← All episodes
Nov 9, 202651 min

How Generative Engine Optimization Changes the Way Customers Find You with Justin Inman

Episode overview

AI search visibility means little if the recommendation gets your business wrong. In this episode, Chris sits down with Justin Inman, founder and CEO of emberos and a former Google enterprise growth leader, to explore how AI-generated answers shape brand discovery. They discuss generative engine optimization, the shift toward zero-click search, and why accurate recommendations depend on signals across your website, reviews, publishers, and creators.

Justin explains how his team approaches AI brand management by tracing the sources behind inaccurate answers and identifying which information to correct. The conversation covers testing customer questions across chatbots, the limits of visibility scores and content shortcuts, and the challenge of measuring AI-assisted purchases. Justin also shares how Granola, Claude, and HubSpot support his meeting preparation, with human review remaining essential.

Listen to learn what to check first and how to assess whether your brand is being recommended for the right reasons.

Key takeaways

  • Justin Inman says AI search answers draw on a business’s wider digital footprint, including its website, public relations, social channels, creators, and publishers. Conflicting information across those sources can lead to inaccurate descriptions or recommendations.
  • He recommends checking AI answers from a potential customer’s perspective by testing relevant questions across several chatbots in an incognito session, then tracing inaccurate details to their sources.
  • Because a buyer may research with AI and later purchase through a company website, Inman says customer surveys can help capture AI’s role when referral analytics do not show the full journey.

Show notes

Chapters

  • (00:00) Introduction
  • (03:10) How AI Recommendations Change Brand Discovery
  • (12:00) AI Brand Management and the Brand Knowledge Graph
  • (13:39) Generative Engine Optimization Beyond Traditional SEO
  • (21:18) When AI Gets Your Brand and Products Wrong
  • (24:29) Tracing and Correcting Inaccurate AI Answers
  • (26:33) Running an AI Search Visibility Audit
  • (30:32) Why Internal AI Agents Need Fresh Context and Human Review
  • (37:04) AI Meeting Preparation with Granola, Claude, and HubSpot
  • (40:49) Testing AI Recommendations with Real Customer Questions
  • (42:36) Measuring AI-Assisted Purchases in a Zero-Click World

Find Out More About Justin Inman

AI Tools and Resources Mentioned

Questions answered

How can a business check what AI says about its brand?

Justin recommends using an incognito session to ask realistic customer questions across several chatbots. Compare the descriptions and recommendations for accuracy, then investigate the sources behind any errors.

How can a company respond when AI gives customers inaccurate information?

Justin describes tracing the answer to the influential pages or channels behind it, correcting information the company controls, and addressing other important signals across its digital footprint. He does not promise that any one change will control every AI answer.

Full transcript

This automated transcript may contain recognition errors. Listen to the audio for exact wording.

Read the full transcript

Speaker 1

If you're not part of that recommendation set or or know how to like optimize your brand for what this world is going to look like, you could effectively disappear.

Speaker 3

You mentioned a timeline of 12 to 24 months. Is it urgent today though?

Speaker 1

Absolutely. And these chat bots are becoming rapidly the front door of the internet.

Speaker 3

But you're telling me that people are now, they're not necessarily going to quote unquote Google it anymore. They're using the models kind of as a search engine to find the answers. Correct.

Speaker 1

And my mom, who's like in her 70s, just booked a trip on Muse without even me prompting her to do that.

Speaker 3

Yeah. And it tells me that I'm going to have a lot of work to do over the next 12 months to make sure that we're squared away.

Speaker 1

Right now, and it's not all doom and gloom. I think you still have some time. What's happening is most people are getting their information from that single prompt or multiple prompts and never having to go to your website. And by the way, AI is lazy.

Speaker 3

Hey everybody, and welcome to the next episode of Using AI at Work. My name is Chris Daigle, and I'm your host. And I think anybody listening to this has certainly heard of SEO. And more recently, you've probably heard of AEO or GEO, which is essentially the equivalent of SEO, but for search engines. When people are looking for a solution that your company provides, how do you get the large language models, Claude, ChatGPT, Gemini, whatever you use, to make sure that your company is one of those preferred recommendations? Because when a potential customer asks Chat GPT, hey, who should I do business with? You need to know if your company is coming up. And if it does, is the information even right? So today I'm joined by Justin Inman. He's the founder of Emberose. And we're going to be talking about how AI describes your business, what you can influence with that description, and how to tell whether improving those answers is bringing you customers. So, Justin, welcome to the show. Anything that you think that the uh the guests need to know before we get started with uh digging in on this?

Speaker 1

Um nice, uh nice uh uh be here, Chris. Thank you for having me. Uh let's see. Any anything that I think that it's it's definitely the wild, wild west right now. Uh and that the internet is going to look very, very different in the next probably 18 to 24 months. And I think we're already starting to feel that. So I think uh it's it's an easier kind of understanding of like what the space is looking like if you could put a label on it, whether that's AEO, G O L L M O, like there's all these acronyms. But I think the most important thing for most uh listeners to to understand is that this is a fundamental shift of how people will discover uh your brand, get recommended, and and actually make purchases. So um similar to kind of the shifts to mobile and social and YouTube, like this is one that's happening quite rapidly and uh and it will have bigger implications than any of uh those other kind of larger consumer shifts that have happened.

