B (3:29)

Awesome. And when you think about your journey to webflow, what are some of the highlights of this experience? I think it's a really uncommon for founders to stay for a long time after they get acquired. But what made this acquisition really rewarding for you guys but also like exciting in terms of the work that you get to do today?

C (3:55)

When webflow reached out, we came to understand as we got to know each other over time, a couple of things. One, webflow does what it says it's going to do and is clear when it makes mistakes or can't do something. In telemize, we ran it the same way and I as a former CEO is used to everybody lying to him all the time. Like I appreciate that there was actually cultural fit which is so rare in M and A. And the the ability to drive revenue through ABM personalization testing was literally the number one ask from prospects. Which is why webflow went out to go find a partner to acquire in. And that value prop of like go build your brand, go drive revenue. Coding optional. You can code heavily if you want, you can code not at all if you want and everything in between. That's really powerful and it's been exciting to have the goods to bring to market.

B (4:49)

Cool, that's awesome. And congratulations on being able to find that cultural fit. I agree with you. I think that's just really rare in any industry to be able to find that connection point. Let's talk a little bit about what's happening in the market in the shifts that we're seeing. I want to start with our first deep dive question that really I think is capturing a lot of the conversation that's happening in the search and in the organic space. And it's this concept of AEO versus SEO and really this whole name or nomenclature around what it is that we're doing to optimize for answer engine results. So what sets answer engine optimization apart from traditional SEO? Why does this even matter today? I think this is a very controversial question. I think a lot of people in LinkedIn and on YouTube and in all these different channels, they're all like trying to label this thing and there's this real, you know, this Real theory of, like, well, isn't just SEO the same thing as aeo? We're just, we're just, you know, trying to give it a new name because it looks slightly different. But would love to get your perspective on that guy and what your, what your thoughts are in between this naming convention and what these two things are.

C (6:06)

It's a great question, and I think it's one that will matter and does matter to a lot of us for a long time to come. At its core, you know, SEO was showing up in a variety of search engines and then eventually just, you know, Google, maybe Google and Bing, having your content directly pointed to, using your words with blue links. We all know that with AEO, the goal is to appear in LLMs as people are asking them questions. ChatGPT, Copilot, Gemini, you know, Anthropics, Claude, Perplexity, and more and more. As you mentioned earlier, usage is skyrocketing. I mean, ChatGPT reached 100 million users in two months like no other product's ever done that before. And it's now at 800 million users. Perhaps more importantly for many of us, Forrester surveyed a bunch of B2B buyers and 95% of them said, yep, I'm going to use Gen AI in my buying process this year. Okay, so it's something to pay attention to. What does it mean for us? In my humble opinion, it's both risk and opportunity. The risk is our brand narrative. We kind of lose control of some of that. It was intentional to call out that. SEO uses our words. LLMs, as we all know, reframes the answers, not through our carefully crafted copy. And so it's totally reasonable for an LLM to be 100% correct, but present us as a list of capabilities rather than this, like, carefully differentiated story we spent so much time working on, or for it to be confident and wrong.

B (7:40)

Right?

C (7:41)

So that's one, brand narrative control, two, traffic, which helps us all deliver the pipeline and revenue we're accountable for. While I know there have been a bunch of headlines on catastrophic drops, I think more broadly the drops are there, but not catastrophic. Bain estimated that overall SEO is down 15 to 25%. That jives with data I've heard from others as potential risk. But the upside is that when people engage with an LLM, they tend to be further down the funnel, they tend to be more qualified. And so firms that we know very well found that quantitatively, Semrush said they are 4.4x more valuable. Ahrefs found they convert 23x better than typical SEO traffic. And I think the last piece of what is kind of a perfect storm for all of us is we're all drawing content from the same LLMs. We sound even more indistinguishable than we have forever. Right? We're all the best, all the first, all the leader. And when LinkedIn ran a bunch of our B2B ads in front of prospects, only 19% were able to correctly attribute the ad to the company doing the advertising. That didn't make me super happy to.

B (8:54)

Hear, right, go spend more money now.

C (8:56)

Yes, exactly. So to your point on hey, isn't SEO the same as aeo? I've heard that. I've heard SEO is dead. I heard anecdotally through someone else that a Fortune 500 CMO fired his or her entire SEO team and said we're doing AEO now. I could not disagree more. I think AEO is close to what good SEO always should have been.

