Top 10 Shifts That CMO and Marketing Leaders Should Plan for in 2026 and Beyond

Welcome to the Marketing AI SparkCast, where AI meets modern marketing. Hosted by Abby Varma, a global leader in marketing, AI, and innovation, each episode brings you real conversations with leaders putting AI to work in strategy and execution. This episode is brought to you by SparkNovus and the Marketing AI Pulse community. Here's your host, Abi Varma. Hello, everybody. Welcome to the Marketing AI SparkCast. Super excited. Happy New Year. I don't know how long we can keep saying that, but I'm gonna say that for this January episode. We are going to do something different with this podcast in twenty twenty six. We are obviously gonna continue the industry and the martech leader conversations that we've been having all this time, but we are going to introduce a new episode every month, which is going to track what is actually happening across the world of marketing and AI. And we are calling these episodes Marketing AI Pulse, which is a natural extension of the Marketing AI Pulse community. For those of you who are not familiar with it, we at SparkNobus host the Marketing AI Pulse community, and you can learn more about that at marketingaipulse dot com. Well, I can't do this alone. So I have a friend and my partner in crime who is going to co host these episodes with me, the lovely Matt Seer, who is the founder of Loop dot ai. Matt may also collaborate at SparkNovus. Matt is MIT Sloan certified AI strategist. He's got lots of hands on experience in leading digital marketing and AI initiatives across major industries in healthcare and higher ed. Matt, welcome to the show. Well, thanks, Abi. This is great to finally make it to the big time here. I feel like I've arrived. I don't know about that, Matt, but tell our listeners a little bit about yourself, my friend. Yeah, you got it. Very glad to be here. Yeah, my name is Matt Seer and I founded Loop Consulting in twenty twenty four after having spent really my whole career, as you mentioned in health care and higher ed and then actually working for an agency supporting those two industries. And I like a lot of folks including yourself caught the AI bug a few years ago, took that MIT Sloan course that you mentioned and came back to the agency I was working for at the time really with sort of a new enthusiasm for what's possible with these new technologies. But I really needed help sort of focusing on what was possible with them, what was not possible. And so I have spent the last few years really getting to know these technologies and really trying them all out. I have spent a lot of time working with clients and with each other to understand this tech and what it's capable of. And so Loop really is meant to be a place in a way to help others start to adopt AI in reasonable and tangible ways. Think I both agree that these are large changes, but they need to be done sort of incrementally and strategically. So that really is the focus that I have, working, yes, with higher ed, yes, with healthcare, but also increasingly with marketing agencies themselves to help them with the transition over into AI. Awesome. Thank you, Matt. And thank you so much for being my partner in crime for these episodes. Hopefully you and me are still friends by the end of twenty twenty. Hey, even if we're not, Abby, we'll act like we are. It's okay. Yes, there you go. There we go. Right, so today's episode, we're gonna get into the top ten shifts that CMOs and marketing leaders need to plan for in twenty twenty six and beyond. And since you mentioned agencies, Matt, earlier, the very first sort of shift is the CMO and the agency relationships, which are going to fundamentally change. And I know, Matt, you were going to speak at the Marketing AI Institute's agency summit coming up on February twelfth. But unpack this topic for us. What do you think? How do you think this relationship is gonna change? And let's get into it. Yeah, absolutely. This is a topic super near and dear to my heart. The change, you know, and those who are watching here today are already feeling it. But this change is well on its way to happening and there are lots of factors that are driving it, but one of the primary things that I'm seeing in terms of how the relationship is changing is that increasingly I think clients are expecting their agencies, one, to have AI capability, expertise, knowledge and experience, but also to invite them in to that learning process and help the clients understand really how to start utilizing AI in their own businesses. And one of the interesting things that I've seen is, agencies are adopting AI at different rates. Some have gone really far down an adoption path, others are just getting started. And those who are just getting started feel a little reticence to act as if they are experts in AI for their clients. And yet the clients are expecting some guidance. And so the relationship there to me is very different in that, agencies are used to being the experts in these things. They're used to leading their clients through strategy, through brand, through creative, all of the many things that agencies provide. And in this case, actually agencies in some cases are still learning themselves what AI does and how it can benefit their businesses. And so it changes the nature really of the relationship between agencies and clients. And that it's not, hey, I'm the expert and I'm gonna train you and teach you, it's, we're learning as an agency, let's learn together, let's give you the expertise that we can provide, let's help you learn about what this means for your business. So that's one of the major shifts that I'm seeing. Another one is coming from a lot of the large holding companies who have invested literally hundreds of millions of dollars in their own AI platforms and are bringing the capabilities of those AI platforms directly to their clients. One of the examples in the last few months is WPP in their open platform. A few months ago, they announced that they were starting to invite their clients into the open platform, allowing clients to actually do creative and strategy within that platform. And WPP acting really in that instance as an expert and as a guide in that technology, but not necessarily having to do all the creative work. And then even just last week, WPP opened unveiled an agent hub, which is really meant again to sort of put the power in the hands of clients with some highly trained agents. So the nature of that relationship is really different. What I think is really critical for CMOs going forward is to engage your agencies, understand how they're using AI, where they are in their journey, and come to them as a collaborator and not expect necessarily that they're gonna have all of the answers to all of the AI questions, but figure out how to partner with each other in really exciting ways. And the last thing that I would say is to me, the big shift here is that AI is really putting a greater emphasis and focus on human strategy, creativity, and expertise. So the more that CMOs can really focus on helping their team members and their agencies provide that expertise and that humanity, I think the more effective they're gonna end up being. Yeah, no, very, very true. I almost feel that there is so much pressure for brands to bring some of the work in house and leverage AI as opposed to agencies. I mean, as we work with clients who are, who have explicit goals for actually not, you know, continuing on with