Welcome to the Marketing AI SparkCast, where AI meets modern marketing. Hosted by Aby 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 Spark Novus and the Marketing AI Pulse community. Here's your host, Aby Varma. Hello, everybody, and welcome to another episode of the Marketing AI Pulse brief. This is our episode for July 2026. If you don't know, the Marketing AI Pulse is one of the fastest growing communities focused at the intersection of marketing and AI. If you are not a member and you wanna learn more, you can visit marketingaipulse.com. You can sign up for free. And this episode will not be complete without welcoming my friend and my cohost, Matt Cyr from Loop dot ai. Hey, Matt. How's it going, Hey, Aby. How are doing today? Doing good. You tell me. It sounds like you have a a lot of new things brewing, you have a new product. Tell us a little bit about that. What have you been up to? Yeah, it's been a busy summer. We've been doing the consulting work for now about over two years and just last week actually established a new corporation. It's called Signal78 Incorporated and it is for an AI platform built specifically for marketing agencies. It's something that we've been working on for a while now and we've gotten to the point where we have a lot of great momentum. We're seeking outside funding. We're working on a pilot project with client agencies, and we thought that this was the right time to establish a corporation and take this product and project to the next level. So it's an exciting time. We've redone our website to reflect better the work that we're doing as consultants, but also the platform that we're building and would love to talk with anyone who's interested in learning more. You can go to signal78.ai or loop agency dot a I to learn more or just hit me up on LinkedIn and I would love to tell you more. That's awesome, man. Thank you. It looks like you've been really busy. So very appreciative of you joining in today. Digging right in, we have four very exciting stories. So the first one is all about the MCP explosion in Martech. MCP has been making a lot of noise. A lot of people are not familiar, so I am going to break that down. You are going to cover... Think Forrester's come up with some new agency research, so that's their second story. Then we're going to talk about AI slop on LinkedIn. There is a lot of that. So can't wait to dig in, and then we're going to wrap up with some cool things that Unilever is doing in terms of how they're leveraging AI, without going generic, and Matt's gonna unpack that. So with that, I'm gonna dive in. I'm gonna start talking about MCP. So in July, 6sense, Sprinklr, Wistia, Adobe have all launched MCP capabilities. Right? And it unlocks a lot of different capabilities. You can now look at buyer signal, customer intelligence, video editing, paid media capability, and you can do all these things using your standard agentic platforms like ChatGPT work and Claude Cowork and so on. And a lot of people, I assumed everybody who is within the world of AI is already familiar with MCP, but that is not the case. So what is MCP? So MCP or Model Context Protocol is basically a shared standard that allows AI assistants to connect with software and use the information and capabilities inside of it. So think of it as a USB connector for AI to various smart tech tools. Without MCP, companies may need to build separate integrations between every AI assistant in every marketing platform. And MCP provides one common method of doing that. Anthropic introduced MCP back in November 2024. OpenAI, Google, Microsoft later added support for it. And in December, MCP became a founding project of the Agentic AI Foundation under the Linux Foundation, giving it industry wide support. Scott Brinker's State of Martech 2026 report, came out early May, identified about 29,000 MCP servers across registries within roughly eighteen months, nearly twice the number of commercial martech products tracked in that report. And this report was in May. I'm sure there's a lot more. So in other words, MCP is pretty much the de facto standards where your AI platforms are connecting to other platforms. So how does that really work? So a marketer connects an approved marketing platform to an AI assistant and makes a request in everyday language such as which as a underperforming or what should we change. So this prompt is something that you would put into, let's say, Claude Cowork. And then assuming Claude Cowork is connected to your paid media platform via MCP, that MCP is going to go fetch all that information and send it back to you. So you're not really using the product interface. You're using everyday plain speak language within your AI platform of choice and connected to these third party tools, which then can bring that information back. Now, the AI assistant uses a connected platform to retrieve information. You could use it to analyze it, recommend action, and they've permitted even complete actions. Right? So the level of access matters. You're treating a campaign report creates less risk. It's read only, and then you can add things that where you're doing publishing or if you're changing budget or activating an ad comes with more risk. So, obviously, there's an element of governance that comes along with it. So what does this really mean for Martech? So AI assistants are becoming really the new way to operate the stack. Marketers can begin in ChatGPT, Claude, or Copilot instead of opening multiple platforms on the computers all at the same time. Martech vendors, they... I think they're gonna compete differently. The quality of the data, the intelligence, or the actions may matter as much as the native interfaces for these Martech platforms. And the integrations become way easier. You don't need a developer