Speaker 1chat GPT and the other tools are taking a small amount of traffic so then Google released AI overviews which is the kind of answer snippet at the top of the search result so basically designed to answer the user's question without them going to chat GPT so when an AI overview is present there's a 58% drop in clicks to the position one in Google so if your business previously rang position one in Google now there's an AI overview present you're seeing a reduction in clicks of around 60% welcome back to the
Speaker 2Smarter Marketer podcast I'm here today with a smiling Joe Older welcome back to the pod thanks James good to be here why are you
Speaker 1smiling I'm smiling because I'm very excited to talk to you for a
Speaker 2minute oh very good very good Joe has been on the pod before as mentioned he's head of SEO and AI optimization at Rocket under his leadership of SEO Rocket has picked up some awesome industry recognition best integrated agency at the 2026 APAC search awards best SEO agency at the 2024 APAC search awards and Joe you're very much at the forefront of Rocket's response to how AI is reshaping search what our position is how we're helping clients navigate the changing landscapes you and your team have been doing some awesome research into what's happening in the landscape and what's happening in the landscape which I think is awesome for the industry within Australia and also obviously great for our clients we the reason we're having this conversation today is we held a webinar two days ago it was the second biggest webinar I think we've ever held at Rocket we had over 800 in-house marketers from around Australia register for it and the topic was AI search in Australia in 2026 so on the back of how popular that session was we thought it'd be good to have a conversation on the side we won't we're not going to run it like a webinar I think that'd be kind of pretty dull so we just want to have a conversation around the key themes from the webinar and the session itself was very much broken into two core areas the first was the state of AI search in Australia in 2026 which was on the back of a lot of research that Rocket has been doing as to what's actually happening with Google and chat your BT and all the major platforms and then the second part was to give in-house marketers assistance and help on how to actually rank and drive visibility in the large language models chat your BT AI I've used it's Google and Google's AI mode so I think Joe as a logical starting point I think it'd be good just to kick it off with the state of search report that you and the team have put together if you could just talk a little bit about the methodology and the kind of process you and the team have gone through to try to actually give Australian marketers a better understanding as to what's happening I think and don't want to kind of speak for you but I think often a lot of the research we get comes out of North America to a lesser extent Europe often the data is not quite right but even a lot of the reports that we kind of get access to they tell part of the story or they're vendor funded so they're coming out of platforms that have a pretty heavy bias in one way or another so it'd be good just to if you could speak to the research that you and the
Speaker 1team have been doing yeah sure so the research was basically split into two different sections so the first section what we did is we have access to over 150 Australian businesses and we actually have access to their Google Analytics we can get really accurate data it's also across a very wide range of Industries so once we get an average from that we can really see how the overall metrics are changing so the data we got from that that I thought was really valuable number one specifically on the AI front we were able to see how much AI traffic has grown on average across over 150 Australian websites and with part of that as well we can also see the split between the different tools so there's been lots of like murmurings and chat GPT is no longer the important one now everyone's using core Gemini etc so we have a lot of information on that a lot of the studies that have been done currently use click stream data which basically monitors people's web usage and has works for a bunch of extensions so that misses out on a lot of information when people are using apps on their phone so we definitely do not have the 100 accurate but I think it is a really good look of the data that is available so that was the first side and I will share some data on that I'm sure I'll go for the call the second was looking at the overlap between LLMs and Google so primarily AI overviews and Google and then also chat GPT and Google so there's been a lot of talks again of people saying that the chat GPT search and Google search it's all the same thing it's all the same techniques there's been lots of studies that have been put out there where the overlap has been pretty low so I really want to design a study that looked at all different angles here and also took into account how people search on each platform so how someone searches on Google I think the average is around three to four terms that they will be putting in where chat GPT is six to eight so if we were looking for a digital marketing agency we might just put that into Google for chat it might be what is the best digital marketing in C for a B2B business so there's a lot of kind of work in the backhand just working out what the terms we track on each talk and then from that we basically have all of the overlap stats I think if
Speaker 2you could just speak a little bit to the sheer volume of data I think that there was and obviously the the part of the study that focused just on our clients obviously the weakness of that is it's our clients in this you're investing in an agency the quality of the whether it's good or bad quality work I like to think it's very good quality work they're going to skew that basket but then also I think speaking more to the other kind of data points and volumes of data that we
