How Brands Can Prepare for the Post-Human Web

This is the Everyday AI Show, the everyday podcast where we simplify AI and bring its power to your fingertips. Listen daily for practical advice to boost your career, business, and everyday life. This podcast is sponsored by Google. Hey folks, I'm Omar, product and design lead at Google DeepMind. We just launched a revamped vibe coding experience in AI Studio that lets you mix and match AI capabilities to turn your ideas into reality faster than ever. Just describe your app and Gemini will automatically wire up the right models and APIs for you. And if you need a spark, hit I'm Feeling Lucky and we'll help you get started. Head to ai.studio/build to create your first app. For the last few decades, you've had great control over how others view you and perceive you on the web, right? You put some information out on your website. Hopefully over time, if you're investing in this search engine optimization thing, brands are going to see you exactly how you want to be perceived. But that was pre-generated AI. And pre what we're really experiencing now is large language models turning into answers engines. And it seems like brands are maybe starting to lose a little bit of that control over exactly how their potential clients or customers are going to discover them online. So how can brands prepare for the post human web, right? Maybe in the future when a lot of content is maybe written or created by AI and scraped by AI and presented by AI. How can brands still have a little bit of control over how they show up, if they show up and how that may ultimately influence their potential customers or clients? It's a topic I'm excited about, but I have an actual expert on the show today to help walk us through it, and I'm excited to get to it. Welcome to Everyday AI. If you're new here, what's going on? My name is Jordan, and this is your daily live stream podcast and free daily newsletter, helping everyday business leaders like you and me, not just keep up with everything that's new in AI, but how we can make sense of it to grow our companies and our careers. If it's your first time, welcome. It starts here with the unedited live stream podcast. But if you want to take it to the next level, our website is where it's at, youreverydayai.com. We're going to be recapping the highlights from today's conversation, as well as keeping you up to date with all of the other AI news. Well, let's just get straight into it and help me welcome my guest to the show. This is going to be a good one because I've actually used the product before in the past, and it's really good. So help me welcome to the show Michael Walrath, the chairman and CEO of Yext Inc. Michael, thank you so much for joining the Everyday AI show. Hi, Jordan. Thanks for having me. Great to be here. All right. So for those that aren't familiar, can you explain a little bit what Yext is and what you all do? Sure. So Yext is the leading brand visibility platform. We have a set of products and services that help brands to make sure that their products and services and offerings, particularly with a localized bent, are constantly discoverable and available. And that's, you know, often we talk about this in terms of SEO, content creation, reputation management, and social media management for major categories, in a singular data led platform to optimize your organic brand visibility. Yeah. And I'm sure we're going to uncover a lot of that here throughout the rest of the conversation. But I'm going to skip straight to the end. Michael, what is the right answer? And brands, prepare for the post-human web? Get your data right. I mean, you know, the problem that brands are going to have with this is that, you know, much of what we've done to create content is in ways that human beings appreciate and machines don't. And what machines like is really solid structure data. You know, they're not big on heavy images. They're not big on dropdown navigation boxes. They're not big on a lot of the things that make the human web usable for human beings. And so we need to go through a second wave digital transformation that starts with, you know, all of your authoritative business data has to be structured and prepared to be distributed so that your brand can be discoverable full stop. And, you know, you've been leading Yext for a decade and a half, a little more than that. Walk us through, you know, kind of what was going on in your head, right? We had the chat GPT moment of November 2022. But obviously at that time, it wasn't connected to the web. You know, we saw perplexity and probably over the last year or so. Every single large language model has essentially turned into an answers engine. So walk us through from your perspective how that's kind of unfolded in terms of your work. Yeah. Well, so I've been here for 17 years. I was chairman of the company for 12. I've been, you know, for about the last four. So those numbers don't add exactly, but it's over. I'm as an English major, not a math major. I, you know, we've seen a really interesting shift. And I actually ran the search business at Yahoo before I invested in Yext. So we saw Google rise up and basically consolidate the entire discovery market into a single model