anymore to do API based integrations, which was the world we all lived in prior to AI. So now it's pretty much one clinic authenticate and you're in. However, data quality becomes more important. And then it also opens up governance and risk to the forefront. Right? Organizations need to have role based permission, human approval, and getting a very good understanding of what actions your MCP connection enables and make sure there's an audit trail. So there's a lot of things to think about in the world of Martech. Though, Matt, I know you mentioned Signal78 earlier in the episode. I believe you guys are leveraging some of the MCP capabilities in there. Maybe talk to us a little bit about that as a real world example. Yeah. It's... It is really a critical thing for the functionality and success of the platform, And you described it well, which is essentially opening up a connector between different AI servers. And the specific way that we're gonna be using it is our platform is meant to be an aggregator of all of the business context for our clients. And we are connecting via MCP to our clients software solutions of record and getting the data from those platforms, bringing it into Signal78, making sense out of all that data, and then having a two way ability to push updates and information back into those source platforms via those same MCP connectors. So to me, when I hear Model Context Protocol, the really critical word in there is context because it will allow us and anyone who's using them anywhere to bring all the context from all your different systems together and make sense of all of that. So you're really starting to get some of the compounding value that AI has been promising for a while. So it's a powerful set of tools to your point, it doesn't require a technology developer, you know, to build up. To your point, it doesn't require a developer to build a traditional API, it just often happens with the click of a couple buttons. So there are risks and rewards to that of course to your point, but we have found it to be an enormously powerful and potential creating technology. Even for us internally, the amount of time savings where you start orchestrating multiple pieces of technology through ChatGPT work and Claude Cowork, we use both internally to support our clients. You can connect. You can say, hey. Look at emails from said client. Connect to Fathom Recordings and connect to Slack and synthesize all those messages and convert that into an actionable to do list and go log it into Monday. Right? So that entire process is pretty incredible that something like that could end up taking one day to go synthesize, sometimes weeks to go synthesize all that information. So it's a pretty powerful capability. And CMOs will often ask, so what should we do? So here's my advice. Three things. So one is assess MCP readiness, and that's gonna start with making sure you have a very good understanding of your Martech stack. A lot of CMOs don't. So, obviously, see what's the key piece of Martech within your marketing team and understand which of those pieces of Martech have an offer official MCP connections, get a very good understanding of what the connection can see, what can change, and what it can record. So that would be baseline, just getting on capability understanding. Once you get an understanding of it and then making sure you have a strategy in place and what are your marketing goals and business goals, then pick out the right pieces of Martech and MCP which can help you achieve that. So that way all your MCP efforts are really laddering up to a larger strategic goal, but pilot one use case. So begin with a read only workflow such as campaign reporting or account research or video analytics and that thing, and then start measuring quality and business value. And then when you get from that point on to more analysis and execution, that's where... Don't forget about governance. So the third recommendation is make sure role based access. Not every Martech user needs to MCP into the tool of choice. So making sure that you have really good understanding of that. Where is human in the loop and human approval processes that don't remain static? Make sure that MCP does not circumvent that, making sure that there is some observability and the ability to audit what changed and who changed it and what MCP was involved as part of that. How is that tracking in place and... And then making sure that what kind of access MCP has to your to your data sources and the type of data and whether that... That's allowed or not and those things. So involve your IT organization, involve your compliance organization for business systems that are mission critical. But in general, I think as a CMO, you can't really ignore that anymore. And, Matt, I'm sure you're seeing that with your clients as well. Yeah. I think the the power and the potential are there, and I think people are really just starting to understand implications, both positive and potential risks. So it's definitely something that marketers and leaders should really pay close attention to and identify to your point which of their existing technologies are already using MCP And those that are not, I think it's worth trying to figure out if they're planning to add those capabilities in the future. But again, I think as you said a moment ago, the industry seems to be settling on MCP as a sort of cross industry protocol. And so I think there's a lot of value in understanding what's possible. Alright, Matt. Moving on to our next story. You are going to tell us a little bit about some research that Forrester's come out with specifically around how agencies are using AI. Unpack that for us, please. So yeah. So Forrester released a report in late June. As I mentioned a moment ago, my focus