Do Rankings Still Matter with AI Search?

Heat. Heat. [music] [music] Hello and welcome to the dojo, a search marketing podcast by Exposure Ninja. My name is Dale Davis. I'm the head of marketing Exposure Ninja. I'm joined by our CEO, Charlie Merchant. Charlie, I understand that you've been fielding a lot of questions from our clients recently and kind of digging into what are the problems they're uh coming across. And one of the primary questions I hear is that they're not sure whether rankings still matter for AI search. Yeah. And I I would like to dispel some myths around this. So we've had questions around whether Google matter Google rankings matter at all whether being in the top three positions is important how AI search is changing all of this and what do we do like what's the most important thing is Google search the most important thing is chatbt search the most important thing AI overviews matter all of this bubble of search questions going on and I think there's been some myths around whether now that AI's chat bots like Chat GBT, Gemini, Perplexity and all of those exist. Do we even need to be thinking about Google and how important that is? So, first thing that I want to address here is how big Google actually is. So, Google search is still by far the most dominant search engine, the biggest search engine, the most used, around 210 times bigger than Chat GBT. Some of the recent stats are around 14 billion Google searches versus chatbt processing around 66 million prompts that are searchike and that's daily. So that's that's a huge difference. Google has billions compared to just millions. We're talking in big numbers, but if we actually compare those, it means Google's dominance is is still incredibly important. So when most businesses now are looking at their Google Analytics, looking at their organic, the majority of traffic, leads, conversions, ultimately revenue from organic is driven by Google search still, which means that there's no we can't just completely abandon the Google SEO ship and move all of our eggs into the chat GPT basket. We are very much in a in a mixed playing field where Google has the dominance and we've got these other AI search competitors coming in chatbt Gemini of course owned by Google's AI mode perplexity claude and co-pilot plenty of others as well. So when we're thinking about do the top three positions actually matter though in terms of your actual Google traffic they matter. We've always known that being in the top three, being in those poll positions means that you are more likely to get traffic than anyone below you and they take a disproportionate share of the traffic that comes in the search results. However, we also now have AI overviews which are the AI generated summaries at the top of the Google search engine results page. And what we also know is that for many many businesses, these have meant a drop off in traffic. So they're getting less clicks even though they're getting a good amount of impressions. And this gets talked about all the time, the rise of the zeroclick search, particularly because AI overviews tend to show onformational searches. So that a lot of the time that's traffic that used to go through to a business's blog, now just hitting the AI overview and getting the answer immediately. However, these AI overviews do site pages from Google's organic search results. And though being in positions 1, two, and three doesn't mean you definitely get to show up in the AI overview, there is a correlation, a strong correlation that the higher you rank in Google's organic search engine results, the more likely you are to show in the AI overviews with positions one and two, particularly much higher. It's about a 50% likelihood if you're in position one that you're going to show up in the AI overview. And there's a great study by HRES that I I recommend looking at. Very lovely graphs that they've got that show the very clear correlation from position 10 on page one to position one and how much it increases in percentage likelihood that that page will be cited in the AI overview. So plenty of reason to still be focusing on those top three positions and at least page one rankings for Google. So one of the more recent uh studies that AHFS has put out is that zero click search you know for a AI overview specifically they're seeing across like 20,000 websites or something like that a 53% drop in clicks through to websites. So regardless of whether I'm ranking first, second, or third, do you feel like it's still worth it? I'm going to get 50% of the return than I used to a year ago. This is a great question. So AI overviews, because it has driven a drop in clicks, the reason that you would want them, there's a couple of things going on, right? The first one is huge, [snorts] huge brand awareness. Seeing the same website answering queries on a similar topic is incredibly important if that person is likely to be moving from top of funnel through to a buying journey. So, for example, if uh you are a cosmetics company, you're selling skincare solutions and someone's thinking like, "Oh my goodness, my got really dry skin. I'm using moisturizer. It's just not helping. I don't know what to do." at the top of the funnel, they're going to be searching things like causes of dry skin, products for dry skin other than moisturizer. Then they're probably going to end up a little bit deeper into types of products and types of ingredients. Lots of those searches are going to be eaten by the AI overview. Someone doesn't necessarily need to go to a blog to read all of the causes of dry skin because they can get a bullet point list in the AI overview. But if your brand is referenced, they're immediately noticing that even if they're only subconsciously taking it in. And particularly, I say that because a lot of AI overviews have visual images for the cited brands. So, if you have a very strong branding or look and feel, and one of my favorite examples is The Ordinary because they have a really striking white uh labeling on all of their products with really striking black text