How to Think About AEO with Brett Domeny (Director of Product Management at Webflow)

You're listening to The Dave Gerhardt Show.

Exit. 1, 2, 3, 4. Exit. 1, 2, 3, 4. Exit.

Hey, it's me, Dave. That's me.

I'm Dave.

Before we get into today's episode, I want to give a quick shout out to our friends at Webflow who are sponsoring this episode. Webflow is the agentic web marketing platform for high-performing brands. Their platform lets you build fully custom websites that rank in AI search and make your website your biggest growth engine without being dependent on a developer. I know that because we use Webflow at ExitFive. No developers here. This episode is part of a 3-part AEO series that we're doing with Webflow over the next few weeks. I've been having a bunch of conversations with practitioners who are deep in the weeds on AI SEO, and web optimization. We talk about what's working, what's changing with real examples you can go and apply to your job right now in marketing. Webflow also put together this free AEO assessment. You can check it out after this episode. We'll link to it in the notes here. AI is changing how buyers discover and evaluate your brand, and this resource will tell you exactly where you stand. You'll get a personalized score showing how your site performs in AI search today, visibility into how likely your brand is to be cited by answer engines, and a breakdown across your content, technical foundations, authority, and how to measure where you show up so you know where to focus. We'll put the link in the notes here. Thanks again for Webflow to, uh, create this series with us all about AEO, one of the hottest topics we've had in marketing, a big appetite for it, and I know this is gonna be a great interview. So let's get into this episode.

All right, so Brett, tell me about, so you're on the, you're on the product side at, at Webflow working on. AEO, what does that mean? How would you explain that to just a marketing guy like myself?

Sure, sure. So I will say my background came out of marketing. So I went to my undergrad was in marketing. My MBA was in marketing. I started in marketing as a, to date myself, I took a class in college called e-marketing.

Love that.

Absolutely. So as I came out of school, it was right when companies were like, to market online, we need to build products online. Uh, so I say all that to say my, like, my approach, my mentality as we're building, obviously we build tools for, um, for marketing teams, for development teams. Um, uh, but the, like, what do marketers need, uh, is near and dear to my heart. Um, and so, you know, when I talk to, when I get to talk to marketing teams that, that we talk to, at the end of the day, my job in working on building out product is making sure that the product capabilities at Webflow are the, the toolshed that they need to get their job done. And so obviously as the world is continuing to flip on its head, a big part of that is making sure that we don't lose and I don't lose that pulse of what's actually happening. What are the real problems that marketers are facing? And so ultimately that's, that's often, you know, how I would describe it on a day-to-day basis in product is like, my job is to make sure I'm making your job easier. That's it. That's in a nutshell, Dave. Okay.

All right. So I got you. So we are really close to the the flame on what's happening, given that Webflow is, you know, a website company. Search happens by, you know, searching for something, and there's a connection between search and website. So how would you explain the state of search right now? And so I'm coming at this from like, I run a community of marketing professionals, especially among our CMOs. Most of them are all asking like, what are you, what are you doing right now? Like, we're, we're losing search, but we're not— it's not as easy as like add a page to your website and show up in clawed search result because people don't use the, you know, the way an LLM works and the search query is totally different. And so I kind of want to use this, use this, use your background in AEO and on the product side of Webflow to give marketers like first principles to think about.

Yeah, absolutely. And this is right at the heart of what we have been working at and what we've seen on those front lines over the last year. And, you know, so you know, let's say a year-ish ago as we started seeing the acceleration of not only the, the data, but even us as our, as consumers when we started saying, you know what, actually, like, it's easier for me or more effective to go answer this question or ask this question in one of these LLMs. Oh wow, that was effective. Like, you start to have new behavior evolve.

Even like, so I'm, I'm in a heavy user of, you know, everything since this came out. ChatGPT, then it was Claude, then Perplexity, then Gemini, and I've almost found myself now, I just use a Google Chrome browser and you know how like if you search for something, it shows up in the AI overview, but if you have another question, you can click into that and the chat experience is like that. And it's like completely changed how I search for things now.

Totally. And so, you know, so back to your point of like, what are the first principles and like, you know, how do we approach this? That is the question that, you know, we obviously, because, because while I build products and get to work with teams to build products at Webflow, we also have a marketing team that has to market Webflow. And so we're working right in line with them. And, and we started spending a lot of time as a company trying to break down those first principles of what we were discovering, what we were studying, what we were seeing. To your point, we are at kind of the, the tip of the, uh, the tip of the spear, so to speak, because we are making sure that we are enabling customers to build discoverable sites. And, and we started developing up something that, that we called this AEO maturity model. Um, and basically the idea was, What really matters? What even is AEO? What even is, you know, other, you know, the GEO? I had one customer I remember said, I don't, I don't know what I need to do. I just want to do all the EOs.

My joke is EIEIO.