their agencies, because AI can do some of that production oriented work internally. Are you seeing that as well, and how should CMOs sort of deal with that? Yeah, I definitely have seen that, and I think there's, you know, it depends on the organization, right? There are, obviously we know, right? There are marketing organizations, you and I have worked very closely with one, that has a really mature internal marketing function and capability. And I would encourage those large, complex, very capable internal marketing agencies to really double down, invest in their people, invest in training and education and technology, and really build up that internal muscle to continue to build the capabilities and value of their internal team. But if you are relying on agencies on the outside, again, I would really encourage you to have open conversations with either agencies that you're working with or thinking about working with to really understand how they're looking at AI, how they're utilizing it. Everybody is calling themselves an AI expert out there today. And I think CMOs really owe it to themselves and to their teams and their organizations to really understand if the agencies truly do understand these technologies or whether it's more lip service. So ask some of the tough questions, doesn't mean you have to reject them if they're not already experts, but I do think there's value in really understanding what's real and what's lip service at this point. I think you and me could have a dedicated episode just in this one topic, but Matt, tell our listeners a little bit about this session that you're doing with the Marketing AI Institute. Yeah, thanks. It is February twelfth, as you mentioned. It is an all day event being put on by our friends at the Marketing AI Institute, also increasingly being called SmarterX. And for those of you who have ever been to MACON, which is the annual Marketing AI Conference in Cleveland in the fall, This is that same organization. Paul Rates or Mike Caput, Kathy McPhillips, just a really incredible team and this event is a free event and it's gonna have a variety of talks from Paul and Mike and our friend Will Reynolds among others who are gonna be talking about the impact of AI on agencies specifically. And my talk is going to be about AI agents. I'm gonna be doing a little bit of demonstration work showing a couple of real agents that I've created and how agencies can use them today to help with the work that they're doing for themselves and for their clients, but also really setting the context for what's happening with AI agents. I think you and I hear that buzzword all the time and there's a lot of confusion about what they really are and what they're really capable of, and so part of what I'm gonna do is unpack what's possible with agents today and where they're headed. So if you haven't signed up yet, folks, again, it's a free event, and we'd love to have you join us. That was very cool, Matt. Good luck, I will be tuning in. All right, what is next on our list? Well, a minute ago, Abi, you mentioned brand, right? And I think in the age of AI, brand is so important. I would love to hear your perspective on brand discovery in the rise of generative search. So much has changed in this area. We'd love to hear how you're thinking about it. That is one topic that you can run, but you cannot hide today. CEO, LEO, AEO, EIEIO, I don't know what to call it. All of those things are legitimate, But really, at the heart of the issue is that users are increasingly skiffing traditional search altogether, and are going straight to LLMs, to your chat JPTs and clods and perplexities, to really get more synthesized sort of quote unquote cook answers, which are amalgamated from these various sources to really learn more about, you know, your brand and your product, and are then using that for decision making. Not only that, I think LLMs are also getting embedded inside of products. I mean, most notably, Alexa and Siri soon to have ChatGPT. All of these things is just going to sort of amplify the effect, right, for LLMs to be embedded and be sort of used as a primary interface for folks to start starting to find things. And that is a stark change from the days of people Googling for stuff, and you're being served up with the ten blue links on page one, and you're, you know, trying to find things. And really, for the first time, I feel that marketers and CMOs and brand managers are going to struggle with really making sure that they are not in control of their narratives. Right? So in the traditional way, you would search for something, it would come up on Google, you would click on the link, go to your website, and now you're in control of that narrative. You can now position your brand, say anything about your brand in whichever way makes sense, but that is no longer the case now because LLMs are taking data from your website, from third party sources, and all those sort of things and serving it up. So a completely different paradigm. I mean, don't get me wrong, Google is still dominating search, like in terms of the, you know, the pie chart, if you will, but that number is changing dramatically fast. Right? And not only that, I feel that the conversions that are happening from NLMs to, you know, to decision points is those conversions are much more, have faster and more authoritative when it comes from LLMs versus, you know, traditionally Google. Whole other sort of paradigm and whole whole other way of really thinking about that. And I think from a CMO standpoint, I feel, you know, every CMO should really in twenty twenty six focus on kind of three things when it comes to this. So the first is start at ground zero and really understand, audit, monitor how your brand is being represented today, right, in the LLMs, both in terms of coverage and in terms of sentiment, right? There's ChatGPT, Plexity, Google AIR, it'll be overview, excuse me. There's a lot of that stuff. And there is a whole host of tools that we follow along pretty closely, and that market has exploded. I feel like there was a new tool. Two years ago, there were three tools, and now there's, you know, lots of tools. We keep track of them, but increasingly hard to do that. But some of the prominent ones are Evertune or ProFound and Gumshoe, Revere. These are some of the common ones where you can really plug in your brand, and then see how these LLMs are kind of representing you in terms of share of voice, and then sentiment. Right? And these are all, all of these tools provide that information based on just random sampling of your brand within the LLM. So that's number one. The second thing is make sure that in addition, yes, for a lot of brand folks and CMOs, you're focused on SEO, but also make sure that your website is AI readable. Right? So there's specific methodologies of how to write key pages in vernacular that is understandable by these LLMs, make sure that the schema models that most websites should have had all this time, and a lot of them don't, or people ignore them, now are playing a really important role in how these LLMs sort of synthesize your website. So obviously not getting into the weeds in this podcast, but it is widely documented, lots of conjecture, and a lot of sort of ideas on how your website should really be made, you know, AI readable, and what's the impact of that in terms of content strategy? Right? Like, how do you write how you write in a manner which is suitable for humans, but also you're giving LLMs what's the