Is ChatGPT Killing Google? Optimising for LLMs, AIO and the Truth About AI and SEO in 2025

James Lawrence: Welcome back to the Smarter Marketer Podcast. I'm here today with Joe Alder. Joe, welcome back to the pod.

Joe Alder: Thank you. Good to be here.

James Lawrence:So Joe is Rocket's head of SEO. He has appeared on the pod before. For those who didn't listen to Joe's first episode on the pod, essentially he owns the SEO product at Rocket.

So in charge of the team, the product design, how we adapt the product to the changing landscape, responsible for driving performance for all of our clients and for us. AI as it relates to people using AI to find products and services firmly sits under that umbrella.

Yesterday Joe and I ran a webinar with AdNews Australia. It was one of the more attended webinars that we've run over the years. We had hundreds of marketers from around the country signed up and I think because this is a topic that is of just such interest and it's so topical and fast-moving.

And we did lots of research before the presentation. Joe did a lot of research on independent data using Rocket client data. So I think it'd be good for listeners to the pod to be able to access some of that research and the conversation that we had. So I think, Joe, just in terms of setting the scene on the webinar yesterday, we had a section all around the macro landscape as we're seeing it. Maybe if you wanna talk a little bit about what we're seeing.

Joe Alder: Yeah, sure. So I think there's a lot of information out there about AI at the moment, and it's all very doomy and SEO is dying, , that's definitely not the case, at least not yet.

The impact is surprisingly really, really small. Even if you look at like top line figures, right? I think it's like. Maybe 50% of Americans between age 20 and 45 have used chat, GPT, which is, it's still a huge number, but it's not everyone. There's still a big portion that would be using Google.

So I guess like the biggest top line figure I could provide is the impact we are seeing. So AI compared to Google on average is less than 1%. So that means less than [00:02:00] 1% of people are visiting websites through AI sources versus going direct to Google at the moment.

James Lawrence: Yeah, it's so interesting, and in the session yesterday , I put some slides together to try to paint that picture, and had some, , articles that had been published two and a half years ago. And one of them was the rise of the Bot. Scary AI Chat. BT could eliminate Google within two years. Um, another one from the New York Times talking about this new chat bot called Chat, BT being a code red for Google search business.

And we talked a lot about the tremendous rise , in all lms, but it had been this tremendous rise. It was the fastest platform to kind of reach a hundred million users much faster than TikTok. You've now got, depending on the estimates, 400 million, 500 million, potentially seven or 800 million users worldwide.

Um, two weeks prior to recording this podcast, there was a, an apple, , executive sitting in the witness box in an antitrust case in the state to let's slip the traffic. On the iPhone in the Apple [00:03:00] ecosystem had dropped month on month for the first time ever, which sparked an 8% drop in Google's, , stock market price.

And Google's revenue is definitely down in Australia. But then we put forward that question, Joe, didn't we? Which is, is Google dead? And if you actually zoom out and look at the macro landscape, I think what we're hearing, . On our LinkedIn feeds, on podcasts like this, , industry publications, a lot of the hype is actually not really being born out by what we're seeing.

And, , the big studies being done in the US last year , the takeaway step for me was what, for every 373 searches in Google, there's one search taking place in, um, in chat GBT and it was , almost identically mirrored the data that you pulled from all of our clients over.

Um, Q1 of 2025, and I think the listeners to the pod, you couldn't have two people. Probably more bought into , the idea of AI and what it does and our adoption as an agency and as white collar workers and the future of, so it's not like we're skeptics on ai, but in terms of the impact we're actually seeing on search, it's just not there yet.

Is [00:04:00] it?

Joe Alder: No, no, definitely not. , And also , just to the chat GPT point, I think with a, we've also released data on that, right? And it was like 70% of AI traffic was chat, GPT. So it definitely is still the dominant force. , But I was expecting , the numbers to be a lot higher. I have seen a lot of other higher numbers.

Put out there, we've tested it to multiple sources. So we looked at overall top websites in Australia. We also looked at our own client's data. , We have a very broad range of clients across different niches. Um, we don't have any news or publications, which I think are maybe the websites being most affected.