and a very similar memorable font. you start to kind of get that stickiness when you see it coming up on lots of searches in that topic. So for me that's incredibly important. There's also a sort of future focused reason as to why you may want those AI overviews and that's that AI overviews are more and more trying to push people into AI mode. So quite often if you unfill an AI overview at the bottom, it's trying to get you to read more in AI mode or to start a conversation in AI mode. And what we know about AI mode, very similar to chatbt, Gemini, and other LLMs is it gives that back and forth conversational element. And the goal there is that the buyer journey, the salesunnel is being compressed because you can have that conversation without having to navigate to lots of different websites, which means someone can move from very top offunnel searches about dry skin right down to starting to make comparisons between different brands. So that dry skin search then eventually turns into, okay, maybe I need to look at The Ordinary versus Sarah Ve. Which one of these brands is better for me to solve this dry skin problem that I have? All of that can happen in a under five minute chat conversation. And the reason I say this is important is not because AI mode has suddenly taken over. Google search results, the traditional 10 blue links are still the dominant search result at the moment. They're still the default. But this direction of travel from Google is extremely clear. Dale, you and I have talked about it on past podcasts as well. We know that Google wants to make default search much closer to AI mode and it's using this as a test bed at the moment to see how people interact. We've seen AI overviews becoming much much closer to AI mode in the way that they actually show information as well. and this clear link together trying to get people to navigate from the main search results, the traditional page through AI overviews into AI mode to get users used to AI experience in their browser. I had a conversation with uh Gemini the other day that went on for like a four like 15 20 minutes maybe longer and I don't think I went over I didn't go over to the Google results like at any point. I think I visited one website um off the back of that conversation. There's just so much happening that it makes me wonder whether okay we do want to be ranking so we do appear in there but does that mean that as I as a marketing leader need to change my perspective on what counts as a result for doing all this work like proactively trying to appear in AI results. I'm surely not going to be tracking traffic as much this anymore or you know do I need to shift to some other kind of core uh performance indicator or this is the perfect question and the one most businesses are thinking about because ultimately we're not in it for the traffic. We're in it for the conversions, the revenue and how that impacts the bottom line of a business. From SEO's perspective, SEO's future is in owning answers, not chasing clicks. We're going to see a huge shift in the metrics that we expect from SEO. There's going to be far less focus on traffic quantity and much more interest in the quality of that traffic because what we care about ultimately as a business is the quality of the leads that we're getting or the quality of sales if we're in e-commerce, the quality of conversions, the average order value, how many people actually come to the site and then check out and then check out with a good amount of stuff as well. So the traditional metrics of conversions and the conversion rate of that organic traffic on a website is going to be incredibly important. There's going to be less emphasis on traffic. Right? Ultimately the business doesn't care if you get 100 people visiting the site or 10 people visiting the site if ultimately you still only get one person who becomes a lead or who becomes a buyer from the business. What we actually care about not the volume of traffic, the quality of what's coming. So I think SEO metrics, we're going to be looking more at topline metrics that are around owning answers, the visibility in those answers as well. And I'm talking about AI overviews. And I'm also talking about visibility in LLMs like Chat GPT, Gemini, Perplexity, AI mode as well. We're going to see a shift in that direction for now because as I mentioned at the beginning of the podcast, those traditional search results are still dominant in terms of where people are searching and the amount of organic traffic and leads and conversions that they drive to a business. We still have to be looking at rankings for now. There is no we cannot just suddenly get rid of tracking rankings. is still going to be incredibly important to the bottom line, but we're also going to be thinking more about the commercial keywords, the middle and the bottom of the funnel, less so about the top of the funnel and particularly the top top of the funnel, which is things like what is X? Would my approach to metrics uh monitoring as well as my approach to AI search optimization differ depending if I was a service-based business or an e-commerce store or maybe a retailer? Will the difference in metrics differ? I mean ultimately both businesses are still going to be looking at the revenue that's coming in. Hopefully hopefully that's what they're concerned about. e-commerce going to be looking directly at sales revenue. Yeah, go on. Sorry. Part of my the reason for my question is we're obviously going to be seeing more sales completions within AI platforms as more platforms take on, you know, shopping functionality like Chatty has and and some others. Do do you think that that will change the way that I as a head of digital or whatever would approach approach things like how I measure my results or even attack the problem itself? Yeah, because what you're alluding to here right is in the US chatbt chatbt AI mode is testing direct checkout as in you would be able to check out in the AI mode search results without navigating to a website. But this is very early stage, so we don't actually