Speaker 3

And for the listener, what qualifies Justin to talk about this is he spent roughly 10 years at Google on their enterprise sales teams helping the largest names in the industry navigate these kind of shifts. And like you mentioned, whether it was search or YouTube or mobile or programmatic advertising. So um his background certainly qualifies him to share kind of what's happening in this new shift. So you mentioned that things are starting to look different and that they'll look a lot different in 18 to 24 months. What will be different?

Speaker 1

Yeah. I and yeah, as you mentioned, like I spent I spent about uh 10 years at Google kind of like helping marketers through these like larger transitions, and I saw where this was headed, and that's kind of the re the core reason why we started Embros. Like, um, because it is it is your entire digital footprint. So it's picking up on uh the PR, the social, your site itself, uh, you know, the creators you're working with, like could be everything. And so um we've really kind of put the the uh put a platform together um that's called the operating system for AI brand management. Because if you think about it in that sense, like it is brand management across the board, across all of these different signals. To your point, and just around what the what the um what the what the what the internet and kind of the consumer experience is going to look uh like in the next uh you know, anywhere from 12 to 24 months, you know, AI, AI overviews and AI mode and these chat bots are are becoming rapidly the front door of the internet. Uh what's happening is most people are getting their information from that single prompt or multiple prompts and never having to go to your website. And so uh it breaks a lot of different areas of the marketing funnel because if no one's actually going to these websites, what is the role of advertising and programmatic advertising if there's not a not a human to actually see those ads? But it also begs the question of okay, what's the role of the website in the future? Because if no one's actually going to your website, they're just getting the information directly through these chatbots and uh AI mode, and and you know, we're in a very much in a zero-click world. And I think that you're gonna start to see that kind of rapidly change the way that brands and and marketers are thinking about the role of their website because no one's actually going to it or small, a small amount of people that are going to it. So, my hypothesis, uh, and I think they're already starting to test some of these things at Google, but that the website is going to look very, very different for you and me, depending on kind of like the consumer experience of like a customer experience or like um signals that we bring in when we hit the website. So uh what I mean by that, it's it's gonna be a highly customized um view, meaning it will be like uh very kind of like uh very different views, not just like uh product feeds, but like what is specific that I would want to see, and websites will dynamically update based on the signals of a uh of the traffic that's coming to that website. So I I think that's where we're headed. I think it's a little bit more of a futurist, but I don't think it's that far away. I think it's probably you're starting to start to see some of that in the next 12 to 18 months. But that's strictly because of this situation, which is no one's going to the websites. And so if they do go to your website, they want to be able to like have a customized experience to get people to go to it versus just getting the information through these chatbots. Does that make sense, Chris?

Speaker 3

Yeah, and it tells me that I'm gonna have a lot of work to do over the next 12 months to make sure that we're squared away. Um, so just so I understand it, because like man, I'm heads down in this stuff. I'm not really paying attention to what because I saw an A16Z report last week, maybe, that said only two percent of US households have a paid Chat GPT or clawed account, right? I thought it was higher than that. So but you're telling me that people are now they're not necessarily going to quote unquote Google it anymore. They're using the models kind of as a search engine to find the answers.

Speaker 1

Correct. And just because only 2%, that means 98% are still using the free model. Yeah. Those consumer, those consumer shifts are happening rapidly. Uh, you know, HP Morgan has come out and said one third of all digital purchases are gonna happen through agenda channels by 2030. So that's literally right around the corner, like three years from now, right?

Speaker 3

Does that mean that an agent will be I'll tell some input environment what I'm looking for? It'll come back with options and I'll say, yeah, option two, and the agent will actually go and complete the transaction.

Speaker 1

Or you are prompting in an LLM and you make that purchase directly through it.

Speaker 3

Is that for B2C, or would that also be for like professional services or B2B? Really?

Speaker 1

Yeah, yeah. I mean, I thought that number was like really high until probably about three weeks ago when when Muse popped, and I was like, oh wow, this is the number one app store. Yeah. My mom, who's like in her 70s, just booked a trip on Muse without even me prompting her to do that. Like, hey, you should download. She just did it organically, and I was like, oh wow. Okay, this ship has left a station. Like, and honestly, one third might even be low. Um, you know, depending on kind of uh how quickly people start to adopt these things. But it's kind of like similar when you first got the iPhone or like a smartphone, you know, like you you were like, Do I need it? Do I need it? And then once you've got you had it, you're like, okay, my life has changed forever. I I think once people start to get familiar, like see how kind of these agents are making their lives easier and like the the some of the decisioning processes are sped up and there's more free time to do X, Y, and Z, I think it's gonna be a floodgate. And I think we're already starting to see some of that.

Speaker 3

So you you kind of referenced it, like we can do that now, but the human still has to initiate that activity. Um, is is I guess the human friction going to be an impediment to this, or do you think that it's going to like everything else with AI, it's just gonna catch and take fire?