But across our client set, we are seeing less than one to 2% impact. Like overall, our searches are up across all platforms. Chat, GPT and other AI tools make up around 1% of that. When we looked at the data, when we were looking at top websites in Australia that had seen a full, there were some websites that looked like they have been severely impacted by ai, but they're very niche or [00:05:00] specific and very B2B content driven.

So for example, I think Men's Health, they have seen an 80% drop in traffic year on year. And that's people. Now maybe going to ai, it could also be from some other factors, some other websites like Wiki, how, which is obviously no surprise, probably to anyone as well as Stack Overflow. Um, any developers now have the best coding assistant of the hands.

They're not gonna go to Stack Overflow to get those answers. So those are the websites that have seen an impact. And, but yet in terms of kind of broad client, we haven't seen much.

James Lawrence: Yeah, that's right. And , the study. Spark Toro did, which was published in March this year, had traffic to Google up 20% year on year.

We're finding the same thing at Rocket, where if you aggregate all of our clients across a broad range of industries, , traffic is up year on year, across the whole gamut in terms of Google search and Bing search. , So , the noise out there would make you believe that. .

Chacho BT is eating Google , search market for breakfast, but we're actually strangely, [00:06:00] almost really given the noise, still seeing Google grow. And Google responded to , that apple, , legal case to the evidence it was given there, which they don't really ever do. But they responded with a press statement the next day saying that it's just not true.

And the traffic across the Google, , search networks, continues to grow month on month and year on year. , So , if. We hadn't have heard of this thing called chat bt, and if we weren't using it in our, you know, daily lives, I don't think we'd be looking around going.

There's that much different in terms of what's working for our clients, the sales and leads they're generating from Google, from Bing. , With probably the exception of some clients where blog sections might be not getting the traffic that they once did, which I think there is what you said there before about there are some extreme losers from.

The rise of ai and it is, you know, men's health sites, uh, that information's now probably being returned in AI overviews within Google. People are probably turning to Chatha BT to get answers to those types of questions. , And then actual websites that used to perform , an action or , a result that [00:07:00] now people would absolutely use chat GBT for.

Joe Alder: Yeah. Yeah, informational terms make up 90% of what people use, AI for at least in a traditional sense, but also I think important to note, the traditional sense of searches only account for 30% of all usage of LLMs. So what that means is 70% of searches performed on LLMs are for everything else.

Write me an image, proofread this. What do you think of this? Um, so. Even there, it's not a one-to-one comparison of how they work. I think the other thing from that as well is that even though they are kind of fundamentally different, how they scrape the information is very similar. So chat GPT uses a Bing API.

Gemini obviously uses Google Perplexity also uses a form of Bing API. So how they're pulling the information is very similar. I think the only information we wouldn't have, and it would be very hard to get the data on, is how this is affecting consumers before they make a purchase or service and that kind of research piece, [00:08:00] which would've typically been done by either like forums or other websites that rank their content and now it's done by chat GPT, which basically takes all of that information and serves it back to the user to allow them to make a decision.

James Lawrence: Yeah, we definitely don't get the visibility, do we, . In the LLMs in terms of how users are interacting with and what those brand kind of touch points are. , And , we made the point yesterday where . Google is continuing to grow. We're saying this is not , a digital marketing agency, an SEO agency telling you something that isn't true.

Like all the objective studies, all the objective data is still, , showing that Google is the behemoth and if you wanna get traffic to your website, that's the place to get it. But there has been over the last 10 years, , a real fragmentation of digital, , where you do see cohorts of younger users turning to TikTok to search for.

Things that different demographics , will turn to Google to, and , the vast majority of users of chat GBT are also Google users, right? But they are turning to chat GBT , to find the answers to things that they once would've browsed the web for or search Google for.

Joe Alder: Yeah. Yeah. And I also think people are so [00:09:00] embedded in the habits. Even if they are using chat GPT, they're likely to kind of close it, going to Google anyway. Um, I think click through rate it, it's really hard to measure the data. That would be very low on chat, GPT. And even just like the usability at the moment, like it's great for pulling up information.

