38. AI Advertising: Stop Guessing, Start Understanding

The rise of AI advertising isn't just changing how companies reach buyers. It's also creating a completely new set of rules for who gets to influence those buyers at all. Over the last 25 years, digital advertising has transformed by paid search. Why? Because a keyword revealed expressed intent. It told companies what someone was looking for in that moment. There was no interpretation. You matched a keyword, you bought a click. But AI changes the equation totally. Today buyers are not just expressing interest and clicking on things. They converse, they compare, they challenge, they evaluate, and ultimately they decide right there inside the conversation. And inside those conversations is something far more valuable than a search term. It reveals what the buyer is actually trying to accomplish and who they trust. At the exact same time, AI engines are becoming the new interpretation layer between companies and their customers. They're interpreting buyer intent, understanding categories, evaluating credibility, deciding on relevance, and determining which companies belong in the conversations. Beyond that, before a buyer even sees your ad, AI has already formed an understanding of your brand and shapes the buyer point of view, which creates a fundamentally new challenge for every business leader. Do you understand the conversations your buyers are having? And does AI understand your brand well enough to put you in those conversations? AI advertising is no longer just a new ad channel. It's a new influence layer. The future of AI advertising is not simply about buying attention. It's about buying brand authority. And without first establishing brand authority with AI engines, advertising becomes an expensive road to nowhere. Welcome to the next frontiers of AI. I'm Scott Hebner, and thank you for tuning in. Today, I'm joined by my friend Mick Holison, a former CMO and founder of Redline Advisors. And together, we're going to explore how this all works, what to focus on, and how to get started now rather than later. All right, Mick, welcome. Let me uh bring you up here. There he goes. How are you doing, Mick? >> I'm doing terrific, Scott. Great to see you as always. >> It's great to see you, too. So, um this is a fascinating topic. It's sort of new on the scene um a little bit here, right? Um >> you know, I think in 2023, Microsoft in Copilot was the first to actually put ads and conversations. And uh while many are out there worried about, you know, it's going to compromise the the quality of the answers and it's going to, you know, mess everything up. I think we can put that all aside and and just focus on how do we help people be successful in advertising. Um, but uh, love to get your your take on this whole advent of AI advertising right in the conversations. >> Yeah, for sure, Scott. It's definitely moving quickly and I think it's uh, you know, to to quote, uh, someone from the Avengers, it's inevitable. Um, so, uh, I, the reality is that the companies that are making these frontier models today are going to have to find a number of avenues for revenue generation to offset the incredible capital expense that they're all um, you know, undertaking at the moment in order to have a thriving, successful, profitable business model. Now, exactly what forms and functions that takes and which one of the frontier models will jump on the bandwagon and which ones don't. A little hard to say. You mentioned Microsoft's already gone. Chat GPT in the past six months is clearly indicated that they're jumping in the ring. Uh you can see it coming with Perplexity and uh and Gemini, you know that that one's there. Google's been making money this way forever. uh and the the loan hold out might be the guys at Anthropic, but we'll see how long they can hold out too. So, uh in the end, I think that it's something that the industry is going to have to accommodate and all of us as users are going to have to uh tolerate, especially at the free or very low cost tiers. I think you'll probably be able to buy your way out of it much like you do with a a streaming service or something. you don't have to see all the commercials unless you uh unless you really, you know, want to. >> Yeah, certainly in the near term, I think you're right. And um it it it really changes things up quite a bit, you know, and huge industry. Let me show you a a chart here. Um and and you know, your point about inevitability, I think is right on, right? Um and by the way, I think this is going to benefit consumers too. all of us that use these. It should allow these providers to make the services even cheaper because right now you're helping to, you know, cover all the costs, right? >> Yeah. Yeah. We're all uh, you know, token maxing is uh um, you know, a a term that took shape when everybody thought the prices were inevitably going to go down at a a stratospheric rate and they really haven't, at least not yet. So, um, you know, I think advertising will be a way for the the makers to offset some of those expenses and, uh, deliver a little bit more affordable product to to the end user. You're exactly right. >> Yeah. So, I think it's something that just is going to become reality and if done right, it may be actually good for consumers and for B2B buyers if done right. And like you said, Anthropic's holding out. They're saying they're not going to