can call it classic search, to deliver that initial search experience that people have gotten so used to.
speaker-0: Hmm. because I I I only mention it because it relates basically to like what one of the well, the main topic that we're talking about today, because you know, we're very keen on knowing what where Google is moving. Now, particularly from a publisher point of view, right? Because now we know and and that now there's a funky new bit of functionality in Google Search Console where you can block Google from showing their content in the generative AI. experiences, right? So AI overviews, AI mode, and also in the AI generated areas in Discover. So the question is, who's going to do this? Because I think, you know, with the with the we I think from the various different sort of like newsrooms that we speak to, Barry, we know that there's some form of experimentation happening. But it's it feels like it's limited at the moment because it feels like it's limited at the moment because it's only at sight level, right?
speaker-1: Yeah, it's an all or nothing kind of thing at the moment. so you have your either your entire site included as a citable source in AI overviews and AI mode or nothing. I d have not yet seen examples of publishers who've opted to opt out of AI overview in AI mode. And I think the business case for that is very clear. Whether you opt out or not, the AI overview is is gonna sit there. It's gonna sit there, it's gonna be there. And yes, we know when there's an A overview, you get significantly less traffic, but you still get traffic. You still get clicks. And I think there is still some value in being a cited source in AI overview unless you are a very hard paywall website and you think, you know what, any engagement with the content, you need to come to a website and pay for it. Which can work if you're a very powerful brand that people will actively seek out. Not a lot of publishers are have that luxurious position. So I think for publishers there is still a case to be made to say, you what, let's stay included in AI overviews and AI mode is a cited answer, because, you know, some traffic is better than no traffic at all. Because whether or not you're included, the A overview is still gonna sit there. And if they cite your competitors overview, you know, they might get that little slice of traffic versus others. It does point out the p systemic issues, of course, that AI has has created for the publishing industry in that they take all our content, summarize it, and then don't send traffic back. and there's just a a deeper issue there that has been playing for years that hasn't been solved yet. Google has started to show signs they might be willing to solve that. by making noises about compensating publishers if they are a a substantial part of an AI overview. That's a beta project they're running at the moment. So we'll see how that pans out. But I wouldn't expect that to become a major revenue source for any publisher in the future. the the opt out mechanism exists at the moment because again, Google is legally forced to do that. I think they don't want to give e publishers that, but competition markets authority in the UK have said now you need to allow Publishers to opt out of AI while still being included in search. So this is what Google has built in Search Console, rather than like a robot.txt rule or some something else. And it's just one of many things Google is now trying to provide publishers with slightly more granular controls for those that want to make use of that. But it's probably not granular enough because it is site level, it's not page level. I think if it were page level, I think with like a meta tag or something, I think more publishers would do it for specific pages. For example, paywalled articles versus free articles.
speaker-0: Yeah. I think I think the the page level may be coming around March time, although that's always subject to change. And yeah, so at the moment it's it feels more like a a reaction to regulators to offer something. So it's like a false acquiesce in in that it appears to offer publishers more control, but I I'd really genuinely haven't seen evidence of anyone that could use it. because the threat of competition, right? They're they're gonna if you've if remove yourself from the space, someone will fill it. but I guess what we could discuss is really, you know, if publishers were considering to this kind of like talk in advance of let's say March is the time when they get the page level and publishers may be considering removing substantial parts of their content, what kind of considerations they might want to put in place before before making that business decision, right? So I thought we could t talk about a number of different points on that. So obviously I c I think like with any kind of like business decision there, it's really about creating a baseline and working out where where is your visibility to date, right? So it's definitely trying to work out do the publish does the publisher or the brand really understand what they are currently allowing Google and AI platforms to use, right? So you mentioned it before, but you know, doing things like checking your robots TXT file, what information in i i i is in there, many companies may be unaware of what that file actually does. obviously you can put in your own hosted solution, there'd be controls at that level, Cloudflare offer very specific controls in the area. But You know, it's looking at things like no snippet, what instruction instructions you'll send into Google extended. You've obviously got your paywalled content as well. And I know that you've written extensively about things that you have to put in place, Barry, about paywalled content. but yeah, it's definitely around working out what is your visibility and and that's almost like that's always worth doing a stock check, right, Barry, in that kind of regard?
speaker-1: Absolutely. And also understand that the AI chatbots use different mechanisms for different use cases. And so they have different bots for training data versus retrieve logmatic generation versus user initiated actions. And the user initiated actions are very interesting because they ignore robots.txt rules. because they're basically mimicking a browser session. So you can use ChatGPT to go and fetch a web page and then summarize it. And it will not obey robots.txt allow rules because far as chat GPT is concerned, that is a user mimicking a browser session. And the page, regardless of robots.txt, will always be browser accessible. So again, you know, there's different context for different blocking rules. For robots.txt versus CDN level blocking versus metadata on the actual web pages. And there's different mechanisms for different use cases. And you like you said, you start with with understanding what the bots are currently allowed to do in which context and then you have to make business decisions about each of those contexts and decide do we want this to happen or do we not want this to happen. And and that is a tough challenge as well, because there's often not a very clear value exchange for allowing AI access or disallowing access. Because we just don't understand the landscape yet. Yes, the traffic implications are pretty well understood by now, but AI isn't just a traffic generation tool. It's not like a search portal. or a social portal. It's of course a very different context in which users use AI and for what they use it. And what kind of value is there in being a cited source even if you do not get the click? What kind of brand power does that generate for you? What kind of recognition? What kind of complex funnel is that a part of that might eventually lead to a subscription or or whatever sort of conversion you associate with that? And that's the tough challenge and that's something that, you know you can't just pull out of a dashboard and and and slap onto a report. And that's more about understanding where you fit in in user journeys, what the value is of your output in a broader context and how you can monetize that. And yeah, I don't think anybody's really cracked that particular nut just yet.