Speaker 1

I think it's gonna be catching and taking fire. I like there's a concept that we kind of are building towards, which is uh, you know, we're talking about a lot of like futuristic things right now, but like I actually don't think it's that far in the future. But you know, we are, you know, typically, you know, it went from search, right? You would raise your hand, make that search. Now it's prompts, and you're kind of writing the prompts in there. Pretty soon, and we're already starting to see this with Gemini and with your kind of like daily briefs and things like that, is that you're gonna be served. And so I, you know, they're gonna the uh the chat spot du jour, whichever one of you you you use, is gonna say, Hey, Justin, you live in Los Angeles, you have two kids, you're free this week and go see this movie. And so that is a whole other thing, right? That the marketers and and brands haven't really even thought about, uh, because that that's the future. That's where it's headed, where it's like it's gonna understand your preferences, what you like, read your Gmail receipts and all the other stuff, which is everything else, and we're gonna be all served. Like, hey, you just moved, like use uh this particular product or go buy this this uh couch because you're looking for a count, whatever it is. And that is a space that, like, if you're not thinking about this now, like you, it's an existence problem, right? Like, you need you need to be thinking about this. So it if you're not part of that recommendation set or or know how to like optimize your brand for what this world is gonna look like, you could effectively disappear overnight because these agents are gonna be looking at, and by the way, AI is lazy, like it wants to go to like the easiest source, right? And just try and find the easiest ones, it's not some sophisticated. Obviously, there's a lot of amazing advancements, but like when it comes to like recommendations and things like that, I would say it's pretty lazy. Uh, that's why it goes to like reviews and things like that. Uh and and so I would say, you know, everyone that's listening needs to be thinking about this because it is it's going to shift the way that marketing and brands actually exist and probably the next like 18 to 24 months.

Speaker 3

So the lazy comment makes sense because obviously they're looking to conserve compute, and me as a consumer wants to conserve my usage limits and that sort of thing. So rather than have it do deep research on every query about the best bagel shop in my neighborhood or whatever, it's going to be looking for those kind of cached or or review environments where it's like doesn't have to do all that compute to give me the answer that's probably a match for what my query is. Is that kind of active?

Speaker 1

Typically what it does, like if we're like that type of question, it would just do a grinding search, very similar to how Google search works. So just scan the web, like uh the internet, like get all of the different signals then determine that with that recommendation list. Now, if you're asking like deep, kind of like technical questions from a B2B standpoint, it might use some of its already kind of learning, um, you know, training data, maybe a mixture of both. Like that it just depends on kind of like how the the advancement of what that prompt looks like. If it's like what time does Home Depot open, like it's just gonna go straight to the web, right? It's not gonna, it's not gonna use the compute power to be able to do that.

Speaker 3

So, and I guess before we go too much further, um, I think it would be helpful for people to understand why you're qualified to talk about this. We talked about the Google background, but what are you guys doing at Emberose?

Speaker 1

Yeah. So we uh we help uh brands and uh large agencies um like through this transition. And this transition is really making sure that not only are you showing up and getting recommended, but you're getting up recommended accurately, because I think that's a big, big issue. Yeah. And so we have created uh basically a technology, it's very similar to how Google Search works, but we have a uh a graph-based uh technology which looks at things from a relationship standpoint. So we can see basically all the influencing factors online of how it relates to that specific entity that we're tracking, whether that's a brand or product or movie title or streaming title or you know, you name it. We then um can understand how all of those like relationships interconnect uh and see which ones have the kind of uh highest influencing factors, like to be able to change. And so we track everything over time. We say, hey, this surface is something that you need to really be thinking about um because it is driving negative sentiment or it's driving uh X amount of results, change this, and here's the exact steps to be able to change it. Um we have predictive capabilities, so we we predict kind of that where the answer is heading. Um so we say, uh, before you even make that change, we have a predicted lift to it with confidence levels, time cost associated with it. And then once we uh the actual change happens, we track it in our platform and show you the tributed lift by each one of the changes that you do.

Speaker 3

Okay, so it sounds like that the time at Google kind of taught you the foundations, but but what what have you had to rethink with the search engine environment as compared to the large language model environment? What's different?

Speaker 1

It's totally different. I mean, like let's let's call what it is. Uh SEO is 98% Google. You know, it's not it's not good. So you're optimizing one surface. This space is very dynamic. Um, you have model updates that are happening across the board from different engines, like some from different labs. Uh, you've got you know um wrappers that are basically uh now consumer facing that you need to optimize. So it's like there's just there's there's thousands of different AI services that you need to be thinking about. It's not just Google anymore. Um and I come from Google, I have a great relationship with Google, like they're they're a partner of ours, like uh, but it is it is it is it's they're not the only game in the town anymore. And I think it's rapidly, you know, um changing. And so, like, you know, to your to your questions just about what have I learned, you know, SEO was pretty much about your own site too. It was like, hey, how much can we get our own site kind of optimized to uh organically show up? A lot of people start playing in that game. You had to kind of pay your way into it then with with the paid search ads, and yeah, you know, it is what it is today. Uh maybe a little bit of PR and like some of these other things. This is a totally different beast. Like you are looking at the creators and even potentially even this podcast, right? So the podcast that you're on, the the creators, the the YouTube channel, the site itself, the the uh the paid signals as well. So like there's a lot of like paid stuff that pops up, which is like connected TV can drive Gen AI search. There's just a lot there. And so um really kind of thinking about it across like how we we look at it is cross your entire digital footprint. And so it's paid, owned, earned, and shared.