B (9:18)

Right?

C (9:18)

Original content that is genuinely valuable throughout the buyer journey. Especially data that gets rewarded. Okay, Minus one thing, plus a couple. The minus is keyword ranking as a concept doesn't exist. You're more focused on like mentions and share of voice and things like that. Fine. And the pluses aren't replacing but one. Rather than optimizing for counting keywords in a basket of keywords, you're now optimizing to comprehensively answer a basket of questions at various levels throughout the funnel. So from keywords to answers to the the structure, the metadata on your site, helping an LLM understand content meaning structure is more important. It's actually hard to overstate how important it is. And three links still matter. That doesn't go away. But plain text dimensions on other sites, they count more than they did when SEO than they did in SEO. And at Webflow we're seeing a big impact from this. 8% of our total self serve signups come from AI search and they convert six times better than SEO traffic. Wow.

B (10:31)

So there's a lot to unpack here because I want to start by. Yeah, I want to start by asking one really specific question around this topic of AEO versus SEO. Foundationally, there's probably a lot that you feel is similar in terms of the conceptual work that's being done by this organic channel or channels. Right. Technical work, building good websites, building good experiences. But when it comes to the disruption in the funnel, you brought that up that hey, this traffic is converting a lot better. I get really puzzled by that. And I want to get your perspective on is the transition of how consumers are going through the LLM funnel versus the SEO funnel, going to foundationally change SEO versus AEO or AI discovery. Because ultimately SEO is going to become your awareness channel and AEO or AI discovery is going to become your conversion funnel.

C (11:37)

Opinion of one Very much looking out in the future because I don't think anybody knows, right? I think the real answer is I don't think anybody knows. I think it's the early days of a new medium. Just like for those of us like you and I who were around in the early days of search, it was not super clear where it would go. John Battelle I think it was 22 years ago in oh three was like, Google is the database of intent. And we all were like, wow, yes, there was great intent here. And we saw, you know, ginormous businesses built on that.

B (12:05)

Sure.

C (12:06)

I also hear things like, hey, you know what, the LLMs, they're more likely to go to third party sources for upper funnel content like top 10 lists and G2, and your website for mid and lower funnel content to get the details. And while that may be true, in my humble opinion, both are going to be for both. I do not think that SEO will be tofu only and that LLMs will be, you know, mid or lower funnel. I personally, I don't believe that at all. I think if we really get speculative, again, opinion of one not based in data, sure. I think over time they will come together because they each serve a purpose. Just like I'll draw a parallel and cut me off. This is not where you want to go.

B (12:47)

No, I love this.

C (12:48)

I think websites will be important forever, but they will now they have two audiences to speak to. They're going to speak to humans. You need. People are going to want to hear your story your way forever. That will continue to matter. And so you need visually stunning, emotionally evocative, engaging experiences that matters. And they need to speak to machines. Now we've quietly been doing that for a while by doing SEO, right? And speaking to Google. Okay, now there's a bit more to do where, you know, the data we provide them will be used more broadly. So I don't think websites are going away. I don't think SEO is going away. I think the balance will shift. So like, you know, for the human, you're going to engage them. For the LLM, you are serving them concise, direct, to the point, well structured semantic content. For SEO and aeo, I sometimes just want the data and sometimes I want the synthesis. The fact that AEO is probabilistic and not deterministic, that it can hallucinate, knowing that there's huge progress being made on that, like every month, every quarter. That's a thing. Sometimes I want to see it the way that brands were talking about it. So sorry, long winded answer. My opinion is that no, SEO will totally not become just upper funnel. In fact, as we know, it was well known for being lower funnel with great intent for a long, long time. And my bet is if you roll the clock forward in our prep, you were saying three years. Three plus years. Opinion of one. I think they come together.

B (14:18)

I love it.

C (14:19)

I think just like we need software to be probabilistic and magical, like all the demos we see today, we also need it to be deterministic. I don't want my pay being probabilistic right there. There's a place for sass and deterministic things. I think the same will be true in this experience where you've got upper, middle and lower funnel. There are times when I want it and there are times when I want the synthesis. That's opinion of one. How does all that land for you?