Yeah, that's right. That's right. And so, so we started breaking it down really into 4 groups of things that really matter when you think about first principles. And those were content. So the content that you're building on your site, technical, how do you technically structure your site so that it's discoverable? Authority. And that's really a lot more about what happens off the site, which is obviously a very traditional SEO concept. And measurement. Are you actually seeing and discovering what's happening? And we started to break down as we looked at sites that were doing better, sites that weren't doing as well, doing our own experiments, talking with customers, starting to unpack what are the things that we start to see spike? Because early conversation, as you would, you well know, is like, it was this like, oh, SEO is dead and it's AEO. And it's like, well, it's actually this like, blended evolution of these things. Some things matter more than ever, some there's like new disciplines. And so what we started to break down was like, one, just a clear way to think about it. Because so many marketers that we talk with, and some still today, are like, I have so much going on. Just how do I even think about this? What are the things that really matter and that are important? Starting to break it down there. And within each area, there's more pieces that you can really start to hone in on.

I saw this, I'm trying to look at it real quick. I saw this report last week from a SEO company, uh, AI citation ranking factors. And basically the summary, which was interesting, was it said, uh, while the evidence— while the evidence suggests that specific practices can boost AI citations, most of the critical, critical factors align with traditional SEO practices. This is important because there's much debate in the AI SEO world about whether optimizing for AI visibility requires a different tool set. Has that been your experience?

Yeah, yeah, absolutely.

Can we actually go into each— I think you have thoughts on each one. So These are great. So content, technical, measurement, uh, content, technical authority, and measurement. If we could just kind of like crack into each one, I think that'd be awesome.

Yeah, that sounds good. Let's start with technical because I think, um, technical is where, you know, so, so, you know, like I mentioned coming from—

this is what I— this is the area also that I know, like the other stuff I think people intuitively like good content that makes sense. But this is, I think, something that there's a lot of misunderstanding about.

Yeah, yeah, absolutely. Um, and what I'd say is like, there's a lot of misunderstanding that's not too dissimilar from the last 20, 25 years of misunderstanding technical SEO. When it's like, what actually matters? What are the things that I need to do? Um, and so from a technical standpoint, much like you mentioned, much of what goes into being a discoverable or answer engine ready site is a lot of traditional SEO. Because when you, when you break it down, what's happening in the AEO front. At the end of the day, there are crawlers that these LLMs are deploying, whether they are their own crawlers or whether they are relying on existing search crawlers. Um, and they are going and accessing pages and bots and agents are basically trying to read and surface pages. And so the traditional SEO world was basically create a guide and create a method so that a bot or an agent can find you and understand what you're writing, um, in the most performant and efficient way possible, which isn't always go look at all the content. So a lot of this, if you think about it, like there's two big buckets here that matter here, which is the site, which is the overall how the site is structured, and then the individual pages. And so much of when we look at the authority model is really like meat and potatoes kind of, kind of pieces. Like, are you structuring your sitemap in a really performant way? Or if you're a localized site, are you handling all of your, you know, hreflangs? And do you have canonicals that are set up appropriately? And how are you modifying your URLs and how are your URLs structured? Kind of broader site-level pieces. Do you have good, you know, do you have non-optimized things in your robots.txt file? Obviously this is where, as you kind of go up the maturity model, there's things that we know work. And I think one of the things that like has always been true with SEO and for folks that now they've not really, you know, especially at different parts of the org that they've not been involved with SEO or maybe kind of been like, well, the SEO team handles that. And now all of a sudden, they're like living up close and personal with it. SEO has often been a lot of, there are things we know and there's things we experiment and we see what works. Where you say, well, I've tried this and let's see if this impacts. That's not everything, but there's a lot of foundational pieces. So a good example is like LLMs.txt, which is a new interface method in addition to robots.txt. And there's like mixed data on like how much LLMs are using it, how much you should do it. But there's things when it's like, should we go ahead and try it and see if that helps us? Absolutely. There, there's something like, we have a great partnership with Cloudflare to handle, you know, all of our hosting for, for, for published sites. And we just shipped something recently where it's Markdown for agents. And the idea of basically like, in addition to a semantically represented HTML page, what about we can deliver Markdown versions of the page along with it? Because we see early evidence that some agents are wanting to do that. So there's like this evolving set of pieces, but a lot of it, honestly, like goes down to start with the basics. Um, as a, as another data point I'll give you, Dave, like as we started working on this, we, we recognized this, this, this challenge to Webflow. And so we started shipping some basic tools even last year where things like schema markup, um, you know, if you think about the schema.org kind of core, uh, implication of how to semantically represent the page, that is not necessarily something net new. Um, but like what we saw was early evidence that LLMs were prioritizing or reading schema more effectively. But the average marketer was going, I need to write schema.org markup, which is a JSON-LD structure. I need to go figure out how to do that, how to inject that in code in the page. And so what we built was just like some AI tools to scan the content of the page, suggest it in. And one of the things that we saw, which was, which was great, but not necessarily surprising, was that early adopters, we actually, uh, published a blog not too long ago on this, early adopters of those tools saw 75% more, or 75% more organic traffic growth in that time period as compared to those that didn't, which is great, but not surprising because it's like if you semantically structure your site and the page so that the bots actually know how to read it, know how to access it, that's the first part of the battle. And that's where a lot of teams want to go to content, which is great. But if the, if the site can't actually be accessed in a way, or the pages can't be accessed in such a way. That's a big losing part of the battle or a big part of the battle where it's lost.