Speaker 3

Okay. So there's certainly some listeners here that say, oh, we have a managed service provider, they but they also do our SEO and they said they're doing our AEO. What's the difference between what they're probably getting from either their internal team or an MSP that says, Oh, yeah, we can do that for you too, with what you guys do?

Speaker 1

Yeah, I would say like um AI visibility and AEO GL that like it's become commoditized. Like there's just a million different vendors here. We have people like vibe coding in their basement, some sort of tool like mix like uh is incredibly um not sophisticated. And um, you know, for me, it's like it's great that they're getting exposure to it, but it's not enough. Um, because a lot of these tools are basically like either SEO tack ons that are like, okay, well, we show you we show you like a AI visibility score. All of that stuff is in the rearview mirror, which is like, okay, well, now what I do. Um a lot of a lot of them are also like giving you basically laundry lists of optimizations of saying, here's all the 5,000 things you have to do. And you're like, I don't have time for that. Like, what what what is this? Um and then I would say the third kind of bucket that happens quite a bit now, uh, which is uh similar to like early days search keyword stuffing and like all of these like hacks and tricks that people seem to be doing, Reddit and like all this other stuff, but um, which is basically saying like the solution to this problem is more content. And so what we'll do is we'll create like not us, but like other kind of like tools out there, we'll basically create uh uh AI generated content and push it out to like either your site or syndicated publishers, um, and it's basically spamming the entire internet. Yes. I know, I know my my my my time at Google, like you do that enough, uh it not only destroys kind of br the brands, but it also uh destroys your organic search, which is still important today, which uh could get you into big trouble with Google with spamming and everything else. So, like there's just a lot out there, there's a lot of noise, and I think for for us, like we um we look at things and with foresight. So we we try to predict where things are headed. Uh, we try to predict like where sentiment is heading and like what the citations that are driving this, what where how long we think that will actually last. And like um, I would say we're more sophisticated in that sense because we're trying to give foresight uh back to like a lot a lot of our uh customers versus saying, like, hey, let's make decisions on what has already happened, uh, which is already frankly too late in a lot of cases, because then you're basically just looking at uh, you know, weeks or months old's data and saying, okay, well, let's make an informed decision, but that by that time the models have already updated and like the the the the uh the there's gonna be some sort of model update that's already happened. So you're you're basically making decisions on things that have like uh from a rear view mirror like viewpoint.

Speaker 3

Sure.

Speaker 1

Uh so for for us, like, but I would say that's the biggest difference for um that we with all the other stuff, we're closed loop, we we give you predictions, we give you kind of like all a lot of other things on top of it, but I would say if it distills down to one thing, it's foresight.

Speaker 3

So this is kind of like in that saying, you're skating to where the puck is going. Correct. And everybody else is missing the puck completely because they're kind of standing still, or as you mentioned, looking backwards on what worked a month ago. Okay. Correct. And it, you know, I I'm I'm there's a message uh underneath what you're saying, and that is this is a lot of work. Your MSP or your SEO people, they're not putting nearly as much effort into this as you are. They're probably using the models to say, hey, what can we do to optimize this? And the model, I mean, I don't know when the last time Chat GPT got updated, but it wasn't yesterday, right? I mean, the the models got updated, but not the data sets that you'll be getting the information from. I think I last time I checked, it was like April of this year, was the last time they they updated the data set that it's working on. So if you as a business owner listening to this, you're like, oh, we got an SEO guy, but he also does or she also does AEO for us, they're probably doing that. They're going to Chat GPT and saying, hey, do an audit. What should we do? Does that sound accurate?

Speaker 1

I don't want to speak to all the SEO like an AEO people because I don't know. I'm some of them are very sophisticated and like uh understand the importance of this, this, this space. I will say I a lot of the tools in the market today, or even if you just did that example that you just mentioned, it gives you optimization recommendations, like maybe they're dated or whatever, but it looks at um, and let's just say your brand is Nike, but it looks at only Nike and tunnel vision. Like this is what Nike needs to do because it's just Nike. How we look at things is from a relationship standpoint, because that is what this world is. It's it's the how the competitors are related to the entity or Nike, how the PR and the creators and every everything interconnected. And that that's like that is what gives you the infrastructure to play in this like meta muse world in the future. If you're only looking at things in a silo, it it might look good, but you're basically grading your own homework because you're only tracking certain things that are matter to Nike and like all this other stuff. So I think in general, uh a lot of these like vibe-coded tools and um maybe like SEO tackons uh are very limiting because they really look at um things in like a single tenant. They they basically just like look at with blinders on as well.

Speaker 3

So okay, this is good. And for the listener, like we've kind of gotten to the point to where hopefully you've got context on what what what's happening at Embros and and across the the market in general and why it's important. But when we uh you know, when we initially did our pre-interview a while back, you had mentioned some scenarios about AI getting it wrong. Um Yeah. Like walk me through what was, I mean, if you have if you have an example, you don't have to name the company necessarily, but like what was what does wrong look like normally?