Speaker 1went through yes we basically did every possible variation so we checked every single AI tool to see if there was correlation between chat GPT Gemini Claude for chat GPT specifically because everyone's search is different and each result returns is slightly different we did anywhere from 10 searches purple to up to 100 searches so across the entire data set it was across I think around 4 500 sub-industries and core Industries in Australia and the total number of data points was over yeah it was I think it was 700 000 just responses alone that we looked at yeah I think it's really
Speaker 2interesting I think it's interesting to be able to compare the changes over time and then to see and there's definitely some real overlap with some of the studies that we have traditionally relied on overseas I think it's a point into the fact that there is a lot of kind of veracity to the data so I think just jumping into the into some of the insights when we framed this the webinar session I spoke to the macro landscape and it is it's legitimately the biggest change in digital in 20 years right the way the behavior is moving and the data that we're looking at has 49 of Australians using large language models on a monthly basis which might come as a surprise to marketers because I think on our LinkedIn site we presume that it's every single person but then it is 78 of knowledge workers right so 78 of knowledge workers are using a large language model on a monthly basis and then when there's some awesome research I think it's the APAC b2b buyer's guide that was done very back end of last year that 93 of b2b buyers are using AI on the path to purchase so this is a you know probably not not saying anything that most listeners to the pod wouldn't know but it is a tool that Australians find the answers to things and it's heavily impacting the buyer journey right yeah so in terms of what we're saying Joe market share between large language
Speaker 1models yeah sure so we have kind of two different metrics from this the first we have from that counter where it looks at all of the different metrics that are available and I looked at a number of third-party tools I also did some of my own research and this was the one that I felt best represented the current market share so chat GPT was around 68 percent market share and that had fallen from this time last year on 85 percent so it has definitely been losing market share to some of the other big players so co-pilot had seen a massive growth Gemini perplexity and cord have all grown massively I think the one that had grown the most was called but is styed as such a small base they're still only around six percent of the market share but that doesn't necessarily mean they're the ones sending traffic to websites so when we looked at our own data set really the main one that is sending website traffic is chat GPT by a huge margin so it's around 90 percent of traffic is being driven by chat GPT I think 88 percent Gemini was around 6.6 percent perplexity three percent Claude hardly even registered was like 1.5 percent so even though the market share is much higher overall the ones actually driving visits to the web is a lot lower from some of these other
Speaker 2tools that is the main rationale for that that we use different tools in different ways where like Claude has become pretty so interesting and I think we're so guilty in the marketing industry of assuming that the average Australian is using technology and tools the way that we do and Claude in the last six months or so has had such a huge ramp up in in the industry but is it fair to say that it's horses for courses and the reason why people might be using Claude might be more around vibe coding and analytics and whatever else versus searching for the best running shoe yeah 100 I think it's a couple of
Speaker 1factors I think the main one is they're definitely different different uses so people might be using chat gpt and then they're also using claude or microsoft co-pilot for doing more stuff at work i think we're also seeing a lot more companies invest in ai this year so we would have seen a big uptick in co-pilot even claude to some extent whether companies getting their employers to use it at work then they still might be using chat gpt from a personal level and then i think the third one is some of these tools are pretty bad as search functionality like even gemini which is obviously google backed google product if you search the products on there it's really hard to even see where the link is if it's going to the right place so the amount of people that are then actually using it for search is pretty low where chat gpt is probably the one doing this the best obviously we have complexity which is a search engine but i've actually seen a bit of a drop when we compare the data year on
Speaker 2year hey it's james here before we continue i have a sincere request if you like the podcast it'd be so appreciated if you could leave a review for us on either apple or spotify it'll mean that the pod ranks better and that more marketers listen to the content that we share thanks so much now back to the pod that's interesting i think the big question for marketers is what's the impact on google and am i getting less traffic to my website is that because people aren't using google anymore is it something different and i think you've done such i think it's a good question i think it's a good question i think this is the area where i find some of the research that rocket's doing to be most interesting so i think it'd be good if you could just talk a little bit about that dynamic because it is something that marketers are getting questioned from non-marketers about i think having some some data to support what's happening is really useful