But if you wanna browse different products, different information, that's pretty limited. They have made some updates in the past few weeks. , Where they're gonna improve the usability, pull in shopping results, et cetera. I've tested that. Still not great.

So I think that's kind of the other thing as well, right? Is we're still very much in this testing phase where it's seen a huge impact in terms of user behavior, but it's still not perfect and not there yet.

James Lawrence: Yeah, that's right. And , having been in this industry for so long, I've seen so many things come and go.

And so we've always been quite cautious about, , the latest shiny toy and gen ai, LLMs chat. JBT is not that , this is real, this is here to stay. , This is one where marketers, I think have, you've gotta have a, you've gotta kind of straddle the divide, [00:10:00] right? Where if you wanna win the quarter, if you wanna win the year, if you wanna win the next two years.

Like SEO Google traditional search is where , you're still getting the absolute vast majority of your traffic from, but , there's just no doubt, , that open ai, chacho, bt perplexity, Google's own AI response, gonna get better and better, smarter. And there are gonna be, , wins to be had by, knowing, , how to optimize within those frameworks.

Right? So I think moving on from that, , um. AI overviews launched, I guess , six months ago or so in Australia? Um, what impact are we seeing?

Joe Alder: So maybe just to explain to anyone who doesn't know, AI overviews is when you search on Google and it pulls back an immediate response, which is generated by ai. So this information is actually mostly pulled from the top Google results, . So it's kind of like top three links and there's an 80% correlation, some of the more deeper links.

Um, and this is changing. Um, in terms of rollout, it's actually been very inconsistent. It's gone up to. [00:11:00] 15% average, then 1% and they've had issues and removed it entirely. I'm starting to see this data settle and now it's about 10% of all searches have an AI overview in Australia. Some countries haven't even, um, received AI overviews at all yet, but there is some industries that are much more affected.

So health I think was like 48% of searches. That's insane. Amazing overview.

James Lawrence: Yeah. , Just to stop you there, something you said yesterday I found really interesting is that it's almost a replacement to the knowledge graph and the featured snippet. Do you know anything more about that in terms of like kind of correlation between the two?

So ,

Joe Alder: when Google was first rolled out, it was actually appearing both, and it still has it, so it's not a direct replacement, at least not yet. And you still will get knowledge graphs, you'll still get AI overviews. Yeah. But it's basically what we call a zero click search. It's where Google answers.

Without needing a click, which is the same as as featured snippets.

James Lawrence: Yeah. So even if we don't have the data to support it, it's kind of similar function. Probably back in the day on health sites, there would've been a high, uh, [00:12:00] percentage of featured snippets appearing and now they happen to be AI views instead.

That'd be fair to say. Um,

Joe Alder: yeah. Yeah. Maybe there'd be niche dependent. Um, . 'cause the level AI overviews can be a lot more specific, where before it would've been taken from one result, this is actually coming from 10 different results. Cases.

James Lawrence: Yeah. Okay. So AI views settled down a little bit.

You have extreme, , swings, right? Health science, people in society, those things where you really turn into Google to get an answer immediately. Very high percentage of I views appearing. Then when you move into more practical things, real estate news, arts and entertainment, it's very low, right?

It's like. 1% sub, 1%. Um, what do listeners have to think about, like if in their category how much of an impact are you seeing , on site traffic? Does that actually matter as long as they're getting referenced in the A view?

, What's your perspective on that?

Joe Alder: So in terms of impact, so if you are position one and there's an AI overview present, you can expect around a 35% [00:13:00] drop, , across like the clients and the data I've looked at, it kind of averages out maybe to eight to 20%. If you're in position one to 10, they gain very dependent on those specific industries.

So , that's kind of the impact you can expect.

James Lawrence: Yeah. Okay. Okay, that makes sense. So in terms of Google and then in terms of, um. Large language models I think it's challenging for marketers who I. I presume are getting a lot of pressure in their business around ai and can't we just AI it and get everyone in, you know, everything, things that used to take people get, do it through ai and why aren't we appearing when I search for best insert, you know, what we do as a product or service in chat or perplexity?