information that they can grab and include in their answers as people are trying to look for things. And then finally, I feel it's a great time for PR folks because, you know, LLMs are obviously, you know, Google was indexing your website, but now LLMs is taking not only your website, but also taking all these third party sources, right? So obviously you have the prominent ones, the Wikipedias, the Reddits, the Quaros of the world. But then there's a lot of these third party websites where LMs are synthesizing that information to really understand how authoritatively is your brand speak to those, whatever that topic area may be. And that has a big influence in how AI describes your brand. So those are some of the things that, you know, if you're a marketing leader, you should be doing. And making sure that, you know, you're really looking very closely at, you know, at brand discovery, now that people are increasingly looking at LMs for these answers. Yeah, it's such a huge shift that is taking place right now. Quick question, follow on question to that, Abi. You talked about the importance of the website and how to make content AI readable. You and I both have seen clients see pretty precipitous drop offs in website traffic. Should CMOs be freaking out right now about the traffic that's disappearing from their sites? The way I really think about it is as long as drop off is in alignment with the general industry trends, you shouldn't be freaking out. But if, you know, generally speaking, people are seeing a referral, Google referral traffic drop off by thirty, forty percent in your traffic that is like ninety percent, then yeah, so you should be worried about that. Not only that, I think there is, increasingly, I feel web traffic is becoming almost like a vanity metric. It's making sure that you are attracting the right people, the right traffic, which would then, obviously, the goal would be to convert that traffic. So again, different ways of thinking about it. In general, can't ignore brand discovery on LOIs. Number three in our top ten list is agents. Now you had mentioned agents earlier. There is this, you know, dramatic shift. We've been talking about agents for a while, but I think there's a dramatic shift from agents being, you know, almost like helpers, quote unquote copilots. I don't mean the Microsoft copilot, but true sort of assistance, if you will. But from there to kind of workflow owners or workflow co owners, Matt. You know, you take any major platform, Adobe, HubSpot, Salesforce, you can't talk about Salesforce without hearing about Agent Force these days, but there's a lot of that, but break it down for us in terms of what does that really mean, and how should CMOs really think about that for twenty twenty six? Absolutely, so first and foremost, I blame Matthew McConaughey for the AI agent situation that we are in right now. The aforementioned Salesforce agent force hype has been, that drum has been beating so loud now for well over a year and McConaughey and Woody Harrelson are very happy to sell us agents. But the fact of the matter is that really agents are in many cases really not yet ready for prime time. They are getting better, they are getting more capable, but they are far from infallible and certainly not set it and forget it. In fact, I just read a study this morning that talked about the fact that there are people in the AI world who believe that agents will never be one hundred percent capable of completing tasks successfully on their own. Now whether that proves out or not, who knows, but from my own experience, having built a number of agents for clients and for myself, I can tell you, and I know you've experienced the same thing, that these things are not simple. They require a lot of clarity, a lot of focus in what you're building and how you're building it and what you expect to come out of it. And it certainly, as I said a moment ago, not set it and forget it. You have to really pay attention to what's coming out of these agents. So the big thing to me in this area is yes, there is enormous potential here. And you talked about this ability for them to really own entire workflows. That to me is absolutely possible and there are examples of that already working well. And that really is in my mind the future, which is how do we string these agents together in a way that lets them do what they're really good at individually so that collectively they become exponentially more powerful. The goal to me, if you're a CMO or a marketing leader of any kind today, is really understand where AI agents can add value, be clear in what you're building, how you're building it, what you want those agents to accomplish, have proper expectations for what's truly possible with them today. But find some of those workflows of things that your team is doing that are maybe not great for humans to be focusing on and allow the AI agents to do some of that work so that your human employees can really focus on what they're best at, which is creativity and strategy and human discernment, all those really important things. And so yes, lean into agents, find opportunities, but they are no magic bullet and certainly require some heavy lifting to get them going and to make them successful. Do you think twenty twenty six is gonna be a year the AI agents, as a lot of people claim it would be? Well I think twenty five was supposed to be the year of AI agents, so I think if we keep pushing the, If we keep moving the goalposts, we'll get there eventually. I think this is the year of AI agent reality setting in. What I would say is that it's, I think what we're coming to is this moment of really understanding that AI agents specifically, but AI in general is not magic. It's not like you can wave a wand and it's gonna solve something for you. It requires work, it requires planning, it requires forethought, it requires investment of time, energy and money. All of those things I think are leading to this moment of like, okay, there's enormous potential here, but we really have to invest in making this happen. So do I think that this could be the year where we get real value from agents? Yes, one hundred percent. And in fact, I'm working on a project with AI agents very much at the center. And so I'm very hopeful that this will be the year of AI agents. But I think there's, again, that reality setting in if we have more work to do to get the real value that's possible here. So you're saying that in the future I can have a MAT agent do the podcast with me, don't need to work with you, that would be the Yes, exactly, yep. I will record an avatar, and you will not have to deal with me directly ever again, Abi, I promise. So I've got a topic for you. This is something I know is near and dear to your heart, having worked with you closely on client work. Interested in your view of how AI really is accelerating the creative side of things, right? We've all seen massive strides in images and video and audio and all the many things, But where is this headed and what should CMOs be thinking about? How do they sort of parse through all the many moving parts here? I know, mean, just the advances in this twenty twenty, twenty twenty five, the advances in video generation and image creation have just been phenomenal. Right? Notwithstanding, if you've seen any on Instagram or TikTok, you've seen the Maduro and Trump dance. Not with standing that. I would say the, you know, the impact of all of these models has just been phenomenal. Right? Our image models