And I'm not a technical website guy, but the, that's different than like the llms.txt file and the robots. Like those are all, those are 3 different things. Yeah.

Yeah. Those are all different things.

And there's, there's like, it's just a good example of just like how many layers there are behind the scenes of like, totally. We wrote this great article. We did all this research. Why are we not ranking for any of this stuff? Even if it was like 10x content and it's like, damn, there's so many things behind the scenes. Like I'm sitting in my kitchen right now. It's like, yeah, it's a pretty looking kitchen, but you have to have the right foundation. You know, you have to have the right, you know.

That's right.

But is the, is the thought that like, if you're using a website hosting, like whether you're using Webflow or you're using HubSpot or you're using WordPress or whatever you're doing, like. Those vendors are all now shifting to like, do it for me in there. Don't, don't make me think about this. Or is there still, is there still some like level of technical, like I can't just, if I'm gonna home grow my website, I need to know all these things. Like, why do I need, my thing is like, why do I need to know all these things when I just use Webflow and like, that would solve the problem. Like how much time should the marketers that listen to this be spending like thinking about all these things versus can you solve it with would say, like, using, using a product?

Yeah, it's a great question. I would say, like, the ideal here, especially, you know, if I were to go back and put my marketing hat on, I would absolutely be in that camp, which is to say I need to, like, to use your kitchen example, like, you want to be able to come in and just cook and not be worried about, like, are the, like, the foundations, is the electricity wired up the right way? You're like, I just want to know that works and I want to know I did the right thing. And traditionally, that's where marketers hire agencies to, to do all that. But we also know, one, that that can still be a good solve. The other challenge that we see though is there's this compression on do more with less and how do we get more bang for our buck and all of that. And so what I would say is like, I think that's the ideal when, you know, so for the marketer listening here, it's like, know the details that matter and familiarize yourself with what needs to happen. But that's not the same thing as now have to go home, grow it and do it yourself. Um, like that's our approach at Webflow. Like we, we just announced a product that we're shipping, which is more of a holistic AEO suite and a big part of it is Agentic Recommendations so that we are, you know, a big part of the AEO maturity model, uh, which you can still use today. So if you're not a Webflow customer, go put in your URL, um, uh, to the Webflow AEO maturity model and we'll go do an actual evaluation of the site and tell you all the areas across this maturity model where you could start to lean in. Well, the reality is, is like, that's how marketers really want to work on a day-to-day basis is like, hey, just tell me the things that need my attention. I still want to, you know, for many of them, I want to still have a view into what's happening so I can approve, make sure it's on brand, make sure I'm comfortable with, you know, kind of the implication, but I don't want to have to be constantly maintaining it. A good example of this, I remember a conversation with a large marketing team at one of our customers and something like broken link maintenance, which is like really matters in technical SEO and really matters now in technical AEO and the idea of like, oh, a content marketer unpublished these 5 pages and we had no idea all of the ripple effects of where it was linked up. And now I gotta go work through the tedious process. It's like, that is where we want to, you know, our, our, our team often, you know, talks about like, we wanna fight the, the impacts of AI with AI. And so it's like, this is where AI-driven tools can also be your friend. And so the technical part is like such a foundational element that it's not always the most exciting area. It's not always the most innovative area, but a lot of the foundations of where you can get a lot more performance that you can control. And I think that's an important part of it. Cause like content, you don't know what's going to resonate. You kind of, kind of try and you do all your research authority. It's like, there's a lot of external factors, but like technical, you can control, you can make sure that the site and the page that I built.

That's like the one that you can have almost like a checklist of things to go and do, right?

Yeah, absolutely.

Okay. And I think the takeaway from this is like, don't, it's not like, listen to this in your car and then like immediately You gotta— it's just like, here, here's a, like, okay, am I— have I thought about the technical side of this and what's the checklist in there? And then also, we'll, we'll link to this, but I'll just— if you go to, if you go to webflow.com/solutions/aeo, you can see this, um, assessment thing, but we'll, we'll link to it. So, okay, that, that kind of scratches the itch on the technical part. I, I understand that. Where do you want to go next? Content, measurement, or—

yeah, let's go to techno. Or content.

Content.