Speaker 1

Yeah. Uh there's a lot that's happening, uh, whether that's uh a hallucination. Okay. And the the the hallucination is Is done because uh there isn't coherence across all of these different surfaces for a brand or product. And so um site A will say it's it's uh uh 399, site B will say it's 499, site C will say it's 699. So it's like it just doesn't know, and so it will just make it up. That's what a hallucination is. Andaccuracy is like uh we see it all the time, like um an accurate CEO for the company, like uh when was it founded? What do they actually do? Like all of those that those are like factual like issues that need to be cleaned up. And then there's this like this larger kind of concept, which is uh, you know, we work in a lot in entertainment. Um we're we're live across like kind of almost, I think, 10 different verticals now. So like we kind of see it all. Uh, but I would say uh the most interesting one I think is grounded because everyone knows entertainment and loves entertainment. Um, but uh LLM's missed genre of films about 20% of the time. And so if you're a studio, or if you're you know, it's the same kind of issue across different verticals. So like I'm just using the entertainment one, but uh if you're a studio and your your title is labeled as a violent thriller, but it's actually a dark comedy, you basically alienate a whole population of people that said, What should I go see this meeting, uh, this uh this meeting, this this weekend, because uh you it's a very specific audience that would go see a violent thriller versus a dark comedy, right? And then you you also have the the the the the negative effects of this six months down the line when people are like, I like this movie, and you're not part of that recommendation set because you don't have the right genre. And so, like a lot of people are saying, like, what should I watch? Like, what should they give me recommendations on a book or music or whatever, right? It's a big recommendation engine now. Uh, and studios, entertainment companies in particular are waking up to the fact that, like, there's a lot of issues with this, and we need to get our kind of core what we do and what our product is and what our title is. And another great one is like, where where can I watch this show?

Speaker 2

Right.

Speaker 1

And a lot of time that's at that's interactive. It's it's it's attributed to a different streaming platform or different channel, network broadcaster, whatever. So it's like cleaning up those those things, I think, has been really interesting. Um, but yeah, I mean, like I said, we see we see it uh pretty uh much across the board for for every vertical that we we touch.

Speaker 3

Okay, I'm gonna throw you a little bit of a curveball then. Let's talk about maybe like a $60 million manufacturer. Let's let's say they're manufacturing, I don't know, some sort of a chemical, right? And AI misunderstands what they sell or who they serve. If they came to you, walk me through how you uh would fix that.

Speaker 1

Well, it's not really a curveball because that's exactly what we do. But I appreciate I appreciate you thinking that's like a that's a hard one. But um uh it is so they would come to us and we would basically say, here is the here's how you're showing up across all the major LLMs, Chach PD, Gemini, Grok, Complexity, whatever. Here's what your AI identity is. Here's exactly kind of what uh the the AI surfaces are kind of recommendations are coming back with. And here is uh maybe the inaccuracies, right? It's the wrong chemical. Then we would show exactly what is siting, what what sites or what surfaces are driving that inaccuracy. So is it a Reddit thread? Is it a old blog post? Is it a PR thing? Is it what what is it? And then we would we would work with that company to say, here how it here's how you fix it. So maybe it's something they can't fix because it's uh it's like a uh Reddit thread. But uh we can then go and say, okay, well, but if we change these other surfaces that we know drive the answer, then we it that Reddit thread will not be as pro prevalent, right? Like we can then kind of uh make sure that things are cleaned up across the board. Um, so like, yeah, uh it will be across kind of all your different kind of channels, whether that's your own site or it's PR or it's your social channels or the creators that we're working with, and not for uh like a manufacturing company, but like just depends on kind of the vertical we we touch. But we'll look at everything that it goes in from a relationship standpoint and then give you the best surfaces to change because we know that that will have the biggest impact across the board. Got it.

Speaker 3

So, okay, somebody's listening to this, maybe you know, they're heading into the office or whatever, and they want to find out what does AI knows about us? What like how how would they do kind of like a light and dirty audit?

Speaker 1

Yeah, I mean, it's like the easiest thing. Uh I would just make sure you're going incognito because there's a lot of like uh memory right there. So um go across all of the different chatbots and see what comes back. Um, and I think you'll be surprised. Probably, I mean, out of 90, I think 90% of the brands that we look at, at least one of them has uh uh sorry, over 90% of the brands that we have. So we have about 20,000 brands in our uh uh brand knowledge graph, 90% of them have at least one inaccuracy. So uh, you know, there's gonna be something there that I think is gonna surprise you and something that you should obviously try to fix if it's um important to you.

Speaker 3

So, okay, in in that B2B environment, and again, let's let's maybe let's move to commercial construction, let's say. Um the developer is looking for a uh you know to expand into a new market. They've never really done business in there. Are they going to AI to start to say, hey, we're looking to expand into Dallas, let's say. Um, this is our type of project. Uh, who are the top performers in Dallas? Is that like so that's kind of replacing to some degree a Google search?

Speaker 1

Okay. Any type of research like that, that's all happening through. Yeah, I know. I mean, okay. It is the shifts are happening quite quite rapidly now across every vertical. I would say highly researched ones were the first ones to fall. Um, so you know, uh, I think it's something crazy. Like 89% of BDB buyers now use AI as the first and final touch point. Uh, you know, like uh legal, massive. Um, you know, financial, like credit cards, travel is now becoming huge. Auto is a massive one as well, where people are going there and saying, hey, I live in Los Angeles, I've got you know two kids. Uh what's the best car for me? Like there's just there, it's any type of like highly researched ones. Now it's like starting to fall on the other pieces too, which is like movie recommendations, like it's it's it's lifestyle stuff, lifestyle stuff, um, you know, best diapers for my my like sensitive skin baby, like that all of that stuff, like that's all happening across the board.

