AI Leadership Insights: Reshaping SEO and Online Discovery with Chris Brownlee

Hello and welcome to the latest edition of the Techstrong AI video series. I'm your host Mike Bazar today with Chris Brownley, who's senior vice president of product for Yext. And we're talking about AI search and how it may be changing the way we discover things and interact with content and maybe just turning our lives upside down.

Hey Chris, welcome to the show. Oh, thanks a lot for having me, Mike. We've all seen, I think by now the AI search results that you see, and sometimes they're generated alongside Google and other times people are just going in and using AI search tools without even going anywhere near Google.

Um, how prevalent is this at this point? I mean, are people not looking at the links at all? Um, 'cause and how accurate or how, uh, comprehensive are those results?

Yeah, I, I mean, being in tech, we probably we're more on the leading edge than, than anything do I think my mom is going in here right now and, and doing this? Not, not just yet, but I think it's very, very close. Like we all know Apple's baking it into to phones.

Google has it with Gemini, and so to me that's where it kind of crosses the chasm. But, um, but obviously the big distinction here is for the last 10, 15 years when you do a search, you get a bunch of, of blue links essentially, and you gotta go hunt and find what you think is the most appropriate for you. And what changes with AI search is that you ask a question and you get a very direct answer or you get prompted to give a very direct answer.

And so there's a lot less exploration happening, um, when you're looking for an offering or a product or a service or, or information even. And some of these tools are now purporting to do research for you on specific topics and uh, you know, they're searching a bunch of different, I'm assuming websites that come up with some sort of conclusion. And I think that's based on the reasoning engine that they provide.

But how do we know that they're kind of comprehensive enough to generate something that's accurate or, or most people just gonna say close enough? Yeah, I, I mean especially with the, the last two years, I think the biggest fault that everyone would point to with any of this generative AI tech is it's prone at times to hallucinate, uh, and to give inaccurate information. And so I think that's obvious.

One of the biggest areas that these, uh, companies are focused on is how do we get rid of these, these hallucinations? And so if I'm a a a business online, I want to think a lot about like, how do I make sure it has the most, um, information about me, my services, my offerings, and, and that it's being consumed as as accurate as possible so that when it does get pulled from, it has far, far less chance of turning into a hallucination that gives someone bad information about my business or, or how I could show up And some competitor that doesn't even exist. Right?

A hundred Percent, yes. So I think there are a lot of people who are having this gut check moment 'cause they're looking at the way this shift is gonna happen and they've spent years on SEO and buying Google AdWords and all kinds of things over the years, and they, you know, they, they kind of understand how that system works and how to drive themselves up a response list. But, um, how do I do that in the age of ai?

How do I kinda get myself inserted into an AI search that may not, you know, generate the level of links or, you know, I mean, I don't think most people make it past the 10th link anyway, but at least I was in the game. Yeah, a hundred percent. Like there was a dashboard that you could log into to try and like update your information to know how you were even ranking in the first place, to know what that was turning into for business impacts.

And none of that exists. You can't log into chat GBT or perplexity and manage how you're showing up there. It's a complete black box.

Um, but going back to the last point, these LLMs are still feeding off of the web. They're still crawling this information and they're doing it as frequently as possible so that they are getting as accurate information as possible. And so there's still ways to impact how you show up to have, ensure it has lots of, uh, good, rich, accurate information, but where you used to focus on just Google as your single pane of a most important single pane of glass for ranking, now you need to think about it across all of these different LLM AI search backed interfaces because they're crawling data and giving import to different sets of data than what Google used to do.

And so AI search for chat, GPT versus perplexity versus Gemini might use very different citations when they're actually citing this information as well. So knowing what data is being pulled and used, it becomes really important for someone trying to manage their performance online. Are there formats and ways that I should configure my data so that it is more consumable by these AI search tools?

I mean, what, what are we talking about there? Absolutely. Like, if, if anything more of the game is not just about creating content for your end consumers or people who's gonna view the content is can you even make sure it shows up in the first place?

Um, and so this is a lot about making your content easily readable by these crawlers. So it's simple things like having good schema markup on your pages or being as as clear as you can in the data, um, to help them target it to the right customer. Um, I think about opening, I just said that they were gonna go more into product catalogs and being able to show what, what offerings you have to for more retail goods.

To me, I know that chat GPT builds up a memory about who I am. It knows I'm, uh, who I am, what I'm interested in. And so when it makes product recommendations going forward, it's going to take that into account.

So if I'm a marketer, I wanna try and add as much information about who my product is suitable for, like different use cases, different ways that we might want it to be shown up because it's going to be looking for such precise answers to give back to its end user that are very tailored to them. So I need to know, I need to provide that information that makes it have that best chance to show up to meet those criteria. I think back in the search days of when it was just straight up Google, we were always trying to figure out, well, uh, should I change my content often or less often?

Because if I don't change it so often, maybe Google will like it better that way, but then recently maybe once more updated content. So it tends to favor that with, um, AI search. Um, it feels like it's continuous, but you know, how frequently do I need to kind of update my environment to keep that position that I might ultimately get within the memory of the AI search engine?

Yeah, these are all of the experiments that we're trying to run today to figure out like we know with pretty good high accuracy what the things you are you need to do to rank in Google today. Um, and it is like frequency of updates, it's consistency. Um, the biggest things that we're seeing right now is, is just having really well structured data that's easy to be consumed and crawled so that it can be well indexed to even just show up in the first place.

Those seem to be, uh, from all of our findings, the things that will give you the most chance of showing up. If, if you don't have any of that in the first place, like good schema markup or um, well structured data, um, you, that's just step number one, uh, in the battle. What's your feeling about, you know, looking into your crystal ball?

