Latest / Found in AI: AI Search Visibility, SEO, & GEO / How Should You Measure GEO Success in AI Search?

Transcript

How Should You Measure GEO Success in AI Search? · Found in AI: AI Search Visibility, SEO, & GEO

• 0:41
  (Transcribed by TurboScribe.ai. Go Unlimited to remove this message.) Only one percent. Seriously, one percent. That's how much overall website traffic is currently coming from AI tools, and that's according to a recent BrightEdge study.
• 0:51
  And yet every marketer I know, myself included, is trying to figure out how to measure success in AI search. Hey, if we haven't met yet, I'm Cassie Clark. I'm a fractional content strategist and the host of this podcast, Found in AI. It's a podcast about getting your business found in AI search. In today's episode, I'm joined again by Christina Fruense of WebView SEO. Last time we talked about whether SEO was dead, and spoiler, it's not. But this time, we're digging into the metrics that matter for GEO, from AI traffic and citations, to shared voice, needs, and sentiment.
• 1:29
  Here's part of that conversation. Since last time you and I talked, you kind of pushed back on the SEO is dead theory. Has anything shifted since then? I don't think so. I think more than that, I might have been, like, all these SEOs might have been proven wrong with the fact that SEO is dead, just because, you know, I've seen a study recently from BrightEdge, and I think, yeah, either from BrightEdge or Amazon, I have to look it up. But basically, the study says that AI traffic, referral traffic is still standing at just 1% of, you know, overall traffic to the majority of the websites. Obviously, they had just, you know, a limited sample of websites.
• 2:22
  But that is to say that organic traffic is still kind of driving the majority of traffic and the leads at the same time. And what's more interesting is that before I was convinced, and it was kind of logical also, you know, with the fact that people were claiming that AI traffic converts better than organic. But again, the same study, or actually the study from Amzev, what they found is that statistically, it's not really true. In fact, they convert about pretty much the same around 4.5%. And that was interesting, because I was, you know, assuming that AI traffic would convert better, because, you know, people kind of vet you before they land on your website, they kind of make their decisions before landing on a website, they would come a bit more prepared to convert. But statistically, that's not true. So I would say SEO is still pretty much, you know, strong, going strong. So definitely, it is the foundation still.
• 3:31
  Which is interesting. And like, why I started this podcast, because it seems like everything changes with the wind. So I had heard too, from a bunch of experts that those that are coming to your website, I mean, I even thought this, those coming through your website through chat, GPT, or whatever AI search engine that you're using, are more qualified and ready to make the purchase. So that's interesting. Yeah, I was surprised. Maybe it's going to change in the future again. And maybe that's going to prove true, as you know, AI traffic kind of picks up the momentum. And then, you know, actually shows when it's going to be at the same level as organic traffic, maybe that's when we can come here. But at the moment is it's the same. Yeah, I wonder, I wonder how that changes things for other marketers? Like, do we keep trucking ahead with our current strategy and thinking like, hey, those that are coming here, probably more qualified? Or do we go back to whole, our tactics before AI search became a term everyone's been trying to figure out? Yeah, it's a, it's a mixed bag. Definitely a lot of people kind of don't know, you know, what to do, whether the question is like, do I do SEO? Or do I do AI SEO? Do I do both at the same time? And I stick to doing both. Like, I think at the moment, any SEO strategy should just come with SEO search landscape in mind. So you can't just exist in the in a vacuum.
• 5:00
  So I think it should be a bit of both. Yeah, I think you're right. I do think you're right. So you made a LinkedIn post this morning, where you talked about geo metrics, like five, and I want to get into each one. But of those five that you listed, which one do you think is gonna emerge as the number one metric that most companies need to keep track of? I think it might be easiest to track at the moment will be the AI traffic, I think, just because it kind of feels familiar with organic traffic, you know, you already know what it is, you know how to track it, it's pretty easy to set it up. The problem with that is that not all of the elements are actually passing, you know, the proper parameters to be able to track them. So there is definitely a lot of uncertainty, whether that's you're getting the full picture or not. So I think that might be one of the most popular, but I don't think that is the most, the most important one. So which one do you think is the most important one?