really has become exclusively on marketing agencies. And so my interest was peaked when I saw the release of Forrester's State of AI Inside US Marketing Agencies report. They partnered with the 4As and released this report that showed how agencies are doing with regard to AI adoption, what kind of value are they able to create, and where things are headed. And I would say the headline here is that nine in 10 US agencies are using AI, no surprise there, but what they identified was that most are using AI to do work cheaper and Forrester is saying is quietly eroding the creativity that can drive real growth and distinction. So just to pull that apart a little bit more, nine of 10 agencies are using Gen AI, half are using agentic AI for marketing execution. I would actually really like to know more what that actually means because in my experience, not half of agencies are using agentic AI for execution. It's much... In my my experience, much lower than that. But what they have found is that the majority of the purpose behind the use of AI is for productivity, and that really isn't a surprise, that's how AI has really come out onto the market which is be more productive, be faster, all of those kind of things. But Forrester is saying is that the typical things that are being used that AI is being used for building creative content, doing media and SEO strategy, productivity, document summarization, research competitive briefing, all of those things are great and valuable. But what Forrester in the 4As is not seeing is this then turning into a real positive in terms of creativity and strategy. And so what their warning is then they say this pretty much verbatim in the report is that agencies are mistaking efficiency for effectiveness. And they're saying that short term productivity and cost are certainly beneficial, but it's stifling creativity and long term brand growth. 61% of the agencies that they surveyed were using AI as a cost center, calling it a cost of business, whereas only 31% are planning to monetize agentic AI in the next twenty four months. So I would say to the agency owners and teams out there, if you're not yet thinking about ways to monetize AI in your operation, now is the time to do that. There were some expected barriers for these agencies to go deeper and use AI in more impactful ways. Accuracy and bias, 63% said that they were concerned about that, legal concerns 62% flagged that, and privacy and data security concerns 55% flagging that as a concern. All of those are expected, I would say the accuracy and bias concern is becoming less of a challenge over time. Think, I don't know, Aby, if you're seeing the same thing but I have noticed that these models are becoming more effective, more accurate, more capable, and it seems to be happening pretty quickly. Definitely model capability over the last couple of years has skyrocketed but agency capability can't say the same thing. Right? Like, were a lot of agencies where... That are still using AI to improve their efficiency, kinda like what the research says. But the the financial model for agencies to help with their clients is is something else, right, where that has, you know, the the t and m, how many hours and billing by the hours, that is... That model is sort of broken down. And I'm not sure... I think value based pricing is great to say, but it's really hard to execute. I see that, you know, I see that a lot with agencies where agencies sort of struggle. And, honestly, clients struggle. Clients, there's an expectation of putting so many hours in and people were paying you for the hour and not for the value that you guys were bringing. What is your take on all of that? Yeah. I think you you have... I always give you credit for it. You've got a great phrase, which is AI can help you do things differently, but it can also help you do different things. And to me that is a really powerful way of looking at it because I think most agencies are using AI to do things differently. And it's how do we make the same things faster, the same things more efficient, which is fine and there is value in that, but it's not as powerful and potentially valuable as looking at AI as a way of doing different things. And I think what this report is calling out is that agencies need to really take the opportunity to think more deeply about how to do different things with their AI not just do the current things, the current models, the current processes, the current approaches, all just faster and more efficiently. So it's a real transitional period and what I think this means for marketers and that the Forrester team calls out is that AI is shifting from adoption to impact. The phrase that I use is experimentation to systemization, right? And I think we're in very much in the early stages of that transition. Everyone has the tools, everyone's spending money, time and energy on having all of this AI technology. Some of it is adding up to value but much of it is not, right? And so just being able to go faster is not a differentiator. It's table stakes. Their clients are expecting it and they're expecting you to be able to do more and bring more value. Cost cutting has a hidden bill. Using AI just for efficiency is going to erode that creative distinctiveness that is driving brand growth, that's driving the ability for you to go in the market and share the things that your agency does than others are not. And I think what I would say savings today, sameness tomorrow. The last thing that this means for marketers is how your agency treats AI signals its ambitions. Right? If you're just using it to do the same things faster, more efficiently versus doing new things, doing different things, I think that ambition is going to start to become really clear both to your and to the