
Show Notes
#269 AI Search & SEO | In this episode, Dave is joined by Andrei Țiț, Head of Product Marketing at Ahrefs, a leading SEO tool trusted by marketers around the world. Andrei has been on the front lines of how AI is reshaping search and what that means for marketers trying to stay visible in an AI-first world.
Dave and Andrei cover:
- Why branded search volume is now a top indicator of visibility in AI-generated results (and how to grow it)
- The new playbook for SEO in 2025, including what metrics to track beyond traffic and backlinks
- Actionable tactics to get your brand mentioned by AI search tools like ChatGPT, Perplexity, and Google’s AI Overviews
Whether you’re a marketing leader or an SEO newbie, this episode will help you rethink your approach to content, attribution, and brand in the AI era.
Timestamps
- (00:00) - – Intro
- (02:48) - – Why this was Exit Five’s most-registered webinar ever
- (05:08) - – Meet Andrei from Ahrefs
- (06:48) - – Is SEO dead? Not quite, but it’s harder than ever
- (08:28) - – AI traffic is growing fast (63% of sites already see it)
- (09:18) - – Why brand is your best SEO defense
- (11:48) - – How Google measures brand impact (keywords, mentions, clicks)
- (14:48) - – Calculating content ROI with traffic value
- (17:18) - – Why branded search volume predicts AI visibility
- (21:18) - – How to track brand mentions in ChatGPT, Perplexity & AI Overviews
- (24:18) - – AI traffic = fewer clicks, better leads
- (29:23) - – How to improve your visibility in AI-generated answers
- (36:23) - – Why backlinks matter less and PR matters more
- (43:23) - – New rules for writing content LLMs can surface
- (52:23) - – Final takeaways: metrics to watch and content to prioritize
Send guest pitches and ideas to hi@exitfive.com
Join the Exit Five Newsletter here: https://www.exitfive.com/newsletter
Check out the Exit Five job board: https://jobs.exitfive.com/
Become an Exit Five member: https://community.exitfive.com/checkout/exit-five-membership
***
Today’s episode is brought to you by Walnut.
Why are we pouring all this effort into marketing just to push buyers to a “request a demo” or “contact sales” button?
Come on, today’s buyers don’t want to talk to sales right away. They want to explore your product themselves, see how it works, and understand its value before booking a meeting.
That’s where Walnut comes in.
Walnut empowers marketers and GTM teams to create interactive, self-guided product experiences in minutes. Embed these experiences on your site, in emails, or anywhere in your funnel to let buyers engage on their terms, from awareness to close and beyond.
That's the beauty of Walnut - you're getting a platform that your sales and CS colleagues can use to showcase the product too.
And the best part? You get real intent data—see which features prospects love, where they drop off, and what’s actually driving pipeline. Demo Qualified Leads are the new MQL.
Over 500 companies, like Adobe and NetApp, use Walnut to drive 2-3x higher website conversion rates and 7 figures in pipeline on a yearly basis. So do you want to drive more leads, shorten sales cycles, and actually show your product instead of hiding it behind another typical B2B CTA?
Go check out Walnut.io. And if you tell them Dave from Exit 5 sent you, they'll build out your first demo for free!
Transcription
Dave Gerhardt [00:00:00]:
You're listening to B2B marketing with me, Dave Gerhardt. All right. Hey, everybody. Bonjour. Dave here. I'm the founder of Exit Five. We got an awesome session for you. This is the most popular.
Dave Gerhardt [00:00:27]:
Don't call it a webinar live session we've. We've ever done. We haven't even done it yet, but we almost have 2,000 people that registered for this session. And I was just talking with Andrei, who runs product marketing at Ahrefs. He'll be on in a second. I was saying what's funny is, like, as marketers, which you all are in here, we bang our heads against the wall trying to do all these creative things for promos and then to get butts and seats for something. And then when you have this huge conversation and topic and issue, which is AI and search, and then you match the content and the right audience. And we sent two emails and we have 2000 people or something, 1800 people registered for this.
Dave Gerhardt [00:01:04]:
There's like 300 people here already live, which is amazing. So I'm super excited for this session. This is truly one where, like, I'm trying to learn as much as I can. We've seen a bunch of discussions in Exit Five about this. I see it on LinkedIn all the time. Andrei from Ahrefs, they have an amazing tool that you can use to do this, but they have a ton of data and using it for myself, and it's. It's really interesting. Even in our own business.
Dave Gerhardt [00:01:25]:
Over the last couple months, I just looked. We had 115 AI searches like that. Exit Five showed up in the results for, which is more than we had three months ago, and there's something really interesting there. So this is going to be a really informative session. You're going to learn a lot. At least you'll come out of here with more knowledge. I can't promise you that this is going to immediately solve your AI and SEO strategy overnight, but I think you're going to leave this session with the right things to think about and the right questions to be asking. So before we bring up Andrei real quick, I'm just going to tell you how this goes down.
Dave Gerhardt [00:01:54]:
So sometimes we just do a conversation. Sometimes we have some slides. Andrei has some slides and some awesome data and real facts and learning that they have. We're going to show you that. We're going to take all your questions, do lots of Q and A. We want to help you out. Obviously, it's all marketers here, so we want to help as many people as we can with their AI and SEO questions. This will be recorded.
Dave Gerhardt [00:02:12]:
We run this as a podcast episode later, if you want to re listen, we'll send out the recording to you if you're here. But if you're in the chat, you're here for a reason, right? You took time out of your day, and so I want you to pop in the chat right now. There's already people shouting out from everywhere from San Diego. Grace is in Pittsburgh. Sarah's in Charleston, South Carolina. Anita can't hear me, but other people can hear me. So I think this is working. We got somebody here from Sri Lanka, Seattle, Washington.
Dave Gerhardt [00:02:37]:
I'm assuming y' all can hear me. Right. But first, before we bring up Andrei, I want you to put in the chat real quick. Why did you come here? What do you want to learn today? What is it about AI and SEO that is that you need to learn about? Why spend an hour of your day with us? Put that in the chat real quick, and then throughout the the rest of the session, help each other out. Put your comments and questions in there, and then we'll take all the questions in the Q and A. Matthew wants to get ahead of the curve. Destiny says AI is changing everything. We need to say current.