speaker-0: No, I mean it's it's obviously harder for news publishers where their content or their it is the pages, right, for some of the brands that outside of news that I'm advising, you know, it's more a port important, so brands that might have multiple products. It's it's more important about the relationship between the parent and the products, you know, Do are the products competing with each other as opposed to complementing each other? So it's more around the misinformation or sending clearer signals to AI so that it can g so it can have a clearer idea about the clump the complex nature of your of your business. But you know, ultimately a lot of those brands are actually selling a much more expensive, either a physical product or maybe there might be selling information which is locked down in a tool which is not open to crawlers. so premium data offerings, for instance. So but it's different it's a different ecosystem than than journalism. yeah, I I guess like we I mean I mean I a absolutely this whole process of should we remove our content from AI overviews? Really, it it it does boil down to well, have you done an audit on how original our content is in the in the first place, right? Because it's not worth I don't I don't think it brands should kind of investigate this if their content is already what they call the ugly word commodity content, right? Because What's the point in in doing all that if the bot's gonna go elsewhere anyway?
speaker-1: Yeah, I mean, i if as a publisher you're still relying on commodity content and I'm sort of starting to get a bit tired of that phrase myself, to be honest. But why are you in business in the first place? You know? you have to have something unique. And I feel like a a broken record. I say this almost every podcast, don't I, Stephen? You and I are so sick of talking about it. But if you're not doing anything different and unique, why are you in business in the first place as a publisher? and I think some publishers overestimate their own uniqueness to an extent. They're not really aware of of what their added value is. or aren't honest with themselves about it. but others are very clear about what their brand is. And it doesn't necessarily mean that you have to report something different than everybody else, but it can mean that you report the same news with a different angle. The different slant, a different perspective. And that too is valuable. And this is what you see with a lot of the UK press where you have the same news being reported on the Guardian versus the telegraph, just they have different slants and different angles on them. And they cater to different audiences. And that's fine, that's great. That's how you build a brand and that's why both the Guardian and the telegraph were reasonably successful with the Their their own respective forms of subscription models, because they know their audience and know what the audience wants, and they feed that into that audience. And yeah, I I work with too many publishers, especially in specific niches, for example, Formula One or Football News, where the site just doesn't really take that unique box where the site feels that they do something different, but when you actually start reading the content, it it's just the same reports that everybody else is reporting. and that the the site sort of overestimates its own importance and value. yeah, but we have paddock access to Formula One. Yeah, everybody in Formula One has paddock access if you write about it. That's not new. you know, w w what can you do that is new, that is different? And and that is challenging for some websites. it's not easy to distill that and find out what makes you a a unique proposition. And sometimes you need to go outside of your own constraints as a written word news publisher and try it something different. Okay, you might not be able to write unique articles about what's happening in Formula One, but you can start a podcast, like we are, or you can do a video series or or you can do something else that that makes you stand out from the masses of of news reporters in your p particular niche. That just adds that layer of differentiation to it. and, you know, you have to really put effort into that now and you know, the journalism itself is is just one part of the the overall picture that you're presenting to users and in the end you're still, you know, wanting people to read and engage with your content. And that means you have to have something that that is worth engaging with.
speaker-0: As we keep saying, yes. I guess the the other step absolutely when considering pulling your brand out of these generative AI experiences in Google is just working out how dependent your brand is on Google anyway, right? Because I'm just thinking about a a scenario where let's say somebody outside of an audience specialist team might say on principle We should definitely opt out of Google's generative AI experiences for these for these reasons, without actually knowing how dependent that your brand is on there. So that's definitely worth sort of like measuring where you are in that space before you pull the plug.