marketplace. So it's an opportunity for agencies to step out and really challenge themselves to to use AI for what it's absolutely the best at. So what do CMOs need to know? What should they do about this? First is redirect the efficiency dividend. If AI is freeing up capacity and capability, turn that into... Funnel that into talent training and creativity. Don't just use that to salt back into your EBITDA, Use it to really make an impact in the parts of your business that your clients are gonna choose you for. Change what you brief and what you reward. Stop rewarding the same output faster and cheaper. Do briefing on differentiation and measuring business outcomes, not speed. And finally, interrogate your agency's AI approach and strategy. Are you treating AI as a cost center, as a capability? Are you see... Using it as an opportunity to help you become better at the things that you have always wanted to be best at? I think those kinds of decisions, that kind of thinking is really critical and this is the moment to have those conversations. The agencies that are continuing to adopt it just to move faster, that's easy. But the agencies that have moved into a systemization that allows them to do things more compellingly, more excitingly, think that's a different kind of a business decision. Aby, I know you work and talk with a lot of agency folks. Are you seeing that e two systemization transition? Oh, 100%. 100%. I mean, one of the agencies that we collaborate with was recently talking about, hey, we're... They got a brand new $10,000,000 paid media account. And in the pre AI world, they would typically add two or three people specifically to the account, and they don't wanna do that. They wanna add, obviously, one or two people and not the three and four they would have traditionally done and really leverage that, and not just for this client across the board. So in terms of just pure internal operations, definitely agencies are getting more efficient, delegating the groundwork to AI. But I also see a lot of AI being used for ideation and concept development and all those things. And it's hard because when you're looking at creative process, what makes things creative is the rigor. And without that rigor, that's the... That's what sandpaper is used for, to polish things. And that's where the... In my sandpaper analogy, that's where human beings have a very critical part to play because you were coming up with ideas. You're bouncing ideas off of each other and all those things, and people can do that with AI, but that level of polish is often lost in in the in that process. You lose that rigor. So I certainly see agencies doing that. It can show up in their work if they're not careful. And like you said, there's certainly an element of sameness or that's the risk that ends up happening. Absolutely. Yeah. And hey, speaking of sameness, Aby, this is a really good segue into your next topic which is all about the content on one of our favorite platforms. Take it away. It's LinkedIn. Right? So AI has made content infinite and nearly free, but not quite. Right? So the bill is coming due and it's paid in sameness. So what is happening? There is this new study from AI detect... Detection firm, Pangram, that scanned a million posts that people actually scrolled over in the last two months. So real feeds, not spam sites. Right? I'm going to read out some of that data. So 41% of LinkedIn long form posts, which is two fifty words and plus, were flagged as fully AI generated. That's like 41%, and 30% of short form posts. So, in LinkedIn, it was only about a third of the post scanned, but nearly two thirds of all AI content flagged. So, most saturated platform measured, well ahead of x, which is about 24, 25%, with a fully AI long form, roughly four x Substack, which is about 10%. And even LinkedIn comments were more likely to be AI written than top level posts on any other platform. So the twist and the irony of the whole situation is LinkedIn has this feature called enhanced post. Right? It's a little bugging that helps you use AI to write posts. So, well, people have been using that. So what ends up happening is people are using that to have AI generated content. Now LinkedIn is saying that it will detect and down rank content that reads as AI generated. And this is from a post from the executive editor, Laura Lorenzetti, calling out AI slopping polished on the surface but lacks any unique perspective or substance. So what is the translation of that? It's really same platform that train everybody to underwrite is now turning down the reach on it. What does this really mean for marketers? So the reach you thought AI was buying is being quote unquote taxed back. So AI posts that are just generic are now going to face algorithmic down ranking, and I'm sure it is not just on LinkedIn. It is going to be on other platforms. So more and faster and cheaper can suppress your own distribution. We've spoken about this in previous episodes, Matt, where authenticity becomes a scarce asset. Right? When 40% plus the long form feed is slop, a real human point of view stands out more than it has in years. In fact, I just ran into an agency at an event, and they are positioning themselves as uniquely human written content, like a 100% human written content. That is their USP. And I was like, wow. They're definitely early in this space, but sounds like they may be onto something. And then the thought of just have AI writer LinkedIn is a losing strategy. Right? If it's polished but generic and it's the default way of how everybody is producing, it sounds like there's gonna be a lot of sameness, such as way we started this episode, and it sounds like it's gonna be in failure mode. Matt, you've been pretty active on LinkedIn yourself with your posts and your