I made up a new one, Dave, the fifth one somehow. Content. Let's go to content. So I think the, you know, the content side, as, you know, as we think about how to evolve from just solving for traditional SEO to AEO, a big part of the emphasis is like even starting at a first principle standpoint of like, what are you doing that's changed? And I think you gave a great example as you were talking through how you use, I think you were mentioning Gemini Search or Google Search where like, AI mode just gets you in a different, different mindset. You know, 2 years ago, if, uh, my family was going on a, on a trip, I would start with a very open-ended, like keyword stuff search, get to a page, you know, go try to, um, um, then get to a specific page and find my answer on the page. And so the idea was like in traditional search, the idea, or, or, you know, former, you know, kind of the, the, the old world was as a brand, you wanted to get found. So you could deliver the answer on your site. Now, when we think about how we as consumers interact with AI search, as a brand, you want to be included in the answer, which is happening away from your brand, so that then when a customer is even more engaged, hopefully they actually make it to your site. And so the big change is like, I'm doing all of this, you know, again, very traditional top of funnel kind of, you know, if you think in the, in the traditional search world, like long-tail keywords almost. And many of the questions that are being asked in LLMs are probably net new questions that may have never been asked before because you're not just saying, you know, like, you know, we were just talking, Dave, about golf. I'm not just looking at like best golf clubs. I'm looking for like best golf clubs for this specific use case and the way that I'm doing here, what should I be considering? And that is a very different content strategy.

It's almost like the AI search, like everyone, and it's always been like, that, you know, long-tail keywords, you know, bottom-of-the-funnel content, that stuff is the highest converting. But what that meant in the past was like your search query into Google is, you know, has more target— it's, it's deeper than just like, yeah, best golf clubs. It's like, you know, you'd have to search, you know, best golf clubs for, you know, a 6 handicap that wants to improve and whatever.

Yeah. I also think if you're a golfer, Dave is just subtly dropping his handicap here on—

Oh, I'm not a 6. I'm a 0, but that's a separate—

Oh, wow. Even—

Wow.

That's a separate topic.

That's a conversation for another—

Absolutely. Yeah.

We'll have to see. I'll have to come. We'll see if the handicap travels, but—

Okay. Okay.

Now though, it's like the long tail, bottom of the funnel, it becomes a conversation. And so like you're at all of search is now much more conversational. And so it's like, Even I search stuff after I go to the gym sometimes, all the time. I'm like, what's the best way to target, you know, this, this act, do this exercise better. And then the future would be like, yeah, serving up the product recommendation there, which might actually come after I've traded, you know, 5 or 6 messages.

Yeah, totally. And I think like when you, when you think through like, okay, how do I as a brand, as a marketer show up in that world? Well, what's actually happening behind the scenes and starting there. So the idea is like. If I want to search a very specific, for a very specific question, the reality is, is that these LLMs are really powerful. They're really amazing, but they're doing basically relationship matching. And they're looking for, for this, how do I, what can I reference that answers this? And how do I best put together an answer based on this? And it's, it's going and looking at source material on this. And so the idea, if as we keep having these newer and newer questions and more and more specific questions, because I think That's really the implication of these like long tail pieces. It's very specific questions. A good starting spot for brands is to really hone in on, am I answering the questions that one, matter to me, two, that my customers actually care about, and three, that I actually care to be an answer for? Because if we think back to the goal on the content side is no longer just make sure that someone get to your, can get to your site. Then you can give them the answer on your site. But it's now the answer is happening away from your site. So you want to make sure that you can control and influence that narrative of what the answer actually is. So one of the things we've seen as you move up that maturity model is a big part of it is like, are you answering, moving from just optimizing for keywords, but for answering for questions or optimizing for questions? And so one of the things we've seen work really well with a number of our customers, with our own internal team, is like start at the, at the ground source. What are your customers already asking you? Go audit your site for are you answering that? So a really good way is like if you're using like a customer service call transcript platform, go access that. Look at all of the top questions that people are asking when they reach out to your customer support team and determine, am I answering those on my site? Because if not, there's a great opportunity to start. Go look at your social platforms. What are the top things that people are asking about your brand? Not necessarily just on your brand pages, or in the forums, you know, that you're controlling, but across the web, what are they asking? And then go audit that against what is your site. And oftentimes it's a really great starting spot of going, wow, we want to say this.

Yeah.

Our customers are asking us about this. So to answer it, they're relying on someone else to answer it rather than us being able to actually contribute to that conversation.

Do you think it's all just going to be like, because I'm thinking about like the very, you would need so many variations of all those searches, like of all those things and like a human human couldn't possibly come up with those. Okay, so my brain goes like, okay, I get what he's saying. Let's come up with 1,000 different variations of content. And then naturally you're gonna use AI to— I'm gonna use, you know, Claude Code and I'm going to generate 1,000 pages that rank for that. Is that where this is going, where like we're just making AI content to show up in AI search and my AI agent is talking to the AI agent of the company that I'm trying to buy? Like Is that, is that where this is going or is there any, any role for humans and creativity? 'Cause my brain goes to like, I wanna write the like killer, there's so much AI slop. I would rather write one article and make it awesome and make it the thing that everyone shares and everyone talks about versus like auto-generate a thousand versions of a landing page. But like, who, who's right?