Speaker 1yeah sure so like in terms of is google dying i think the answer is no like they just released their q1 report and their search and other revenue is up 19 percent just 2026 the first quarter when we look to all of our clients obviously we're doing seo work for some of these but not all of them we saw an increase in impression so that every time someone performs a google search so by running shoes how often is your brand appearing not being clicked we saw an increase around 43 which normally means we're appearing for more but also there are more people searching than ever i've seen other stats the search is on google or up around 20 so very much in line with that so i think it's a good question i think it's a good question i think it's a good with that that being said traffic from google search is falling so although more people are searching than ever businesses are likely to see less traffic actually coming onto their website so there's been quite a few reports on this there was a general one that was done by mart early this year they found that the average website has lost around 35 percent in traffic when we looked at our client pool and we'd obviously seen a 43 percent increase in impressions our positions were really good so we're doing really well i would say we had still seen a 2.5 percent drop in clicks year over year so we are still seeing all in and then when you look at anything publisher related they have definitely seen a huge drop so daily mail australia lost 2.5 million readers year over year they were actually the cmo was writing in some seo forums asking for help and there's like business insider and a few of these other publications have also been really hit as well so we're seeing a 2.5 percent drop in clicks year over year so we're
Speaker 2doing really well yeah it's it was obviously a leading question is google dying and i can't stress enough that we are platform agnostic at rocket right where our kind of north star is how do we essentially how do we put our clients in front of their perfect clients and i was meeting yesterday i was in google australia yesterday and meeting with a pretty senior member of the team and they are super buoyant it's really interesting i think the feeling in that organization is very different to probably what it was 18 months ago when there was a lot of the news media was code red for google etc etc but it feels like the moat around their products and the fact i think they've had a lot of this technology for a very long time but they probably weren't able to roll it out because it is it's it's not a safe right that large language models have a whole bunch of errors and mistakes but once open ai left the lid out it was it allows google to rip in as well but there's no doubt that the growth of ai overviews are appearing on a broader range of terms is having a massive impact on businesses that relied on kind of content and publishers published type kind of content to drive traffic right
Speaker 1yeah 100 and i think on that so where the traffic drop is actually coming from i think chat gpt and other tools are taking a small amount of traffic so you might have previously gone to google to research something now you're going to chat gpt and then as you said google's response to this was how do we get people back onto google so i think that's a really big question and i think that's a really big question and i think that's a really big question and i think that's a really big question so then they released ai overviews which is the kind of answer snippet at the top of the search result so basically designed to answer the user's question without them going to chat gpt but then also as a result means they're not going to websites as much i actually am surprised like i would think google is probably somewhat hurting their traffic and their value by doing that like if they just have to keep up with these large language models and i actually have some stats on that so like when an ai overview is present it's a 58 percent drop in terms of traffic and clicks to the position one in google so if your business previously rang position one in google now there's an ai overview present you're seeing a reduction in clicks of around 60 percent and now ai overviews have also ramped up significantly so we did a big study on this last year now we've updated the metrics when we did it this time last year ai overviews were present in around 10 percent of all searches and now they're sitting around 30 so it's been a huge uplift yeah i think you're
Speaker 2right i think the them rolling it out from first of all rolling it out at all and then taking it from 10 to 30 to me that's not necessarily google taking a short-term view on this i think it's a protective move to keep users within the google search ecosystem when they're trying to find the answers to things that they started to move towards open ai for and it's a complete sample of one that i had my little boy in the lounge the other day and he just turned to me and said i love this the way that google generates these review these are these reviews for me and you can see that it's changing the way that he's engaging with the search talk but it is it has prevented i think a lot of leakage into to chat gbt yeah yeah i'm representing it great really interesting so i guess the next point is then so if google's kind of managing to protect a lot of mindshare if the large language models aren't pushing heaps and heaps of traffic out why does this stuff matter yeah