like the Nano Banana producing, like, high fidelity imagery. And then you have, you know, BO3, Google BO3, Sora, Cling. There's so many of these tools where you can really now create, you know, ad films. People are, brains are already doing ad films, and people are creating documentaries and moving towards full length, you know, movies. And it's pretty evident, especially given movie studios and these video model providers has been sort of a kind of a merger of sorts where there is people definitely movie studios see tremendous potential in this, you know, in this technology. And this is not something that is limited to, you know, smaller brands and people tinkering around with it. You know, this is clearly being adopted by more and more brands. And, you know, a great example of that was, you know, Coca Cola. So they produced their twenty twenty five holiday campaign. And then they released, interestingly, a behind the scenes video, which I think they used Notebook LM to do the audio for that video. So AI again. But it was they sort of shared what their process was, and essentially, it was a team of about five specialists who generated and refined about seventy thousand video clips in thirty days by using AI to design the realism, the motion, the brain details upfront instead of fixing some of these things later in post production. So essentially, the post production overhead became almost like a pre production exercise for them to create all these video clips, which was then kinda used in the final commercial. And that shift turns, you know, that shift sort of turns the process where you can start to rapidly iterate, right, as opposed to a traditional slower linear process. So really, I feel this is game changing, and this is, there is a lot of video slop that you see on all these channels. I mean, several of that is the, you know, slop central when it comes to video, but of course, you know, even on Instagram and TikTok, and also social media, you'll see a lot of this stuff, but really, for serious applications, I think this is becoming, you know, a serious contender, AI generated video and imagery. And, you know, for CMOs, I feel, you know, what does this really mean for them? So this means, like, a creative standpoint, it is your creative process and imagination and storytelling is not limited by production or production budgets. Right? And so anybody can do anything, but I think for CMOs, what this really means is there is an obligation to sort of set the guardrails. The next thing is for CMOs, speed becomes an advantage, right? So you should really invest in workflows and model access and things like that, and not focus on the actual asset because the asset is just a one time thing, but focus on sort of institutional systems to kind of go along with the guardrails that I referenced earlier so that you have the ability to create assets in kind of an accelerated fashion where speed can truly become an advantage while you're staying on brand. And then you can have guardrails, you'll have speed, and then that increases so that that scale that comes as a result of that will really increase, has the potential of increasing your brand risk, right? You don't want a situation where, yes, it's super easy, and you can do it on a dime, but now, you know, you're creating things that is going to put your brand on risk. So that's where I feel something near and dear to your heart, human in the loop. You want to make sure humans are in charge of taste. Right? That's where, you know, make sure you're resigning your senior creative people to really approve your, you know, creative direction and, making sure that this brand is consistent and storytelling is authentic to your brand values and those sort of things, which are really hard to pull off with AI. AI can really help with accelerate production, remove the barriers in terms of production budgets and whatever. Had a client who wanted to create a commercial for their brand. It was a B2C brand where they wanted the environment to be on Mars, you know, four prompts, and that's easily attainable. So think about that. It's pretty wild compared to what it could have been. Not that they were sending a crew to Mars, but just the the fact that that was their idea, right? That's how they wanted their brain represented and they wanted to be done in sort of a brain compliant manner is pretty incredible. Like, so all those constraints are disappearing. And I feel that sort of human factor of taste and judgment, we have, which we'll talk a little bit more about later, is such a crucial aspect of it. Moving on to number five. So I want to talk a little bit about Martech. Right? So with marketing, in in the world of marketing, now that we have AI, there's a real opportunity for marketing departments to truly leverage all of their systems, have historically been disconnected point solutions. Now, they may be integrations and whatever, but they're still point to point integrations. But I feel AI will increasingly allow for these systems to be connected so that marketing leaders and CMOs can really think of that almost holistically. What do you think about that and what is your take? You asked a minute ago whether I thought twenty six would be the year of the AI agent. I actually think twenty twenty six may be the year of AI as infrastructure Because what I'm seeing and what you're experiencing with your clients is exactly what you just mentioned, right? Which is a lot of point solutions, right? I need something that does media planning and buying, or I need something that helps me analyze my data, or I need something that helps me be more effective from a project management perspective. And it's that classic tale of buying a bunch of different tools for individual needs, where what's really needed is a sort of systems approach and thinking about how all of these things come together, or should come together for maximum benefit. Because I think right now AI is in such a shiny object syndrome place that it makes sense. People are like, oh, I need to get into the AI brain here. But the truth is that almost every technology that folks are using now has got some AI in it somewhere somehow. And what I think you and I have both seen in our client work is that a lot of clients are not even aware that the technologies they're already paying for have AI capability. And so, you know, I always recommend that people start with understanding what's going on in their current environment, what AI capabilities are currently there, what gaps are already being filled, and what if anything is left over in terms of gaps that need to be filled that maybe then you need to buy another technology to fill in. But increasingly, I think what's going to be happening here in the coming year and beyond, is that these technologies are gonna start communicating with each other. If you're familiar with Anthropic, they're the maker of the Claude LLM, they last year, actually late twenty twenty four released what they call their model context protocol or MCP. And MCP really is essentially a way for AI systems to talk to one another in an easy, much more streamlined way than sort of traditional digital integrations. And these MCP servers are increasingly being found across technologies. For example, I use HubSpot as my CRM for my business, and Claude has a direct integration with HubSpot that allows me to essentially through natural language speak to my HubSpot database, that is done through an MCP connection. And so these technologies