Dave Gerhardt [00:03:03]:
I hear you. AEO tips. I love that. What kind of content we should be. Oh, that message already went away. We got our first lead via Chat gbt. And we want to say ahead of Rapid Developments, AKA we got our first lead from Chat gbt. It was good.
Dave Gerhardt [00:03:14]:
How do we get more? I feel that. So, Allison, why don't we bring Andrei up here to say hello and we'll kick things off.
Andrei Țiț [00:03:19]:
Hey, folks.
Dave Gerhardt [00:03:21]:
Hello. There he is. Andrei, look at this response. Man, this is amazing. This is amazing. I know you're a pro. I hope you're excited to be with. It's great to have you.
Dave Gerhardt [00:03:29]:
It's Andrei. He runs product marketing at Ahrefs, and he's going to be our presenter today. Andrei, I'll let you introduce yourself, say hello, and we're off to the races.
Andrei Țiț [00:03:37]:
Sure. Thanks, David, for having me. And, yeah, I've never seen such an engaged audience. Everybody's, of course, trying to optimize for AI search SEO optimization. Now it's G O L M O. But I'm here to kind of, like, more or less light a torch. And it's quite the new, I would say, industry for us as well. We're at the forefront of SEO.
Andrei Țiț [00:03:59]:
We are at the forefront of AI search. So we have the data necessary to point you in the right direction. And I hope you're going to learn a lot. Hello, everyone, My name is Andrei. I'm a product marketer and I lead the product marketing department here at ahrefs. And welcome to the CMOS Guide to Winning in AI Search with Ahrefs. And honestly speaking, it's not about cmos, just cmos, but also about marketers. And I know you have a lot of questions, so I actually cut a part of my presentation so that we can have a longer Q and A session.
Dave Gerhardt [00:04:31]:
You said it's not about CMOs. I have a private channel with a bunch of CMOs, and I can tell you that this is top of mind. Everyone from, from the marketing intern to the CMO is being asked. I just met with a CMO who was. She was out for like eight weeks and just kind of diving back into the business. And the very first thing she's tackling is like, oh my gosh, all this stuff has changed. What does this mean in this world? So I think it's the, it's why the response for today is so, so strong is that everybody's wondering these things.
Andrei Țiț [00:04:57]:
Yeah, we made it like top level presentation, but then we figured out like, hey, everybody's question at the moment, like you stressed out. So that's why I'm saying don't get swayed from the title itself. It's for all marketers right now who want to get into a new traffic channel. And that's AI search. So without further ado, let's talk about organic search. So we're going to talk about how things are at the moment. And to be honest, they're not so cool. Somebody asked, is SEO dead? Well, we're going to see about that.
Andrei Țiț [00:05:26]:
It depends. But to be honest, organic search is harder than ever. There's zero content, so that kind of content that answered the searcher's query directly on the serp. If we check here, this is a research from Rand Fishkin. I'm sure pretty much everybody knows Rand Fishkin. So According to it, 60% of the searches in the US are zero click searches. Approximately 60% as well in Europe. On top of that, we have Google, who throws at us frequent Google algorithm updates.
Andrei Țiț [00:05:56]:
These are kind of like curveballs. So you either adapt or die. Just scroll through it. It's like one what, how many Google algorithm updates we have here? We have a core update, a spam update, a Google search ranking bug. So they also make mistakes. Be sure about that. And we also have AI traffic as a slowly but surely emerging new channel. So based on our study, we analyzed 3,000 sites.
Andrei Țiț [00:06:21]:
In this study alone, we have more studies like this one. And for larger data sets, 63% of the sites received at least 21 customer from an AI traffic referral. And this is about as much as Reddit. And Reddit traffic just picked up last November. So it's still growing and you'll see why. So given all this, the question begs, what's the mode to all these kind of changes now? The truth is there's always been so many marketing channels that are relevant, but not so many that are relevant at the same time. Time. And even us at ahrefs, we have our product marketing team and we're always asking, like, how can we be top of the mind for our customers? How can we meet them in their preferred channel? Something that we call fire all cannons at once.
Andrei Țiț [00:07:09]:
If somebody's curious about product marketing, just fire me a DM in LinkedIn. But anyway, what's the single constant that can withstand all the changes in this search landscape that's forever changing? And we ask ourselves that question. And the question is quite simple, to be honest, and it's under our noses. It's brand awareness. So ultimately, at the end of the day, a brand is what people are thinking, talking and searching about. And a brand also has multiple benefits. So on one hand, for example, it strengthens your SEO strategy. It also lowers your cost of acquisition across all different channels that you might have, and it also increases the likelihood of showing up in those AI generated summaries.
Andrei Țiț [00:07:53]:
And if you do it right, those benefits compound. Why? Because brand awareness and what more or less entails about brand awareness. So I'm talking about trust recognition, brand affinity. They will most likely protect your brand and bring more profits in the long run. So what I'm going to do in this presentation is. One second. Yeah, I have three topics, but I'm just going to focus on the first two. So I'm going to show you how SEO is changing and how not to focus anymore on traffic, but focus on the right metrics.
Andrei Țiț [00:08:24]:
And I've outlined a few brand awareness metrics. And I'm also going to talk about brand visibility in the AI search chatbots. Now, remember, this is quite new, but we're at the forefront. We're ahrefs, so we have some cool data to show you.
Dave Gerhardt [00:08:37]:
I love the focus on brand, by the way, because this is what's so interesting. I was thinking about this the other day. Like you search something, Google just gives you the answer. People are not dumb. We have Pranav, who's the founder of Power Market. He's been on this webinar. He's like, your customers are not dumb, right? If you provide them value and they end up searching for you, like down the road, you're going to find out how they heard about you. It's just like, it's the internal challenge as marketers of like, if these AI search engines are now like eating our traffic and they're not giving us the credit, that's where it all breaks down internally.
Dave Gerhardt [00:09:08]:
It's like people still got the answers. They're just, we're not able to track them along the way. So I, I love the focus on brand. I think this is going to be a huge topic as we get deeper in AI. Like, let's bring brand back. I love that. All right, keep going.
Andrei Țiț [00:09:20]:
We should make some caps. Dbb. Bring brand back.
Dave Gerhardt [00:09:24]:
Absolutely.