know the results of this yet and what the uptake of it will be. Google would obviously love it because they'd love to keep people in their ecosystem. They'd love to get product purchases done quicker for buyers. And of course, they'll be thinking about ways to integrate ads, which is their main revenue source, Google's main revenue source. So if they can get it working for organic, then there's a a good chance that they'll be able to get it working for ad revenue as well. How it impacts the metrics of the business. It's still going to be important where those people searched and where they checked out from. So in that sense, all of these are still organic, but it is going to make that metric of total traffic to your website significantly less important, particularly in e-commerce. If this does stay the direction of travel, if that's rolled out further, right? I mean, you've I've heard you talk several times before about how you see things going on the commerce side of you we've got agentic shopping basically is going to increase as well. Is rankings going to impact that as well? like with a gentic side of things like a an AI agent in the form of a browser or as an extension or whatever. Is that also going to rely upon where you rank or is it is it a different kind of system? My perspective on this is that it will not matter in the same way that it matters to humans. So when humans are searching, we're calibrated to look at the top of the list. This is true for everything we read. We place more emphasis on what's at the top of the list. Same for top of the search results. Same for what's top of a chat GPT answer. Even though technically those aren't rankings, it's just the ordering within there. When AI is searching, it's able to search much more broadly over a significant number of pages. And it's able to doesn't have to read the whole page in order. It's able to just go to a specific paragraph, even a specific couple of sentences to pick out what it needs. So I think we will see a significant difference with how agents search because they'll be looking contextually. They're going to be looking for authority in topics and then context around the kind of query that has been asked because they'll be trying to even more closely match the intent of the searcher, what original intent they actually had, which could have quite a few nuances. So I definitely think we're going to see a change there. I just had a thought of a LinkedIn post I saw yesterday which talked about like the measurement devices behind um AI search optimization or answer engine optimization something like that and the conversation was about do we need to change the actual like way we report this and track things in in the system itself you know with we often talk about you know first click and last click attribution and so on. Do we need how do we go about the um attribution process and should we be focusing more on brand tracking versus keyword or page rankings? Firstly, I think attribution is going to become even more difficult and murky than it currently is. There's a couple of ways around it and there's no perfect solution. So, this is the imperfect solution. The first is Google Analytics does still show us what comes from where roughly and that's lastclick attribution but it will show us traffic sources. So if we create separate channel groups we can see the amount of traffic and conversions from chatbt Gemini etc. What it doesn't show us that's just showing us right the referral traffic that's coming through. What it doesn't show us is if there was a touch point earlier in the journey, but actually that person later came through a Google um 10 blue link instead then we don't know that they were originally in chatbt. So there's what what's called dark traffic going on and that's really really difficult to attribute and it would take a lot of uh a lot of work to get anywhere near it. So, we get a rough and ready result based on the Google Analytics data that we have at the moment, which is not the end of the world, but it means there's probably a lot more traffic hitting other touch points that we don't even realize are part of the the whole journey at that point. The other really useful way of attributing some of the leads that you have is self-reported. So having a a box on your website that's that how did you hear about us? Particularly if you're a business that generates leads and those leads are about to go into something like a sales conversation or a demo or some type of pipeline off the back of it. That self-reporting you will start to see more and more people actually saying I came from ChachiBT or I came from some type of AI search. and it probably will give you a better idea and it's it's not 100% accurate because people will make mistakes. Some people will just very quickly input into that field, but it does give you something to work from. And then you can even have a follow-up conversation to ask about how they found you. Tend to find that that data is useful to have. just thinking about like is there any kind of metrics or attribution or performance stuff we should be focused on or should we be really focused on actually how to increase our visibility should I mean the reason I ask is that we we need to sit in uh in meetings with other people and justify why we're doing what we're doing should we be actually spending more time on doing the doing or you trying to find a perfect way to articulate that with our key stakeholders. I feel like you need a balance because you need to be able to prove to stakeholders the value of any marketing channel that you're working on. So, you need to be able to show at least something for it. The other side of this coin though is you could end up so in the weeds just thinking about your tracking and attribution, which there is no 100% perfect solution to that you spend all of your time actually working on that. and very little time having any impact because you're not driving traffic. A marketer's main problem, right? This is the problem that all marketing managers experience is how do I