working together is gonna become something that makes life a lot easier and it reduces the burden of sort of traditional tech implementations and integrations. And I think there's just an enormous amount of power potential there. Should CMOs think about technologies differently, more tech differently moving forward? So I think they should be thinking about technologies from a problem solution mindset, right? I think probably most CMOs are doing this already anyway, But I think AI again has been this shiny object, let's go buy the magic trick, that I think there's been a lack of understanding of this is just technology at the end of the day, right? Just like any other technology, you're making a decision on purchase based on specific business needs. And so I would just invite CMOs to step back and say, not necessarily, hey, let's go get the latest AI technology, but instead let's understand what we're currently paying for, whether we're getting maximum value out of that, do an evaluation really of our workflow and if there are things along the way that can be benefited from AI, then certainly fill that gap but I would take a much more holistic and a much more stepwise approach to it than a lot of what I think we're seeing so far. Another topic near and dear to your heart, something you and I have worked on with a client is in the advertising and paid media space and there's obviously an enormous amount of disruption that's taking place in the media platform area. I'd love to just hear your thoughts on both where we are today, but also really where we're headed in twenty six and beyond, and what CMOs should be thinking about in this really critical but fast moving area. I think what this sort of, I'm gonna talk about two things. So one is, paid media platforms are letting AI do more of the work, and not just help with tasks. And this has been going on for a while in the form of machine learning and that sort of stuff, you know, Google Performance Max or Pmax as folks call it, and Meta's got Meta Advantage Plus and so on. And a lot of these tools are already automating, you know, targeting, bidding placements, creative combinations across channels, and things like that. And there are these whole sort of AI native tools like Albert AI out of Israel, where they can sort of independently optimize campaigns across platforms using AI signals. So that's one. That's the current landscape where all of these paid media platforms are moving. On top of that, the advertising landscape itself is moving into sort of our AI answer environments. Like recently, OpenAI actively announced that they're testing ads inside of ChatGPT, and, you know, Google's doing the same thing, and I'm sure all these LLMs are doing the same thing, where it was just a matter of time that they would monetize it. And, yes, they would do it gracefully, and things would be labeled, and there would be choices and exceptions and all those sort of things. But advertising is going to move into, you know, LLM Answers as well. So from a CMO standpoint, what does that really mean? So I think the way I really think about it is sort of three folds. So first is, you know, shift from kind of campaign control to almost input control. What do I mean by that? So assume that all your paid media platforms like Pmax and what have you and Albert will leverage increasingly will will leverage AI to optimize automatically. So really focus on, you know, the quality of the audiences, creative inputs, and kind of the conversion data that you feed them to make sure that AI is doing its job well. So give it the right input, and let AI do sort of some of these automations when it comes to campaign control. Right? Obviously, test and iterate and see if that is working well, but I think that's where it already is in a lot of ways, and it's going to be increasingly getting to a point where, you know, you and me can just go in and saying, hey, here's these are the type of people I want to reach in natural language. And based on that, here's here's my services. This is what I wanna sell. And then let these platforms are gonna start generating ads and make try to reduce the friction as much as possible for folks to for folks to start advertising. So that's number one. Number two, I feel, again, this is already happening, but I feel the emphasis of AI enablement when it comes to creative. There's going to be more emphasis on that, you know, in a way that you want that creative to be modular, right? So you're going to be able to provide headlines and visuals and video clips and call to actions and links and all these things to your AI, and let AI assemble it in the right combination with the right guardrails so that the right message, you reach the right audience in the correct context to maximize your conversions and ROAS. Right? So that modularity you've created becomes super important. So you're not thinking of ad units as sort of independent units, but almost like components, which then you can let AI assemble and deliver to the audiences, the intended audience in some sort of a, you know, automated manner. And then at the end of the day, you want to make sure that all this is great, but you got to reassert that accountability through measurement, and that's where you got to make sure that you're using the right kind of data, you're doing incremental testing, business outcomes, the right kind of KPIs to make sure that AI based automation is clearly driving value and is just not platform level efficiency. In other words, make sure that you are getting the output of why you are doing that paid advertising, you know, in the first place. So don't focus on vanity KPIs, focus on business outcome KPIs, and hold AI accountable. That's sort of a different different way of thinking about it. And like I said, none of these things, people who are deep in the world of, you know, paid media, these things should resonate with them because these things have already been happening, you know, over the last two couple years, I'd say, and I just think there's gonna be more and more of this. Yeah. You know, it reminds me, Abby, when you and I were talking with the CEO of Albert, this was probably, I don't know, eight or ten months ago at this point. And one of the things that he said really has stuck with me, and it's very much around this idea of how much and how completely this field is changing. I remember I asked him, I said, What makes Albert most successful? What scenarios are the best for Albert in order to be successful for a team? And he said, The ideal scenario is to turn Albert on, train it, give it a clear strategy and guardrails, and then leave it alone and let it learn and act on its own and you're gonna end up having probably a more successful campaign at the end of that. I think that's completely the opposite of what most marketers would be comfortable with today. Do you have just a quick thought on how marketers can become comfortable with that change of power structure? It's gonna be a result of sort of testing, and I mean, I think succeeds like success. So I think it would be that incrementality where you trust AI, you let it play with it. So that's the paradigm that you provide the goals, set the guardrails, give it a little bit of budget to go play with it, and then you see the outcomes, and you see that, hey, compare that to what your kind of human paid media specialists are doing, and how is AI doing, do that comparison, and see the areas that it excels. So I think AI is going to have to earn that