Yeah. Yeah. It's a, it's a great question. And I think like somebody we work with often at Webflow, um, Ethan Smith over at Graphite, he's been a big voice in AEO. He and his team actually did a breakdown on this where they were looking at basically what is the success rate, if you will, what's the visibility success of content that's, that actually passes the AI sniff test, if you will. And his basically overall summary was, no, this isn't where we are all going to, to your answer to your question, Dave. Um, because both sides of the, both sides of the equation are actually incentivized to work through it. And so he compared it back to like when some of the early SEO kind of strategies evolved, um, when it was like, oh, if I just go, um, keyword stuff and like, you know, like if you think back to the old days of like, if I have a white footer at the bottom of my site and I keyword stuff in white text and put all these links, I'm going to hack around it. And then like, you know, companies evolved models to make sure that it was actually rewarding the right answers. So I do think there's enough, there's enough incentive to like make sure that the products are actually delivering the right answers. I think we're going to see blips and ebbs and flows as new models change how they're doing something. People and tools find spots to exploit in that, and there's some self-correctiveness. So I would say as a brand, I wouldn't go to that extreme and say, well, what's every variation? Let's go auto-generate it and sit back. Because I do think we can see there still is a place, there still is a rewarding, if you will, for actually creating authentic content, for actually answering core questions. The other thing you were mentioning is like, wow, this can really get into— I've got all of this overhead of like, what are the thousand permutations of this? And that's often where too it's important to kind of connect into how these LLMs are handling things, which is like, again, they're continuing to move the target of how they build. So this isn't like a declarative, we know it's always this way, but as a general standpoint, what they're trying to do is they're taking these long-tail questions and decomposing them into smaller keyword search style breakdowns. And so there is, this is often where it's talked about of like query fan out or like prompt fan out. So like if you think about like the heart of the question, there's probably 50 different ways that you could subtly reword a question. From what we've seen in our experiences, it's less important to like build answers to every 50 permutations of that thing, but like have a, you know, basically answer the question. And there's, you know, like LLMs will generally do a good enough job working against those permutations. But the big gap really is honestly a lot more foundational than at the very edge, which is like there's often a core set of questions that actually your customers really care about that are really being answered and relied on being answered by someone else because You're just focused on these are our brand keywords and your customers are like, that's cool. But I have all of these detail questions. I remember we were working with a, with a, with a, with a company and they had a whole, you know, content strategy really around we want to like basically accentuate these brand campaigns with every time we have a brand campaign that pops up, let's go actually evaluate what are the things, what are the secondary types of searches back against our generic profile that you would actually want. Then let's go focus on the highest priority ones, build a few, uh, versus like, let's just go start at scale, build a few. Let's start to see what works, what doesn't work. So then you're not just like, well, I've made so much inundation that I don't really know what's happening. And now I've got all of this and you're moving so many variables at the same time can become difficult because then you have all of these metrics of going, well, I don't really know if it did the right thing. So I guess I should continue. Versus like being able to focus on, you know, again, kind of some core things and evolve from there.

Is this the right point to transition into measurement? Because, yeah, it doesn't make sense to talk about— is it okay to talk about authority last?

Yeah, absolutely. That makes sense. Yeah. Yeah.

My brain goes to like, now we've like, we've done all the technical stuff, right? We have all this content, but still I'm Dave, Dave the consumer. I'm going to buy Perfect example is right now we're like, we're moving banks for my, my company. I'm doing 99% of the research about the bank without ever actually hitting the bank's website.

You bet.

How do they know? And then the marketing question there is like, and the people who listen to my podcast is marketing people, marketing teams.

It's like, you bet, you bet.

I want credit for that.

Like, where's my attribution? How does all that work? Like, how do I know?

Where in the past it'd be like, oh, we create great content, we start to see organic and direct and brand, you know, that those, those things start to move the needle. So what, what matters in this world?

Yeah.

Yeah.