ethic platform, right? And by the way, this made the world really easy for brands, relatively speaking, because hey, if I'm discoverable on Google, I'm winning 92% of the time, right? We started talking about this a couple of years ago, and I think people thought we were crazy at first. We said, look, Google's going to be disrupted when it comes to answer engines. There's a better mousetrap and that better mousetrap is going to be the combination of a natural language engine with context and memory. And what we've seen over the last two years is an unprecedented rise of obviously chatGPTs leading the way here, but there are so many of these and there are going to be so many more now. And they might use overlapping technology, but the consumers are going to choose answer engines that make the most sense for them. And I think we're just at the very beginning of what's going to be a very rapid, very broad fragmentation cycle of how consumers find information. And ultimately, kind of like how I opened up the show, organic SEO, right, even though it's always been like, oh, there's these 200 factors from Google and everyone's always trying to fight against the algorithm. But for the most part, it's been a little bit more understandable, right? How are you seeing brands tackle now this big gray box of generative AI? What's working? What's not? Yeah, so right now, actually, it turns out, you know, a lot of the same stuff is working, right? So a lot of what we were doing, what brands have been doing for SEO for the last 15, 20 years is very much working. And a lot of reasons that's because behind the sort of grounding of a lot of these AI experiences are the same search indexes, right? So, you know, it's tempting to say, hey, SEO is now just answer engine optimization, right? And we've seen this huge crop of startups going after like basically saying this is something totally different, right? And look, I think there's value there. I think, you know, measuring the visibility across these AI engines specifically, we have a product that does this, you know, there's lots of startup products out there. You know, we refer to it internally as it's admiring problem, right? And this problem is going to get harder and harder to solve. And so what we anticipate is that the ultimately for brands, the most valuable thing is going to be, okay, let's admire the problem. Let's understand what is it that's causing us to be not as visible as we want to be, you say answer engines, but then you have to be able to take action. You have to be able to do something about it. And it turns out that a lot of the same tools that you use for SEO can be applied in addressing the opportunities when it comes to AI visibility. So I want to get slightly dorky, but not too much, because one thing I'm, you know, always talking about is kind of the concept of how obviously large language models slash answers engines are very different, right? You know, Google or search engines for the most part, aside from personalization and localization are largely deterministic, right? Large language models, well, they're generative, right? You can ask the same question or query 10 times, you might get nine different answers, you might get one or two, right? With that in mind, how should, right, if I'm the CEO of a small or medium business that's traditionally just been writing blank checks because of good SEO, but now you get this generative nature, how can they tackle that and even understand it? Yeah, well, I think it's really interesting, right? Because I think it depends on the question you're asking, right? So if you're saying, you know, what's the best sort of off-the-shelf artwork for me to hang in my new apartment, right? That's not too expensive and looks nice, right? That's a very subjective question and like there's probably lots and lots of answers. And so if you ask an AI engine that question, then you're going to get, and you do it 10 times a row, you're going to get very different answers, right? If the question is, and actually I think that's value, right? That's the probabilistic nature of it and it's valuable, right? I think where it gets frustrating sometimes with answer engines and this is getting better is when you say, look, I'm really just want the best burger that's closest to me right now that also has a milkshake, right? And if you keep asking the question, it will keep delivering you often different answers, but those answers rarely get better, right? And so I think that when it comes to fact-based, when it comes to what we would call structured database queries, the more localized it is and the more specific it is, the more important the memory element of the AI engine becomes. Because it's what is, how can this answer be delivered better because of what the answer engine knows about? And why is it, it's a really valuable assistant if I don't have to tell it that I want gluten-free or I want vegan or I want carnivore or I want, you know, I want electric cars, I want gas cars, whatever your sort of your preferences are. And so I think that's the huge hurdle for marketers is they are going to have to understand that like every question asked of an AI is not a question, right? It's much more than the words that are