trust. It's not going to be set it and forget it sort of thing. It's real money at the end of the day, so you don't want a situation that your budgets are blown just because you trusted AI and AI did not deliver. So I think that it's going to be an evolution, but it to me, it's inevitable where you're gonna get to a point where AI is going to really handle all the grunt work. And then the input that the humans are gonna provide is audience, goals, modular ad elements, and, you know, expected business outcomes, and then let AI do its own thing. So, and then obviously, like I said, make sure that you're holding AI accountable from a measurement and analytics standpoint. So that is sort of a key output that you wanna make sure that humans are holding AI accountable for. Yeah. And I love that, what you just said about what AI needs to earn the trust, and I think that sort of can be generalized across everything related to AI today. So that's good stuff. Let's talk a little bit about analytics. Because that's to me is one of those things which is so key in all aspects of marketing. So I always joke about it, you know. I love an inner SQL join said no marketer ever. So marketers want insights. They want data. But historically, that process has been, you know, not very marketing friendly. You got your data sources, and you got your, you know, data warehouses and data pipelines, and then, you know, dashboards and all the whole sort of processes, very traditionally IT oriented, whereas marketers really want insights on what to do when it comes to campaign execution of what not to do, or understand their audiences better. And I feel that AI is really changing the way marketers are consuming these insights to where now you can quote unquote top tier data, like you mentioned in your case. Yep. When you were talking about HubSpot and, you know, Claude earlier. And then that's one part of it, you know, natural language, talk to your data to extract insights. But then also, for a lot of marketers, the holy grail is not only understanding what is happening, but understanding the why, the narrative, You know? And I feel that both of those things are getting opened up on account of AI and you know, the AI's ability to analyze all this information. What are you saying? Do you agree with that? I do. Yeah. And I would say that I've got good news and I've got bad news for you here, Abby, when it comes to marketing analytics. But the bad news that I would say is that this is not gonna be easy for most marketers today, right? That to really get greater value from your data is not gonna be the proverbial push of a button or flip of a switch. It comes back to something you mentioned a moment ago, which is that there's all these data sources and there's historical performance and communication, systems that don't communicate with each other, and so I think for a lot of marketers, they've got kind of a hot mess of data. And as nice as it would be to just be able to say like, let's go get more value from all this data, the bad news or at least the truth and the reality of it is that there's a lot of work that needs to happen in terms of data hygiene and cleaning in order to get that value. The good news is that I think we really are close to being able to get exponential value from the data that we're all sitting on top of because of the power of AI. But it requires that upfront work and labor of cleaning the data, getting everything into a consistent format and then and I think only then will you start to really get that exponential value. You mentioned the idea of talking to the data and I talked about my MCP connector experience between HubSpot and Claude. And for me, it was this sort of eye opening moment where I connected Claude through the literally through the toggle of a switch within the Claude interface to my HubSpot database, I had to give it some permissions of what it could and could not look at, but once Cloud was in my HubSpot database, it really was able to give me different kinds of insights and value that I was able to get otherwise. In fact, I asked it just as a test, said, go into my HubSpot database and give me a list of my top ten most engaged prospects and that's all I asked it to do. Was purposely very general and generic with what I was asking. And what Claude did was go into my HubSpot database, evaluated everybody in my database and created a top ten list of my most engaged prospects based on website visits, email opens, all the usual metrics of engagement, so created or looked at all those data, but not only that, it actually created a scoring system on its own, it then scored and ranked my prospects based on their interactions with me, and then it created a custom marketing plan for each and every one of those specific individuals based on their interactions with me. So the point about the data being a powerful tool that we're all sitting on, it just needs a little bit of an unlock, is that we can truly start to understand the motivations at a more granular individual human level of the folks that we're interacting with and that we're trying to market to. And so I think if folks can invest time and energy in cleaning their data, but then really starting to understand or think about strategic ways to benefit from the data in new ways, I think the potential there is just enormous. I am super excited about, you know, this one aspect of it, because as much as I love data and insights, I just don't like the sausage making portion of it. This super is super exciting and kinda near and dear to me. Probably most of the folks who are listening to this podcast have heard about ChatGPT Atlas or Perplexity Comet, or these AI enabled browsers that have become increasingly popular in the last maybe three to six months. So I would really love to hear your perspective on them. Have you used them? What are they good at? What are they not good at? And what are some of the watch outs that folks should be aware of? Love this one. I have played a lot with ChatGPT Atlas. So essentially, what is happening is, you know, AI is moving sort of up the chain into the browser layer and really changing the way how people explore the web. So like ChatGPT Atlas, for those of you who have not played with it, would highly recommend you play with it. It really acts like an agent. You can delegate that task for it to go sort of navigate things, click around, give it a goal, and it is going to do it for you. And this is something that we've played with it directly inside of SparkNobus, where we give it, you know, a blog, where we give it a copy of a blog post and then have it go actually post the blog on the website, right? Not publish it, but post it in draft post so we can have a human in the loop, and it works really well. Now, is it a hundred percent perfect? Does it get stuck sometimes? Yes, it does. But to me, it's pretty incredible where you can just upload your, you know, your your blog copy, log in to your CMS, give it a task of going to post it. Obviously, your instructions provide the details on what to do and where to click on and that sort of thing. But I can just do that. Go, you know, grab a cup of coffee, come back, and it is done for you. You know? So that is pretty incredible. And then you have another one, Comet. You know, so there's all of these tools. I'm sure, you know, Google has DISCO, which is a little bit different. Microsoft, let's not even talk about those guys. But there is just a lot of opportunity. And, you know, what this really means