What every CMO needs to know about Agentic AI

Cmos need to up their game and upskill their teams. When it comes to agentic AI, there's a critical lag between understanding and utilizing agentic AI. And my guest today, Ravi Shankar, has been looking at this very closely. Ravi, welcome. Great to see you again. How are you? I'm good, Chris. I'm happy to be here. So let's just dive straight in because I'm guessing you mix in circles of other CMOs. And I'm just curious to know your assessment of the adoption of AI or agentic AI and do they know the difference? What's the general rule of thumb out there? Well, the adoption is actually lagging a little bit, I should say. There are three things that are actually leading from the AI perspective. One is the investment, the second one is the innovation, and the third thing is the hype. These things are running ahead of the adoption. I would say probably about 25% of the people when it comes to generative AI, they are adopting it maybe once a week or so, or maybe even more frequently. But the overall adoption, what I have kind of seen as people about, you know, 10% of the people using it on a daily basis. So they are not very actively using AI. And now it comes to agentic AI, that awareness is even low. That percentage is probably in the single digits, I would presume. And it has a long way to go. It has a very promising potential and people should adopt and learn and use it for, to be more effective. Do they generally get the difference? You know, because obviously AI, agentic AI, AGI, the former has been here in the marketing realm for some considerable time. It is agentic AI. That is where the breakthroughs are now happening. That is correct. I don't think there is a lot of awareness about it. There are some people who are educated that when it comes to the agentic AI, it is more autonomous. They are still treating it like the generative AI, where it is more conversational. That is the way they are still treating the agentic AI. But the promise is in the autonomous capability of it. And I think that education needs to still happen. And that's the interesting point, the autonomous aspect of it, because I was reading myself the other day and also just before coming onto this discussion, that agentic AI is going to be able to plan, execute and optimize marketing campaigns without humans in the future. What does that mean for the role of the cmo? It actually means blending both the workforce with the agentic AI. So the way I kind of see it as, once the agents are embedded within the marketing team, you need to think about it from three different ways, areas where the humans will basically lead a lot of the functional processes. So such as which ones would they be though would be related to, for example, anything related to the branding side, all the creativity, coming up with a concept, all those things you cannot outsource to an AI. AI is something that can take and that can act, but it cannot really create something new. So that creativity still has to stay within the marketing team. The other two are primarily like ones that are business processes that are enriched by AI. So humans and AI work side by side in order to accomplish them. And examples of those could be, for example, on the content side. So you use AI to produce the content that is actually needed. And the third thing is where it is purely AI driven, humans are not there. They have given the instruction to the AI agents and the agents are running independent, autonomously, trying to do it. And that to me is like hyper personalization. So being able to understand the intent at the time it is coming from the engagement with the audience and be able to customize whether the content, email, outreach, engagement, everything the AI does on its own. So that can be. So it needs to be divided across these three different types and then move forward with it. So let's just describe this. What does the future marketing team look like? What people, humans do you need? What role will the CMO play in particular? I'm guessing as the conductor of this orchestra. But tell me more about the structure of the team. It's less about who does what, but it really comes down to the business processes within the marketing. We have really end to end business processes that cut all the way from creating the plans, being able to execute the plans and tracking the plans and so on. So in all these cases it requires these two coming, the human and the AI agents coming together and working, working together. And that's where the CMO should be seen more of, as you mentioned, a conductor of an orchestra be able to orchestrate how the humans kind of work. What are the processes where humans can really excel? The ones in which AI can excel and together, how do we complete the task in a much more timely fashion and then be able to automate and produce the results? Because there are certain processes that are really heavy lifting in terms of a campaign creation and evolution, execution, especially when it relates to content. Different types of Persona, different types of industries, different types of buyer journey and the languages can complicate things. And it requires humans to do all these things takes a lot of time. AI can really act really very fast. And the agents, once you kind of program them and provide this, they can actually go about doing this without the humans much more faster. And that way together we deliver things, the results, much more faster. Now, I know you're from the tech sphere, you're pretty savvy on AI and agentic AI because I know you're passionate about it. But what about the rest of us, the laggards, the people that are behind and listening to you now, this sounds fantastic on one