It's a great question. You know, I'm a big like bucket thinker, like think of the bucket. So like the 3 buckets that I think, and you can again really picture them as a funnel when you think about measurement, it's starting all the way at the top, which is really the agent discovery. So if we think about how this is evolving is if now if you want to be represented in an answer, by an LLM, you have to actually be crawled, you have to be surfaced. They have to actually discover that. This is where, you know, agent traffic, bot analytics, like, you know, there's a number of names on that. That's the first part. And what's really interesting here, I guess, kind of funny, interesting is, you know, over the last few years, I remember Webflow, you know, we worked on a visitor analytics product, which we'll get to visitor analytics in the funnel in a moment. We worked on that a few years ago and we were first building it. You know, I, I spent a lot of time and then again, even from a background standpoint, talking with marketers and, and the idea of like, I don't want to be looking at bots, like make sure that like, I want to focus on the real visitors. And now we're flipped to a world where we're like, I need to know what bots are hitting my site and which types of bots. And like, there's now because of this change. And so that's really that first part, which is understanding, are you actually getting accessed, crawled, and discovered by these LLMs, by search engines, by crawlers in the first place? Because if, if so much to your point, like using the bank example, so much of the real, of the experience, um, and so much of the, um, the representation of the brand is happening on another platform, um, making sure that they are actually accessing the pages as frequently or the right pages so they can be even included is a first important part. We can dig into that in a second. I'll finish the funnel first. The second one is really this, this fast emerging area over the last year and a half of prompt analytics or AI answer insights or LLM, um, visibility. But basically the idea of, hey, for questions that are happening, am I showing up? Um, because that is this new domain in search. You could traditionally go tap into these search engine, you know, APIs and search volume and understand like, hey, when people are looking for best bank, am I showing up? Who's showing up? How many people are searching for it on a monthly basis? And now when there is this new LLM spot where I basically say, what are people searching for? Am I showing up? It's a black box. And so there's a lot of companies basically trying to crack open that black box to basically say, solve for when people search, are you showing up? Um, so there's a number of tools that are emerging in that space, uh, which is like the, the prompt visibility.

Are they accurate? Are they good? Like how, how representative of Again, some guy left me a rating, one review on my podcast is like, how is this guy a CMO? He doesn't know anything. And I'm like, I don't, I really don't know that much. But I, my limited experience with like Search Console with Google was always like, okay, this is awesome. But so many of the, so much of the keyword data like never actually got passed through to HubSpot, which is our marketing automation system. So I actually have no idea.

Yeah.

So I just go into incognito mode and I just Google it myself and I see like, oh yeah, we're in the third ranking here, any reaction to like how accurate some of these tools are?

Yeah. Yeah. So what I would say is all, all of the tools, Webflow included, we're building tools to help you in the prompt visibility space, are, are doing their best, doing our best to give you directional insight into what's happening. Because the, the fundamental challenge is, um, what is the truth set? So if you think about two different dimensions, what are people searching for and how often, and am I showing up? Those are the two big answers that you want to, you want to understand. What are people searching for and how often are they searching for it? This is where in the traditional search landscape, um, Google has obviously built a marketplace around this because, um, if a, if a keyword matters to you, you want to know, are enough people looking for it that I should actually go pay to run ads there? You know, I know there's more nuance to it, but, but in a nutshell, it's that. Um, and so they built a business model around it, around giving you visibility into a lot of people search for this. You need to be in this conversation. The challenge in the AI space is right now that is still a black box. So these LLM tools are not delivering official LLM prompt volume APIs, if you will, to say, oh, how many times a month is somebody looking for what's the best podcast for CMOs?

Yeah.

And then all of these LLMs saying, well, here's how often that search, should you go look for that? So the first part is all of these tools are basically needing to try to create some semblance of, is this something you should look for? Which is basically they're trying to build their own dataset so you can have directional insight. And the second is, am I showing up? And this is the tricky part because, you know, from a, you know, as, as, you know, as we've spent time with LLM platforms and partners too, it's like non-determinism or like the vari— variadic nature of an LLM is a feature, not a bug. So like their goal is to make sure that they can give you a unique response for the unique detail they're giving versus being able to create this like strong determinism. Like, that's why an agent's involved. Oh, bidding on that. Let me go do this. Oh, instead of this. And so the idea of, am I showing up? A number of variables actually matter. Who are you? Which, which client are you using? What, what history do you have? What skills do you have? What context do you have? So the idea that you can just generically hit an LLM or ask an LLM, do I show up? Is, is challenging. So what I would say is like, it's impossible to replicate.

It's like making an image with like Nano Banana and I'm like, go back and edit, just change that to black and it creates an entirely new image. It just doesn't work work that way.

Yeah, you're like, ah, and now I lost the, the upper half of the image. Yeah.

Um, okay, so, so give me like, not like a one-liner, but like a— how would you be the therapist to us marketers? You as a former marketer listening to this, like, given all of that, how would you talk about measurement with content? Yeah, because of all this, how would you, how would you think about it? How would you present it to your boss? How would you frame it inside of the company?

Yeah, absolutely. Well, thanks for coming to the session, Dave. You're welcome.

Anytime.

So what I would say is having visibility matters. So first off, so the idea that it's not perfect should not be the blocker from using it as an input, but it doesn't necessarily mean that it is perfect. So take it for directional insight. What it does help you give visibility to and that you should use it for is there are, going back to the content strategy, there are a number of questions that you want to show up for. Knowing whether you show up or not, even if it's directional, is a really good place to start. And knowing if your competitors are showing up as compared to you is also another good insight. Um, because as you're investing in technical strategies and content strategies, what you can look for is, am I directionally moving in the right point? So it's, it's kind of like CSAT or NPS. Every day that NPS dips 2 points, you don't necessarily call a 5-alarm fire and go, what happened? You're like, oh, okay, there's some natural movement that happens there. But like, are we generally heading in the right direction with NPS and CSAT? I've invested in these guest cert, you know, experience programs, and we've improved over the last quarter. That's the same way I would use prompt measurement. Start with the questions that you want to care, want to care about, because bad data in is bad data out. That's, that's still true. Which is if you're trying to show up for all of the wrong questions or you're like, these are all aspirational and they're not actually grounded to things you're doing, that's not going to be good data. So start with the questions that you care about, that you want to show off, that you, you feel matter for your business.