there. It's the words that are there, which can be much longer than a search query combined with everything that that AI already knows about me. So it's the words that aren't there that makes this such an interesting marketing. It's such a great point and it's something that I think makes it a hard target to lock down for marketers, content creators, business owners, et cetera, right? Especially when it's changing, right? Like as an example, Claude just rolled out memory last week, right? So, you know, probably the level and the depth in the context of what is now being, you know, kind of risen up by Claude, just as one example is much different. How can, you know, business owners, marketers, SEOs, et cetera, start to learn more about their demographics, right? Because it used to just be, all right, well, you know, we're going to use these tools and, you know, we're going to optimize for search engines. But now these answers engines know everything about us. So how can they go about learning more about who their customers or clients are? Yeah, and, you know, I think that's an interesting question, right? Because marketers have so much data now about who their customers are, right? And what we, you know, the world we've been living in with SEO is one where you have to hope that that data is being transferred in some way through the query, right? So the near me query establishes location and it makes the query much more targetable, right? And advertisers, marketers who are trying to be visible, they will either really care about being part of the answer or not, depending on whether or not they have a store nearby, or they sell insurance products in that area, or they have, you know, a medical office nearby. But I think, you know, in a lot of ways, so much of that data is going to become harder for marketers to target specifically. And so what they're going to need to do is they're going to need to inform, right? And the way that they're going to, so it's like, how do you talk to an answer engine if you're a brand, right? And the answer is you talk to it through data, right? And everything you tell it has to be in the form of here's structured information about my product. So that when you know more about it, because you the answer engine, you the AI agent are going to know more about the consumer than I'm going to be able to target. So what I can do is I can tell you, hey, I have a, I'll give you a silly example, right? It's not enough to have a menu anymore. You have to have different versions of that menu that make it really easy for the AI to understand. Okay, like, I know my human is a, is a, if I'm thinking as the AI, I know my human is a vegan, right? So when they ask me, where's the best place for me to go, go for lunch for me? They're not saying vegan, but I know they're vegan, right? So I'm going to look for places that have, that have delivered me structured information that says, hey, we have a vegan menu, right? Here it is. And that's like a really simple example of how you feed structured data into the AI engines so that they service your brand more often. And we could do that for any, any industry, any, any company. You know, I'm curious, right? Because even though you are ultimately providing these types of services, you know, for clients around the world with the X, how are you even, you know, your company? How are you, you know, going out there and kind of, you know, dogfooding it so to speak, right? Totally, it's the same thing, right? So the, you know, it's, and it's, you know, I think sometimes we're so focused on doing it for customers that we actually forget to do it ourselves. But we have to do the same transformation. We have to make sure that, you know, every piece of content that we've ever created that's relevant to our products and our services and our teams is, is, is distilled down to its most elemental structured data form. And is, is available for, for, for the, for the AI answer engines, just as we've always attempted to make it, you know, it's like, you know, and this is why I say like a lot of it's still the same, right? Like, you got, you got to have schema, you got to, you got to have the page has to be structured, right? It's got to, it's got to be rich, you know, now it's all about snippets and chunks and things like that. I don't think those practices change, but I think the way we think about how granular we need, we need to get and how real time we need to get is, is really important. You know, this is maybe a question super specific, but I even know a lot of small business owners. Because, you know, I used to kind of work in this field, they know which pages, right, are bringing them in traffic, bringing them in clients, bringing them in customers. You know, is there blanket advice that you can give to people. So, like, should they just be looking at that one page and expanding it out for large language models, should they be taking that page and, you know, making 20 different versions of it, you know, spreading out, you know, the quote unquote link juice or SEO juice. Like, is there a best recommendation for something like that yet? Yeah, look, I mean, I would defer to, you know, sort of, we have a lot of hundreds of SEO partner agencies who are really, really