Speaker 1so i think it will be interesting to actually share the stat because i think if you look in some places like if you look on linkedin it doesn't make it seem like the google is completely dead and chat gpt is driving now all the traffic to websites we looked at this in a number of different ways and when we compared the traffic coming from google and traditional search engines being etc and chat gpt and all of the other ai tools generated about 1% of the traffic to the website and we looked at this in a number of different ways and when we compared the traffic to sites and any of the ai tools and i think then you're saying why does it matter i think that is really the tip of the iceberg the one percent of the top is what we see coming through ga fork we mentioned before that i don't think all of these ai tools are really designed for search but it doesn't mean people are not going to them from search so the average click-through rate if someone is searching on chat gpt for find me the best running shoes how often they actually click on a link from that is only around one percent so it's really low click-through rate coming from the tools the interesting data that we have and also a report from scm rush is i think people compare it to it's either google or chat gpt that you're using but that's really not what we are seeing we are seeing people use basically both tools hand in hand so it's just report the 77 percent of users are using both tools basically together so 68 percent will go to ai first and then they will validate it with google so they're doing the kind of research piece the before the search they're getting the information what the best running shoes for me and then once they've landed on an answer they're then going to google to finalize that pudge so it's having a huge impact on influence even if the traffic is a lot smaller and that's the really
Speaker 2important takeaway which is it's funny i think it's almost backwards where this time last year marketers started to see a drop in traffic to their websites and they were starting they were seeing a lot of content and commentary around the growth in large language models so i think the assumption was the traffic's down because the large language models are taking that it's actually not what's happened the traffic's down because of ai overviews and what google has done itself in the last couple of years is that the rise of the large language models is completely legitimate and it comes down to influence and the issue is that we've always measured digital by traffic traffic coming to my website what's the conversion rate is it good is it growing and that's been the narrative for the last 20 years but it's just not how influence works in the large language models and that's why it's really important to go back to those headline numbers at the beginning which is 49 percent of australians are using them on a monthly basis 68 78 percent sorry of knowledge because you're using them on a monthly basis and when b2b buyers are moving into market 93 of the time they're using a large language model to help them evaluate so it's if you're not in there you're really at the mercy of something that you can't control and i thought it was interesting joe through the webinar but they're not a client of ours but using the example of optus and just looking at a lot of and we'll jump onto measurements i think this is the bit that marketers need to solve but looking at the narrative about optus within the large language models it's so negative right maybe if you could talk a little bit about
Speaker 1that joe yeah sure so we basically did we have a deliverable where we look at how chat gpt and other llms are perceiving brands on chat gpt so if someone is looking for a mobile provider or a network provider then what is the kind of sentiment that's coming back and when we were around optus comparing optus to competitors or even just looking in that space they were mentioned negatively 72 percent of the time so if someone is doing a research on who should i use for my phone plan in 72 percent of the time is basically coming back with do not use optus they have poor customer service they have network reliability there's issues with trust and regulatory so if more and more consumers and i think there's a staff in australia it's like one out of three consumers are using ai in total across all industries to make purchases so one out of three if they're then going to chat gpt looking for a mobile provider optus is losing that market share because how they are being perceived by chat gpt is bad and there's another
Speaker 2brand we're looking at which i'm not going to mention and it was like would you recommend them and it was simply no it's literally that black and white so i think that leads to the next question joe like how do you how do businesses or maybe speak to how we're helping our clients right what are the practical things that we're doing for our clients to help give them insight into performance within chat gpt germany etc yeah
Speaker 1so i think basically everything we do has to start with measurement in these calls it's really challenging to do any type of measurement so we've mentioned a few different metrics here and the reason for that is because we don't have any accurate data coming from all the tools it's a reason we have to do all of these in-depth studies so there is no way to see how many people have found your brand on chat gpt and then gone to google unless they have click through which is only around one percent of the time and there's no native tools provided it's also really hard to know what people are actually searching so that's where we start we basically come out with a modeling on this is how people are searching through seo traditionally look at the search volume and then we come up with based on some data and research that we've done this is how we perceive the user to most likely to be searching on chat gpt there's a few tools available where we can extract that and again on google it's around three to four words chat gpt six to twelve words so yeah i used an example earlier of like digital marketing agency versus what is the best digital marketing agency from b2b so we once we have those terms we then track them basically on a daily basis so we would run these search terms a hundred times per day there's a number of ai tools out there that can do it they are quite expensive and the reason they do it 100 times per day is the response basically changes every single time so it's not the same as google are we ranking in top 10 or are we being mentioned it's out of the hundred times that we perform this search how often is rocket agency appearing within these results and trying to get a market share that's larger than your competitor the top line way you obviously then can have ga4 set up and a few other tools but i would say that's a gold standard of the month