• 6:08
  I think AI share of voice is something that I like to track and keep an eye on for the clients that I work with. And it helps to put things in perspective, because if you just, you know, look at your AI traffic, just to look at your AI citations is great, you know, you understand which elements are kind of setting you and mentioning you, but you're kind of missing out on the bigger picture, what your competitors are doing, and how you stand against the competitors in the LLM search. So it's kind of showing you, you know, your share of voice versus your competitors for the prompts. And it's a little bit tough to track at the moment. But there are certain tools that I've tested so many tools, honestly, but there are certain tools that actually allow for AI share of voice. So which one is your favorite tool so far?
• 7:07
  It's again, very hard to say. Some of the tools that I tested that I really loved are not live yet. That's the thing. Okay, gotcha. So tight lipped for now, but hopefully, hopefully soon enough, we'll be able to share. I like ziptidev. This is actually one of the first tools that I tried. And I really like it because it shows your AI share of voice and shows the competitors and it breaks it down by citations and domain mentions. So citations is when you know, LLM is just citing your brand versus domain mentions when they actually link to your website, which is pretty cool to see as well. So I really like the fact that it helped me with a lot of pitches what I was doing, you know, putting together the audits for the company. So it's a pretty cool tool. Yeah. So when you measure share of voice in AI search, how are you communicating that to like non marketing stakeholders and clients that come to you? Right, it's, I kind of go back to, you know, showing them how the landscape has changed right now, for example, before a potential customer would type something in Google, and they would get 10 blue links right now, that's not the case. Majority of people actually are typing in Google and getting AI overviews for some queries, or they're going straight to their favorite tool, AI tool and then typing in there. So when it comes to AI share of voice, I just, you know, I kind of put it in front of them as how many times you come up versus your competitors and how much of that real estate they take in the LLMs versus your brand. And usually, it's also very hard right now to measure because another interesting thing is the fact that if you put the same prompt in GPT, three different times, you will get different answers.
• 9:17
  And this is just crazy to think about. And the fact is that the tool that I tested the last, I got a demo of it, and it's not live yet, what they're doing is they actually are testing it three times. So give you the average of how many, how many times your brand shows up. That's really interesting. Yeah. Yeah. I have noticed that when I put in a prompt and then it changes and that's frustrating. That is so frustrating. But I guess from a consumer standpoint, if you're researching, it gives you different options to consider. I don't know if that's any better because I'm one of those people that just, just give me the answer. Not three different ones. Like lately I've been looking for a new running watch because my Apple watch is like, it's like last leg. And so I've been looking at the Garmin and it keeps giving me different answers to the different models that I should look at. I'm like, I just want to see the best one. I don't want to see four different ones. That's frustrating. Yeah, it is. And another interesting point is that I've also also read recently that a lot of AI tool users, when they perform a search on let's say chat GPT, a lot of times they will go to Google and double check that, you know, they will get separate, you know, search on Google and then they will double check them. But it's the other way around people that will search in Google, they rarely will go to chat GPT. So that's also pretty interesting observation. That is interesting. I don't, for my own behavior, I think, well, when I do research on a company or something that I need to purchase, it's like, let me, let me see what chat suggests first. And then let me go to Reddit.
• 11:05
  Reddit, yes. Yeah, that's true. So that's, that's interesting. So the third metric that you shared was about AI citations. Do you treat paraphrase mentions without direct links as a win?
• 11:18
  Or is it only a citation if the brand is explicitly mentioned?
• 11:24
  Um, what do you mean by paraphrase citations?
• 11:28
  Um, so some of them that just kind of like, summarize without actually hyperlinking it anywhere, like they might, like they might mention HubSpot in passing if you're talking about a CRM, but it doesn't actually tell you what it does. Just, you know, kind of a light mention.
• 11:45
  Right? I think, um, for the purpose of AI citations, for the majority of it is just, you know, actual just the mention of the brand name, or actually the main citation. So these are kind of, this is at least what I've seen in tools that are being tracked properly. And still, honestly, the tools are evolving. And it's, you know, every day, something changes in them, and they are not accurate 100%. But this is, at the moment, the best we can rely on. So I think for the moment, you know, just the brand mentions direct brand mentions is kind of the best way to track it.
• 13:20
  So because I talked with Chris Ventelli earlier, he was talking about one of the ways to do that through your backlink strategy. Like, are you still recommending that to clients? Like you need to really work on getting these mentions elsewhere? Oh, you mean, getting your brand mention across third party websites? Yeah, 100%. This is probably going to be one of the, it's kind of the backlinks of SEO with, you know, SEO would try to get as many actual backlinks to the domain. So with AI SEO, I think links don't matter as much as long as your brand is actually mentioned on, you know, whatever, communities, third party websites, directories, like getting your brand really kind of spoken about and getting it right, I think is very important.