Speaker 3

So, okay, let's say I go do this, I get to the office, I say, well, let's just find out. I go into the incognito, um, and I say, let's find out what's going on, like where my company's showing up. If if I find something that's that's wrong, like how much control can a company realistically have over an AI answer?

Speaker 1

A lot. Yeah. I mean, we we um, you know, that's why we're in business, because we we can actually give it control, right? We we give agency back to what feels very much like a black box. Uh and so, you know, depending on who you're working with or how how that like how that company is set up or you know, the size of the company, I think there's little things you can start to control, right? Like so um, you know, seeing kind of what are the inaccuracies are, uh depending on what those are. Yeah. Yeah, just like just publishing those on your site, cleaning things up. Um but yeah, I mean, there's a there's a lot of control there uh that you can you can start to really impact kind of at least the most basic stuff. I think what it gets a little bit more complicated when you when you think about all the stuff that's out of your control, which is PR, uh maybe like um maybe not your own PR, but like publishers talking about you or Reddit threads or UGC or whatever, whatever it is. Um, that's where I think it's gets a little bit more. You need like uh a better understanding of like how all of these like other signals are coming in. But the site is that's typical SEO stuff, right? Like you can see what the inaccuracies are, and then you just go to your site and publish and and and make sure you update or for accuracies across the board.

Speaker 3

Okay. So for the listener, I hope you understand that like at this point in the conversation what you've done in the past for SEO and having you know blog writers or whatever so that you rank well for keywords, it is not the same thing. Um, so I want to kind of shift gears a little bit here, um, and and talk more about your experience with using AI as an executive, as a builder, and that sort of thing. Um, and in particular, maybe like what is something that you guys tried that you were convinced that would work? Like AI is gonna have a big impact on our operations, not necessarily on developing the product, but not the the algo or the rubric that you guys use. What was something that that you were convinced that would work that maybe it just didn't?

Speaker 1

Uh I would say a lot of things. Yeah, yeah. Uh, you know, it's like it's uh it's it's not a silver bullet. And I think a lot of people think like, oh, this is gonna um uh take away all the jobs immediately and all the you know, I think there's gonna be larger kind of economic issues that happen, but uh for for us, like you still very, very much need a human in the loop. Okay, um, and I think you know, we build internal agents and do all these things, but like if we're not getting the most updated context, it's like it's not as helpful. And like then they kind of kind of collect dust, and like you have to constantly be updating these things. Uh and I think for us, like just a learning as we go through it, you know, it is uh it does it takes a human to like make sure that these things are running correctly. Uh and it's not just like a set it and forget it type thing. Now, for basic stuff, maybe scan LinkedIn or scan articles and like do that stuff. But if you're talking about like truly like workflow and you want to make business decisions and all of the other stuff, you absolutely need someone that's managing those databases to ensure that you are actually making informed decisions, especially in a business like mine where it's constantly changing. Uh, you know, the the models are getting upgraded, like there's just competitors that are rolling out X, Y, and Z. And um, you know, we work across so many different verticals and like understanding different, like, the nuances between those verticals and like how what are the use cases? Because I would say every vertical kind of uses this in a very different way, right? Um, so it is it requires a uh a lot of patience, but also a lot of maintenance. And um, I think that's something that we learned pretty quickly and we're continuing to kind of learn from.

Speaker 3

So I imagine you're a lot like me that you you know stuff about using the tools and the capabilities and that sort of thing. And you kind of think, well, like some of these things, you're like, well, every everybody knows that, right? What are some of the things that that when you're talking to clients that you're kind of surprised that, oh, they've got a complete misunderstanding, or they weren't aware of this basic capability? What? Are you seeing any trends?

Speaker 1

Yeah, I would say it depends on the vertical too. Like uh, you know, it really depends on the vertical. I would like the the like the some of the um, you know, uh I would say the the most researched verticals that are already feeling the pain because no one's going to their site or uh that they know that they know this world exists because they've been they've been feeling a lot of pain. Ones that uh maybe haven't felt the pain yet, but it's definitely gonna come to them pretty soon, which is like organic trough tropic just falling off a cliff. I don't fully understand it yet. And I'm I think we're kind of blowing their minds when we come in. They're like, oh whoa, like you're right. Like I my my my own consumership, like my own habits have changed so much. Like, why are we not even thinking about this? Like, yes, I I get a lot less of that now. I would say six months ago, it was probably a mixture, like a mixed bag. I would say yeah, majority of people are kind of like understanding that the shift is happening and that we're in the eye of the storm right now.

Speaker 3

So when you're talking to them and they're talking about, like, you know, the not just about specifically about what you do, but generally about how they are using AI within their organizations? Are you comfortable like, okay, these guys are on top of it? Are you noticing that wow, most of these companies, and again, it it depends on the vertical, I'm sure, technology would obviously be more advanced. I would imagine entertainment certainly under threat with AI video, so they're probably paying attention to this stuff. Yeah, but are you have you also seen that shift over the past six months to where maybe six months ago they thought, oh, everybody's got a chat GPT license, we're good? Or are you seeing that they're now uh maybe asking more sophisticated questions, or you're hearing just anecdotally how they're using it, and you're going like, oh, okay, this company, like they get it?