Andrei Țiț [00:09:25]:
Okay, awesome. So let's focus on the first topic, brand awareness metrics. And to be honest, brand has always been this kind of like fuzzy, hard to measure channel. On one hand you have a few brand outcomes like brand sentiment, reputation, affinity and so forth. Those are things that are quite subjective in matter. So if you ask people like, hey, do you recall this kind of brand? It's a subjective answer. On the other hand, you also have some types of channels. So this kind of branded content can exist on other channels that you do not have control over.
Andrei Țiț [00:09:58]:
So you don't have data and analytics. But if done right, you can actually reap a lot of benefits. And in our research we found out that Google assigns site a site quality score based on brand metrics. So we have your website and Google is looking at those three metrics. Brand keywords, brand mentions, those are rich brand anchor texts and also website clicks. Based on those three, it assigns a site quality score. And the higher it is, the higher are your chances to be recollected in Google. This is traditional organic search.
Andrei Țiț [00:10:32]:
Let's see how it ties to AI search in a few minutes. So based on that, I want to show you a few of the brand awareness metrics we found in our own study. And the first one is organic share of voice. Now this is quite important because it shows you what's the percentage of your organic keywords, organic traffic amongst your whole niche. So in other words, it tells you what's the popularity of your brand across your niche. Now I've listed here the definition and also the formula, but as I said, it's quite boring. Just remember that share of voice is how popular your brand is in your category or industry. And that's what it matters at the end of the day if you're using Ahrefs, because I assume there's some people who use it.
Andrei Țiț [00:11:14]:
You can pop your own website into our site explorer tool. That's our competitive intelligence tool. That's what I did here. I popped in a website and also three competitors. I'm a runner, so I went for Holca on cloud, Asics and Nike and I simply clicked on share of voice and I can see what's the share of voice across that niche. So at the end of the day, what you want to do as a marketer, not necessarily as a cmo. I cannot do this here because it's not app ui. I would probably glide over the chart and see who's winning and who's losing.
Andrei Țiț [00:11:49]:
Because if somebody is dipping, then most likely somebody else has taken over. It's not just Google algorithm updates, those that I've mentioned in the beginning, but it's probably somebody, a competitor most likely that's doing a better job than you. And this is like a top level metric to see what's happening and then you just go on into the weeds and check out what's happening.
Dave Gerhardt [00:12:09]:
This would like share a voice as it relates to a handful of competitors that you pick. Is that more of a useful metric than like for us, for example, the company is called Exit Five. If I look that up in, in Google or Ahrefs or whatever, I often see a lot of searches for Exit Five that actually don't relate to the company. So the better exercise would be like who are the three or four perceived competitors and how are we showing up in search compared to them?
Andrei Țiț [00:12:33]:
Yep. As long as you know your competitors you can see and your niche in general, you can see what organic keywords, this is more or less the share of organic keywords and the traffic it brings, the traffic it's brought by those keywords. And yeah, it's way to measure, as I said, the popularity of a brand in a particular niche. This is just one of them. I have three more. The next one is more or less closely tied to this one and it's called organic share of traffic value. And what's traffic value? Probably asking. So I'm going to explain that Traffic value.
Andrei Țiț [00:13:04]:
We thought of, okay, measuring content roi. It's quite hard. There are several methods to do that, but we're like, okay, what if we show people how much money you've saved if that content were to be paid for in Google Ads? So that's what traffic value shows you. What's the value of my organic content in Google if it were to be advertised via PPC in Google Ads. So a meta extension to that is the organic share of traffic value, similar to share of voice, but it shows you the value of your brand relative to your other competitors. And what I did here was play around again with the same tool, but this time I changed the metric to share of traffic value and I got here some dollar values. So hawkeye.com for example, their organic traffic, if it were to be advertised on Google Ads, it would cost $1.6 million. Compare that to Nike, that's $18.7 million.
Andrei Țiț [00:14:01]:
It kind of helps you as a content marketer and also as a marketer in general to evaluate what's the value of your organic content right now in Google. And if you have a CFO at the end of the day who's like, hard pressing, you're like, hey, what's the ROI of your content? And this is also something nice to show them without having to scramble a bunch of Excels.
Dave Gerhardt [00:14:21]:
This is like the ultimate answer for like, what's the ROI of content? Why are we writing all these articles?
Andrei Țiț [00:14:28]:
Yeah, exactly. As I said, there are several methods to measure that, but we felt like, okay, let's rethink the problem. Like, how much money would actually pay for this content to be advertising Google and reach at the height that you're currently at. So, yeah, those two things are quite important when it comes to this new brand world, as I said, because it gives you an idea about, okay, where's my brand position in Google search? We're still talking about Google search. And also, what's the value of my content if we were to buy ads in Google? All right, we covered those. And the next one is branded keyword search volume. This will tie directly to AI search. I'm going to answer in some of the questions later on, why? Because basically this shows the search demand for your business.
Andrei Țiț [00:15:11]:
If this is not growing, your brand is more or less dying. So there's no pipeline, there's no revenue, there's no profits. So essentially, you want to see this branded keyword search volume growing over a sustained period of time. And to illustrate even better, we actually conducted our own study and found out, like, hey, more like 50%, almost 50% of searches in Google are actually branded. So this is even a better reason to track it. And we also looked at other studies. So this is from Kevin Indig. I'm assuming most of you who work in SEO know him.
Andrei Țiț [00:15:43]:
He has this substack newsletter, growth memo and he actually found out that branded search volume is the biggest predictor of mentions in ChatGPT. More on that later. If you want to check what's the global search volume for your own brand, just go to Ahrefs Keywords Explorer, that's our standard keywords explorer tool. And type in your brand followed by a few misspellings. So first it's Ahrefs whatever. However, you might misspell it in your own language. Then go to the matching terms report. This will extend, this will expand the base of initial seed keywords and then look at global search volume.
Andrei Țiț [00:16:17]:
So for us, that's 1.7 million the volume in itself for the keywords that it brings. And there's also a metric, it's called growth forecast. For us, it's always growing, so we know that we're doing a great job. This is more like a north metric star in terms of, okay, are people thinking about us? Are they remembering our brand when it comes to SEO marketing in general? So that's the kind of thinking you should have as well. Are people remembering our brand in the long run in the past 12 months and future 12 months? If yes, you should see some positive deltas. If not, then you should probably ask yourself some questions. But this kind of metric helps you focus on this and not just necessarily only on traffic. The next one, and this is the last one, it's called branded Keywords Traffic.