What are those steps to actually get a brand appearing? , And what are the considerations there? Because obviously it's quite different to just ranking first in Google, right? , It's a more elusive kind of target.

Joe Alder: Yeah, so just uh, to round off the AI overviews one as well, like that has a very, very strong overlap between what works well in Google.

It's also got very low click-through rate [00:14:00] and very much mostly on blog traffic. So if you have been impacted there, , there's not much you can do. To be honest, if you appear in an AI overview, it isn't gonna generate more clicks or any SEO benefit. It's more about education there, like educating your stakeholders that this is why we have seen a drop.

Then with LLMs, there is also a very high correlation between traditional SEO. It's definitely less than AI overviews, but from the studies I've seen, it's around three out of 10 of Google. Top 10 will correlate with chat GPTs top 10 and complexities top 10. If you remove some biases, like, uh, Google loves Reddit, for example, and if you remove that.

Chat. GPT really likes big brands as well as like very, very niche and specific. It's probably around a more of a four to five overlap.

James Lawrence: So, Joe, , you've done a research piece on this topic, right? For Rocket. , Part of that was looking at all of our client data, all anonymized, , across a very broad range of industries, , verticals and on a website [00:15:00] purpose to look at what impact.

We've seen in this space. And then you also use EMR data and run , a big analysis of the top, I think it was 50 or a hundred, um, trafficked websites in the country. What are the kind of key takeaways from that data? 'Cause , it's really hard to get genuine research in Australia.

Like we get a lot of stuff out of America than to a lesser extent the uk. But I think it's so good to have Australian data. That's up to date, right? It's all Q1 this year.

Joe Alder: Yeah. Yeah. So I think the first point is the impact is a lot smaller than you would expect. , I looked at all of our clients and not all of them would even do SEO.

, So it's very much an average across all brands, and I think there was 0.6% traffic is coming from ai. I. Versus from traditional Google at the moment. That's probably , the big key takeaway. The biggest impact we saw for any client was around 3%, and they are very B2B driven and very content heavy. , For Rocket specifically, we have seen about, I.

2.2% of our SEO traffic has come from AI tools. And then what's [00:16:00] really interesting as well is if you look at where they're going to, so 52 or 53%, we're going to blog and resource pages. So very much that informational context as the

James Lawrence: first click from. From, uh, ai.

Joe Alder: Yeah. So what LLMs like chat, GPT are actually serving about 50% to blogs.

Yeah. And then 30% was to the homepage. , So it's really more brand driven in terms of people searching for a service. Here is Rocket agency versus here is Rocket Agencies, SEO page. And then service pages had six or 7%.

James Lawrence: Yeah, we'll include a link to , the webinar presentation that we did, just so I think if anyone wants to see the slides, they can, and I think you made a, good point yesterday, which was, , well some of you might be thinking you guys just aren't very good at optimizing for chat, BT and a io , if the numbers are so small and, , I don't wanna sound too self-serving here, but we , we are very prominent.

When people query in chat, BT Perplexity, best digital agency in Sydney. Best digital agency in Australia. Best SEO [00:17:00] company. We have done all the right things in terms of SEO over the years , you know, being featured in legitimate industry publications, pumping out lots of content ourselves, good quality, offsite reviews, et cetera.

And. , We believe in this stuff wholeheartedly, but the idea that only 2% of our traffic comes from AI is testament to that. So it's not like we're laggards , and ranking hopelessly. And I thought what was interesting though was that equals like a nine or a 10% higher conversion rate.

The traffic that came from Chacha BT to the Rocket website and we have landed, , two clients from. From chat BT, right? So it's not like it's, an, , unimportant source for us, but just making the point that compared to, , our other traffic channels , it's kind of a drop in the ocean, isn't it?

Joe Alder: Yeah. Yeah. And it's, it is really cool as well, just having clients land from there. So it shows that there is, is definitely value in it, even if the impact is small.