advertise. Um, and then companies like OpenAI have made it crystal clear that the ads do not influence the answers to questions. And [clears throat] keeping that boundary and being able to create trust in the industry that that boundary really exists is going to be key. Um, but for everyone that is out there that does that, you know, that uses these AI agents to do, you know, go through their buyin journey, it really changes things quite a bit, right? Obviously, you're asking questions, you're having conversations, the AI answers. Now, you're going to start seeing advertisements show up, right? And those advertisements are clearly going to help buyers, you know, make some decisions or investigate new options and things like that. So, it's it's an opportunity for the sellers and perhaps maybe an opportunity for the buyers to get new ideas. Uh, but the market the market growth projections for AI advertising are massive. 37% compounded for a little under a billion now and as you correctly pointed out Microsoft um perplexity I think and Google are doing this already to some degree and uh I believe Amazon is playing around with it more from a you know more from a um you know a purchasing perspective right >> um you're in a cart or something like that um but by 2028, which is not too far from now, $13 billion industry. So, a lot of people >> Yeah. And I I I would suspect that that's going to be on the low side, Scott, because I think as you start looking at individual entities, you know, individual um you know, e-commerce sites and so forth that all have their own Aentic AI capabilities built into chat bots and so forth on their sites. Now that that number could be even bigger if you sort of look at the the total addressable market in a larger sense. >> Yeah. And that that is for that those numbers of 13 billion is for ads showing up in chats and conversations. >> Right. Exactly. Yeah. >> There's also other kinds of other forms like um sponsored prompts for example. Yes, there's there's using the intelligence from the conversations to fuel other types of advertising perhaps paid. There's there's some other, you know, corlaries to to to that. But, you know, right now the focus is, you know, the ads show up in your chat, which they're not really doing today. And you're right, for today, at least from Open AI, if you're paying for Open AI for chat GPT, you're not going to get the ads yet. But I think over time that will change. >> Yeah. Yeah. for for right now, I think it's the the go version and the free version or uh you know where you're getting the advertising there. But uh but but you're not above that. But we'll see over time. Just again, I think the parallels with the streaming universe are pretty interesting, right? Like you know, you can you get tiers of buying if you want to see ads or you don't. >> Yep. Right on. All right. Let's talk a little bit about the big differences here because kind of set this up in the intro and I just want to get your take on it but you know my view here is the biggest difference is even before a buyer ever sees an ad the AI engines have already you know have an understanding of your company your brand your products what category you belong in what problems you solve you know how credible and trustworthy are you you know and how you compare alternatives it already knows a whole bunch about you which is so different than the world of a placement of a paid ad or paid search or anything like that. >> And so I created a little chart here that kind of does a comparison. Let me bring this up and then um we'll get your take on it here. Um so in the world of paid search on the left obviously you know it reveals interest, right? The keyword is basically saying that someone has an interest um in this at that moment like AI factories if you did a search on that. Um, in a conversation though, it's really revealing what decisions the buyer is trying to make. Um, very very different, you know, mindset to go into if you're advertising. Interest versus decisions to be made. The paid search is also contextf free, right? There's little or no context >> um with that search. You have no idea what the buyer is doing or why they, you know, decided to search on that term. Where in paid advertising, it's all about buyer intent, right? What's the buyer's needs, beliefs, the trade-offs, the constraints, you know, the journey stage that they're in? Are they in discovery? Are they shortlisting? Are they trying to, you know, you know, validate a ROI before they can, you know, convince people to sell it? Like what are they actually doing? That's crystal clear in the conversations, not clear with a search term. Mhm. >> And then finally, the key to success with paid search is the placement. Obviously, where in paid advertising, it is not your message. It is not your um history. It's not where you're placed. It's your authority because you're not going to show up unless you have brand authority for that conversation, for that buyer intent. Um and I think ultimately what you're buying with paid search is clicks, >> right? >> Yes. And what you're buying with advertising and conversations is you're buying influence on those buyers and whatever they're trying to um decide or do. So, it's a very different world to think about if you're an advertiser. >> No doubt. No doubt, Scott. And I think you said the critical things up front there, at least in in my mind, you have