Speaker 2and we'll include a link to the webinar in the show notes for this podcast so you can find that on the rocket website under the podcast drop down because in the webinar joe's shows examples of the types of dashboards that we're building for clients to allow them to better measure what's happening within the large language models and we never overly promote rocket through this podcast we're always trying to genuinely give great information to in-house marketers but if the measurement piece is something you're struggling with with within ai or you're struggling with actual optimization which we call geo ai search optimization feel free to check that out and i'll see you in the next episode bye bye reach out to us james l at rocketagency.com.au because it is it's something i can see marketers struggling with and i think a lot of them either just don't know what's going on which is why the measurement piece is so hard or are struggling to advocate within their organization because the way that we have spoken about digital for the last 20 years is different to what we're seeing here and we've got some really good aggregated data as well that shows the impact of when you're running campaigns within large language models tracking like other metrics around like increases in branded search potentially increases or decreases in direct traffic to try to start to build a case showing what impact large language models are having within your particular category because it's definitely not as easy to track as traditional meta campaigns or google ad campaigns where you can spin it up through analytics it's pretty easy so joe i think it'd be useful now just to talk about actually rank right so how do you improve visibility within large language models both from a prominence viewpoint but also making sure that the narrative being told is what you actually want it to be right in the optus example or in the one when we didn't mention the brand where it's would you recommend this business and it's no what are the types of things you need to do to start i was going to say manipulating that's not what we do we're trying to put the best of the best case forward so is it just simply just do seo or is it different to that
Speaker 1i guess i maybe we start there is it the same as just seo because that is like one of the big things i wanted to get out of the research that we did so i'm looking at ai overviews and i think in like the kind of seo geo space ai overviews and chat gp really merged together where i almost see them as a bit separate because the influence is very different i'd say ai overviews are a lot more informational content you're still in the google platforms and naturally if it's in the platform the algorithm is likely to be very similar so we actually found if you ranked number one on google 60 of the time you will be included in an ai overview which is pretty high like it's not 100 top 10 was around 43 so it's very similar and i almost see it as like an update on like knowledge bra featured snippets where previously you would search for calculator and get instead you're just getting the answer when it comes to chat gpt however it's a very different story this was based on doing eight calls on chat gpt if you just do one the numbers are less when we did eight calls on chat gpt and we compared the brands that came back in chat gpt versus a 10 on google it was around a 15 overlap and there was 80 of the time gpt was using urls that were not even mentioned in google's top 10 at all so the overlap between the tools is very different that's something
Speaker 2which is a more traditional or conservative part of the seo community has basically spun this narrative which is they're the same thing optimizing for google is the same as optimizing for chat gpt and it's just not true like the idea that if you're on the first page of google you're only a 15 chance of appearing for the same term or a similar term within chat gpt that's it's chalk and g's and the idea that 80 of chat gpt citations don't even exist in google's top 10 it just proves and demonstrates that there is such a divergence in the way that those two platforms are synthesizing data so it does warrant first of all a different measurement framework but then it warrants a different approach to the approach you're taking to optimize within google
Speaker 1right yeah yeah a hundred percent and i think there is some foundations we'll maybe get on when there is some seo put in the skill set is maybe similar but how you actually do it the metrics you're looking at where we're reaching out where we're creating content is very
Speaker 2different yeah and i'm not at all because i've been sparring a little bit on linkedin around and i've never suggested that the vast bulk of the tactics are necessarily different there's a lot of overlap with the approaches that you take content and off-site and brand and all that kind of stuff but for me it's the same as you're not going to take this same approach to the facebook because you're going to take the tick tock yes all the levers potentially are the same great content well managed etc etc but it's a horses for courses right it depends on your industry depends on what you're trying to achieve how the two channels interlock it definitely requires a different approach yeah yeah