James Lawrence: Yeah. I think the other thing I liked about your research, hard to get data on what Australian users are doing, and you [00:18:00] put together, , a depiction showing percentage of AI traffic by tool.

So chat GBT was roughly 70%. Of AI traffic to our client websites in the country. Perplexity was 20%, which probably surprised me. I was surprised it was so high. And then Gemini plus Google kind of the other 10%. But you know, clearly plus perplexity equals the absolute vast bulk of traffic coming , from ai.

Joe Alder: Yeah, and Claude didn't have any search function till very recently. 'cause I'd say they're one of , the bigger ones. , So we might see an increase there. I have definitely seen Gemini increasing, although it's like, yeah, I think it was like 6.7% to Gemini versus seven to 2% from chat GPT. Um, so yeah, I'm definitely seeing Google increase a little bit.

James Lawrence: That's really, really interesting. Um, whenever we produce content. So the, you know, the podcast is an example of that. , But if we do webinars, when we send emails out, all those types of things, we try genuinely to present content to listeners and readers that will allow them to do better in their [00:19:00] job, right?

And to get better performance , for their business in the upcoming period. And. In the webinar yesterday, we \ set a lot of the macro landscape stuff. We went through a lot of the data and then we had a section which was kind of Joe, your thinking. But essentially the six key ways to rank in LLMs.

So trying to help. People that are on the webinar and , doing this now for listeners to the podcast, but based on everything your team's doing, what your learning is working and what's not working based on the stuff we're doing for Rocket, what are the six key ways to rank in LLMs? And the first one of those is , almost a funny one.

I don't know how, depends on your sense of humor, I guess, as to, you know, this one. But essentially do good SEO.

Joe Alder: Joe's not laughing.

James Lawrence: Now he's,

Joe Alder: I mean, all of the data really just points to that, right? Like it's a very high correlation. And even if the correlation is in 10 to 10, all of the main ranking factors are no different to what Google has been asking for for the past few years.

There's a few more specific ways so you can [00:20:00] encourage it, , like specific techniques, which I'm sure we'll get into. Overall, like for Rocket to perform? Well, it was us doing good SEO. It wasn't us doing good a EO.

James Lawrence: , It's just so funny, isn't it? , All the studies I read from abroad and people with these kind of big breakout statements about how to, , rank an AI and it's like.

This is all the stuff that we've been doing for the last six years to actually get clients ranking well in Google. It's great quality content, it's well structured, it's technically sound. It's getting the rest of the web genuinely to talk about your brand. It's about getting great content onto third party websites.

, It's reviews, it's all those things. And it's kinda like, yeah, okay. But, , we digress. So, we'll that's the first one I think is probably a bit more a, the theoretical point. But the second one, which is to optimize technical for LM lms.

Joe Alder: So I think like first off, if the technical SEO isn't good, you might not be ranking an LLM.

So that's like really the groundwork. So point number one there is to make sure you are not blocking LLMs. There was a [00:21:00] lot of websites when chat GPT first came out, they didn't want accessing their data. If you block it, it's not gonna appear. There's now also an LLM txt, which is basically like a directive page where you tell LLMs what section of the site they can and can't crawl.

, It's not really being used too much at the moment, but if you're very like, precious about this or you're worried about accessing, it's worth setting that up. Um, one of the biggest things is for JavaScript Heavy site. So if you turn off JavaScript and it can't serve any content, um, that means chat GPT is unable to read the content on your site, so you would need to speak to the developer.

There, there's a few options, scale or around it using things called like pre-render. Um, but yeah, that's one of the, the main technical ones.

James Lawrence: Nice one. , Point number three, content is still king, so obviously the very nature of how LLMs work, , it's very logical. But I think just jumping into some of those more specific takeaways and things that listeners should be doing around content to, increase the chances of ranking really, really well and [00:22:00] appearing prominently in charge of PT and perplexity.

Joe Alder: So first is like kind of the EAT, which again, Google has been saying. So, expertise, experience, authoritative trustworthiness. And what that means is basically any content you produce should seem like it's something that comes from an expert's opinion. If you work in digital marketing like we do, the content you produce should be all around that really act as these kind of content pillars and let chat GPT and other LLMs know that you are really the expert in this space.