to show up on I I've coined a name for it. I call it the silent short list. And if you don't show up in the first place, then none of the rest of it really matters. Uh, you know, there is no influence to be had because the the buyer never even knows that you exist. And in in order to work in today's world, you have to talk to sort of I think of it as surfaces, two different surfaces. You got the human surface which is moved by emotion and by the impact of humanto human communications. And then the other surface is the machine. And the machine itself, its definition of authority is is based on numbers and reviews and hard facts. And so you've got two different surfaces that you've got to hit if you are uh trying to sell a product or a service today. And it makes your job of building out your digital content that much more complicated. Quite literally twice as hard as it used to be. >> Yeah, it's true. And you know, I I think it all starts with something that you're passionate about, right? You think about what's happening out there. More and more companies, more and more submarkets, more and more access. You know, companies are producing more content, more messaging, more digital uh signals out there than ever before. And it's really becoming hard to distinguish because they all sound similar, right? What you call the sea of sameness. So yes, talk a little bit about what you think companies need to do to really differentiate their message. Um, so that as you said, it's not only meaningful to humans that get to it directly, but you know, humans that get through it through the AI systems as the arbiter. >> Yes. No question, Scott. So yeah, we've built a a system uh for both monitoring what your current signal is in the marketplace and not so much do you show up at all on the silent short list, but does your message as you originally designed it show up intact after it's gone through the compression that inevitably happens through all of the the models that are out there today? And so we've devised the system to create content that will talk with really really high efficacy to the LLMs that are doing the searches. And that means you've got to create content that is very digestible to the machines. In particular, we've created something called the checksum statement. And the checksum statement is usually 25 30word type of a piece. And it'll establish the following four things. uh Scott, it'll it'll identify who the buyer is, what category that buyer is in, what are the key differentiators of that particular good or service, and then what's an outcome like 30% faster, 20% less cost, whatever it might be. And if you can really compress that and get that digital footprint, that message, that check sum statement out through a number of your digital channels, the likelihood of your signal being high and the whisper as we would call it being low is really really good. And that's uh that's what we believe at Redline and we've put it in practice and a bunch of our customers and it has been a fantastic outcome for them thus far. But you do have to build your content with those two surfaces in mind, the the human reader and the AI reader. And more often than not, the big thing that's happened is the AI reader is first. So, they're going to be the ones that uh that read your message first and then decide whether or not the buyer ever gets presented with your value proposition or not. And so, uh their judge, jury, and executioner on the LLM side. >> It's true. It's true. And one of the points that you had made to me offline in the past that really resonated was if you don't get your message right, it's not just about potentially being absent, but it's it's you're running the risk that AI is going to misunderstand and misrepresent you. Um, which could be even more damaging, right? >> Yeah. You could have the wrong message make it through and then have that get propagated very very rapidly through other systems and it becomes sort of a a negative cycle if you will with your message. Last thing anybody wants, right? And uh the great thing about ads is you are in control of the message. The bad thing about ads is people may or may not, you know, engage with them in any way, shape or form, right? So, uh that's the the the two sides of the coin with advertising, >> right? So, I think where we're at right right here in at least in our conversation is two things, right? The AI models that you're going to be in the conversations that you're going to be advertising in already have an opinion about you, your products, >> for sure, >> your trustworthiness and so on and so forth. And they understand they understand your message whether it's the message you want them to understand or not, right? Which is the signal watch uh point for tuning that. >> Um, yes. But given that kind of um background that people are going into as they start to think about advertising, it begs the question um how do you actually do brand advertising correctly in these conversations? How do you get the right ads target at the right people at the right time with the right message with the right call to action? And if you went to right now OpenAI's um ad ad uh console to do some ads, um they effectively ask the advertisers to arrive at a theory of buyer intent and then encode that with into what they call context hints, you know, and obviously your campaign structure, your creative variations, your