Andrei Țiț [00:17:01]:
It's quite simple before I showed you that, okay, you can check the effectiveness of your brand awareness campaigns. How much search volume do I bring? But actually how much traffic is converted? Like we're talking about traffic at the end of the day, branded keywords do exactly that. And we have a very easy widget so you can see a breakdown of branded versus non branded organic traffic. And there's some deltas here, quite negative in our case, probably because of AI overviews. But I did this presentation, I created the slides like three or four months ago. So I'm sure these things have changed. The way to look at it is with the mindset of again, are my branded keywords and the traffic coming from them increasing? If not, I should probably make some due diligence and see why. Now, of course I didn't talk about all those brand awareness metrics that you should track.
Andrei Țiț [00:17:48]:
My colleague here, Luis, she wrote a very interesting article on it. I didn't talk, for example, about social media traffic or direct traffic, people remembering your website and typing in your main domain. So if you want to check it out, then you can scan the QR Code and do it on your own.
Dave Gerhardt [00:18:04]:
One quick follow up on that. By the way, someone in the chat is asking, where's the Q and A? At the top of this chat box there's a Q and A tab. Just if you put it in there, then we sort by like most upvoted later. And then someone asked in the chat, just quickly, Steven said, did I hear you correctly that brand searches are the highest correlation relation to showing up in.
Andrei Țiț [00:18:21]:
ChatGPT brand search volume, which ties directly to brand searches? Yes.
Dave Gerhardt [00:18:26]:
All right, keep going, sir.
Andrei Țiț [00:18:27]:
Yep, no worry. We're on to part number two. We have a lot of questions so I'm going to keep it short. I think you're going to get much more value from those questions being answered live. But I also want to prep the terrain for that. So we're on to our next part which is brand visibility and AI search chatbots. And before going into that, I actually want to make it more engaging. So I'm curious how many of you are already tracking AI mentions? Just a simple yes and no.
Andrei Țiț [00:18:53]:
I think there's a poll should be on the screen. Yeah. Do you already track brand mentions in AI search? So probably go to the polls tab and say yes. No, I'm just curious. A lot of no's, a lot of nos. Wow. I thought people would be more into it, especially because it's mainly a US based audience. But yeah, a lot of people are not tracking it.
Andrei Țiț [00:19:16]:
Wow, interesting. So out of those who are tracking it, the 31st 4%, again, I'm curious, chat GPT perplexity, Google a overviews. What's your poison?
Dave Gerhardt [00:19:27]:
Yeah, so I recently did this for us for Exit Five and we, we just started tracking AI mentions, AI search, same thing here. It's basically been like 95% ChatGPT and 1% perplexity, but the last two months that I've looked at it, it's more every month and I, we're definitely crossing the, the chasm here because even somebody like my wife doesn't work in tech. We all work in tech. Obviously we're on the early edge of this. Like she's now using ChatGPT for her basically Google replacement. Right. And there's, there's certain things you go to Google for, but a lot of that search is just consumer beat. Attention has shifted to use ChatGPT we're seeing in our business and you know, I haven't seen much in the other sources.
Dave Gerhardt [00:20:09]:
So this is interesting to match this up with everybody.
Andrei Țiț [00:20:11]:
Cool. Yeah, I'm happy to see Copilot Here, I mean, the top three I'm going to talk about them are obviously ChatGPT, Google AI overviews and perplexity. Claude. Quite useful for content writing. I use it myself, but also for coding Copilot, probably some people who work at big orgs and they're forced to use it, but I'm happy to see Copilot among those as well. All right, I think, Alison, we can go back to the presentation. Thank you. So you might think like, okay, Google is actually losing some market share in favor of those AI chatbots.
Andrei Țiț [00:20:42]:
Right? And that's why they actually released AI overviews. Well, not quite. So Google is not necessarily losing market share as you might think. In fact, three months ago, when I created these slides, the top three AI chatbots, which are German Perplexity and Google AI Overviews and ChatGPT were bringing in 345 times less traffic than Google alone. Right now, ChatGPT is bringing 100x less than Google alone. So probably by the end of the year, if it goes in the same rhythm, it's going to be more or less on the same footing. But what I want to stress out here is the fact that AI search sends just 0.1% of total referral traffic. So right now it's quite small.
Andrei Țiț [00:21:25]:
And if you check it right next to email and paid, and when you compare it to search, it's probably, yeah, just dwarf. So my point here is that even though it's small, one does not simply use AI search primarily for clicks. And I'll tell you why. A search has its own advantages. For example, it increases brand awareness as we already talked about. And to be honest, brand awareness, trust and recognition are the byproducts of good content. Doesn't matter if that content is surfaced by an AI generated summary, a video or a blog post. It's also going to bring you more brand positive brand sentiment because again, people recall your brand and it's much easier to do business with you.
Andrei Țiț [00:22:06]:
And we've also noticed that these are more qualified traffic leads. I'll give you an example. For Ahrefs website, we have 0.5% of our traffic. Our traffic is around 10 million. If you check it in Ahrefs, 0.5% of 10 million is coming from AI search, but those people are responsible for 12% of the signups. So those people are ready to buy, but they also have other caveat. So probably they have higher bounce rate. They're not going to search so deeply across your website.
Andrei Țiț [00:22:35]:
So these are people who are ready to buy, but at the same time, you might lose them in a second if you don't feed them the right information. How to track brand mentions in AI search chatbots well, our answer to it, this is the AHREFS answer. It's called Brand Radar. It's an LLM visibility tool. And this is how it looks like. Essentially you can check your mentions in Google AI overviews, but we also can check them in ChatGPT and Perplexity. I need to update my slides because we already offered them. Microsoft Copilot is next and Claude as well.
Andrei Țiț [00:23:06]:
So anybody who answered all the other queries, I think there was 2% Microsoft pilot, 3% cloud. We're also on track for that. But essentially this is our way of covering this gap. And to be honest, everybody's more or less searching or developing an LLM visibility tool. You had the other guys that you mentioned, Dave, previously, you had them on the talk. But we are ahrefs, so we have pioneered SEO and we are still doing that for the AI search industry. How to show how it works, just going to show very quickly. I know you don't like product plugs and stuff like that, but this ties directly to what you've come here for.