• 14:17
  And it actually ties with the last one as well, which is AI sentiment, which is pretty crazy as well. I'd say interesting, because we haven't had that with SEO.
• 14:28
  Yeah, which brings me to a question. So we'll skip metric four, I'll come back to that in a minute.
• 14:33
  So metric five sentiment, how much control do brands actually have, like realistically, over the sentiment in AI generated answers? It's not a lot, unfortunately. And AI tends to hallucinate a lot. I've seen AI hallucinating some features of products that don't exist.
• 14:55
  I've seen AI, you know, putting out incorrect information, negative information, like this, there is some tools that actually track your AI sentiment, like what AI is talking about you, in terms of if it's negative or positive. And there is some ways to influence that.
• 15:17
  I think one of the ways I deal with the AI sentiment right now is I push all the clients to have as much clear information about their products, have product pages, make sure that you, you know, highlight all your benefits, who you work with, who you work for. So basically, make it as clear as possible. And it's not only beneficial for AI SEO, but also for traditional SEO. And then, you know, making sure that you also syndicate that information across different directories. So sometimes directories have outdated information or plainly incorrect.
• 15:59
  So, you know, reach out to directories or, you know, the software aggregators, make them correct that information. And another big one is negative reviews. If you have some negative reviews, you have to address them. This is a very big one, because for local services, for local businesses, this is very important. That's why Google My Business Profile is also actually getting pulled into AI overviews right now. So working on your GVP is another big one, having it properly set up, completely filled out, and up to date is important.
• 16:38
  Yeah. Yeah. I mean, just going back to thinking about Reddit, like I've seen a couple of brands, like really take a hit because everyone kind of dog piles on it when it's, Oh yeah.
• 16:46
  Something's not great. So that does seem like it's kind of hard to control.
• 16:50
  Um, if that, if that were to happen, like what would you tell a brand to do?
• 16:54
  Like, especially if it's like on Reddit, where everyone's talking.
• 16:59
  Yeah. I think Reddit is kind of a very rough, honest, brutal community that just, they don't tolerate any, anything, you know, related to self-promotion, things like that. So brands definitely need to be very careful when it comes to software promotion.
• 17:16
  If there are certain things that you don't like, you know, on Reddit, I think the best thing the brands can do, obviously you can't take it down. The threads are impossible to take down, especially if they're truthful. Perhaps if it's, you know, not truthful and it's, you know, hurting your image, then there is ways to deal with that. But if it's, you know, just some negative experience, I would suggest brands reaching out to that user and, you know, just trying to make it right by that user, either compensating them, like doing anything in their power to have that user take the post down. And maybe that's how they deal with that. Because it's an analogy with how you would deal with a negative review on Yelp or, you know, on GDP, the same thing. Yeah. Yeah. And I mean, anywhere that that negativity is, it seems like it could really influence what AI is talking about, especially with how customers are feeling about your brand.
• 18:16
  Yes. I suppose customer service is going to play a huge part in all of this. Yeah. Yeah.
• 18:22
  This is Reddit, honestly, like one of the most cited domains on ChatGPT. So definitely Reddit.
• 18:31
  I've actually seen people like Reddit marketing strategies. I don't know how feasible they are, because it's very tough to do any kind of promotion on Reddit. So I just downloaded a guide to Reddit marketing a couple of days ago, but haven't had the chance to read through it. But I'm very curious.
• 18:56
  Yeah, I've seen a couple of those too. And then I've seen a couple marketers immediately after like trying something like, I just got banned.
• 19:04
  The rules are all different, whatever subject they're different. So let's go back to metric four, which was AI traffic leads. Are you seeing any patterns in which like the kinds of content that actually drives conversion from AI driven traffic?
• 19:19
  Yeah, I think with AI leads, it's the content that drives them the most is the content that is kind of bottom of funnel content. And that's why I try to push our clients towards creating that content, like for example, you know, XYZ versus the competitor or XYZ, like your competitor alternative. And these ones are usually people typing in this kind of prompts when they are already in that, they are past the awareness stage, the interest stage, now they're already evaluating their options. So if you happen to have that piece of content sitting on your website, and you present your brand in a better light, versus that would be a competitor, you know, presenting that. So this is one of the things that works pretty good. And like integration pages as well. And I think it's the same, the same things we're working in SEO, but right now, more than anything, I think they are also showing up in AI search engines.