messages. Um, it then uses that to interpret where it may place your ad, but it's not going to tell you about buyer intent. It's not going to, you don't align to a specific prompt or anything like that, >> right? So, you got to go in knowing a lot of this stuff. Um, and the challenge that a lot of these companies are going to have, I think, and we've talked about this a lot, is almost all the AEO platform and and tool providers out there will tell you what is going on, what your metrics are, right? But it's not telling you how to fix things or why things are actually happening, which is critical. Understand those root causes to be able to actually fix things, right? >> Yes. Yes. It's uh their their view is largely in the rear view mirror, Scott. So, uh, it's it's great for telling you what already happened. It's not so good at telling you what you need to do in order to correct it moving forward. >> Right. Right. And so, that kind of brings us to um, what I think we likely will agree on here. What are some of the key success factors for people to think about as they start to think about AI advertising, which I'm sure almost all B2B businesses will eventually be doing at some point in the future, just like they're doing paid search. Um, the first one is you got to know the conversation before you actually buy into it. >> Otherwise, as I said in the intro, you're throwing money at nothing. It's going to, you know, you're just wasting money. Have to know that conversation. >> The second thing, which is where your passion is, is you have to know how much of your story, you know, your message is, you know, survives the algorithm, right? The AI algorithms, right? I mean what what is actually being obtained and then communicated by the AI engines. You have to know that. That's not metrics, by the way. That's understanding what the hell's going on under the covers. And then you really do need to know buyer intent. You know, again, the the search word, the keyword was an expression of interest in that term. Um to be successful with advertising, you need to understand the exact buyer's intent before you even target them. Are they shortlisting? are they discovering? Do they have a preferred vendor already? Um are they looking for use cases? Are they looking for ROI? You know, so on and so forth, right? You have to understand the buyer intent if you're going to be successful. And in the end, these three things add up to brand authority, right? You have to build brand authority with your with these AI engines. Um, otherwise, one, you're not going to be successful in getting your message out uh in general, but you certainly are not going to be successful uh with advertising. Um, what do you think? >> Yeah, I think you you you nailed all three, Scott. Um, and I think the thing that probably surprises me the most in running a little boutique consultancy that that assists companies in figuring all of this out is how many of them don't know how to answer even the basics of some of those questions. And I mean, look, this is as old as time. If you don't really know who your buyer is or what their intent is, then how can how in the world could you expect to produce a good ad or good copy in general for your website or anything else? And yet, I'm uh I'm constantly astonished as I walk into clients and they've got seven different definitions of their uh ideal customer profile. They're not certain about what the the their intentions are. They don't necessarily know the kind of questions that are going to lead somebody into the category that they service. And without all of that information in hand, putting money uh into the advertising machine with any one of these vendors might as well be putting it on a roulette table, right? So, um you know, unless you're you're that certain, you know, you're going to have to be quite thoughtful about the way that you direct your spend here with any of these providers. It has the opportunity to be far more directive than anything we did with uh with Google search in the past. Uh and and that type of thing, but but not if you don't know the answers to those questions that you posed on the previous slides, Scott. >> Yeah. I mean, it's one of these deals, I think, where you know, you can just let things happen to you >> and let these models form their own opinions, >> which again could either leave you absent or can misrepresent you. I don't know what's worse. Um or you can actually take control here now and everyone should understand when it comes to how these AI engines work when they generate the answers and have conversations. It is definitely a compounding effect. >> Oh, it most certainly is. And I, you know, I think it's one of the reasons, Scott, where we why we define something called a signal brief, which is kind of like a modern-day messaging source document, but anybody that's creating company, creating content for your company, should unequivocally be using some kind of a canonical document to feed to the LLMs in advance to put in your claw project or whatever it happens to be, so that your content is self-reinforcing that it is providing the AI engines what they need to know and understand about the business you're in and the buyers that you service. Because if you're not using that kind of a model, then it's it's highly likely that the search engines, the modern day search engines uh are not going to find what you'd like for them to find