I'd say that's point number one that encompasses it all. And then some more like really like, I guess more low hanging fruit of how you can actually encourage ranking. , So there's a few things to like. They check GPT and other LMS use URLs a lot more. So having your keyword in the URL is really important.

, Same as meta descriptions. They sometimes scrape meta descriptions, so if you have any facts or data to actually add them or answers and questions in your, your meta descriptions, obviously you only have a small snippet. And then in terms of like actual content, so [00:23:00] like listicles, so what are the top 10?

Any unique data, . Really performs well on chat GPT and other sources and also just being very, very relevant and very specific. Although Chat, GPT did favor, , bigger brands, they were serving brands as well that had like less than 50 traffic, , just because they were super relevant to what the user was searching.

James Lawrence: Why do you think, , the LMS like listicles and that kind of structured content so much.

Joe Alder: I think it just maybe groups everything in a very easy way. Easy to read. I think it's the same as users, right? Like if you are looking for something, it's much easier to digest something that's top 10 ways to do SEO versus this is how you do SEO.

Like it just really breaks it down, gives it priority and makes it easier to, yeah. Yeah.

James Lawrence: , We had a q and a section at the end , of the session yesterday. And, there was a lot of questions coming in around, can I use AI generated content on my website?

And, Google came out very, very early. Like I think we're talking kind of [00:24:00] February, 2020. Three or four like, but , three or four months after chat, JVT kinda launched , in a popular sense and said, yeah, absolutely. Like we don't care how content's created, as long as it's awesome , for the visitor and for the reader.

So definitely feel free to use AI to help craft content, but then obviously you've gotta make sure that it is. , Unique and interesting and has a perspective and all those things. And I'd still be paying more attention to how users are interacting to that content, bounce rate, exit rate, those types of things, which we know Google takes into account when ranking content, , over probably some of the more AI , tools to see, , how unique it is.

, Would you agree with that, Joe?

Joe Alder: Yeah. I think as well, like with AI content, it's like , it's fine to use, but if it looks ai, I would say that is the point where it's not. Right. So if you can serve AI content that is very similar to human written, there is basically no different. But if you open content, it looks very fluffy.

Yeah. Very ai. A few. Key call outs there that I'm sure we've all seen like the

James Lawrence: um, dash es. [00:25:00]

Joe Alder: Is it if anyone even tries to find that on your keyboard, you can't even do it on the Australian keyboard. I just keep, I

James Lawrence: keep telling chat bt I say like, stop using it. This is not my writing style. It goes, no problem.

I will no longer include it. And then the next time I do it, prompt comes straight back again. Response? Yeah. Point number four, once again, still about content, but more around structuring content. You had five , points around that, so maybe just summarize that and listeners, if they wish, they can get a copy of the deck or a copy of the presentation rather.

Um, yeah, in their own time.

Joe Alder: So I think similar to what I said, just making it very digestible. So having a summary of the top of what the content is about. Having very clear, like heading structure, , especially if it's a longer article, so using H ones as well as like using bullet points and tables, like if there's any data making that really easy to understand.

Then there's a more technical side. , All LMS still used some form of structured data. It's the same as Google, some marking up that this is an offer. These are my reviews, this is my price. And making that [00:26:00] all really easy to understand, um, the content itself.

James Lawrence: That's awesome. This one's really important.

A huge one I think, which is build Brand Authority online. And I, I feel at the moment that, , in some ways there is that kind of overlap. As to where this technology is headed, this is just gonna become more and more important for marketers in terms of.

Getting brands promoted consistently. So point number five was build brand authority online. If we could speak to that for a little while.

Joe Alder: Yeah, so I think like what really encompasses this point really well is imagine you were purchasing a product, maybe like an expensive product, like a car. How would you actually look for what car to buy?

You would look online, you would look at publications related to cars, see how they're talking about these cars. You would look at reviews, you would look at forums, and if you ever do a search in chat GBT, and you can actually see the searches performing, it's very similar to that. It will go and it will search for.