B (14:47)

I fully agree. I think today we're in a very turbulent period where we're seeing these channels through a very myopic lens and these are only gonna evolve very, very fast. And I think that even AI, discovery and SEO have already merged a ton in the last 12 months. If you would have asked me this time last year if I would have thought Google would incorporate this much AI into the search results, I would've told you no.

C (15:15)

No. Yeah.

B (15:15)

And we've already seen that just accelerate exponentially. And I anticipate it to happen more. Talking about AI and the utility of it, we see a ton of AI in helping brands and companies build content. Now. One of the interesting things is I think that content as a structure and as a form is coming through a lot of change and it's evolving as well in this current period of time. As you think, guy, about content, what types of content are going to be the best at performing, Answer results and getting into answer results and how should we be thinking about adapting our current content? And I want to even expand this a little bit to the visual expectations on our pages. How should we be thinking about expanding our visual expectations and the content to meet these types of results from LLMs and even AI overviews that happen in Google?

C (16:08)

Great broad question to the first thing you said. We all have access to LLMs to create content. Mercifully, thankfully, creating a zillion pages just with LLMs is not ranking, it's not doing well, it's not showing up. 85% of the AI content on Google is human generated or human with a little bit of AI. Like the part that's pure AI or heavily AI. By that I mean LLM generated is not doing well, which makes me as a consumer happy. I think the importance you brought in the visuals at the end of emotionally engaging with your prospects will only increase because that more qualified traffic. The LLMs in some sense they have a better read on what's working and what's not because they see the follow up questions. They see right, we are opinion of one not based on data. I think we are more likely to go back to an LLM to ask follow up than we might be in Google. May or may not be true. I don't have data for that. What I do have is the data on like hey, people are more qualified. I also see that as we all learn to speak Google, our queries were four words on average. With LLMs they're 23 words on average. And that's growing. And if OpenAI and perplexity as they are rumored to be doing create browsers, then they'll have phenomenal data on what's resonating and not. And so I think the although to your point, this question almost always at the moment is focused on the words. The overall experience is hugely important. Okay, on the words I think we shift from hey, how can I create content that will be relevant to my buyer and make sure these keywords are in it? To clearly and comprehensively answering questions asked throughout the funnel. And so I think our call recorders, like we happen to use gong, I think we should go to GONG and say hey, what are the questions being asked at every stage of the funnel? And let me then go answer those comprehensively. Longer, not for the sake of being longer, but longer for the sake of potentially having better cosine similarity, better alignment with the questions being asked and then answer a basket of questions around them. What are the clusters of things people might ask? Answer them at multiple levels so that like topic clusters which were baskets of keywords now become baskets of questions. 1, 2 structure where the easiest part is the stuff you do on the page. Like you can signal to the LLM structure by using phrases like in summary or making it easy to parse with bullets. Or consider we often put the takeaway at the end. Consider putting that up front to make it the cost for the LLM to crawl lower. Three, I think we need to personalize content Even more. Right. That's always fundamentally the goal of AEO SEO. And all the stuff we're doing is to attract people's attention and get them to buy stuff or do things. Creating content by industry, by company size, by Persona, by product, by use case, whatever that may be, that's relevant to you, that is even more important. And localization, being able to do that at scale is hugely valuable. I think that one of the biggest changes, and I'm curious your take on this is the metadata. You know, we all did a bunch of stuff for Facebook's open graph. I think it's hard to overemphasize how important that is for LLMs. And it turns out I know many of us know. But for those that don't, Google, Bing, Yahoo, Yandex and others came together a few years ago to create one. They called it schema. If you go to schema.org, you will see that. And being able to it's got structure on like hey dear LLM, here's a blog page, here's a bio page, here's an event page, here's a product page, here's a case study page. And we happily make that really easy to do dynamically in our cms. We also make it easy to auto generate a sitemap XML robots Txt we were the first to support LLMs Txt opportunity wise. The last thing I'll share is that there's so much opportunity here. 88% of top sites haven't implemented schema markup seeding visibility to those that have yet at the same time, 73% of first page results in Google use schema and content with structured data gets 42% more answer engine citations.