about you and your business. >> Yeah. Yeah. And I can tell you just as you know, I I have this AEO advantage index that we're >> Yes. starting to partner around and bring together with the signal watch stuff you have, which I'll talk about in a second. Um, but having done 12 of these, including big companies and smaller startup type companies, it is absolutely fascinating what you learn when you do the actual diagnostics. Um, and I think a lot of people would be really surprised about what is happening, what they need to do, how these, >> yes, >> how these interests really operate. And like a good example would be buyer intent. What does not work very well is taking keywords and and trying to transform them into prompts and then test against those prompts or even combinational matrices where you do more algorithmic approach. The way these engines operate is they create architects, >> right? >> Um which are you know which represent thousands and thousands and thousands and thousands of questions right and you get caught up you need to understand the architects because they represent what people are really asking at what stage. And there's real diagnostics that you can do to understand that for your market and your products. And you're not going to get that from monitoring metrics. >> No, >> there's a lot there's a lot of science mixed in a little bit of an art form, >> right? [clears throat] Especially as everything is evolving to get this all right. >> But that's a general statement when we talk about the advertising. Um, let me just bring up a chart here on what we're working on. >> Yes. So, I've been working, as I just mentioned, on the AEO advantage index and doing assessments to help people optimize their visibility in these AI engines. Um, not by helping you track metrics by but by showing you why evidence-based why you either show up or don't show up and all the insights around that so that you actually know what to fix. You don't know why, you can't know how >> to fix. um more recently created a set of indices that are for advertising. So um the opportunity right how do you identify the highest value conversations to actually target for your business um competitive you need to know who's advertising where they're advertising what messages they're using right you have to understand what's going to influence [snorts] buyers so what message narrative is going to get the buyers um attention and to take action at that moment in the buying stage um you have to understand what slows or accelerates on these buying journeys and then you got to measure how you are performing not just for you can see it but against your competitors. So that's what these indices do to help people better understand how to optimize the advertising and actually as open AAI asked you to do come in with a theory of buyer intent that you want to target. Um then I think you ma match that with the overriding step one which is your message right which you can talk about a little bit here on on on that right yes no doubt Scott I mean look one of the interesting things that we've you know all learned over the course of the past couple of years is that these models are perpetually changing. So no matter what it is that you build out in your storyline and you put out in the form of digital content, you can know that the way that the models are going to um uh take a look at that is going to change o over time as they continue to get better and smarter and look at new data sets and and so forth. So, we've developed something that gives a client an ongoing way of not just monitoring what is happening, but also making the uh fundamental changes to the message that are necessary to ensure that your signal stays high. And that's incredibly important in today's market because you can build a great message, but it doesn't get to last for a year or two or three anymore. you're lucky if it gets to last for a month. And so you have to take keep an eye on it and be flexible enough in the message architecture that you've built out to make minor modifications to ensure you're going to make that silent short list that we talked about earlier and that your advertising, if you're spending money on it, is going to hit the mark with your particular buyer. without having some kind of system in place to monitor and modify that message on an ongoing basis, there's just no way you can keep up with it. There's too many moving parts. It's moving too fast and and you're going to get stuck in the same sea of sameness everybody else is. So, you got to have a way to both monitor and manage the message over time. And that's what we're trying to build at at Redline. And I think it's a big part of what you're trying to build with the uh AEO advantage index as well. >> Yeah. I mean, you think about the AI engines are they're they're becoming a key buyer for you. They're an it's like a channel for you, right? It's like your channel partner >> and uh people need to take that incredibly seriously because >> Well, and as we said with the advertising, just like channel partners, they're going to want to take a cut. >> Yeah. Exactly. All right. So, last question before we wrap here. >> Yeah, we step back and think about this a little bit >> and we've both been through all this over the decades, right? >> Every major advertising transition has created new winners, right? Search rewarded companies that understood keywords and