is think about the future. This means that we should all assume that AI sits between us and, you know, the user. So us, in this case, brands. So you have brands and then you have users, and AI browsers and agents are increasingly going to read, summarize, and filter your content before a human ever visits your website. Think about it. This is already happening in the email. Like if you're getting email, whether it's Gmail or Microsoft Office or whatever, you have those little AI summaries unless you've turned them off. You have the AI summaries up there where sometimes, you know, I don't know about you. I'm still reading the full emails, but I'll definitely glance, you know, at the summary in some cases. I think extrapolate that to your web browsing experience, we're going to get to a point where you start relying on your browsers to summarize all this information. So that has profound implications from a brand standpoint because you suddenly have brands with, you know, again, website presence, whether it comes to, you know, creative or narrative or whatever, that sometimes, you know, users will never see. So how do you deal with that reality? The second thing is you have to make sure that the message on your website is sort of survives that interpretability of AI, right? You want to make sure you structure a content in a way that positions your offerings, allows for differentiation, it provides proof points from, you know, those sort of things. And makes it friendly for AI based browsers to provide that information to the people who are browsing it. And then finally, you know, we are going to get to a world which I think is going to be a whole new industry where you are going to have websites which need to be made agent friendly. So we all have to prepare for agent first interaction models, where, you know, as agent driven interfaces sort of mature, and there's early protocols for AG UI, which are already sort of taking shape, where brands need to make sure that the website usability, which was all driven off of, you know, human usability and the ability for humans to interact with your website in an easy sort of way is going to be appended because now they need to be usable by AI agents. In fact, in my experiments where, you know, we had, you know, ChatGPT, Atlas postings, some website control, it got stuck because some website controls are not just not conducive for, you know, agents to go interact with it. A human had to interact with it. Right? But I think increasingly, we're going see more and more of that. And I think all our websites are going to have to cater to that. Right? So again, they're completely different paradigm. As AI based or agentic based browsers are gonna proliferate and it's just inevitable. And I think there's gonna be more of that before the end of the year. Yeah, and it is a fascinating thing to watch these browsers work on your behalf, and again, to your point, I would really encourage folks to get out there and give these a try. Number nine, Matt. This is a very interesting one where, you know, marketing leaders are gonna have to manage humans and AI systems or AI agents, right, as part of some people say as part of their org charts. Right? Yep. And so what is your take what is your take on that, and what does that mean in terms of future marketing orgs? To me, what's fascinating about this is that the idea of AI employees, it's already here. I've got a client who already has AI employees on their org chart reporting into human beings. Those AI employees have very clear tasks that they're responsible for, very clear KPIs that they are being held to and they're sitting right alongside other human beings on the org chart. And so the people who are leading these organizations are increasingly having to think down two very different paths. One is the traditional human path of management and motivation and empathy and maximizing potential and all of those things, and the other is managing technology to a set of very clear expectations and goals, and as we think about AI agents, you talked earlier about sort of running workflows beginning to end, that's really where a lot of this is gonna be going, which is you're going to have agents that are living in the org chart that are responsible for things, and whether it's the CMO or the marketing leader or a marketing manager is gonna be responsible for making sure that that AI agent does its job. And if it doesn't do its job, it sounds strange, but you're gonna have to fire that agent and create a new one so that it actually does what it's supposed to do in the way that it's supposed to do it. And so it's just a really different way of thinking about managing of resources, motivation and management become very, very different things when you're bifurcate between a human and a machine. And I think marketers certainly have sort of a technological bent to them very often, but I think there's going to need to be a real push for marketers to understand these technologies at a deeper level, specifically because they're gonna be asked to manage them and their outputs. No longer can you just outsource the knowledge and understanding of these technologies to the IT team, for example. You need to really understand what makes these technologies work and not work so that you can manage them effectively and have them help you accomplish your organizational goals. To me, as we look forward to the rest of twenty twenty six and beyond, it's really about leaders understanding the different ways to manage resources in the most general sense, human and machine, but I think really critically around trust, and I know this is a topic you're gonna talk about here in a moment, but I think we all know that AI is not in a place yet where it can just be trusted to do the right thing and to do it well and to do it repeatedly well, and so these managers who are guiding these technologies and these humans really need to know when to trust automation and when not to trust automation. Again, it's a different kind of a skill set, it's a different kind of a mindset. I think it's really exciting and interesting, but it will take a different approach to management for sure. One sort of follow-up question for you is, you know, technologies have existed, right, in the world of marketing and all functions for that matter. Right? What makes AI agents different in a manner that we are all sort of anthropomorphizing it? So that that's that whole thing is a little bit different. Right? Like, so in your mind, what is what is triggering that? Yeah. I think it's I think it's exactly because these agents are increasingly capable of doing things without us, right? You talked about that a moment ago, that these technologies are increasingly capable of running on their own. I actually just saw an article yesterday about a company that enabled AI agent driven web browsers, and as a test allowed them to run entirely independently for a full week. This got to all sorts of things, not least of which of course is how long are these AI agents able to operate on their own? But really like what are we handing over to them? And so I think we're in this moment of saying, okay, these agents are capable of doing things, increasingly doing things on their own. What are we gonna give them to do on their own and what are we gonna make sure that we as humans are still very firmly in control of? But I think there's also this natural human interest in just sort of giving over a task to someone, think about when you have a