B (20:24)

Right. There's a real piece that you said that stands out to me and that is the story, right? If we can't build good stories, it doesn't matter what the content has on it, right? Like it has to have a good story. And then if you can layer in these other components of utilizing schema and data to convey that story, producing the right visuals and the right experiences around that content to enhance that story, those are the things that I think a lot of SEOs are missing the mark on. One of the most eye opening studies that we've seen is we work with a healthcare client. Their largest source of traffic from LLMs goes to their patient stories. These are real stories about a human being that went through a very specific health crisis and how they overcame it and what happened and what their story is. Those are the ones that get them LLM traffic. It's not the silly disease page that's 70,000 words long and has way too much context and at the end of it just tells you that you're gonna die anyway. But that's where all mortality heads anyway. It's a crazy thing to look at the data and realize that these real human narratives are the ones that get LLM traffic and not these dictionary like pages of content. So that's what I think is really revealing about where LLMs are going and how they're taking traffic today.

C (21:52)

To your point, I think they are absolutely valuing sort of authentic human content. More to your point about the case studies, with the discussion about traffic decreasing, Google went to defend itself. Their head of search publicly said, hey, organic clicks from Google are flat now. She included AIO AI overviews in that, which in my humble opinion is legitimate. That's organic traffic. And she said, hey, there may be a mix shift towards what you just said. Authentic human content where people have case studies, they have stories, they have original opinions. And so I think that's super interesting. Yes.

B (22:32)

Let's take this into this next question, which is how do we measure success in this new world, this new era? And you mentioned Liz's article or presentation on traffic being flat. You also mentioned Bain's study that SEO traffic is down 15 to 20% in a world where there is seemingly fewer clicks. We can debate and discuss that how should success be measured when it comes to these new AIO or AI experiences in search and discovery?

C (23:04)

In my humble opinion, ultimately we are driving pipeline or revenue. And so yes, the things we were measuring on our sites, we should continue to measure, right? What's the conversion rate, what content's working, what's not. But in terms of what's happening off our site to what we were saying earlier, keyword ranking as a concept doesn't exist. And so I think the first place to start is thinking, what questions do I want to be part of the answer for? Just like we figured out what keywords we want to rank for and then start measuring ad hoc, sort of, am I showing up? Am I even being mentioned? And then, you know, there are 15 + vendors helping us measure systematically. Hey, are you showing up for these? I think you move to share a voice, you versus your competitors, which I think is similar. Start. It's sort of the logical equivalent of keyword ranking. And unlike SEO, where it's just your words, sentiment matters, you know, how, how are you being brought up in this synthesized content and the companies are, you know, ones we know well, like Semrush and Ahrefs, along with some new ones like Scrunch Profound and a bunch of others. And I think that the most sophisticated measurement then becomes guiding, just like sophisticated SEO measurement did. It's it, it, it points you to, hey, where should I invest, you know, time and energy creating content, reacting to things that are happening is where my head goes. The one last thing is I think listening expands to places many of us may not have invested in it as much before because the LLMs have chosen to value that content as original, authentic content and sometimes do deals with them. Reddit being the classic example, Quora, Wikipedia. If your Wikipedia page isn't up to date or accurate, that's worth spending time on now even more than before.

B (24:50)

Yeah, let's talk about that last piece for a second because I think brand plays a big piece in this, right? Whether it's your brand sentiment or information in like a Wikipedia or a Reddit or even more broadly speaking, the way consumers for larger brands try to discover and get to know your brand. What types of prompts are commonly relevant to my brand and is it that people want to understand my pricing is that people want to understand the quality is that people want to see reviews or context from the market and that is all relevant to your brand. How do you think about the brand piece of measurement when it comes to the LLM exposing and sharing context about a brand?

C (25:41)

I personally go to a place similar to sort of brand measurement in general in that I'm like, did my messages get pulled through? Right? We, we spent all this time as we were talking about earlier, carefully crafting these stories and these points that do not cover our entire product on purpose because we're focused on engaging differentiators. Did that get pulled through or am I just a list of features to the LLM 1 and 2, when it does talk about me, meet broadly our brand, you know, is it positive, is it neutral, is it negative, is it saying I'm a good one at whatever category I'm in, I'm a bad one is where my head goes. And I think, you know, some of these tools that we were talking about earlier are good at doing that. They are helpful at doing that.