Andrei Țiț [00:23:43]:
So in my case I've searched for a broader niche, basketball shoes. And then I have a few mentions here. So unlike other tools, you can search for brands. I did search for Nike. Then I had the AI suggest more competitors to my initial search. And by this way, I'm actually checking for these brands across this larger basketball shoe scope. But as I said, we're not like the other tools because we also let you to check products and or services if you have enough search volume for them across those AI chatbots. But coming back again, so I'm searching for nike's presence in AI chatbots across Google AI overviews.
Andrei Țiț [00:24:23]:
You can choose whatever LLM you want. So ChatGPT or perplexity in the basketball shoes niche. And what you'll get, I'm not sure you can see it here quite clearly, but you'll get the number of mentions impressions as well and all those competitive share and competitive reach. So you're knowing, okay, what's the flow? How am I comparing relative to my competitors and what should I do about it? The results, you see them here at lead issue industry and to the right you have the AI overview. In this case, what again can be chatgpt or Perplexity and how the AI generates that and features your brand. Now, cool bonus tip here. You can also track AI visibility gaps because that's what people are doing. And this is I think the lowest hanging fruit you can do.
Andrei Țiț [00:25:06]:
So what I did here is show me keywords that trigger AI overviews that contain Adidas. This could be your competitor, for example, but show me at the same time AI overviews that do not contain Nike. And I'm a big Nike fan. So Nike for the Quiri worst basketball shoes is not mentioned by any AI generated summary in Google. So I guess Nike is doing a great job at this. This could be again your brand. So you could plug in here all your competitors that are not necessarily that are showing up in AI overviews or whatever LLM of your choice and your brand that isn't so you can cover them. I think this is the lowest hanging fruit so far.
Andrei Țiț [00:25:45]:
I'm going to talk about that a bit more. It's a small product plug, don't want to bore you with it. How to get mentioned by AI cert chatbots it's quite new, so each LLM more or less has its own rules and they also have more or less a bias in terms of what kind of results they want to show. But we actually identified two things that are shared across all those chatbots. First one is don't block LLM crawlers. Blocking them means brand suicide. That's a no brainer. A better way to do is to actually direct them.
Andrei Țiț [00:26:16]:
What kind of content needs to be crawled in order for it to be pulled in AI generated summaries. So there's the robots txt file. There's also an LLM txt file. Right now to be honest, it's quite ignored by the LLMs. But we can talk about that in the Q and A. Be sure to mention there all the pages that you think are useful and hide all the private proprietary information that shouldn't be crawled. So pagination, for example, shouldn't be crawled because unlike search, you don't need to show them that. Look, we have so many pages about this topic.
Andrei Țiț [00:26:49]:
No, you just need to feed them the right information. The next thing that's also common across all those LLMs is the fact that you need to create deep, comprehensive content. So keyword stuffing SEO is dead in terms of that. Welcome topical coverage. This is a study that shows that, okay, out of those content related factors, which one is the most promising to surface your brand in ChatGPT and all the rest of the LLMs. And those that do that are word count, sentence count and domain rating. So the authority of your website.
Dave Gerhardt [00:27:21]:
Can you go back one slide for a sec on this one? Don't block LLM crawlers. Is this something that most people on their website like, would they be set up to block them right now and they should go back and change this after. Or is there a default here?
Andrei Țiț [00:27:35]:
There's no default. Your robots Txt file is the file that tells okay, what kind of crawlers can parse and cannot parse information. If somebody from your dev team probably blocked them by default or, I don't know, they were too afraid to have all their content scrapped. You're an art exhibition and you don't want your art to be scrapped or something like that. You might want to check it and say, okay, we allow those to scrap our website. So we show up in the AI generation.
Dave Gerhardt [00:28:00]:
Perfect. So, like, super actionable. Leaving this today would be like, make a note right now and go back and either talk to your developer or whoever runs or manages your website and make sure you're not doing this.
Andrei Țiț [00:28:10]:
Yeah. And also mention that the pages that you need to crawl and the ones you need to hide, because obviously everybody has private proprietary data. You don't want that to be surfaced by LLMs. So a smarter way is to actually direct the LLMs. What kind of information.
Dave Gerhardt [00:28:25]:
Okay. All right, keep going.
Andrei Țiț [00:28:26]:
No worries. As I said, the second most important thing that's shared across all those LLMs is the fact that you need to create deep, comprehensive content. So forget about keyword stuffing and focus more on creating content that answers the searcher's query, searcher's intent. We'll talk about that in the Q and A because this is just the simple stuff I'm going to talk about the actual science behind it. But we're back to the brand world. And as our cmo Tim Solo said, that's what we did at AHREFS for the past more or less 10 years. So we created great tools for marketing and SEO. We then published tons of content typing, tying Ahrefs to the different problems and outcomes marketers have.
Andrei Țiț [00:29:01]:
And then we educated and supported all of our customers so they could become themselves evangelists for our brand. So, yeah, whether it's called Geo LLM or A E O, I think some people are preferring more or less this latter term for now.
Dave Gerhardt [00:29:17]:
Somebody asked me the other day, I had a VC friend of mine messaged me and said, what's your opinion on geo? I'm like, I have no idea what's geo? He's like, dude, how do you run a marketing business and you don't know what GEO is? I'm like, I don't. I just haven't heard, oh, AI search, of course, people are talking about AI search, just more acronyms for us.
Andrei Țiț [00:29:36]:
All right, yeah, Generative experiences or something like that. But yeah, what I want to say is that all those signals that are kind of murky, we're still in a black hat world when it comes to AI search. They're still based on SEO signals. So if you want to read more about this, Ryan Law, content marketing director, wrote about it. You can scan the QR code and read it on your own. But yeah, I'm going to skip this part. Not that important for you. I'm more curious about the Q and A part because I have a lot of data in my head.
Andrei Țiț [00:30:06]:
It's locked here and I want to.