Would you literally like have a, have a list and it shouldn't be infinite? Like there are maybe, I don't know if it's 10 or 100 or 1,000, but like you're trying to like, here are the things that we care about, or am I overcomplicating? It's like if you're Webflow, Is there a handful of things that you care about, or is it like, actually, no, aren't there like endless, infinite variations of what someone would be caring about?

Yeah, I mean, it doesn't necessarily have to be a super short list, but I do think you could compartmentalize the list because even if you say like things that I should do really well in my core brand, am I showing up well for my core brand? And then maybe a whole other category of like experimental new types of things. If you don't subdivide those well enough, like it can get noisy. So there, there could be a really long list, but I do think compartmentalizing is an important part of that so that you actually know, am I moving, are the things that matter most the right way? Because if you're, if you're showing up in 20% of searches for things that are more experimental, but you're showing up, but, but you can't really tell if that 20% is for your core brand or experimental, that's a problem. So what I would say then is like, use a measurement tool because having visibility is important, but know that it is going to be, to your point, it is trying to recreate and represent for your user. So understanding the directional inputs on it and use it for that. Don't be watching it every single day to say, ooh, why did this drop 2 points yesterday, but it went up 1 point today? Use it really as trend information. Are we moving in the right direction over the last month? We started tracking these new prompts. Are we actually improving as we are investing in these pieces? So I would say it matters. It's important to, it's important to track because it actually does give you insight to that black box. But be mindful of what it is and what it isn't, because it is an evolving discipline and it's not perfect.

Hmm. How do I say it to my boss?

How do I say that to my CMO?

Yeah. So, so which part do you think is more important, Dave? The like, the, the like, why should we go invest in it? Or more of the like, why should, um, uh, how do I know that I'm doing a good job? You know, really at, at, at that latter point, it's kind of both.

I think the question that most people are asked, I think the question most people most commonly ask is like, How do I show my work is working? Yeah, you know what I mean? Like, the fact that we have a blog, or the fact that we have a podcast, or the fact that we're— so much of marketing now is creating content because it's 2026 and like everyone's podcasting, making videos, YouTube, social media, all those things I would bucket as content. And in the past it would be like, all right, well, if you do more of that stuff, I would bet that our You know, direct search is going to go up. More people are going to hear about ExitFive because we're doing more stuff. And that's how I've always kind of correlated like LinkedIn and social media and podcasts with that. Does that logic still apply in this world?

Yeah. Yeah, it generally does. And I think that's why having measurement here matters. So to go back to like, what do I tell my boss on why do I invest is like, if we aren't tracking it, it's a, it's a black box and we have no idea if we're showing up, if we're not showing up, if our competitors are showing up. So that is the, like, the P0 or the, the, the, like, the must-do. And how do I show that I'm, I'm still doing a good job? I actually think if marketing teams don't have a way to represent this, it does feel— again, I keep using blind spot— it feels like a blind spot or an area where you can't actually show the, the, the byproduct of your work. I, I think back to, like, going back to the technical element, a big part, honestly, even if you don't go change a single thing on your, on your, um, Content strategy, like there's probably for most teams, a lot of core technical foundations that you can improve. When you do those, you should generally be showing up more, but rather than saying, well, I sure hope that did the right thing. You can actually start to see that and you can represent that. Obviously we, as we, I almost bucketed myself again. I'll, I'll bucket myself as a marketer again, but like as marketers, obviously have to take discretion of how do we represent that? Because it's like, what's the story we want to tell? And so as you're investing, like, it may be the top-level visibility. So across all the questions, we've seen a big uptick. I think of a customer we worked with that did a, like, completely replatformed their site, rebuilt all their pages, just really prioritized schema, structured data, and they saw a significant uptick in their visibility. And specifically what they were looking at is like they could go point to, we've made all these technical investments and look at our visibility at the beginning of the quarter. And look at our visibility at the end of the quarter. I forget the numbers, but like 56%, we were on average showing up for a little more than half of the questions that we want to, we want to track. And now we're showing up for like 72% again. But like a good one.

And for that you would need, you just need to use AI.

You'd like, yeah, you need to use something. So you can again, it's not perfect, but the data is directional. It tells you that something is changing. And generally it should obviously be changing positively as you're making these investments. but you can actually go show that off. And we, I mean, we do that at Webflow, um, and how we're talking about our efforts working, what's happening in prompts that we're, we're trying.

Yeah.

