Latest / Found in AI: AI Search Visibility, SEO, & GEO / How Do Brands Show Up in AI Search Results?

Transcript

How Do Brands Show Up in AI Search Results? · Found in AI: AI Search Visibility, SEO, & GEO

• 0:41
  (Transcribed by TurboScribe.ai. Go Unlimited to remove this message.) So you finally nailed your SEO strategy, and then AI Search came along like, cute, but welcome back to Found in AI, the show about getting your brand found in AI Search.
• 0:54
  I'm your host, Cassie Clark, content strategist and CMO.
• 0:59
  So when you ask ChatGPT, Perplexity, or Gemini, or whichever one that you're using a question, how does it decide who gets mentioned and who gets ghosted?
• 1:09
  Today, I'm joined by Charlie Graham, founder of RivalC, to talk about the next evolution of visibility, AIVO, and how to make sure your brand actually shows up.
• 1:21
  Here's part of that conversation.
• 1:23
  Yeah, I'm the founder and CEO of a company called RivalC, which is in the AI visibility space.
• 1:29
  And our focus is letting you track and boost and optimize how your brand appears in chat engines like ChatGPT, Google AI, and Perplexity.
• 1:40
  Our focus is showing you how you appear for each customer segment and each persona versus just kind of looking at keywords.
• 1:48
  Because the way we've seen AI Search is that it is very heavily skewed with memory of the individual, and it's very personalized.
• 1:55
  And it's a totally different way to think about.
• 1:58
  Optimizing for that is the, it behaves differently than search engines.
• 2:03
  It behaves so much differently.
• 2:04
  So when you're doing like audience personas that's by customer segment, how many are you looking at?
• 2:10
  Like however many the brand has, or just one or two?
• 2:13
  So you can do, in our system, you can do up to 10 different personas, and you can change them all the time.
• 2:19
  We have, you can have up to five different customer segments that actually can grow if you have an enterprise level account.
• 2:25
  But you, it basically, it will generate them for you, but you can actually go and create your own as well.
• 2:30
  You can edit them and so forth.
• 2:31
  So, and we've had customers use the same ones again and again to see how they track over time, or just change them every time.
• 2:39
  Yeah, that's nice and easy.
• 2:41
  That's nice and easy.
• 2:42
  So, when we were talking about scheduling this, you had a new term.
• 2:48
  So on this podcast, we've talked about AEO, GEO, LLMO, and you had a new one, AIBO.
• 2:57
  Can you briefly describe what that means?
• 3:01
  Yeah, I mean, they're all fairly synonymous of what they're doing.
• 3:05
  AIBO is just a different way to distinguish it, AI Visibility Optimization, which is kind of what you're doing, which is trying to optimize how often your brand is mentioned in AI.
• 3:17
  SEO is Search Engine Optimization.
• 3:19
  The AI chats are not quite search.
• 3:22
  GEO, it's Generative Engine Optimization.
• 3:25
  Normally that also is a synonym, but that conflates with geolocation and targeting.
• 3:31
  So there are so many different names.
• 3:32
  No one's really kind of settled on one.
• 3:36
  And Answer Engine, it's more than an answer engine.
• 3:39
  And so the AIBO was just an attempt to say, okay, what is the new way that we should be thinking about this?
• 3:47
  And it really is optimizing your visibility in the AI engines.
• 3:51
  So how would you say that differs from traditional SEO?
• 3:55
  I ask this of everyone that's on, at the time, because most of us have been elbow deep in SEO up until March, when this sort of blew up.
• 4:05
  Yeah, so first of all, you will see on LinkedIn, all these people saying SEO is dead and it's totally different.
• 4:12
  And you'll see the exact opposite on LinkedIn and Reddit, which is like, they're the exact same, don't do anything.
• 4:17
  The reality is somewhere in the middle.
• 4:21
  To optimize, to understand how to appear better in these kind of AI engines, you first have to understand what they're doing behind the scenes.
• 4:30
  But the short-term TLDR is, good SEO is a great starting point and it will get you a really far along the way.
• 4:37
  There are tricks and things you can do if you understand exactly how these tools are working, that you can actually get a bigger, you can optimize even more to appear in these AI visibility things beyond just great SEO.
• 4:51
  But with that in mind, great SEO matters still.
• 4:54
  And without that, you aren't gonna appear.
• 4:58
  But sufficient is not the only way to optimize.