good at that. What we do is we power them with, with the tools to be able to, for example, you know, make pages around offers, make pages around intense, you know, make all these menu pages seamlessly and easily. So, so part of what our platform does is basically if you, if you put any information into the Knowledge Graph, which is the data center, the data, the data center of our platform, you can, you can then seamlessly create a page, a published page using that, right. So, want 10 menu pages with different, you know, carnivore, keto, vegan, you know, whatever it is, you just, you publish those pages and now you, you, so I still think high quality content wins. And I think we're out of, you know, I think, I think AI will do a particularly good job of figuring out that like, look, if you just republished the same article and you put, you know, from 2019 and you put 2025 in it, like, you might trick me today, but you won't trick me in the long run on that. And so, so I think the game, you know, everyone's going to have to elevate their game a little bit in a world where these things are just going to keep getting smarter and smarter about like, what's the right answer to the question? Where's the good content? Where's the good data to give me the answer to this question? Yeah, and the answer to the question, especially in large language models is getting more and more important, especially as we're going to eventually start seeing ads inside large language models, which I want to get to in just one minute, but before we do a quick word from our partners. It's a problem I hear all the time. The gap between the AI champions and everyone else in your organization is sizable. You might have half a team that wants to fine tune models by hand, and the other half doesn't know what an API is. How do you get them working together on AI that moves the needle without creating a security nightmare? That's where area really shines. They built one platform with three ways to work. Your developers can go full pro code and build custom agents with Python. Your business analysts can use the low code tools or your domain experts who've never coded, they can use the drag and drop no code builder. Everyone's building in the same secure governed environment, no shadow IT, no security gaps, and because their model agnostic, you're not locked into one vendors ecosystem with area. You can even A/B test different agents against each other, try different models, and when you're ready, deploy to production on the same platform. Your AI strategy should unite your team, not divide them. Check out area in today's show notes or on our website for a free trial. Go to airia.com. Get rid of the AI gap and move forward with a more resilient AI ecosystem. What a transition, right? Ads. What's going to happen when we look at how brands want to show up? You said earlier in the conversation, a lot of it is still good old-fashioned SEO. What about when ads come? How does that change? Well, so I think we've seen almost no advertising in the AI experiences so far. So, perplexity, I think, is the one who's played around with this the most. And I think, not surprisingly, you know, what they've done is kind of, I'd say, experimental, right? And so, they really just kind of put a panel of ads that are contextually targeted, presumably to the content that's being created, right? This technology is not new. You know, GoTo was doing it in 2000, you know, 25 years ago. You know, I think there's a couple of really interesting things about what's going to happen when advertising comes to AI experiences. So, you know, I'll sort of give you three that I think are really interesting. One is that this will be the first medium where advertisers don't get to create the copy. And a lot of brands are going to be really uncomfortable with this idea of, you know, what am I doing? I'm buying ads, but I don't get to decide what the ad is, right? That is going to be a very uncomfortable feeling. But if you want to be able to buy ads in AI, I think you're going to have to come to terms with that. So, that's thing number one, I'll come back to it. I think thing number two is what we're going to see is that there's always been a pretty big difference between SEO and SEM, right? How do I manage the organic visibility across search and how do I manage my paid marketing across search? That will change because of, you know, I think the first thing. And they're going to look a lot more alike. So, we're going to have this world where you're going to have to decide, am I going to advertise on an AI? And by the way, I'm making things up here and I could look very stupid in the future and you'll have a recording of it if so. But like, I fundamentally believe that every ad medium we've ever seen in the last in the digital environment, the ad mimics the content that's being inserted into. This is why search ads look like search results. This is why magazine ads look like magazine pages and TV ads look like television, right? So, what does a generative AI ad look like? Well, it looks like a generated answer, right? And the third thing is this will create all sorts of really interesting conundrums for the answer engines, who will have to get really smart about, you