Speaker 1

Yeah, it's a it's a it's a mixed bag. Um, I would say some are really, really advanced, uh, and uh uh some are not. But I would say the most recent trend that I've seen is uh a lot of like enterprises are now almost all of them have some sort of AI council where it's like they have different people from different teams that are managing kind of like the AI um and uh kind of engagements across the board, whether that's from like a workflow or a partnership or from marketing or advertising or creative or whatever it is. Yeah, but it all kind of funnels through this some sort of like AI council. Almost almost all big enterprises that we talk to, like have that now. Um, and I would say that's a net new thing, probably in the last six months. Okay. Now um, but yeah, I mean it's it's just it kind of depends across the board. I mean, um, you know, you think about gaming, like gaming, uh, there's a lot of news about that like this week. I don't know if you're a gamer or not, I'm not, but like uh we work up with a lot of gaming companies, you know, gamers hate AI. They like just despise it. But like, so you know, gaming companies need to be thinking about how they can leverage AI for like some of these things, but also not disturb kind of their larger yeah, like the like their larger kind of like gamer base. Uh you know, it just depends, it just depends. It just depends where they are as an organization, how vocal they are about these things, and yeah.

Speaker 3

Okay. Now, uh, one of the conversations that whenever I'm talking to execs or teams or whatever, everybody wants to know the tools, like what tool are you using? What right? That seems to be kind of like the because it's the AI magic, right? There's a thing that, oh, it does this kind of thing. And I I get it. But as a, you know, as an owner of a business, a leader of people, uh, somebody who is in a, you know, there was this quote, I think it was by Charlie Munger. It's like, go find someplace we're gonna have dumb competitors, right? You're in a an environment where your competition is very smart. Um what's some yeah? We won't name names. Um what's something just like in the running of your organization that maybe you used to do yourself, but now AI is doing for you that you found like something that you'd want to tell every exec that you were like meeting with a group. Hey, guys, you you got to start doing this.

Speaker 1

I mean, a lot of admin work. So like uh proposals, uh decks, um uh just yeah, planning, like um, you know, we can move so much quicker now just because we have context, like uh, and a lot of these kind of like more admin type things that you would maybe call preps. Like I get every call, I get like a uh automatically like who's gonna be in the call? Here's what we talked about last. You know, we have granola that's like taking all the call notes and it puts it right into Claude, and like that's connected to HubSpot, that's connected to like our larger kind of ecosystem. So like we MCP into everything, so it the it's pretty amazing what it can do. Uh now you need to check it, right? You need to you can't just solely rely on these things, but that saved a ton of admin time or like maybe prep time that you need. You know, I'm I'm going 50 meetings a week, probably. Um, so like it is it's a lot of like making sure I have enough time to like prep for these calls and things like that.

Speaker 3

So you're showing up prepared when otherwise it your life isn't as busy as I mean, not to say that's not busy, but it's not as busy as it would be if you were not leveraging these for every interaction you're having with a prospect or client.

Speaker 1

Oh, it's busy. It's still busy. I would say like that just sure is up a little bit more time for you know, to just get replaced by five to ten different other things, you know what I mean? Yeah. So it's like it's it's it's very busy. Uh, but it I would say if we didn't have kind of some of these tools in place, it would be hard to execute kind of at the level that we are.

Speaker 3

And just out of curiosity, like how much bigger do you think your head count would be operationally if you guys weren't like an AI native organization? 30%, 50%?

Speaker 1

Probably 50%.

Speaker 3

Yeah. So that that means, and again, I mean, you're in the AI space, so it's a little tougher to compare, but for the listener, like uh finding talent, retaining talent, training talent, paying talent, like that's that's a grind. And if there's turnover, like that's painful. But if that turnover is happening, you know, in in Justin's case, like preparing him for all of these 15 meetings that he's doing per week, like those would be some some shitty weeks where he was waiting to fill those roles. But now it's less of a concern because the human in the loop is reviewing, not doing. And honestly, you're probably doing the review uh because you're the one that's having those conversations. So um very interesting. So I think that that's kind of like a universal takeaway uh for how an AI fluent uh leader is running this business. So I want to ask you, and we kind of touched on it earlier, but this episode it comes out on a Monday. Um somebody's gonna listen to this and they're gonna want to to take immediate action. So let's say they have an hour on Monday. After listening to this, like how exactly would you we talked about the the um the incognito kind of like tactical stuff, but what should what would they want to ask? I guess they'd want to sit in the position of somebody who was doing that inquiry where they would want to show up, but what should they check and how would they justify spending more time on this or reaching out to Embrose to say, hey, like I did that thing that we talked about on the podcast and uh we're scared? Like what what should they do to to make sure that they're um I I I guess ready for that conversation or at least know that they're they're covered.