Speaker 1a hundred percent i think that this example is a great example we would maybe look at reviews a little bit with an seo but we're never looking at the sentiment online that's probably more of a branding task originally and that's starting to now overlap
Speaker 2yeah it's so true isn't it okay can we talk then about ranking what i'd love to frame this in a in the context of what's actually working for our clients we've been running geo campaigns for just on three years now with obviously a massive uplift in the last 12 to 18 months i think you mentioned on the it's our fastest growing service right what and i'm also conscious of not on podcast format not laboring too much around too much specificity on each of these things i think if listeners are interested they can catch that webinar but from a top level i think it'd be good to break it into the core basket so when i start with technical what are the technical things that you're finding rocket clients are doing or not doing that really helps to improve visibility in large language models yeah
Speaker 1I think maybe before we jump in, it would be good to just have a very quick explanation of how I see it working. I really think that ChatGPT is like a user. If you were searching for an expensive product, i.e. a car, what are the types of things that you would look at? You might go, you might look at reviews, you might look at industry publications like what car, the car specific website to make sure it has all of the features that you have. You would compare it to other brands where on Google, it might just be one search, right? It's all of the research that comes behind it. I really see ChatGPT as doing that user part there. So then when we move on to the actual ranking, like first starting with technical, and this is maybe the one that's most black and white, but it has the biggest impact. It's also the most technical, but like to simplify it, you want to make sure that ChatGPT can actually access your website. That is the most basic fundamental path. There's a few things that can happen here. So we've had clients that run cloud there. Yeah. And our flat was automatically blocking all AI bots for a while. So like, basically you want to make sure your AI bots can access your website. Otherwise your content is never going to get indexed. At scale as well, it's just making sure that you, even if AI tools can access the website, there's sometimes security measures in place that will basically limit the amount of time the user can access your website. But these can also then impact these user agents or these LLMs. So we're seeing. And basically get limited from the number of pages they can access. But there's some, a few like technical things like that are really important.
Speaker 2And when you, the, in the webinar, you, you had a case study of a client in the legal space and it was, they've had ridiculous improvements in visibility along with traffic from large language models. And it was some real foundational stuff here, right? And like a really well organized business that does digital actually quite well, but just not actually putting attention into this particular area as it relates. But it was some real foundational stuff here, right? And like a really well organized business that does digital actually quite well, but just not actually putting attention into this particular area as it relates to the website. actually putting attention into this particular area as it relates to chapter bt has had like ridiculous results for them yeah
Speaker 1yeah i think we basically saw a 200 increase in three months and a lot of that were ai tools could hit the website but they were basically getting limited like they were it was basically coming across right from their security and their software that is like an attack on their site but these crawlers hitting the website and i just think a lot of businesses have not looked to that to make sure they are allowing it similar to if you block google you're not going to be indexed in google you need to also consider that for these tools there's also some like software and other things that ai tools are a bit more basic than google that you need to check as well i think to give the users a takeaway on how to do this like the most basic thing is just asking chat gpt hey can you read me the last line of this page on my website and see what it returns it will sometimes return something that is completely wrong and just hold on to it and then it will return something that is completely wrong and just hold on to it and then it will return something that is completely wrong and just hold on to it and then it hallucinates you'll need to check that but that's sometimes a good telltale sign if it is having issues hitting your mind i think
Speaker 2a theme that was coming through yesterday or the day before sorry was that chat gpt is very lazy it in google kind of was that at the beginning but it's become so much more advanced in kind of handling javascript etc but the way chat gpt works it's you have to kind of spoon feed it site speed which has always been a massive signal in traditional seo is a huge issue with chat gpt and large language models
Speaker 1yeah there was actually a study that was done and it was like i think it was like 40 percent of the websites that chat gpt was going to it couldn't actually access but like page speed is one if your page loads slower than four to five seconds then chat gpt will just bounce off and it will go to one of your competitors most likely so it is really important