B (26:23)

Right. And it's a starting place, right. At least to understand what the sentiment or direction is on your brand and how eventually you can change it or craft that new narrative about the brand.

C (26:34)

It's interesting to hear you say eventually because. And I wonder if you've seen the same thing. I feel like different than SEO where You had to build authority measured in links and other things here. Because the LLMs currently are basically crawling every inference time, every time a question is asked. You can change content on your site and it can show up in the LLM, like this afternoon, tomorrow.

B (26:58)

That's right.

C (26:58)

The flip side of that coin is to the discussion we had about Wikipedia. If you've got out of date content, which I know none of your listeners have on their site, this is only for other people out there, but. Or if someone else has out of date content on your site that could be cited for a really long time, very long time.

B (27:16)

And I think that ultimately the piece of the puzzle that we don't have a lot of context on is when you think about authority and you think about your own brand's authority, how much of that is crafted by your own owned assets versus the earned components. Those are the things that we still don't really understand yet. If you look at all the mentions and all of the citations that are being utilized to craft a story about our personal brands, our company brands, that really becomes the new recipe for unlocking how an LLM can craft the response about your brand.

C (27:55)

And to your point, not only do we not know it, it is changing rapidly as we're both talking about how dynamic this space is at the moment. I think that, like the highest level of maturity today is about bobbing and weaving as things are changing. Because it's like there's a reason it makes the news when Google changes their algorithm because it's such a mature space. And that happens, you know, several times a year, not several times a week. Roll back the clock 15 years to the beginning of search. Happen all the time.

B (28:22)

Exactly.

C (28:23)

Which you and I both remember.

B (28:24)

Yeah. So let's talk about technical foundations. I mean, you brought this up earlier. We talked about schema, we talked about like structured data on pages. But really I think technical optimization is going through a set of changes as much as the content components that we talked about before. When you think about answer engines and a website's readiness, whether it's what you guys are doing at webflow, whether it's just generally how websites are creating integrity behind their tech, what are some of the foundational optimization or efforts that webmasters, web designers, Web developers, and SEOs should take into consideration to have a strong.

C (29:04)

Technical foundation, some of the things that come to mind. A couple of thoughts, and I'm curious where your head goes. Like there are some that will say, here are 500 things you should go do for technical SEO. In my humble opinion, it tends to be a much, much, much, much smaller number that matter. And my hunch is that the same will be true here. I think some of the fundamentals also remain true. You signal quality by having a super fast site globally. We Happen to have 250 plus data centers around the world. You need a secure site, secure sites that auto update patches and make infrastructure improvements without me, the webmaster needing to manually update a bunch of plugins like that signals quality. Okay. In addition, we have plenty of work to do on basic stuff. I mean, 25% of the top ranking pages are missing meta descriptions. Like how is that possible in this day and age? We were talking earlier about the metadata. If you can dynamically generate schema, which is easy to do, for example on webflower, sitemap, XML or robots Txt, that's really, really valuable. LLMs Txt I don't see any evidence today. I'm curious if you do that the LLMs are using it. So I wouldn't put it top of your list, but I would put it somewhere on the list because it seems illogical that they never would. You of course need a site that is mobile responsive by default, which is not new news and you want excellent Lighthouse scores. Someone poked me back the other day saying, hey, I don't see any evidence that the LLMs value speed of website. That may very well be true. I can't imagine that it would stay true. I can't imagine that that will continue to be the case even if the interface to the LLMs does not care about latency as much. I have no evidence one way or the other. The humans enjoy responsive fast sites and that inevitably will be part of the equation as they leave the LLMs and do they bounce right back or not? What are you seeing? If you were to guide people on a few things, where would you point them?