• 20:32
  So when we talk about these direct competitor comparisons, we are talking about absolute direct competitors and not like a stretch, right? I asked this, because I've worked with a couple different brands. And like, they'll ask me, hey, do a comparison on this. And like some of the tools that they're trying to compete with are not a bit related. I wonder, I wonder how that plays in.
• 20:58
  It's, I think it depends on a brand. And some, some brands, you know, compete with direct competitors. And with some brands, for example, you know, they compete with against Excel sheets, or something like that. And something that is, you know, or even, you know, on paper or something like that. And I think I've, what I've heard as a kind of objection towards this is that I don't want to give any promotion to my competitor.
• 21:31
  What if, you know, a potential client will come to my website, and they already almost want my solution. And then they realize, oh, there is this alternative.
• 21:40
  Right.
• 21:40
  But it's the way you create this page, you know, you make it almost a no brainer for them that your solution is much better. Like, you know, you definitely have much more, obviously, it has to be factual. This is what I always advise, it has to be factual, definitely don't talk in a negative light about your competitor, because that's going to come back to you.
• 22:02
  But just kind of, you know, sing songs about your brand. And this is the way to deal with that.
• 22:09
  I have also heard those objections when I said that to other brands, like you really need this competitor pages, but we don't want to advertise.
• 22:16
  Exactly.
• 22:17
  But you still need it. So of these five metrics, do you think eventually geo success metrics will eventually standardize like SEO did, or will they kind of stay fragmented?
• 22:29
  I know, that's like, look into your crystal ball and tell me what you see.
• 22:34
  I personally think I even SEO metrics weren't really that much standardized. I mean, everybody was tracking, you know, their kind of a traffic leads and the rankings.
• 22:45
  But I think with AI SEO, there's just so many touch points that and it's much harder to track at the moment, perhaps later on, you know, it's going to be a bit better. But it's also interesting with the fact that with organic traffic, we only had Google and maybe like Bing and, you know, some other ones. Here, we have so many LLMs. And it's crazy how all of them, some of them are different, you know, some of them actually do web search, some of them are just going back to their trained database. And it's hard to track.
• 23:25
  So I think some of them might stick, like, for example, the Azure voice, traffic, but I have a feeling it will also depend on which industry you're in, like, you know, what you care about the most, whether you know, you care about the traffic, or you care about the leads.
• 23:47
  So it's, it's tough to say, but for the moment, at least I tracked this five, and then I don't want to make it too, too much, because then it just gets overwhelming. Even five is too much, I guess. So I'd like to narrow it down to three, to be honest.
• 24:03
  So which three would you choose?
• 24:05
  It's tough. I think Azure voice to have a bigger picture with competitors, AI traffic, and then perhaps the AI leads, just because it leads is something that I typically work with clients that are not publishers, they just don't care about the traffic, but the leads as well. So, you know, the form fields, things like that, probably these three.
• 24:32
  That's it for today's episode of Found in AI. Huge thanks to Christina Fonse for breaking down the five geo metrics. If you're a marketer wondering how to measure success in AI search, I hope this gave you a clearer picture, and maybe a few ideas to test out yourself. Okay, quick before we wrap up, we are officially in beta with ThoughtTree. It's the workspace my team and I are building to organize your prompts, reports, and templates into repeatable workflows. If you'd like to test it out and help share the platform, you can sign up for early access at thoughttree.io. And again, that's thoughttree.io. Okay, and here's this week's experiment. Let's pick one of your priority keywords and run it into three different AI tools like Google AI Overviews, ChatGPT, Perplexity, whichever one, just three of them. Make a note of how often your brand shows up, if at all. Then repeat that same search a few times and see how the answers shift with different prompts.
• 25:26
  If you want a structured way to track those results over time, I've built a free AI Visibility Tracker on my blog at cassieclarkmarketing.com. I'll link it in the show notes so it's easy to find.
• 25:37
  And if your team is ready to go deeper building the kind of content engine that gets you found in AI search and driver's pipeline, that's exactly what I do as a fractional content strategist.
• 25:46
  You can learn more and get in touch at cassieclarkmarketing.com.
• 25:51
  Thanks for tuning in. I will be back next week with Romana Cutts of SassStorm, who's been helping clients add FAQ schema to their blog post and seeing great results with this tactic. She's shared exactly how and why it works, so don't miss that episode next week.
• 26:07
  I will see you then.