Best cars to buy. What are the best car reviews? How is this car reviewed? Right? So I think that kind of encompasses the whole brand piece. Um, but it is really, really [00:27:00] important for LLMs to see you listed. So if there are any directory or review sites, you wanna make sure you are listed there, if there's any content speaking about, , your product.

And you can get listed there, especially very relevant and recent content. That really helps Also, , link building has been a big part of SEO for the forever. Basically, I think since it first came out. Links don't matter as much on LLMs, but brand mentions do. So. Although people might say link building an important anymore, getting your brand mentioned across the web, which is virtually the same process anyway, that is still the same.

James Lawrence: Massively. And we're putting so much effort into building out that digital PR product at Rocket. Right. And because , it is the future of this stuff, , it's been important in the Google space. , As you said, since the beginning,

google was looking at what the rest of the said about a page. , But more recently kind of the growth of digital PR and things that, how that then merges in with. LLMs and how they work. And then once again, tying into social search , and how , [00:28:00] the landscape has changed. It's just a, it's a space that is gonna continue to grow and continue to become more important for brands.

Right.

Joe Alder: And as well as owning any other entities on the web. So , I think Google is mostly focused on your own website, right? There are also other websites that mention you. So you might have a Wikipedia page or review page, or your social media page, like LLMs. Instead of focusing on just your website, they're looking for this.

They're gonna look at multiple different sources as well as things like Reddit and online communities. Um, so yeah, owning that space and just being engaged with community is really gonna help there.

James Lawrence: Yeah. Nice one. , And then the sixth and last one was monitor and adapt, which is obviously quite broad, but within that there was actually some very practical takeaways.

I think it'd be good if you could talk , to those for listeners. So

Joe Alder: I think like number one, I would segment your traffic coming from AI sources. We have grouped it with SEO as we believe it is virtually an SEO channel, but we have made it so we can actually separate, , organic traditional sources versus ai.

Um, and then you can see what impact is actually having on your business. There's obviously a [00:29:00] lot of changes happening in this space as well, so just making sure you stay on top of that. Making sure the fundamentals are correct and then just preparing for the future, like all of this stuff is makes good SEO anyway, so I would make sure you are investing in SEO and then just yeah, monitoring, see what Google Chat GPT does next and adapt when it comes.

James Lawrence: Yeah. Cool. Nice one. So I guess Joe, I I think we're big believers in how do we help clients?

If I was , in an in-house team, how am I gonna win the quarter? I've got my boss breathing down my neck. We need to hit numbers that, you know, macro economy's not great. How do I win the year? So we try to get marketers probably focusing not on the , short term. 'cause obviously you wanna plan and have a longer term vision, but probably control what you can control.

So I'm gonna throw all of that into the bin. What does it look like , in one year's time and five years time? , What does the space look like?

Joe Alder: , I think in one year's time, I don't think it's gonna be too different.

I think we're gonna see the tools improve slightly. We're gonna see enhanced results. AI mode has just been tested in the US at the moment, and Google has been, when they have rolled out products, there's [00:30:00] been a lot of issues and they've rolled them back. So , I don't think a year, I think maybe if we look five years.

We're definitely gonna see more integration in traditional search engines. So I think AI mode will become more of a thing. Even the search result is probably gonna have more AI features. You can actually go back and forth, answer questions, tailor the search results a bit more to you. , That's kind of where I see it going.

With chat GPT itself, I think it's just gonna try and keep competing with Google and try and take that market share. So it started bringing in shopping Now, , they're also not running any ads at the moment, which is maybe a plus to some users. , So they're just gonna keep adapting, keep improving their maps, listings, et cetera, et cetera.

And

James Lawrence: what happens to search like in five years time? Do you think that when you're looking to find the best digital marketing agency in the country, or you're looking to find the best ERP software for your needs, how are you doing that? I.

Joe Alder: , I think fundamentally it's gonna remain the same, right?