newsletter. Talk to me a little bit about what you were seeing, what your perspective is here. Yeah. It... I think the AI generated content, both posts and comments is really becoming ubiquitous. Right? And I think we all do the natural thing, which is to quickly recognize that it is and as the data showed, we scroll past it. And I think what's interesting to me is that you can't game these systems, right? If you think that you are gonna be able to beat these systems at their game by just pumping out a ton of content, you know, that that pendulum has already swung all the way back. And I think it's ironic and a little bit hilarious that now we're talking about 100% human created content. The new and exciting thing, it sounds like we are living in a bizarro world here. But I think, and you mentioned it, but we've talked about topics like this in the past. I think it just comes down to the obvious answer of be authentic, be human, have a point of view, share that point of view. If you're not a great writer, make a video. If you're not great on video, write something. But people are buying people and points of view at this point and AI is not offering either one of those. And I think it's really good to see that LinkedIn is starting to decrease the value and visibility of the AI content because I think it's just making it impossible to find anything of real value through all the noise. Yeah. A 100%. And I think from a CMO standpoint, what should CMO do? So I would say move AI back one seat. So use it for research, structure, editing, not the author of record. Right? If a human isn't supplying the actual point of view, the post is not worth posting. So making sure that that point of view and opinion and judgment and those things are very critical components in telling your story on LinkedIn or any other platform. So that's one. Number two is... This sounds incendiary, what I'm gonna say, but, like, resource what AI can't fake. So if if you're using AI to fake things, like executive opinions, first person stories, UCG, proprietary data, and believe me, brains are doing that. Right? There's a lot of that synthetic stuff out there. My advice is resource it so you don't have to come up with it synthetically. Invest in coaching your executives and getting real executive opinions and speaking to customers and getting first person stories, that investment is going to go really far and gonna ring true and stand out in, a sea of sameness. Right? Because, again, it is real. It's the real deal. So resource things don't use AI to fake some of those things. And then finally, change the scorecard. Reward engagement quality saves real comments, inbound conversations over post volume. I've seen a lot of teams have artificial guidelines and saying, yeah. We gotta post once a day. And that pressure often leads to brands coming up with agents or what have you for authoring and posting some of this AI content. So rewarding output per week is what is helping you almost accept that AI slop. So stop doing that and really focus on engagement quality. Yes. You wanna make sure you're posting at the right cadence. But if it's at the detriment of your team and they're unable to do that and now they're using AI in order to hit those goals, it is not rewarding you correctly. And, Matt, the good segue to the next story, which is one team has been doing this really well. So it sounds like Unilever is using AI in the in the correct way. Maybe talk a little bit about that, about what is going on in that world. Yeah. It's... It is interesting right on the heels of that LinkedIn study to have seen the story about Unilever and how they're really putting the focus on human created content with a point of view and with creativity and all of that and yes, this really is the sort of other end of that spectrum, right? You talked about the sameness tax a moment ago, AI making content infinite and free. The cost of that is sameness, but what Unilever has done is to invest in content and creativity and using AI in the ways that can really allow for that content and creativity to come through. So at Cannes last month, their CMO Leandro Barreto announced that they had grown their creator network from 10,000 people to 300,000 people. Their products are being sold in over a 190 countries. And so this is a massive content operation at a massive scale. So impossible to do all of this by hand and yet a perfect opportunity for AI to be used for that grunt work that you talked about, right? And so Unilever has done is to use AI for creator discovery, vetting, brand safety checks, content briefings, document standardization. They have a tool that scans social platforms to surface people who are already posting positively about Unilever's products and identifying them for outreach and potentially bringing on board. But through all of this, what stays human, are those creative decisions and brand creator relationships, right, as the, as the CMO said, technology is used for us to augment human choices And that's really what it's all about, right, is how do you let the AI do what it's really good at and thus freeing up the humans to do the things that they are incredibly good at. As additional backdrop, US influencer spend is projected to grow 15.7% by 2027 to a total of $13,700,000,000, billion with a b. And a typical campaign jumped from five to eight creators to now 25 to 30 creators. So the scope and scale of all of this is increasing exponentially. And so of course, think the worry and the risk is what you described in the LinkedIn piece, which is that it just becomes all the same stuff. And as as mentioned in the article about this, what you get is potentially unimaginative. So anything original risk getting flagged, delayed, or reshot. Automation can quietly kill that creativity