• 5:01
  You can do more.
• 5:02
  So I'm happy to go into that today.
• 5:04
  Yeah, just high level, what makes these AI search engines operate a little bit differently than Google?
• 5:11
  I mean, not like technical, just TLDR of what the difference is.
• 5:16
  Yeah, so the AIs, I mean, people call them, so the way the AIs work is they will, for the most part, there's two different types of ways they get the information.
• 5:25
  One is they are trained into this model, and that is usually data that they've scraped off the web or gotten from different sources, and it's usually a year old.
• 5:34
  So by the time the model comes out, it is based on information a year old because they had to spend that year testing and training it.
• 5:41
  The way models account for this when you search for something recent is they do a real-time search.
• 5:47
  So the way most people are seeing these brands appear is you will go onto OpenAI or Cloud or Grok or Perplexity, and you will put in a search, and it will first check its regular model from the year ago of what the information is and realize that, hey, I need to catch up.
• 6:04
  So we'll actually search, and we believe, with the exception of Perplexity, which does its own, that the general belief is that right now they are searching a combination of Google and Bing and sending out their own keywords, figuring out what keywords to ask, acting like a pretend user, sending out the search requests, coming back, reading those pages, and then using that as part of its memory to kind of come up with an answer.
• 6:27
  So it's basically reading in the top 10 pages of Google search results for certain queries, and then using those, summarizing all that, and putting that back into the answer, and it's doing that very quickly.
• 6:40
  So that's, yeah, go on.
• 6:41
  I don't think anyone has explained it just like that.
• 6:44
  Like, that makes more sense than what, than what I had read about later.
• 6:49
  Yeah, so that's generally how they're behaving.
• 6:51
  Obviously, they're doing different things behind, some different things behind the scenes, and they all have their own nuances of how they do that.
• 6:58
  But I can go way in the, you know, I've researched this.
• 7:00
  We've done a lot of studies on this.
• 7:02
  We can go way in the weeds on kind of a little more of what they're doing, because how they do that and how they do the searches, how they read the pages, if you understand that, you can actually do a better job of having your content appear in their engines, because you can make it fit better with what they're trying to do.
• 7:20
  Yeah, so let's go in the weeds a little bit on that one, for sure.
• 7:23
  Yeah, so a couple of things that we have found, that the most, the biggest one, well, not the biggest one, but there's a lot of them.
• 7:29
  The first one that we can talk about is how the AIs chat, do these searches when they do these real-time searches, is different than how a human does.
• 7:38
  So when you are doing SEO, normally you're trying to look for three to four keyword phrases.
• 7:43
  Sometimes you do longer tail, but the longer tail is just kind of very, very, you know, you do a gazillion longer tails to try to get, you know, try to aggregate them together.
• 7:52
  When ChatGPT is doing a search, and Google and these other ones, they are often doing much longer keywords than humans do.
• 8:00
  So human types in three keywords, like best CRM, that's it, two or three keywords.
• 8:05
  OpenAI may do best CRM, small business, you know, pest control agency, Oklahoma, or whatever.
• 8:13
  It may include 10, 12 keywords in its phrasing.
• 8:18
  If you can understand what keywords it is constantly using and aggregating, you can actually build content targeted towards that long tail.
• 8:25
  And the difference is in, you know, compared to SEO when you normally do long tail, maybe each long tail is a tiny fraction of the searches.
• 8:34
  Well, with these AI searches, you can actually start figuring out which long tails are being called a lot more often.
• 9:32
  And so if you can start optimizing for those specific long tails, you can have an advantage.
• 9:36
  So that's one example.
• 9:37
  I can keep going if you want.
• 9:40
  So one of the things, and I didn't mention this before we started recording, when I'm doing these interviews and talking with people, I like for the guests to think, I don't know anything about this, to explain it like I'm five.
• 9:52
  So when we're writing content around these long tail keywords, how often are we using them within the content?
• 10:02
  So for regular SEO, the recommendation is like five or six, depending on how long your post is.
• 10:09
  What's the recommendation for this?
• 10:12
  Yeah, so the goal on these keywords is to appear in the Google or Bing search results in the top 10.
• 10:18
  So you're gonna follow the same rules that you would follow otherwise.
• 10:22
  You don't wanna oversaturate it.
• 10:24
  Obviously, Bing behaves a little different than Google.