B (31:13)

You know, one of the things that we're becoming very prescriptive in the technical conversation is understanding your core audience and what their needs are. You know, and I think that's one of the things that tech has never taken into account. And I think candidly, and it's not to bash Google, but I think Google has created a very, very huge disservice to much of the web community in basically making really broad blank statements. You have to have a mobile responsive site, right? And that is definitely true. But there are certainly some industries where mobile responsiveness is really, really, really important. And there's other industries where it just doesn't matter. If I'm in the news business, you Gotta have a very mobile responsive site because people consume lots of news and information on their mobile devices. If I sell giant contracts for construction equipment, probably not the most important thing for me to be super mobile responsive, do I have to have a mobile friendly website? Probably yes. But I doubt someone is buying a 400 foot boom crane from my mobile device. They're probably going to do that from a computer and there's probably a lengthy process in evaluating that purchase decision. So I think that we don't really spend enough time talking about our technical capabilities and the audience's expectations around that. Even speed is a highly debatable set of layers. And to your point, having global footprint and speed is very important for some types of sites, but then there's others where it isn't the main root purpose or intent of it. Does it have to be within striking distance of responsiveness? Yes, but not some lightning fast world class speed site. I think these are the things that we miss as web developers and we miss that context. And I don't think that Google and much of the community does enough to talk about context setting when it comes to technical investments. Same goes for some of these opportunistic investments like schema and other ways of sharing data. What really matters to the audience is the best starting place in my opinion.

C (33:28)

That I think is super insightful and full stop. And I recognize and appreciate that at least it seems like those broad statements are backed up by the data that they are valuing mobile responsiveness. Superfast sites for everyone.

B (33:43)

Yes.

C (33:43)

Even if to your point, I'm buying a million dollar 400 foot boom crane that I'm not going to buy on my mobile phone.

B (33:50)

Exactly. All right, speaking of planning, let's talk quickly about planning and how we start to build our plans for this AI discovery world. So if we're building a 30 day plan and we're a business that needs practical steps to organize and level up our AEO or AI results game plan, what are the three practical steps we should take?

C (34:16)

If I were to boil it down to 3, I would humbly suggest 1. Go experiment with answering some high website visit intent questions. So go listen to your sales call recordings. If you have a sales team, go read support emails or sales emails. Pick a handful. And by high website visit intent I mean things that an LLM can't summarize for webflow, that's often. Show me example websites. It might be the human case studies you were talking about in the medical example earlier. 2. I'd go try out schema on a few key pages and see what impact it has. Google mentioned that Rotten Tomatoes put structured data on forgive me, I have the numbers elsewhere. It's some 100,000 or 250,000 pages and they got a huge number more. I want to say it was 25% more organic traffic because they were implying because of that. And three, I'd go renew a bit of the effort on go ask ChatGPT or Gemini or Anthropo, your favorite tell me about my company, see who's cited and then go look and see hey, is the content there accurate? Are they talking about us in the right way? If not, consider literally it's human to human going to reach out to them. And I'll add a corollary as a fourth we built a rubric of how to optimize for aeo. You said hey, here's a framework to logically put it all together around content, technical authority and measurement created made up five levels around them. If you go to webflow.com resources aeo that is there it's not a pitch, it's genuinely like intended to be value add and you can put in your domain and we have a ginormous prompt we wrote to LLM that will go look at your site, go look at a bunch of other places and give you an assessment, give you a point of view on here's what we see, here's what seems to be missing and if you're only going to do two things, here are the things we suggest you do and here's a longer list hopefully to help activate exactly the question you're asking for each site individually.

B (36:22)

I love that I love A I love the fact that you're using AI to speed up the process of learning. That's great. B your last point there around how to understand your brand and what your own brand is being, how it's being depicted in the these LLM responses is such an undervalued exercise. And I think in that first 30 days, if an LLM can't get your brand messaging right, if it can't communicate what you do as a product or as a service, as a company, if it can't explain how your pricing is set up, then you're going to go it's going to take you a while to really get traction because that should be the basics of getting the LLMs to respond accurately about who and what you guys do. And so I think that's one of the interesting things in that first 30 days is getting that comprehension so critical and that's a great place for us to wrap up this episode of the Voices of Search podcast. Huge thank you to Guy from webflow for joining us. If you'd like to contact Guy, you can find a link to his LinkedIn profile in our show notes or at the voicesofsearch.com for more information about Webflow, visit webflow.com if you haven't subscribed yet and want a daily stream of SEO and content marketing knowledge in your podcast feedback, hit the subscribe button in your podcast app or on YouTube and we'll be back in your feed soon. Okay, that's all for today, but until next time, remember, the answers are always in the data.