Dave Gerhardt [00:30:08]:
People are chomping, they're chomping at the bit. So I'm going to go to the Q and A for a second. I want to go find this question that I saw in chat. If I can't find it, I'll try to paraphrase, but the question was basically, okay, but what happens if people get these answers from the LLMs and then aren't they just going to go and plug in that URL directly in the site anyway? And that's not the AI and SEO point. That's more of like the why, how all of this attribution stuff is kind of going away in this AI search world.
Andrei Țiț [00:30:37]:
Yeah, speaking of attribution, those LLMs are kind of making it hard for you to track attribution. And we have a study. Patrick Stocks, our tech advisor, already did that and he took all the LLMs, the most popular ones, and checked their referral traffic. Claude is the only one who sends proper referral traffic. ChatGPT and perplexity, they do the same only when you access the web version, but they don't do it for the desktop version. And everybody's more or less having their own way of hiding the information, either being shown as direct or unknown traffic in your Google Analytics dashboard. We do track it natively nature. So we have tools called Web analytics, basically alternative to the traditional, the GA4 one.
Andrei Țiț [00:31:19]:
And you can see they're broken down by channel. But for now, just so you know, it's not you, it's LLMs. So some of them do parse referral, some of them don't. And those that do, as I said, are cloth and chatgpt and Perplexity, the web versions, but not the desktop ones.
Dave Gerhardt [00:31:35]:
Okay, I'm going to summarize a bunch of questions and I'm going to just ask you my I'll get to some of them, but I'm going to ask you kind of the overarching one, which is, all right, you're Andrei. You run product marketing at this company that has deep experience in SEO. Things are changing. You all are learning a lot about AI. Tell us right now, if I'm listening to this and I haven't changed my content strategy, my SEO strategy for this AI world, what are the one or two most actionable next steps that I need to be thinking about? I'm going to use this quarter to kind of like reevaluate our content effort to be relevant in this AI world. If you're my trusted marketing advisor, what would you tell me to do? What direction would you point me in?
Andrei Țiț [00:32:14]:
Awesome. Yeah, very good question. I would stress out two things. So first, that's how I started the presentation. We need to focus on few different metrics than we were used to in the SEO world. We did our own study and those metrics are branded search volume, branded anchor text, and also it was clicks in my presentation. But we found that it's branded traffic. And by that I mean that you need to check whether or not your brand popularity is increasing over time.
Andrei Țiț [00:32:41]:
And when it comes to branded anchor text, the good news is it doesn't need to be interlinked. So you don't need the hyperlink there. You don't need to pay for an expensive backlink. 300, 400 bucks. That's going to be useless anyway in time. So in the new AI search world, backlinks don't matter anymore. But branded co mentions or branded rich anchor text matters. So I'll give you an example.
Andrei Țiț [00:33:03]:
If we write content for Ahrefs and we mention somebody from our team, so I'll say team solo, our cmo. I'm not going to say that. I'm going to say Ahrefs CMO Tim Solo so that I give the LLM some context as to who is this person and how he is he tied to the brand in itself. So, number one, I would probably look for other metrics that I'm currently used with. And I told them it's branded search volume, branded rich anchor text clicks, but also branded traffic. Second, I think I need to be a bit more scientific. I need to explain how LLMs take information.
Dave Gerhardt [00:33:40]:
That's the thing I was going to ask you because it's like, okay, because I remember like coming up and I was reading the Ahrefs blog. What were these other tools? Semrush, Buzz, Sprout, all these tools. I used to use for the 2010 era of content. It was like it was guest blogging or like I'd want to basically get ahrefs to mention me in their blog or HubSpot to mention me. And it was like this game of backlinks. And the point of backlinks was like backlinks were a signal to tell Google that like, hey, you got a backlink from this like high authority domain that's valuable. Is that going away? If I'm a business and I don't have brand mentions, how do I go get placed in different places?
Andrei Țiț [00:34:18]:
To answer this question right now, how do you get those brand mentions again? We're going back 10 years ago how we did traditional marketing. You try to get more brand mentions through PR campaigns. You're trying to get mention in the niche publications in your industry. I don't know. You have a cycling e commerce store. Then be sure to get mentioned in all the cycling niche magazines websites that exist on the Internet. You also want to partner up with influencers. So that's going to cover the PR part in terms of what kind of content you should create.
Andrei Țiț [00:34:49]:
You should create user generated content. So definitely tutorials on YouTube or any kind of video and also be active in communities that can be crawled. So Reddit, Quora, having a Wikipedia page. These all help you to gain more visibility in front of LLMs and also perceive your brand as some brand that has like a history and also authority. Some people might ask, okay, should we write informational content? I'm seeing this question AI overviews. For example, in Google at least you have the zero click content. We call them crocodile charts because the impressions are going up and the clicks are going down. And that's mostly because AI is stealing those clicks from you.
Andrei Țiț [00:35:31]:
We found out that more or less informational content is dying. But this doesn't mean that you should not write any more informational content. On the contrary, you want to actually think about that informational content as top funnel content. You want to insert yourself in the buyer's journey as early as possible so you can actually influence people further down the road. Now there's a caveat here. You don't want to write something that's very generic and can be generated by an AI. And I'm going to talk about how AI generates this type of content. So don't try to create copycat content.
Andrei Țiț [00:36:06]:
I don't know. I'll give you an example. What is project management? I used to work for a project management tool. You're not going to write the very Long definitional article just to top the other ones and make it a bit. Yeah, it's just a word salad.
Dave Gerhardt [00:36:19]:
That was my whole like original SEO play. When I worked at Drift. We just basically wrote like what is for all of the topic maps of things that people in our audience were interested in. And then it was like, let's say you sold like auditing software. It'd be like, here's 10 auditing tools you need to try. And like your company was number seven on that list. Like all that stuff goes away.
Andrei Țiț [00:36:39]:
Claude, for example, has just disclosed is what's called crawling mechanism or whatever, machine learning mechanism, not crawling mechanism. And they have a tag, it's called never search. So basically for very generic topics like this one, it's going to instruct its LLMs not to search on the web or into its own database to search for an answer. So it doesn't make sense for you as a content marketer to write copycat content and to just make sure you write a word salad that's going to be better than the rest. Which brings me back to the second point I wanted to make when you asked my question. I think it's important to understand how those kind of LLMs are generating the answers because it's quite different than SEO. So in SEO, in traditional search, let's call it traditional search, you had Google, there was a query somebody was asking like, I don't know, pickleball, because it's quite famous right now. And then this is pickleball for example.