• 10:27
  So you're gonna wanna kind of, another one thing we've noticed is you wanna put a little more emphasis on Bing than you normally would.
• 10:32
  But you wanna follow the same rules on the SEO.
• 10:35
  It's just the keywords are going to be, the focus of what keywords to focus on is going to be slightly different.
• 10:41
  And that doesn't mean you shouldn't do the other one too to get in regular SEO.
• 10:45
  It just means if you wanna get that extra boost in AI search, this is a way to take advantage of it.
• 10:51
  I'm wondering how that fits into overall SEO strategies.
• 10:56
  Like do we just mix them all in?
• 10:57
  Do we create posts that are like more focused on that long tail, very, very specific?
• 11:02
  Do we do more general SEO on another post or is there a happy medium for all of that?
• 11:09
  So I think we're still early enough that we're figuring this all out to be quite candid.
• 11:13
  The way I have been doing it for our companies is obviously you want some things that appear for regular search because regular search is not going away right now.
• 11:23
  It's still the majority of people using it.
• 11:25
  And then also on top of that, create posts that are for these specific long tails, just like you normally would have a long tail.
• 11:31
  You have a portfolio of search, both long tail and short tail.
• 11:35
  The difference is the specific long tail change, what the actual keywords of these specific long tails change in my opinion.
• 11:43
  So now you're gonna put a little more emphasis on these specific long tail searches because the AI tends to canonically use these.
• 11:51
  The AI tends to use these words to describe things and tends to do these type of keyword searches again and again.
• 11:57
  And so if you know what those keywords that the AI uses to describe your business, then you should take advantage of it.
• 12:05
  Yeah.
• 12:06
  Okay, so I was talking with Usama Khan earlier.
• 12:10
  It was last week, but I think the episode went out already.
• 12:12
  But anyway, I mean, the episode did by the time this goes out.
• 12:16
  So what kinds of content formats or structures tend to perform better in these AI generated results when we're thinking about these keywords?
• 12:25
  Yeah, so you're gonna hear a lot of different people saying different things and it's probably changing.
• 12:31
  Obviously having good technical SEO structure matters.
• 12:35
  So having schema, having things easy to read, having a very good title and description because that tends to be skewed towards that.
• 12:43
  Having FAQs, people have said that that works and both in the schema and on the content.
• 12:49
  Two things that I would say that also matter is when you're getting these real-time searches, as we did a study on this, none of them load JavaScript.
• 13:01
  So if your content is being loaded by an API or React component, or embedded testimonials that are done from a third party, that will not be read in the real-time search results.
• 13:12
  So when they are reading your, there's two kind of avenues of when your content is read in terms of AI search.
• 13:18
  First is it's read when it's getting indexed by Google.
• 13:21
  And in that case, you will get JavaScript read.
• 13:23
  The second is when the AI is trying to read your content like a user to summarize it and give it back to you.
• 13:30
  In that case, the JavaScript is not read.
• 13:33
  And so any content that you're using from third parties or that is embedded in there will not be read by the AI.
• 13:39
  And so that's a difference and something that you wanna think about is all your content should be statically there.
• 13:44
  And that's different than what we've heard from Google SEO where it reads your content.
• 13:49
  And it makes sense because in Google, they have all the time in the world to index your page.
• 13:54
  So they can do it behind the scenes and render the JavaScript.
• 13:57
  When you're typing in a conversation, you want the response as fast as possible.
• 14:02
  And so you don't want OpenAI to go and open up a page and render the JavaScript and wait for the JavaScript to load.
• 14:09
  Instead, it will just read the HTML, which is much faster.
• 14:12
  So that's a different way.
• 14:15
  And yeah, we'll go from there.
• 14:17
  Yeah, that's really interesting because I've been using, and maybe I should do this with a checker, ChatDBT to do like the schema for my FAQs.
• 14:27
  And it always spits it out in HTML.
• 14:29
  But at one point, I think it was giving me the option for Java too, like the JavaScript.
• 14:33
  And I just chose HTML because that was so much easier just copy and paste it in there.
• 14:38
  So that's actually really good to know.
• 14:41
  Yeah, so I mean, if the actual content, I don't know if it parses the JavaScript if it's embedded on that page, but if it's not loading any JavaScript files, we found that, and it's not loading any images.
• 14:53
  So if you also have content that you think could be read by the AI from it, it's not doing that right now for the AI search.