that you hired those creativity, those creators for. And so again, Unilever putting a focus on that human generated content, the relationships that add real value and really understanding their customers. So Aby, again, know you talk to a lot of marketers every day, every week. Do you see any of your clients or contacts who have taken a similar approach, which is to allow AI to do that grunt work and leave the humans for the creativity and the relationship building? Oh, absolutely. I think it is fascinating, especially in creative teams where creative folks are gonna use AI from a production standpoint. So for example, one of the clients that we work with, they're a big paid media shop. They'll have a core creative. Now they need to kinda replicate and add derivatives of that creative in 20 languages and 20 platforms and eight sizes per platform. That's just a lot of image assets that need to be done. So they they love to use AI for that clause where you can just take an image and then replicate that across different sizes. So that's a fantastic use case for that. Obviously, they have human in the loop to review and check the work that AI has come up with. Prior to that, they even have guidelines in terms of breathing for derivative creations and do's and don'ts. But on the flip side of it, the creative people are very uncomfortable for AI to come up with the creative ideas behind it. So they want those to be uniquely human. Not saying they're not using that for creative exploration, but the actual creative that ends up being tested or or tried out definitely is a 100% human. That may have been an amalgamation of five concepts, which... Three of which may have been AI and two human, but the final product before it gets replicated into 800 image assets derivatives is definitely human. So that's an example of where I'm seeing people are getting smarter about where to use AI and where to keep AI out. Yeah. Yeah. Those are great examples and I think as we look at what all of this means for marketers, the first one I think is really directly connected to what you just said which is automation is table stakes at scale. Right? Unilever at an average of 30 creators across a 190 countries, there is no manual option to do all of that and do it all well. So the question is not what you automate. So the question is what you automate, not whether you automate it, and this is where that really critical distinction comes to play. Scale and distinctiveness are actually the same problem. The win is not volume versus originality, it's automating the operations so humans have the room to be original. To your point, it's not about how many derivations of an ad can we create. It's about how do we make the ad powerful in the first place and then let the AI handle the derivations from there. And the relationship is the mode. Unilever is relying on these influencers to help really drive brand growth and visibility and ultimately revenue growth and if you can't build relationships with human beings and keep those relationships strong and find value in those connections, then it's kind of pointless to be using AI frankly for any of it. Keep that creativity, keep that relationship, keep that judgment, and that's what your competitors will have a very hard time copying. And so what do our CMOs out there, what should they do about all of this? First and foremost, draw the automate versus protect line very explicitly. I think it's easy to just say let AI do that, but I think what we saw on the LinkedIn example that you gave was that just automating the creation of this content is gonna, all that's gonna do is turn into people scrolling past without paying any attention to the content because people know it's not real. Make that line very clear, make sure your team understands that line, and make sure that you have systems in place to gut check and make sure that people are doing what they're supposed to be doing. Pressure test your briefs, if a per approval workflows punish anything off template, you're manufacturing unimaginative content that I mentioned a moment ago. Build room for creators to take swings to be creative. Don't just use AI to automate out those oddball ideas, that can turn into the things that really break through. And last but not least, invest the time that you save. This actually calls back to the Forrester topic that I spoke about a few minutes ago. The point of automating this work is to redeploy people towards creativity, strategy, relationship building, human connection, not just to EBITDA or habit or to headcount. Make sure that the free up time, at least some of it is going back to those things that will make a difference for your agency and for your clients in both the short term and the long term. A lot of... I think a lot of really good advice to be learned from these two studies that I highlighted, but if I boil it all down, to me it's human in the loop. It's letting the AI do what it's best at so the humans can do what they are best at. Awesome, man. Thank you. That was that was a perfect end to a fantastic episode. Thank you very much for joining me and sharing your perspective. Absolutely, Aby. You know, love this stuff. So it's good as always to see you and good to chat about the latest. Awesome. And to everyone listening, I hope this conversation sparked a new way of thinking about what AI can make possible in marketing and in your work if you're on the agency side as well. You can follow Matt and me on LinkedIn to continue the conversation. We'd love to connect up with you guys and really understand what you guys are doing. Until next time, see you on the next Marketing AI SparkCast. Thanks for listening. Big thanks to Spark Novus and the Marketing AI Pulse community. You can join the community at marketingaipulse.com.