speaker-1: Yeah. I'd also say that you have to be careful in how you measure that and what your analytics tool tells you. I've had discussions with clients recently and also saw a LinkedIn post, based on chart beat data. Now I have a big beef with chart beat data, because chart beat data has a traffic source they call internal. It means every internal click is is listed as an internal traffic source and gets a separate graph in that bar. In that in that report. And I've had discussions with clients where they look at the chartbeat report and say, Google has only five percent of our total traffic, and therefore we're not the dependent, we can just kill off Google entirely. And that is such a narrow and naive view because it means you don't understand what that data is telling you. Give you an example. You get a user who clicks from Google result to your website, that is one page view associated with Google. Then that same user from that initial page view clicks on three more articles. For most web analytics programs, that would mean four page views associated with Google. But not for Chartbeat. For Chartbeat, that is one page view associated with Google and three associated with internal, because they're internal clicks, not Google clicks. Which means if you kill Google off, you don't lose the one page view, you lose all four page views. Which is what's happened with some of my publishers who didn't take Google or SEO seriously because it's only five percent, and then they end losing 20, 30, 40% of that traffic because they just didn't understand what the data was actually telling them. Yeah. Which is one of the reasons I don't like Chartbeat as a as a tool that the default reports has this internal traffic as the biggest source, but because it strips attribution from that internal traffic entirely. And that is a very, very dangerous place to get in because internal traffic isn't just people clicking. They arrive to your website through a mechanism and Chartbeat doesn't tell you what that mechanism is. It's just an internal click. That internal click i comes from a source. And that source is usually Google. So you have to be very, very careful with that data and don't make rash decisions on a you have understood graph that somebody shows you. You have to actually understand what the data means, what the implications are, before you decide to opt out of Google entirely.
speaker-0: Yeah, I mean I mean that's just around I mean that's a that's a absolutely key case study there. But yes, it's it's a reminder because we are definitely in the land of multiple pretty dashboards at the moment, right? And some of them are very expensive and it they seem to promise the world. So particularly in our roles as consultants, Barry, it's Quite often it's our role to really get to grips with the data because sometimes internal teams don't have the time to go through the ins and outs of of these tools. So if they were to hire us, we can help in that in that area, right?
speaker-1: If I get a penny for every time I get approached on LinkedIn by another prompt tracking tool wanting me to test them out, I would retire because I wouldn't need to work anymore. It's just ridiculous. I mean, stop people reinventing the wheel. prompt tracking has its own issues and I'm not a particularly huge fan of it unless you do it in a meaningful way based on actual user prompts rather than stuff you pull out of your backside yourself. so yeah, don't buy into the latest hype without understanding what it actually means and what the value of it might be to your business.
speaker-0: Yes, absolutely. I advise clients. I think the worst there's nothing there's nothing wrong, I think, with using synthetic prompts if you want a quick snapshot of something. So you might want a quick snapshot of a of how a product sits alongside in jet very broad terms, the landscape or of where it sits in alongside its competitors. But if you want to then apply resource and get internal teams focused on it. Absolutely get serious about it. I mean I I I use I use open source prompt tracking using methods where is extremely cheap for me. but that's only when I want to do some very brush broad brush strokes preliminary analysis. But as soon as it gets serious then absolutely like you say use proper tools, but also you know, get have a governance model around prompt tracking. try to use real questions that are gathered from what your customers are saying and there'd be multiple inputs for that. But yes, this is that is a separate podcast, I think, Barry. but yeah I guess another thing with the this ability of being able to switch off Google gener generative AI features. you know really yes you could do that but it's just a reminder as you say unique quality content but it's also absolutely convert trying to convert your user into you know try try to forge that direct relationship right so for instance okay you could take that app action but elsewhere for instance what are you doing by way of news l newsletters you know I have recently sort of transferred clients from one email database or service over to another and that's interesting in itself to look at w who are actual users and well cleaning up the the the the email database basically and working out who who are the real real engaged audience here and how can we not only appeal to that core audience but also build the awareness of that newsletter. So I think this time's like really, really interesting because, you know, we're all matching the clients' needs because it's not just search that or that they want they want clients want to know all the different sort of like channels that they could switch on and experiment with. So I think that's where we come to the fore on that. So I think you know we'll we'll kind of wrap wrap up there. there's lots more that we that all publishers might want to consider. before they turn off the tap in Google Search console. Hopefully we've gone through some of those points there. I think we would you know obviously, you know, encourage journalists always to invest in original journalism to really challenge themselves to really look at the brand messaging to make sure that it's really, really clear and you know, test themselves against it because You know, it's such a competitive landscape up there. There's so many AI startups. Google is obviously responding to that as fast as it can, but it's a bit whack a mole. so you know, it's our roles really to advise what to prioritize and where to best spend resource. do you have anything to add to that, Barry?
speaker-1: No, except to say that bad data is worse than no data. And don't make important business decisions on bad data. get multiple sources, multiple perspectives. And properly understand your own business. I'm still surprised, shocked sometimes at how badly understood publishers understand their own business. and and not really sure what what their business model is. and how often commercial people are left making important decisions Which generally ends up in bad decisions 'cause commercial people have quarterly targets to hit and don't really care about the long term viability of a of a publishing business. So don't fall into that trap, people. build on longevity, build on long term value, build on reader value and the commercials will follow. And there's been so many examples of recent times of websites who have succeeded by doing exactly that.
speaker-0: And invest in good beer as well.
speaker-1: Absolutely. Probably the most important thing.
speaker-0: Okay, well, please reach out if you've got any questions for me and Barry and we'll feature those in further pods. But for now, is it goodbye from me?
speaker-1: And goodbye for me.