know, where is the line where consumers going to feel really good about a targeted ad? And where's the line where I might not be giving the best answer, but I'm giving the one that pays me the most. And so when you combine all those three things together, we're going to see is just like massive disruption. And, you know, I think from our standpoint, one of the most exciting things about that is that in order to influence through paid advertising, and I answer engines in the future, you're going to have to have really well structured and really detailed authoritative data, because that's how you're going to influence what they say about your brand. So, you know, if I'm Land Rover, I use this example, a lot, but I asked you to be to the other day, you know, what, what, what SUV should I consider buying drive a Tesla Model S I like it, but maybe I want an SUV. And they say, you know, ask me as many questions as you need to and then give me a great answer. And so it, it does this thing. It thinks for ask me a bunch of questions. Why do you want an SUV? Are you going to be off roading? What do you like about the Tesla? What don't you like about the Tesla that it thinks for seven minutes. And then it delivers me five recommendations. Right. And what it says, and then, and it always like now it always ends with a question, would you like me to get you some more information about any of these, these options, right. And so you can go as far as you want with it. Yeah, you know, tell me more about that Cadillac, right. And, and at some point in the future, what's going to happen is to say, Hey, turns out I can have one of these Cadillacs delivered to your house for a test drive at three o'clock today, would you like me to do that. Right. And there's a moment there where you're like, wow, that's an extraordinary consumer experience. Like, I'm, I'm shopping for a car. And this thing's literally going to schedule one to come to my house. Right. And there's another moment that we're like, whoa, like I'm being monetized. Right. Cadillac is paying to have that car delivered to my house. Right. And I don't know, like, you know, sort of, everyone's going to have different feelings about that. But here's what I know. If you, if you're a brand and you want to participate in advertising that way, the only way you're going to do that is you're going to give them structured information about where are those cars, where are those test drive cars, where do I want them test driven, what, what areas, what demographics, all those types of things, and how much is it worth to me. That's all structured data. And by the way, that's going to very much mirror the way that you manage your organic brand visibility today. So, you know, Michael, you brought up something interesting on the consumer side, right, because traditionally when it comes to advertising, well, and even just traditional SEO, right, it's, it's keywords, it's, you know, consumers type them in brands bid on keywords or key keyword phrases, right, broad match, whatever. But this is different, right, like the future when we talk about AI answers engines, whether it's delivering us organic things for certain brands or in the future ads, it's not keyword based, right, it's intent based, it's memory based. How should consumers be thinking about this because like you said, like, oh, is this helpful or is this just an ad? Yeah, and look, it's confusing, right, because on the one hand, it feels really, you know, it feels stalky, right, like, you know, these things remembering everything I tell it, right, like every time I ask it a question, I tell it like, you know, you know, hey, I need, you know, my lips are chapped, I need, I need, you know, so what's the best chapstick, like, you know, it's remembering things about me, right. And so on the one hand, it feels creepy, right, I sometimes I've actually stopped doing this because it scares me, but like, trying to explain this to people, I'll pop open chat to BT or one of the other ones that I use because we play with all of them, you know, in front of people and say hey, sell everybody what you know about me. And like I always wind up stopping it because it like it gets really personal, right. And it's also a great test to make sure you you're not sharing information sharing but but like I think that this this line between, like, you know, I get way better answers the more it knows about me, but it can feel a little creepy is is is probably the, you know, sort of the thing that every consumer is going to need to decide for themselves. And, you know, speaking of decisions, it seems like maybe it like maybe this is no longer a hot topic but you know you're the expert. I think earlier on maybe in 2023 2024, a lot of brands were just blocking right. And they're like, oh, you're not going to take our traffic right. Yeah, but it doesn't seem like that's the case anymore or maybe it is but can you tell us maybe if there are still some some brands out there they're like yeah we're blocking, you know, we're blocking big AI is that a good move is Yeah, no, I think obviously there's been a lot of heat around this topic. And I think, you know, having worked in, you know, and run companies and built and run companies and