Andrei Țiț [00:37:33]:
And Google would check its database to see, okay, what pages closely answered the query Pickleball. It's quite general to be. It's not what is pickleball? How is pickleball different than paddle or whatever? So it would probably surface draws from its index a lot of stuff in LLM world. Imagine that you have a fan out query. So somebody is asking a query pickleball. That pickleball is going to be fanned out just like you have a fan. And what it's doing, it's finding relative terms of pickleball. So it's what is pickleball? Pickleball for beginners, pickleball versus paddle.
Andrei Țiț [00:38:10]:
And it's taking all that context from those relative terms to create a comprehensive more or less article, not necessarily article answer. That could answer that from all the angles when you type in pickleball. So I'm not going to talk about the science behind this. It's just like dumbed down version of it. What matters more now in AI search is to actually have a Topic that's super well covered in depth topical coverage and the way it's going to be retrieved by LLM is how close it is to my actual initial quiry. That's called cosine variable. So that's the mathematical formula. But just remember this, the closer it is to the topic itself, the closer, semantically speaking, the closer it's going to be recalled by the LLM.
Andrei Țiț [00:38:58]:
So we're going away from keyword stuffing. We're not looking into that anymore. We're trying to cover a topic as much as possible through topical clusters from each angle. And the way information should be structured also matters here. So we're not trying to write again word silence. Like you have a recipe and you're talking about how you were taught that recipe by your grandma or something. LLMs in general prefer more this kind of like too long didn't read format. So try to answer maker point in each subtopic in itself.
Andrei Țiț [00:39:29]:
And then if people are interested and want to delve deeper into a topic, then you can further detail the steps in the paragraph itself. But always front load the main idea when it comes to covering subtopics in a particular topic. And for those of you I didn't show that that's how we do it in ahrefs. So we're trying to crunch those numbers and make that similarity, cosine similarity into something more easier to digest and understand. So we have a topical coverage score. We're also going to break down that score on each subtopic and you can see how much closer you are when it comes to being to writing something that's going to be retrieved by the LLMs.
Dave Gerhardt [00:40:07]:
It's not possible to do an hour on this and like give you the recipe. And then like it's just like anything in marketing. There's going to be no playbook out. You're not going to be able to just like, it's not like you're going to get these three paint by number steps from Andrei after this and then three weeks from now you're going to magically double your website traffic because of chat GPT. But I think everyone is kind of readjusting and rethinking our marketing strategy. And I kind of always skew back to like the timeless lessons in marketing. That's why I don't care if it's B2B, B2C, whatever. My big light bulb and I'm taking my own notes during this is like we are going back to brand, we are going back to being interesting and relevant.
Dave Gerhardt [00:40:45]:
And when I Heard you say, I kind of went back to a decade ago, there was this movement in SEO and content and it was like create 10x content. I think that is still true to this day. It's like your goal is not to write these listicles. It's like how can we do original research? How can we create really interesting content? Okay, here's the things that the AI is going to answer, but we can give you 12 really specific examples from our customers who are in your niche and use examples and really go deep as opposed to like everyone can use chat GPT to write junior level blog posts, but we're going to publish two or three metrics meaty in depth articles a week or we're going to do one original research report a quarter. I love that trend. And then also this is what's crazy about all this stuff. It's like I'm out there saying for the last couple years PR is dead because of influencers and like the shrinking media landscape. But now it's like there's actually a stronger appetite than ever for pr.
Dave Gerhardt [00:41:37]:
I think it PR doesn't necessarily mean traditional publications, but I think you mentioned like someone has a substack or someone has a, a community finding the influencers in your world and having interesting stuff that they can write about that also can be surfaced in the LLMs. That's super interesting. One of my questions summarizing some of this stuff back to you is, does distribution matter? Because it used to be my friend Ross Simmons would always say distribution is king. You can't just create content, you got to distribute it. So what if you go and you don't have a lot of website traffic today, but you create this amazing piece of content. How do I, in the old world it would be like get backlinks to that. Do I need to think about distribution? I can't just write it and Hope that the LLMs will crawl this like, is there some play there that I need to be thinking about?
Andrei Țiț [00:42:22]:
I think Ross is quite famous for redistributing content or repurposing content in different formats kind of works. But for us at Atris it doesn't because you have so many different channels, so many audiences, so many frameworks or playbooks and a YouTube how to tutorial is not the same with a podcast on Apple for example, or on Spotify. But to come back to your question, I do think that AI search is more about an omnichannel strategy. So before you have organic search, low hanging fruit, it's like forever dripping. You don't need to invest too much in it. Just people who write content, make some decent content, put some backlinks and it's going to trickle, it's fine. But now that kind of era is over, we have omnichannel presence. So the AI is going to take multiple sources.
Andrei Țiț [00:43:06]:
It's going to take Reddit, Quora, it's going to take a Wikipedia page, it's going to take Listicles from, I don't know, G2 Capterra, it's going to take any kind of PR stunts from influencers, your own brand or events in itself. Maybe you have an event that your brand hosts and it's important. So right now, as I said, there's always been a lot of channels to acquire traffic, but never so many that matters all at once. So AI search, think of it as omnichannel approach and what I would do because I saw some questions in the chat. First thing you want to check the traffic like you said, do I get traffic from AI search? Let's see how much. If you get substantial traffic and you see a positive delta each month, then it's clearly a new acquisition channel and you should probably get into it as fast as you can. Reap the benefits as it's still early inception 2 Next thing is to actually, as I said, forget about keyword stuffing. The good part is that you don't need to buy backlinks anymore because all that matters is topical coverage, how well and in depth you're covering a topic with your own expertise, authority, data, something that your competitors and LLMs cannot generate on their own.
Andrei Țiț [00:44:20]:
And the way to do that, as I said, is focus on branded reach anchor text. You also want to write the front load domain information across various subtopics in your article in each paragraph. So LLMs can parse it very fast and somebody who wants to know more about it can read the rest of the paragraph. And yeah, that's about it as a starting point. So track first. Do I have actually traffic coming from those AI search bots in general? If yes, you could check the AI mentions in themselves. I've shown you Atrius brand radar. You can track it to across AI overviews, ChatGPT perplexity whatsoever.