, All of these LLMs use APIs built by search engines. If you look at Google, it's been around for over [00:31:00] 20 years, and it's been developing an algorithm with one goal, which is to serve the user exactly what they're looking for in the best way. And it might not always get it right. , With LLMs it's gonna get right a lot more of the time, so I don't think that is gonna fundamentally change.

I think it'll still be people searching for a product and then getting a result. How tailored and customized that result is to the user. I think it's definitely gonna change, but the market leading products and the brands that lead in the space, I think are gonna remain the ones that are shown the most because they have the most backing online against that single product or service.

James Lawrence: Mm-hmm. Yeah.

However users are searching, whether that's within Google as we know it now, whether that's within chat j BT as we know it now, or whether it's much more deeply integrated into how we use tech in terms of voice search integrated into our Gmail, integrated into our, Word doc processing that we're using, um, these algorithms, they still have to return based on [00:32:00] relevance, right?

So there's always gonna be a job in. Your particular business or your brand relevance compared to your competitor set?

And Rand Fishkin, who Adam on the port a few months ago, was quite dismissive of chat. Bt as a search engine where he was saying , by the end of 2025, if it has 1% market share, it's still smaller than Duck doco. And I'm not exactly worrying too much about duck doco. Um, I do think that might be a little bit harsh potentially where , half of Americans have played with it, you know, that has potentially jumped from 500 million users to close to a billion in the last kind of quarter. It's definitely, more integrated into our lives and at once was I've, the thing that I feel, , often isn't discussed is just how well set up Google are.

To protect themselves where they're a behemoth and it was much more difficult for them to roll this kind of tech out without taking a whole bunch of brand damage and doing a whole bunch of things that would've been scrutinized very heavily. , They've [00:33:00] had to respond because of , the nimble nature of chat, GBT.

But we're talking about a brand that has Google Chrome, that has Android, that has YouTube, that has maps, has. , Many more, , staff members and employees and scientists that work in AI and have been doing it for a lot longer. And , I think people listening will probably go, yeah, but that's kind of bloat.

And maybe , they haven't got the real smarts. But I think what we're gonna find is that the nature of how locked we in we are to that Google ecosystem, they will. Respond and , they will deploy those AI smarts in a better way. And a lot of the top level reading actually says that they actually have a more evolved and better AI product than open AI do.

And it's just that what we're currently seeing is skewing that. And they've still got a lot of time to respond. I know that there's market share being eaten up, but I don't know that the history of all of these products and this space is that , you end up having one, one provider that becomes pretty big.

Right? And the rest . Struggle to get past that point. I don't know. How do you feel about that?

Joe Alder: Yeah, I agree. I think like Google, even in the most, [00:34:00] the latest model, it seems to be being chat gpt in all fronts. , At least like the stats I've looked at. So I think at the moment, chat gpt is better product to use, but Gemini is a much smarter, so once Google fix the usability issues and make it a bit more, maybe even personable, where I think chat GPT does really well, , I think , it's gonna really, , close that gap.

Okay. I think the, the other thing as well, , we spoke about was, um, Microsoft was an internet explorer where , they stopped even processing any of the information and they actually had to kick people off of using it. So I think people are really just creatures of habit and changing that is gonna take a while.

James Lawrence: Very, very true, Joe. We're gonna finish off , the podcast with a lightning round. You have to do your very best , to answer these in one word. I'll give you maybe a couple of them you can answer in a few words. , One word to describe SEO in 2025. Exciting. What's the biggest myth about AI and search?

It's [00:35:00] similar. , The most overhyped AI tool right now.

AI overviews. Maybe , what's better for search chat, BT or Gemini Chat? Bt is bing cool now?

Joe Alder: Nah.

James Lawrence: And , last one. What's one SEO tactic That doesn't work anymore? I.

Joe Alder: , Buying rubbish links a cop out. That's sort a very old tactic.

James Lawrence: Why, why, why text on white background?

Joe Alder: Yeah. Yeah. Well, easy, easy answer there.

James Lawrence: Uh, Joe, always great to talk all things, SEO and Ai o with you. Thanks for coming back onto the pod.

Joe Alder: Thanks James.