human intern, you're like, okay, I need you to go and create this report for me, and so you give that intern an assignment and she goes out and she does that work for you and she brings something back to you, And I think we're increasingly seeing AI and AI agents as interns who we're sending out to do work on our behalf. And I think that the more we think of them in that way, the more we get comfortable perhaps a little bit too soon, get comfortable with letting them just do stuff on their own. So I think that constant balance of pushing the technology to see what it's capable of but also maintaining that very human in the loop moment of we're not just gonna have this thing go off and do it itself. I think that really is at the sort of core of how we're trying to come to grips with where this is all headed. All right, home stretch here, Matt. Yeah, Abby, we have arrived at number ten, and is, again, is an area that you know is very near and dear to my heart, is human judgment, and I'm interested in your point of view on what is the role of human judgment going forward? Are there limitations? Are there benefits? How should we be thinking about that human in the loop? This is just so critical. So we save the best for last, right? Where I think AI is removing constraints around speed and scale and execution really around all marketing functions and across kind of everything, content, media, optimization, ads, all the things we spoke about, you know, up until now. And so the limiting factor for marketers is no longer production. It could be data, but I feel the limiting factor is really going to be human judgment and taste. And that is what is going to drive, you know, what is right, what is wrong, and what are things that that marketing team should never share. From a CMO standpoint, I feel that decision quality becomes the constraint. You want to make sure that when AI can generate endless content, endless whatever, fill in the blanks, the real bottleneck shifts to the marketing leaders in deciding, you know, what is in alignment with their brand value. So decision quality is sort of key, is a key constraint. Right? Number two, I feel, and I just referenced it, but taste and judgment, right, will start to matter, you know, more than throughput. Can I go to chat GPD today or whatever, you know, LLM or content technology of my choice and have it pump out three hundred and sixty five blogs, a blog for every day that I could post on the website? Of course I can. And it's gonna take, you know, maybe five minutes to pump that out. But that's where taste and judgment will start to matter. Because now I have to obviously read it and understand it and make sure it's in alignment with my brand voice and my brand values. And it makes sense to the ICP and all those sort of things. Even if you can provide all those things as an input, it still requires that uniquely human quality of sort of, I don't know how to explain it, a feeling or gut feel that when you read something, want to make sure that that taste and judgment is in alignment with your brand and your sort of value system. Finally, I feel like governance becomes sort of a strategic pillar, right? So CMOs are going to spend less time in managing execution. They should. They should not spend a lot of time in managing execution. But more time in sort of setting principles and guardrails and escalation pathways and everything when AI output needs that human intervention. And honestly, I feel that one is so important because human beings can be inherently lazy. Like marketing teams have a lot of things to do. And just because the last three times your AI generated on brand blogs doesn't mean it's gonna do it for the fourth time. It can still hallucinate. It can still make things up. So you gotta make sure you have the guardrails and the processes and sort of the inherent systems, which will, you know, protect your brand from getting hurt, and that would be sort of the underlying responsibilities of the marketing leaders and CMOs. So I feel like human judgment is gonna end up becoming that bottleneck, and it's going to require, you know, CMOs and leaders to really not think only in terms of production of volume and speed and scale, but also in making sure that those things sort of make sense. Just because you can doesn't mean you should. So I I know this is near and dear to you, so what do you think about this? I love that. And I you know, it's interesting. I I've spent a lot of time lately thinking about a flipping of the script. Right? We talk an awful lot about where can we use AI in our business to be more effective, more efficient, more whatever, creative, more productive. But to me, increasingly, I'm seeing out in the world and what I'm feeling in my own business is that that script is now being flipped to the point where that unique human set of qualities and capabilities, to your point, you called it a bottleneck. I look at it as a real positive to say, let's let the AI handle the things that it can handle and let's elevate the human qualities that matter so much in these relationships. You talk about taste, right? Taste is enormous, right? We can all tell the difference when something is AI generated. It lacks a soul, it lacks a purpose, it lacks a humanity that we all inherently want, and I think the flipping of this to the focus on the human and all those great qualities that humans have is one of the big changes I think that's gonna start happening in twenty twenty six and beyond, and again allowing that AI to have its place, but using it as a way to elevate the human qualities that make such a difference in the field of marketing. That was a great note to end this episode on, Matt. Thank you very much. To our listeners, thank you very much for listening in. We really hope that this has been a spark for learning. And if you want to see past episodes, you can go to, you know, MarketingAI SparkCast dot com. You can connect with Matt and me on LinkedIn, and would love to hear insights and commentary. We'll obviously be posting this in all our favorite podcasting platforms, and you should see some of these snippets and clicks on social media. Reminder, don't forget to join Matt on February twelfth at the Marketing AI Institute agency event. Abby, wait. You have to promote your own event, which is that same day in Atlanta that day, because I will be doing the Marketing AI Institute AI for Agency Summit, but that very same day, what's happening in the Atlanta Tech Park Avenue? Thank you, Matt. I see you made your choice, but Marketing AI Pulse is having an event here in Atlanta. It's our quarterly in person meeting. We have just kind of a phenomenal, you know, event plan. We have some really, you know, really good speakers. And, you know, overall, I'd say that if Matt was given a choice, he would have really come to the Marketing AI Pulse event, but I will forgive him this time. But yeah, for those of you who are interested, if you're in Atlanta or around and want to drive up, it's half a day event at Atlanta Tech Park starting at one thirty. It goes up to five thirty. We have a happy hour at the end. We have some phenomenal speakers. We have folks from HubSpot, from Citi, from UPS, from Coca Cola, and a couple more brewing. So I hope you can join us. Matt, thank you very much again to join in. To our listeners, till next time, have a good one. Thanks for listening. Big thanks to SparkNovus and the Marketing AI Pulse community. You can join the community at Marketing AI Pulse dot com.