• 15:00
  Which is very interesting as well with the images because a lot of marketers are, put more images on that page, make the infographics.
• 15:09
  For my own website, I don't do that, mostly just out of time.
• 15:12
  So that makes me feel a little bit better.
• 15:17
  So I've also been talking recently with a couple of different guests about author credibility, backlink strategies, that kind of thing.
• 15:27
  So SEO has always been about ranking pages.
• 15:30
  AO, AIBO seems about surfacing facts, authors and sources.
• 15:35
  So how should content teams think differently about this in terms of visibility and getting into those searches?
• 15:43
  I mean, EAT is still very important.
• 15:45
  And authority is important and all those aspects.
• 15:49
  I don't know, it's hard to know and it's all dependent on each which model it is being used.
• 15:55
  I know Google behaves differently than say, Chat GPT on what it's using for sources and how it's doing that.
• 16:01
  And it also depends if you care about being mentioned as a citation versus having your brand mentioned.
• 16:07
  Because those are different things as well.
• 16:09
  Can you explain that just a little more so we're all clear on what you mean there?
• 16:14
  Yeah, so when you get a result back, there's two things, two ways.
• 16:18
  Well, there's actually many ways your product can be a result and come back.
• 16:22
  But the two kinds of ways I think about it is either your product is mentioned in the actual paragraphs or there's a little link that goes to your site.
• 16:31
  I think too much time has been spent on trying to get your site to be the little link and not enough time is being spent to getting your brand mentioned in the results.
• 16:42
  And our hypothesis is very, very few people actually click on the citation link, maybe some.
• 16:48
  A lot of people in the industry will, but people outside the industry are gonna want an answer.
• 16:53
  They don't want to go to a website.
• 16:55
  And so it's just something to think about is you wanna, in our opinion, you wanna optimize to get your brand mentioned versus having a little citation.
• 17:04
  And what we're also seeing is, at least in Chat GPT when we did our research for SaaS, 85% of brands that are mentioned do not actually have links.
• 17:12
  Only 15% have prominent links.
• 17:14
  So when the brand is being mentioned, what's also happening is people are copying that brand name into Google and then doing a brand search for that brand name and then clicking through.
• 17:26
  And one thing to just be aware of, and this is a whole other study thing that we did, which is Google Analytics in particular will not show that as coming from Chat GPT.
• 17:35
  They're gonna show that as coming from Google.
• 17:36
  So you're gonna see a lot of Chat GPT behavior being masked as Google.
• 17:41
  As Google.
• 17:42
  Also very important to recognize.
• 17:45
  Very important.
• 17:46
  Yeah, and on top of that, I'll add mobile.
• 17:48
  If you do mobile and there is a link, it's not being treated as coming from Chat GPT.
• 17:52
  It's being treated as no reference.
• 17:53
  So you also have, it's just very little, it's really hard to know where the traffic is coming from from Chat GPT when it's coming there and you shouldn't be trusting Google Analytics for that.
• 18:02
  So it makes tracking that much more difficult.
• 18:07
  I don't wanna say already is, but it can be a little bit hairy if not paying attention.
• 18:14
  Okay, so all of this to say, and like talking about user behavior, both using Google Chat GPT and then also stepping out to people outside of marketing or outside of SaaS.
• 18:31
  Do you think optimizing for AIVO, AEO, whichever one you wanna call it, is absolutely necessary at this point?
• 18:41
  I know we're still very early in the game.
• 18:45
  So like, okay, I will say it this way.
• 18:47
  Isn't absolutely, no.
• 18:49
  But you know, advertising on Facebook and Instagram is not absolutely necessary.
• 18:54
  So what I do think is it is a opportunity.
• 18:57
  It is a very much what we have seen is growing usage in that space.
• 19:02
  It's still a minority of use cases.
• 19:04
  It's not the majority of people doing it.
• 19:06
  We are seeing Google switch over a lot on their UI from doing answers that are just the 10, well, the six paid links and two, three blue links to having this answer engine.
• 19:18
  And they've really started to be pushing over a full chat interface, Google AI mode.
• 19:26
  You've seen more and more people switching their behavior of search to open AI, but it's still the minority.
• 19:32
  So the way I see it is two things.
• 19:35
  One is not a lot of people are thinking about this space in the real world.
• 19:39
  A lot of people in the SEO world think about it, but not a lot of people in the real world.