track effectively, but also I need to actually get things done to start driving customers to the website to to the business. So, there has to be a bit of a focus on on both. And as long as you have some way of reporting useful metrics that are at least leading indicators of how a channel is performing, I think that that is top level mostly going to be enough for most businesses. We had somebody send us a question live because we do record these live before we publish them. Somebody's asking um how far do we see AI search going and could this be the start of the end for Google search? Which is a good question to end the podcast on. This is a great question and I think Google search is going to be around for a long time. Google is so so dominant that they're not going to just sit on their hands and do nothing here. AI search, however, remember Google has its own types of AI search. It has AI overviews. It has Gemini. It has AI mode. Are we moving in the direction where AI search is going to take over what we know as search? Now when we're thinking about the tender blue links absolutely 100% that is the direction of travel that we're going in and yes there's going to be competitors like open AI who own chatpt um perplexity as well claude all of these different AI chat bots that we know are they going to knock Google's entire market share I think it's extremely unlikely but good luck to them and I expect that we'll actually find they end up operating in different niches that they'll be looking for different target audiences in the long term and Google is of course also focusing on the general the general population because it already has uh the dominant market share. The thing that I would like to pick up though is also whether rankings still matter when it comes to chatbt which is kind of the second half of this question. We've talked a lot about Google and the attribution that we're using for Google traffic, but also we have a separate situation here which is AI chatbots like chatbt, Gemini perplexity, let's include AI mode as well. They all have underlying search functions and most of those are Google search. So Gemini and AI mode, they're owned by Google. They're Google products. They have an underlying search function and it's Google search. No surprise to anyone. Chat GBT, they used to use Bing, but now they're also predominantly using Google search when they live search the internet. And then we've got Perplexity, who uses its own tech. They also sometimes use Bing as well. So these underlying searches mean that sometimes they're going to answer a question just with their own data and they don't need to live search. Sometimes they're going to live search. And quite often I find that the majority of them do now live search when they're asked a query. But does that mean that they're just going to rank the the top three results from Google? Not necessarily. Um I think what we're seeing is that the way that Google ranks content is very different to answers that you'll get in chatbt. So, for example, Google prioritize prioritize like informative content, really comprehensive, well-written content. That's what the algorithm likes. You you often see very detailed articles, first party data, that kind of thing. Chat GBT has a significant amount of short content, best of listical content. And we know that LLMs like CHBT don't always pull from the first page of Google. They can pull from anywhere. They can pull from different pages. But I did see some data that I thought was pretty interesting. I'd like your perspective on it as well, Dale. There was a study from Chat Optic who checked whether brands that rank on the first page of Google then get mentioned in chatbt answers and there was an overlap. So it was 62% of the time chatbt answers reference the websites ranked on the first page of Google. Interestingly, I think that's pretty significant overlap. I think 62% is is quite a lot personally. Interesting. The study itself felt that that wasn't a strong overlap. They felt it was quite weak and that the correlation of what was ranking on Google versus what appeared in ChachiBT answers wasn't so strong because in four of the 10 times they did a search, the platforms would diverge. They would have different answers. But for me, even though that's not directly translatable from positions one to three on Google into chat GBT positions one to three, that correlation isn't there. Actually, 60% of the page one answers showing in the chat GBT answer I think is a fair overlap. Yeah, I would agree with you. It's almost two/3. So, it's it's considerable amount of overlap. Yeah, I think so. And that just goes to show that those rankings in Google still matter. Even if we're focusing on LLM like ChateBT and Gemini who aren't giving exactly the same top three answers, if they even have a 60% overlap for those first pages, then strong, well-written, authorative pages on relevant keywords, relevant topics still showing in AI searches. It [snorts] sounds to me that they very much still do matter. Do you have any final thoughts to share before we finish for the week? I think I've hit everything, but I would love any more questions around how much rankings matter with AI search. And if anyone wants to continue this conversation, please definitely uh jump on into our YouTube comments or LinkedIn if that's where you are. Yeah, if you reach out to Charlie directly, you'll be able to uh ask her. I'm sure she'd be happy to answer your questions as well as you could book Charlie for any future speaking that you'd like to appear at. We kind of like filled up her calendar. I don't think you've had any opportunity to take any time off yourself, but that's okay. Time off. [laughter] No, but I have all the fun of going to these great events and meeting amazing people. So, you know, swings and roundabouts. [laughter] Yeah, totally. Fabulous. Well, thanks everyone for joining us this week for this episode of the Dojo. We'll see you all next week for the latest news and going on in all things AI search and SEO. Take care and see you then. Bye bye.