How soon will it be that, you know, the way we currently use Google search will just fall by the wayside become antiquated or, uh, will this be like a demographic thing where, you know, old, old guys will be using Google search and the kids will be using AI search? I can't see a world where the traditional ways last long and I, I can't even imagine it being a, a, a generational thing. The ease and simplicity of not having to think about what you're searching for in terms of a Google search query where you have to like, you kind of gotta shift your brain to put it in terms of like, okay, how do I make sure I get the results I'm looking for?

Where with these LLMs I can ask a very human based prose question. Um, I can have a conversation to refine that search and know, um, how to get it more, more tuned into what I'm looking for. It's just such an ease of use solution that gives you the answers you want essentially immediately, rather than having to hunt and peck for these things that I, I like for me, in terms of timeframes, I, as I kind of touched on, like as soon as this becomes baked into our daily day-to-day devices, I think that's when it's gonna really cross the chasm.

I think, uh, even just this last month chat, g BT was the highest downloaded app in that month alone, um, I think within Google. So it's, it's already happening though. I think this is gonna happen, but won't I just wind up with a, a lot of AI agents that are anxiously standing by waiting to help me and every time I'm working on something they're gonna say, Hey, do you want me to launch a query and looking for this, that and the other?

And you know, they'll make quote unquote helpful suggestions. I don't know how annoying that might get, but um, you know, that relationship is gonna change where they're not gonna wait for me to do something. Yeah, a hundred percent.

Well it's funny 'cause you think of like sci-fi movies and stuff, you think of it as just being this one all seeing, all knowing ai, but really what we're seeing is the rise of these like agents where it's lots of many, many different ais that are specified to do one job really, really well. Um, I see it like the internet being 99%, these agents just like communicating with each other and finding information before it actually gets to the end consumer at the end of it. And so yeah, what does this look like if you were to go like five years down the path?

Ah, it's hard to, to even predict because you will just be interacting and behind the scenes I think it will be, uh, agents interacting with agents, interacting with other agents and, and what does that look like? And I, I couldn't even forget guess right now in my crystal ball. So what is your best advice to folks about how to kinda get, you know, get down this path a little bit, especially, you know, I think end users will figure it out on their own, but if I'm trying to be the company that, you know, maintains some level of attention in this new age, how do I, uh, get started?

It's citations. They've always been a thing in traditional SEO if anything, they're more important than ever right now. And it's knowing though what sources are being cited by the ais and then making sure you're on top of those places.

So it's interesting, I've been doing a lot of testing around like how do financial services businesses show up when I search for them or search for a specific agent, for example. And it's adding sources like, uh, websites I'd never heard of before, um, for just like, like address phone address. It's, it's, it's using MapQuest all the time.

Like who's, who's managing MapQuest? com, like things that probably me, uh, as a financial services agent and probably paying zero attention to right now, but these are the things that are representing my brand. Um, any information about me.

And so I need to actually figure out what those sites are and get on those to manage all of my information because those are the things that are, are powering how I show up. And so you need to know broadly all of these different citation sources and then make sure they're all being up to date managed and that they also have the same information. Because as soon as you get dissonance between multiple sites that there's less trust in that data and there's less likely the, the LLM is going to surface that information because they don't want hallucinations.

That's obviously one of their biggest concerns. So it is still managing your presence online, but it's almost even more important than ever if you wanna show up in AI search, which almost feels backwards, Will we fall back into the same trap we had with Google search where we all wound up using the same search engine? Um, or will AI search be more diverse and there'll be more tools and more engines and we're not gonna have this, uh, you know, concern about monopolies, et cetera, et cetera.

Uh, it's such a hard one we're, 'cause we're in such early days of it. Um, you know, back when it was NetSuite and Naje Eves and all of those, probably no one forethought going down to a single tool and I think we're still very much in that explosion phase before we get to any kind of, of coming back in there. Also, there's gonna be a lot of lock-in.

I think that starts to happen with, with this memory storage. I think, yeah, it's just too early to say like, I would say plan for more fragmentation for the next five years before I would say like go all in on just optimizing for one. OpenAI has a clear lead in terms of AI search engines and like referral sources.

Gemini obviously being, having a lot of power behind itself too is being with Google. Um, but I I, I personally, if I was a marketer, would not put all of my eggs in one basket for the the next while. Do you think I might see agents in search tools that are optimized for specific, I don't know, vertical industries for salespeople or, you know, if you've got a hobby, maybe somebody's gonna tune up an AI search engine that's optimized for that specific hobby 'cause it's trained on that particular kind of data.

How granular can all this get? Yeah, um, there's lots of tools popping up like that with, with retrieval or augmentation based on like very industry specific data. Um, and even some of those tools are just getting baked in to the tools that we use.

You can create projects now in chat GPT where I can load up a whole bunch of information and then be able to query against it and get, get insights. And this is where I think the agents will, will really have their time and place where they're really deep experts in doing one thing really, really well and then can hand that off and can be reverified and rechecked. Um, and that's why I don't think it is just like one tool to rule them all.

It will be each of these individual things that will, will do their, Their jobs, uh, very specifically. All right folks. You heard it here.

Hey, they used to say, you know, the medium is the message. Well, the medium is changing. It's called AI search now.

So stay tuned and we'll see what happens. Chris, thanks for being on the show. Uh, thanks a lot, Mike.

I enjoyed it. All right. ai video series.

You can catch this episode and others on our website. Until then, we'll see next. Now.