Or like a lot of people that listen to this stuff because of the nature of being in B2B, uh, have a sales motion. And so it could also be like, you know, on the sales calls, the how did you hear about us is gonna show more like, yeah, I was doing a bunch of research with, you know, Claude or whatever.

Yeah.

Yeah.

You bet. You bet. Um, okay.

I, I wanna, I wanna wrap in a minute, but we didn't hit on, gimme like the elevator pitch. On authority, what matters?

Yeah. So, so what we, what we're seeing is the authority piece of what matters is honestly so much of what Google and other search engines really initially set out to be, uh, which is, are other people talking about you positively? Or how are other platforms talking about you? And so, um, one, that's still the general principle. Are you being talked about positively? Are you being, are the questions that you actually want to have answered? so a couple of the biggest feeders of LLM responses are tools like Reddit because it's authentic community interaction. So to your point on the content, like, how do you negate just like AI from telling AI what to think about AI, um, is like community and people talking about things. Are you showing up there? So like, again, I referenced Ethan earlier. He, he talked about work that he's done with customers where one of the biggest things when you say like, how do I focus is like, go do the basics on technical. Make sure you're measuring, go find the conversations that are happening in Reddit and actually contribute to them. And so, uh, that's kind of the elevator pitch is like how people are talking about you externally from the platform. There's a handful of places that have a little bit more outsized impact if you're going to focus your resources, which tend to be social platforms because LLMs are saying that must be a good source for what people think because that's where people are. Um, and so that's why, you know, to a, to a marketer listening, you're like, oh, so it's like our community strategy or it's our social strategy. And like a lot of it's like, yeah, just make sure it's connected back to the other things versus a completely disparate piece.

I think you should come work in marketing. I think you're, I think you're a great, great spokesperson. You've been at Webflow for like almost 6 years.

Yeah.

What keeps you there?

Yeah, it's a great question. So what's fun is as a marketer, you know, I was working in marketing and I started working on product by, by nature because we were like, hey, if we want to market all our brands, we have to build web platforms and we need to build, you know, commerce applications. Um, and I actually found Webflow in like its first year and a half as a user and found it as a like, wow, we, we, this, you know, we're able to move faster as a marketer. You know, I did some classes, uh, again, to date myself, I took Dreamweaver classes in, uh, in college, but it was just like, I'm, I'm not very good at that. So the idea of, I had, I had vision of what I wanted to do as a marketer, but I couldn't execute. Execute it. And I found that Webflow could actually help me with that and, um, became really passionate, loved the product. Um, and so working here today, still today, is like, that still is deeply connected to me, is like, I love what we get to do in that, um, our goal and, and what we're, we, we, we talk about it in a way of saying giving development superpowers to everyone. Um, is we want to, you know, make sure that, uh, you know, the, the original vision of the company was the, the world runs on the web, but less than 2% of the world knows how to build for the web. And obviously the how and everything's changed, um, uh, big time. But like one big part of that, Dave, is like, I love the platform and I love to get to help, you know, marketers like myself and teams like myself in, in the past, like realize their visions and build enterprise-grade applications. The second is, um, uh, as much as you can love a, a vision, like showing up and working with, with great people, um, and enjoying who you work with and, and having a good time, uh, while you do it is important. Um, uh, at the end of the day, it's, it's jobs that we all do. And so, hopefully have a good time doing it. And that's another piece. And so those two things have definitely kept me. It's been a fun adventure. I've got to work on a number of things over my time. And obviously right now is as the whole, how we build for the web and how we work on the web is being flipped on its head.

Yeah, it's an interesting space to be in for sure. Like the website and content space. Anyway, Brett, awesome. I'll come to, uh, where are you at, Missouri?

Yeah, yeah. I'm, uh, it's, uh, it's a beautiful place to golf, Dave.

Okay, I know, I, I believe it, I believe it. It's, it's May 12th here in Vermont and it's like 40 degrees out today. The ball's going, the ball goes nowhere. I'm not golfing today, unfortunately. And no, not unfortunately, I had a great conversation with you, Brad.

Let's be honest, okay?

Let's be honest. Anyone who says they would rather be working than not working is a lot— is lying. And golf is very fun. It's beautiful. Spend 4 or 5 hours outdoors. There's nothing like it.

So you bet. You bet.

All right, man. Great job. Appreciate it. Thanks for coming on the pod.

Thank you for having me. It was a pleasure.

We'll link to all this stuff and, you know, reach out to me, reach out to Brett, send us an email, reply to my newsletter when we have this in there. Useful framework for thinking about AEO, this AEO maturity model, the 4 areas we covered were content, technical, measurement, authority.

Brett, you did a great job and I'll see you soon.

Okay, thanks.

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 1 private community for B2B marketers at Exit5. 5, 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 Exit5 is to help you grow your career in B2B marketing, and there's no better place to do that than with us at Exit5. 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 7 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, exitfive.com, and I will see you over there in the community.