Speaker 2yeah schema which i think you're saying chat gpt doesn't use but the other large language models do and just best practice anyway for yeah for general seo yeah yeah exactly super important any other plain language hints for listeners as it relates to technical setup before we move on to copy the other
Speaker 1one i'd say i gave the example there of just checking it can read the laugh line of text the other thing i would really recommend especially if you're selling a product or service is if you have a chat gpt account a lot of the other tools now have this but go into agent mode so if you click on agent mode and say hey can you make a purchase can you send an inquiry whatever your main cta is via chat gpt and then you can go into agent mode using agent mode you can actually watch how it goes through the site and if it hits any issues so in the webinar i use an example trying to book a flight through quantas using agents and it cannot load the quantas website i think this is going to become more and more important google this month released an article on how to optimize the agent to make sure your website is friendly so i think this is what we're going to start to see more and more of
Speaker 2yeah nice didn't they also advertise a role with geo in it
Speaker 1yeah they did and it ended hopefully ended the long discussion is it
Speaker 2a o is it geo is it a i o so we've uh we've settled on geo and now thankfully we did just prior to google's announcement and then the next i think tips and tricks around copywriting large language models clearly they're built around language they're built around tokens what are you doing with our copy team at rocket to have them writing in a way that works in terms of maximizing visibility for clients in large language models
Speaker 1i think like the first thing when it comes to copywriting is maybe the most obvious but it's becoming more and more important is there is so much content out there that is the same written by ai etc so anything you can do to invest in high quality content and that might be instead of publishing 10 blogs you've published one that is like an industry paper has data has statistics and it's something actually unique like the chances you've been cited are going to be way more so that's like the first place i would start is obviously a lot of people that are you can do around that the other thing is as we mentioned before like chat gpt acts like a user doing research so trying to understand the paths that chat gpt takes you can actually do this if you go to thinking mode on chat gpt and you click thinking or sources you can see what other searches it has done in the background so for the example of the car if you're searching for what is the best car for a family you want an article around that you might want what is the best car for a you have content pillars that address every different angle what is the safest car for families like thinking about the other types of questions that your users would likely have when you can feel product or service and then building out what we call content around that because chat gpt is very lazy it will look so the basically the main heading i think is 70 or so balances generated within the first paragraph or two so also including things like too long didn't read summaries at the top are really useful as well yeah
Speaker 2and i think in plain language it is so important to be adding to the narrative you need to be adding awesome content into whatever industry or some industry you're writing on original research stuff that hasn't been out there before because the and the large language models are so good now with writing awesome emails and awesome white papers and whatever else but they're kind of just spinning up content that already exists in the world and it's so easy for them to identify and they're going to be able to identify and they're going to be able to identify that it's and i remember the first time i played with chat gpt i was like oh what does this mean for rocket's copy team and our team has literally doubled in size since that period and it is because there is this insatiable demand for really high quality content the idea of your pretty average 500 to 700 word blog article with the keyword stuffed in at eight times it's just worthless now like that content you could be writing those articles in the background while joe and i talk on the podcast pretty easily so you really focus on quality not on volume absolutely key takeaway in terms of in terms of content types a little bit different right but what are the actual types of content as opposed to how you're writing it joe what are you saying works well yeah so i
Speaker 1think again like just having your head chat gpt is lazy so any of the work that you can take out of it is what's going to really help so the type of content from that is like listicle so what are the 10 best what are the 10 best ways to do x what are the 10 best businesses in this service and because during this time i've been writing a lot of content i've been writing a lot of content i've been writing a lot of content i've been writing a lot of content i've been writing a lot of content and that you're answering multiple different questions that chat gpt might have same as like pros and cons content like you're already giving both arguments without chat gpt needing to go to multiple sources so that really works well anything like faq focused including faqs at the bottom of articles but even just like what are the frequently asked questions on x service and perform really well anything within the specific location your users are likely to be searching so chat gpt really using memory, etc. So if you're searching for Sydney, it will know that and it will prioritize content that is written for that audience.