intent, right? Um, social rewarded companies that understood the whole engagement model. So now you have AI advertising and as it matures, you know, at least my bet on this is that what's going to differentiate people is their brand authority, which is a combination of your signals, your messaging, you know, your proof points, your third party uh reinforcement, a whole bunch of different things. What builds real brand authority? Um, so I'd like to get your two cents on what is going to separate a winner from a loser. And then even more importantly for the people listening in the next 30 days, what is the number one thing someone should start doing right now to get ahead of this? >> Wow, great questions. Both loaded ones. So, uh, I'll try to hit them both quickly. One, I couldn't agree with you more. The brand is making a comeback, if you will. Uh we we had a lot of years as we were able to get very scientific with digital demand in the sort of traditional um model that we had all grown used to with SEO. You could by click you could count them up and X number of clicks ultimately uh you know turned into Y number of dedicated leads Y number of leads turned into Z amount of pipeline or or even ultim ultimate purchases. um those days are are are largely behind us now. And the way that you've got to go about building um demand uh it actually is with brand. It is about establishing that authority because you don't make the silent short list unless your brand is out there and out there with a differentiated signal. So, if I had to give everybody one simple thing to do, it's to create both language um and numbers that uh will address both the human buyer and the machine surface that we've talked a little bit about. And that means they need to have at least one thing that they can doubt that's truly different about their business, their product, their service. They need to talk about that. And then they need to feed the machine a number about what that special feature or secret ingredient does. It makes people go 40% faster. It saves them 30% on the cost side. And if you can do just those two things, you're going to be way way ahead of the vast majority of the people out there. And honestly, Scott, I don't think it matters whether it's just the content you're generating on your website or on a LinkedIn page or in a blog or even in an ad. You need to talk about what makes you different and then you need to give the machine something about what the outcome of that differentiator really is for the business. If you can do those two, you're way the heck ahead of the game. So, that would be my next 30 days advice for folks. >> Yeah. For everyone listening, I got something you can do right now. All right. >> Go to redlinecs.ai >> and look for the signal scan. >> Yeah, >> there's a free assessment out there, right? That you can go out there and >> you can put your brand, your, you know, website and all that in there and it will give you um a view of how strong your signal is, right? Did I represent that right? >> Yeah, you sure did. And uh you did a better job at doing the plug than I did. I got so excited about answering your question and getting people on the right path, I forgot to plug my own offering there. But uh yeah, the signal scan is out there on Redline CS.AI. You can put in any website URL or just some general marketing copy. Interestingly, a lot of BDRs have found it helpful for looking at email copy, but it'll help measure your copy and determine how well you're doing against those two measures I just talked about and a couple of more as uh as well. So, won't give you all of my secret ingredient uh online here, but uh take take a try to that signal scan. I think you'll learn a lot from it. >> Yeah, I've been using it quite a bit with the clients and the prospects I've been dealing with to get an idea. Um it's a great it's a great little uh tool for people to get started. So, go do that right now. Um okay, Nick, this has been uh fascinating as always talking with you. I really appreciate you taking the time. >> You bet Scott. Absolutely enjoyed it. >> Yeah. And for everyone listening, I think the takeaway is quite clear here. AI advertising is not um just a new channel for buying attention. It's a whole new environment where companies must first learn the way it works and then earn understanding, trust and authority with the AI engines um because that's the way you're going to influence the conversations and that's the way your ads are going to get um placed in the right um conversation for the right buyer and the right intent. Bottom line is the winners will not necessarily be the companies that spend the most money. I think this is going to neutralize to some degree big companies that can spend a boatload of money like where we used to work, Mick. >> Yes. Because it's really going to be about the art of your message and how well it is differentiated and how these AI engines understand it. Um, so that's where everyone needs to focus. And you know, as we've been saying, you need to know the conversation before you buy into it when it comes um to advertising. All right, so everyone, thank you so much for being a part of the next frontiers of AI. Make sure you subscribe so you don't miss future um episodes and go visit Redline Advisors at Redline um CS.ai. Again, I'm Scott Heer. Thanks for listening. We'll see you next time. Bye everyone. >> Thanks. Bye-bye.