advertising like, you know, I have an acute sort of, you know, I'm feeling of alignment with the publisher community who create all this amazing free content and it's all ad supported, right. And so I think there's a tremendous amount of concern in that world about, hey, if I give the AI access to all my content, and then what they're doing is repurposing it. You know, I lose my traffic and I lose my revenue stream and I can't afford to pay the great content creators who are creating that stuff right it just wraps my whole business model. I believe that's true and I think that there have to be ways to do that. I think what's happened and we see this occasionally with our customers and, you know, our customers aren't publishers, right, in the sense of they don't typically they're not typically ad supported. They don't really care in most cases whether the product is bought through a third party service or directly. They don't really actually care if you visit their content as long as you access the content. And so it's a completely different strategy for a brand who sells a product or a service. And unfortunately, I think people are painting this with a really broad brush and telling, Hey, you should block all the crawlers you key should not if you sell a product if you sell a service if you're a local business. If you're a small business like the last thing you should be doing is blocking a crawlers, you need them to crawl your content, and you have to be encouraging them by giving them more content to crawl. Right. If you are a ad supported publisher, I complete, you know, I think if I were an ad supported publisher, I would be blocking a crawlers, because I don't want my content being stolen and repurposed and trained and things like that. And so, so it's, it's, I don't think it's a one size fits all answer. But the first thing we tell brands is like, if you're blocking anything, you have to stop. Yeah. So we've, we've covered a ton on today's show, Michael. I mean, we've gone from everything from, you know, SEO and SEM and how those lines are starting to blur, you know, talked about intense versus keywords and even how, you know, AI, large language models are changing and, you know, how intent and memory impacts that. But as we wrap up, what's the one most important piece of advice that you can give to brands that want to prepare for the post human web? Yeah, I mean, I kind of started with it and I'll end with it and expand on it. So, so if you don't have your data right, right, if you don't have access to your data, if you don't have, you know, a singular distributable set of, you know, a set of authoritative data around your business that's easily distributable out to old and new endpoints, Google and Bing and Yelp and also open AI and Claude and Perplexity and all these, these different endpoints, then you're just not ready. And the pace of change here is going to increase, not decrease. And so, as Apple eventually comes online with a intelligent agent, and look, I think Reddit will build one, I think lots of community driven things will have some form or some version of their own consumer, you know, intelligence or some sort of AI agent or answer engine. We're just at the beginning of a fragmentation period here. And it starts with getting your data right and then you got to roll with it, got to understand that like, you know, you don't need a Reddit strategy that everyone thinks need a Reddit strategy today they're not being cited in detailed queries. Right. It's great training data. It's not authoritative citations that will change if they build their own answer engine. But today, it's a bad use of your of your of your resources to people trying to trying to build citations through Reddit. And this is a highly controversial thing because at the highest level people see that Reddit are showing up as citations but they're not at localized queries they're not at specific queries. That's your own content. That's your listings information that's your your review responses and things like that. And so this is going to, I guess that, you know, long answer, get your data right and be ready for tons and tons of change. And this is kind of what we wake up and help our partners do every day. Getting ready for change. It's all we can do every single day. But Michael, you helped a lot of brands out there who are struggling with that question. Be prepared for some of that change. So thank you so much for taking time out of your day to join the Every Day AI Show. We really appreciate it. Thanks for having me. Great chat with you. All right. And if you miss anything, y'all, it's all going to be in our newsletter. So if you miss some of those gems that Michael was dropping on her head, make sure to go to your everyday AI.com. Sign up for that free daily newsletter. Thanks for tuning in. Hope to see you back tomorrow and every day for more everyday AI. Thanks y'all. And that's a wrap for today's edition of Everyday AI. Thanks for joining us. If you enjoyed this episode, please subscribe and leave us a rating. It helps keep us going. For a little more AI magic, visit your everyday AI.com and sign up to our daily newsletter so you don't get left behind. Go break some barriers and we'll see you next time.