Speaker 1

Put your put yourself in the the lens of a potential customer or customer and ask those types of questions that would maybe if they're in the kind of uh discovery phase and like um you know are are um ready for like intense like uh trying to buy or or uh make a decision, like ask those types of prompts and see what happens. Um, because I think you'll probably be surprised of where you're at. And um, if you're doing well, you're doing well, it's amazing. Um, but I I would just I I like to put yourself in the the the lens of potential customers. Uh also like you know, you probably are on customer calls all the time. Like I try to answer the ask those questions that they're asking you and seeing how what happens, right? Um, so you know, there's a lot of different ways to be able to do it, but I I always start with kind of the lens of like what your potential customer is gonna ask.

Speaker 3

Enter that conversation that's happening in their mind already.

Speaker 1

Yeah.

Speaker 3

Um, do you think that it would be, I don't know, too soon to start to ask clients like, hey, did you use AI to research us?

unknown

No.

Speaker 3

I guess depending on the industry, but yeah, yeah, yeah.

Speaker 1

No, I mean, like um I would say a lot of custom companies um are getting creative because they're it's a zero-click world that's like, right? And so they might be doing the research, but then they go to the site later and make that purchase. So I've seen a lot of creative ways where like um uh they'll put a survey on the site and say, like, hey, uh, was AI ever assisted in this buying to like journey or whatever? And the stats coming out of those surveys is insane because it's like it's showing a lot of it's coming through organic, but it's actually not organic. They started with a chatbot. Uh and and so it's like that that I've seen a lot of like tricky, not tricky things, but just like interesting, smart things that people are starting to do to capture that because it is breaking kind of a lot of the different kind of attribution and um traffic analysis and all the all the other stuff.

Speaker 3

So, and I guess kind of final question here. You mentioned a timeline of of 12 to 24 months. Is it urgent today though?

Speaker 1

Absolutely. Okay. Uh because the like I said in the beginning, like the ship has already left the station. Yeah. train trains out there like I and y you know I I I I it makes me worried for some companies that are not even thinking about this because if you're not right now and it's not all doom and gloom I think you still have some time but like think about that instinct like that sorry the mute the muse like use case where like it's going there and making decisions on behalf and your your product or your brand is not already kind of readable for these machines to be able to pick up uh as more and more consumers going there like you will effectively disappear because you're not gonna be part of that recommendation set. So like I you know you still have time and like I but I think now's the time to really start thinking about this because it is a massive shift that we we're all gonna get whiplash from.

Speaker 3

Wow and so listener even if you weren't aware of this it doesn't change the fact I think the quote is you can ignore reality but you can't ignore the consequences of it. So like it's and I agree I mean I'm I'm using the agents and and Justin and I are like kind of on the frontier I guess you'd say but his mom booked travel through the agent the other day. My mom's probably not far behind. My wife is doing this kind of stuff. So even if you think oh our company's safe probably worth reaching out so it's embaros ai. We'll have links to it in the uh in the show notes here we'll also have in the show notes we'll have some action items that we think you should probably take based on what what I've learned in this conversation. It's been uh a very cool conversation because I've you know I've been trying to kind of homegrown vibe code my uh websites with AEO S C O but I didn't realize that it makes perfect sense that what I'm doing is yesterday's best practice. So um like thinking that you're gonna do this this weekend and have ChatGPT or Claude you know prepare you for 12 months from now it ain't happening. So Justin thank you so much for um taking the time and with 15 meetings a week I hope you know this didn't put you behind on anything. I guess you've got AI supporting you but I really appreciate it. Any kind of like closing thoughts or any um things that somebody's like I'm sold I want to find out more about this. What does it look like to to learn more about how embrose can help a company?

Speaker 1

Come on the website um or you can find me uh on LinkedIn Justin Emmin uh and yeah more than happy to happy to talk to you or someone on my team talk to you.

Speaker 3

Great and we'll have that uh uh link to his profile on LinkedIn in the show notes so Justin again thank you so much um I'm excited to see what happens with your company and with this in general I've got a new perspective on how I'm looking at this not just for my company for but for my clients' companies it's gonna be something that I bring up uh it hasn't really been part of our you know strategy conversation in the past but I think that after this conversation it's certainly going to be on the list. So I appreciate it man thank you so much. Yeah I appreciate it thank you for the listeners thank you again for uh listening to the show I'm gonna encourage you to go check out our website using AI at work and just for kicks try it in your LLM and let's see let me know how we show up. But we've done a lot of work on the site we've we've had so many fascinating guests over the past uh almost three years now that um there's a ton of information that we've been able to kind of synthesize into action guides and to-do plans and all those types of things for the lessons that have been taught on all these episodes. So go swing by the site and if you enjoyed the episode or any of our episodes we'd really love it if you would take a second and you know give us a rating on your uh wherever you're listening to this. So thank you and we'll see you next week with another fascinating conversation about how companies are using AI at work. Thanks for tuning in to using AI at work. Don't forget to subscribe for more conversations about how to use AI at work. And a special thank you to our sponsor Chief AI Officer for empowering businesses with AI education and training. Visit their website for a free AI readiness assessment and AI strategy guide to help you get started using AI at work. That's www.chiefaiofficer dot com. Follow us on Twitter at the handle usingAI at work and visit www.usingai at work dot com for free resources to help you harness AI in your role

Transcript source: Buzzsprout.