Andrei Țiț [00:45:00]:
That's going to surface some visibility gaps. It's going to tell you what's your search demand also because we also allow it to track that. So is my brand being recalled by people? Are my brand awareness campaigns paying off or not? And then when it comes to content writing, do what I said, optimize for content coverage. So topical Coverage that's very in depth and also owns your own data so it's not a copycat or whatever type of content. And try to front load information across each paragraph so it's easy to be parsed by those LLMs. Of course. Also make sure that your website is parsed by LLMs so the LLM txt file doesn't ignore them. To be honest, those guys are kind of playing dirty.
Andrei Țiț [00:45:42]:
As I said, it's kind of like a black hat world. They're ignoring your robots and LLM text files on purpose because each one of them is trying to get more market share, more users, more accurate answers, comprehensive like that in one or two seconds. So that's probably something that's going to be ignored, but it's a good practice. The social contract more or less showing the web okay what's possible, what's not and what you should crawl and stuff like that.
Dave Gerhardt [00:46:09]:
How do you think about measurement in this world? Not specific to. I guess it is specific to content here. For a lot of us, content is the brand investment, creating content in some form. If we can. If we can no longer say, and I think a lot has changed because of this, but if we can no longer say, hey, Here are the 46 MQLs that we generated from this piece of content, I'm just curious, based on your company, the people you all are talking to, how do you think measurement and talking about the marketing's investment in content, how do you think that story plays out in the AI world?
Andrei Țiț [00:46:41]:
I spoke about that mostly in the first parts of the presentation. So we more or less need to shift away from the traditional way of reporting stuff because as I said, we have these crocodile charts. So impressions are going up, that's cool. But the clicks are going down. Why? Because there's zero click content. Mainly because the answers are generated by AI. But there's other metrics. I'm not saying those are functioning for us, but might not function for you.
Andrei Țiț [00:47:07]:
So I'll probably look at what my share of traffic value. So how much would my content b money wise, dollar wise if it were to be acquired via PPC ads in Google Ads? That's one way of doing it. For organic search for AI search in itself, as I said, it's quite murky. They don't provide referral data. But coming back, you said sorry, my bad. You were talking about content, right? The value of content. Yeah, yeah. So that's how I would measure it the other way.
Andrei Țiț [00:47:37]:
Yeah, it's quite easy and cheap to create content, but you cannot Spam the Internet and create copycat content, as I said, because that's not going to be surfaced. And for general queries, there's that never search tag in their own database that doesn't prompt them or it stops them to prompt something that's like common knowledge. So as somebody said, double eat eeat Expertise, experience, authority and trustworthiness. These are SEO signals. They signal to Google and also to LLMs in general that, okay, you're the topical authority on specific subjects, so you're the subject matter expert on this and your subject is the closest to this query that's being pulled. So it's easier for them to pull it before your competitors.
Dave Gerhardt [00:48:26]:
We're going to have to do some type of follow up or AMA or something. There's a bazillion questions on this, which is an amazing signal. Look, I think that my takeaway from this today is like super interesting and informative. I think there's a lot. I've been making a ton of notes. I'm just thinking about like moves that we need to make as the business. I think every business is going to be impacted by this in some way. And I think, look, you can sit on the sidelines and you can say, oh, this is the worst thing ever.
Dave Gerhardt [00:48:49]:
AI is going to ruin all of us. Or you can say, like, I think my nature is an optimist and I see an opportunity here. Just like there was with marketing when marketing became digital and Internet marketing became a thing. I've seen time and time again in my career, the people that create advantages for themselves, the companies and the people are the early adopters. And so I think having Andrei on to share some of this stuff is, is just for you to start planting those seeds. Like, man, you're going to go into one of these planning sessions with your team and somebody on your team, maybe your boss, wants you to just keep doing business like we've always done before. Now's the opportunity to say, well, hold on, look like we're starting to see some changes in search and things are shifting to AI. Like I want to shift our content strategy to match here and this is the way you start to make progress against some of the stuff.
Dave Gerhardt [00:49:34]:
All right, Allison, roll that poll, sister. We got a quick poll just to get some feedback. Measure this event on 1 to 5. So throw that poll. Andrei, great to have you. We'll send you all this. There's a lot of follow up. We should create some follow up content from here, obviously.
Dave Gerhardt [00:49:48]:
Go find Andrei on LinkedIn, connect with him, send him a message. Maybe he can get to be your private marketing whisper in the DMs or the Andrei robot will come in and help you out. But it's great to have you. Huge fan of what you all are doing. Perfect example of a company I think that is going to is leading through education and I think that is also a timeless piece here. How can you be that for your brand? Lead with education, be the expert for your customers, go deep, get specific real examples. You can help your customers win. Andrei, great to see you.
Dave Gerhardt [00:50:16]:
Thanks everybody for coming. This was a record turnout today. Holy smokes. We had over 500 people here live at one point, which is amazing. So great job, Andrei. Go for a run, get the Wiggles out, go for a run and enjoy the rest of your day.
Andrei Țiț [00:50:27]:
Yeah, thanks a lot for having me and you can reach out to me on LinkedIn and I'm also follow up with all of your questions.
Dave Gerhardt [00:50:33]:
All right, sounds good. See you all later.
Andrei Țiț [00:50:34]:
Okay, cheers.
Dave Gerhardt [00:50:39]:
Hey, thanks for listening to this podcast. If you like this episode. You know what, I'm not even going to ask you to subscribe and leave a review because I don't really care about that. I have something better for you. So we've built the number one one private community for B2B marketers at Exit Five. And you can go and check that out. Instead of leaving a rating or review, go check it out right now on our website, exit5.com our mission at Exit Five is to help you grow your career in B2B marketing. And there's no better place to do that than with us at Exit Five.
Dave Gerhardt [00:51:07]:
There's nearly 5,000 members now in our community. People are in there posting every day, asking questions about things like marketing, planning, ideas, inspiration, asking questions and getting feedback from your peers. Building your own network of marketers who are doing the same thing you are. So you can have a peer group or maybe just venting about your boss when you need to get in there and get something off your chest. It's 100 free to join for seven days. So you can go and check it out risk free and then there's a small annual fee to pay if you want to become a member for the year. Go check it out. Learn more exit5.com and I will see you on over there in the community.