Speaker 2Nice one. There's some good practical takeaways there as to kind of how the models themselves are working. If you want more detail, I just encourage you to look at the webinar as we do go into a lot of specificity on that. And I think just in terms of the final big pillar that the large language models are looking at, which is all around authority and brand, just talk, I think, generally around the how and the why of that.
Speaker 1Yeah, sure. So I think like one of the biggest changes between AI models and Google is what happens outside of your site is more important than ever. If someone is searching for a business on Google and like trying to find out more information, most of the time they're going to go to that website. They might go to one or two sources. Within these larger language models, there's no guarantee they will even check your website. So everywhere that exists on the web, you want to own and anywhere that you're using. Yeah, I think that's a really good point. So I think that's a really good point. I think that's as well so like owning your wikipedia page any social profiles making sure all the information is correct there we've had a lot of clients where we looked into this and it's been chat dbt has been bringing in sometimes that even the wrong business like it's being diluted when they get the information and a bit confused and we've had to come up with new content to put on the straight there yeah
Speaker 2it's just fascinating isn't it and i feel that has a step charged like the literally the foundation of google was link building right so that was obviously a huge part of the early and to mid phases of that kind of seo growth through the 2000s and 2010s and it was always important to tidy up citations and just make sure that stuff was consistent but this just step changes it right because we have had so many instances where literally the wrong business is being brought in potentially in a different vertical completely wrong information the importance of that is just magnified
Speaker 1you , yeah and sometimes it's an easy fix we had one client where it was pulling in a u.s brand that's operated in a different industry all we did is built a citation on that website and now they have the correct business and now it's fixed information so it yeah it's just seeing where i'm pulling information and making sure that is updated
Speaker 2awesome joe thanks a lot for coming on today to chat all things ai search and seo and geo i thought i'd put you on the spot here with a rapid fire round of questions for you joe thank you so much for coming on today to chat all things ai search and seo and geo i thought i'd put you on the spot here with a rapid fire round of questions it's all going to be one two or three words okay you ready for this i think so the biggest ranking factor in ai search right
Speaker 1now i would say external sources so yeah like roundup content most
Speaker 2overrated seo tactic in the ai world i would
Speaker 1say trying to optimize the ai overviews because the overlap is just so high for google and it's very non-informational so it's more
Speaker 2important brand what's more important brand or
Speaker 1backlinks i would say brand
Speaker 2one platform leading ai search today a chat gpt
Speaker 1the biggest mistake
Speaker 2marketers are making with ai
Speaker 1not monitoring or having any type of framework or monitoring incorrectly like monitoring the same way you would as google
Speaker 2and other tools one metric that matters
Speaker 1most i would say citation share against your competitors like shareable could
Speaker 2answer that one and then this is more broad but what's your one piece of advice to an in-house
Speaker 1marketer i would say do not ignore this as a channel make sure at least you have reporting and setup and probably the your brand online is going to become more and more important so making sure that yeah how customers are talking about you is positive there's not too much negative press out there etc
Speaker 2great answer joe thanks so much for coming back onto the pod awesome thanks james you listening to the smarter marketer podcast stay up to date about new episodes on linkedin and instagram by searching for smarter marketer podcast you can purchase your own copy of smarter marketer via the amazon website and if you want a second opinion about your business's approach to digital marketing send me an email james l at rocketagency.com.au or visit the rocketagency.com.au website thanks for your time
PodcastTranscript AI