• 19:44
  And so it's an opportunity to get ahead.
• 19:46
  It's kind of like the early SEO days where you can get ahead and get a beachhead and take advantage of it before other people know about it.
• 19:54
  So that's great.
• 19:56
  And also it's good to be versed in this because at least in my belief, this isn't going away.
• 20:01
  It's not a fad.
• 20:02
  More and more people are gonna be using this.
• 20:04
  I personally believe this is the way in the future that people are gonna do most of their searches because it's just a better user experience.
• 20:12
  And so it's better to become an expert and to learn and start learning about this now than wait and be a laggard.
• 20:20
  So does it mean you should throw away your SEO?
• 20:22
  Heck no.
• 20:26
  You should not be throwing away your SEO, but you should start becoming an expert in this because if you believe this is going to be a new way, if you start using it yourself to do searches, then it's something to think about that if you're using it and your friends are using it, it's probably something that's going to become the norm.
• 20:46
  I agreed.
• 20:47
  I agree with you there.
• 20:48
  So how can smaller teams start experimenting with this instead of just using it as a tool instead of overhauling their entire strategy?
• 20:55
  What do you recommend?
• 20:56
  Yeah, I mean, well, now I'm gonna, you know, do it with the caveat of like, this is what our company is about.
• 21:03
  So our company RivalC lets you do this, right?
• 21:06
  So you can use an AI visibility tracker like RivalC, which would actually take your website, figure out your customer segments, figure out the individual persona of who you are, of the ideal customers, what questions or comments they'd be chatting with.
• 21:22
  And then see how you're ranking.
• 21:25
  And then RivalC also shows you which keywords that the AIs are using when they're calling out to Google and Bing and lets you create content.
• 21:32
  It actually generates content for you.
• 21:34
  So the first thing I would tell anyone who is starting on this, start tracking.
• 21:39
  None of the trackers are perfect.
• 21:40
  None of the trackers know exactly what queries are being done, but you can start getting some signal.
• 21:45
  Are you even appearing?
• 21:46
  For which engines are you appearing and which ones are you not?
• 21:49
  And what are the type of searches that those engines are calling out?
• 21:54
  And you can start making small changes that can have a big impact.
• 21:59
  We've seen people increase the amount of exposure, double, triple the amount of exposure that they've had from AI visibility just by making small changes, both in terms of obviously structural SEO, but also just adding things specific for AI visibility, like going after longer tail, doing the other one I didn't mention that we can talk about is bottom of the funnel tends to the AI tend to do brand mentions much more from bottom of the funnel content than top of the funnel content.
• 22:28
  My personal belief is top of the funnel content is gonna have a big headwind because the information gets eaten up by the models and then just summarized without you being attributed.
• 22:38
  So the bottom, site comparisons, product comparisons, product lists, those things are getting read right now and being used to help people consider purchases.
• 22:52
  Yeah, which goes back to what Usama and I were talking about with that bottom of the funnel content would be super important these days.
• 22:59
  So last question, if you were building a new strategy from scratch, would you still start with SEO or would you go straight to like AIVO, AEO strategy, focus less on SEO, what would you do?
• 23:14
  So I would say you definitely still need to start with an SEO strategy.
• 23:20
  Like we built an AI visibility tool, which is awesome, but we're still saying start with the foundations of SEO that's needed.
• 23:29
  If you're not appearing in the top results on Google or Bing top 10 results, you're not gonna be in the messages.
• 23:36
  So start there, you can go a long way with just building a solid foundation there.
• 23:41
  On top of that, I would say you want to dedicate some percentage of your resources and time and mind share into this new growing area.
• 23:51
  And so, getting a tool, getting a visibility tool is a great start, like RavelC or others.
• 23:57
  Getting, starting to understand how these engines are talking is really important and starting to try to boost and experiment because you can, over time, you'll have a headstart over everybody else on that.
• 24:10
  So here's your challenge for the week.
• 24:12
  Open ChatGPT, Perplexi, Google's AI mode, whichever one that you're using, and type in a question your ideal customer would ask.
• 24:20
  Does your brand show up?
• 24:22
  If not, you've got some work to do.
• 24:25
  Big thanks to Charlie Graham from RavelC for breaking down the science behind AI visibility.
• 24:30
  As always, if you need help with your content strategy, connect with me on LinkedIn or visit my website at cassieclarkmarketing.com.
• 24:38
  See you next episode.