level, daunting actually on another. And I'm just what guidance would you give to the CMO who needs to embark those ones that you described earlier as the ones who are lagging behind, where do they start? How do they keep up? How do they avoid falling behind? So I would say the role of the CMO kind of becomes one that enabler, educating the teams, upskilling the teams in order to be able to use a very effectively. You are going to see a spectrum of people who are really going to fast movers, adopt it and use it. Even within my own team, that kind of happens. There are certain people who are not really up to it and then they are lagging behind. It's important to corral all of them to come together and be able to do it. To me, I always say, Chris, is that there is no AI without humans. Humans are needed in this equation to make AI work. So if part of my team is not adopting AI, it is to the detriment of my operation. So it is important as a CMO to enable them so that they can be more productive as the ones who are adopting it as a much more faster pace. But it also scares me at the same time, people who are willingly not getting on this stuff are going to be left behind. So it is in the interest as a cmo, I can bring this horse to the water, but it's ultimately up to them to drink it. So they need to be able to embrace it, adopt it and use it for, for their own career and for their own productivity. But the headlines aren't helping, are they? Because all around them we're reading the shock headline. Similar headlines to digital transformation in the old days is 70%, 80% of these projects fail. How can you avoid screwing up by taking the time to do it right? I think as I mentioned, definitely the hype is overrunning before the adoption. The investment is for outstripping, as you see in the data centers and all that particular stuff. And also the innovation companies are coming up like every day there is something new that is Coming in the AI, you read about it in the news and people are being pushed to kind of get it done much more faster. And we know if you drive down too fast down the highway, you're going to crash at some particular point. That's what is happening. People need to be methodical in terms of understanding the technology, where it is useful, where they can insert it, take the time to actually implement it so that way you can be successful. So it is not like slow and steady tries to win the race, but where would I start? Go slow, but where are the quick wins? The quick wins are primarily going to be first of all, like in the content side, we have actually seen this within my own company where we can actually produce the content at scale, especially on the website. For example, once we implement something in English, it can automatically translate it and push it down to my language websites. It can automatically create the meta tags for people to come and search on my website. So content is one area is the quick win. The other one is on the design side. My design teams are using it to produce videos to be able to customize it for different languages. The AI generated voice so we can produce a video and then have it like six different versions of it for different languages. So these are the areas where the quick wins are. But eventually where I see the holy grail of marketing has always been one to one marketing. If I know Chris, what he likes, what he doesn't like, my conversion is very high. But I do not. And hence I try to abstract with AI, I can hit that precision. It can understand at the time you're engaging with me, who you are, which industry you're from, what is it that you like, and it can give you something that you like that will increase my conversion. That is a heavy lifting from a marketing perspective. AI can do it at scale much more better. And that's where the inflection point is going to be. So do you advocate then just experimentation or do you suggest that there are experts out there now who will guide you through? Because I'm guessing there's lots of other issues around compliance, security and so on. How do we find the right people to guide us along this journey? What is the process? Start with the early adopters. Not everyone is up to speed on this particular one. There are people who have gone and tried and learned about it. There is going to be some amount of experimentation at this phase because the technology is new, you have to try it out in order to learn it. So that is actually going to be there. But lead with the early adopters who have already done this stuff. But more importantly, what he talked about is the AI without guardrails is really scary. It can go do things without you knowing about it. And it can damage the brand and the trust, which is very important. You need to be able to hold the reins on the AI to be able to do the right things for you. And it is important that you put that structure in place with the guardrails before you let it loose. Otherwise you're going to find a point where it is going to make up things that will. That can destroy your brand, which is very important and sacred. I was reading something this morning on LinkedIn about the need for a quality control layer. A lot of brands are moving too quickly, I think, as you pointed out, and particularly in the agency world, but moving too quickly. And quality control, that is mistakes setting in factual inaccuracies. What are the big dangers that you would warn other CMOs to look out for? The thing is, it can misrepresent the brand if you're not actually trying to program it correctly. So that's why it is important to set all these things before you can actually let it loose. And quality control is a very important aspect of. Even in the software, we talk about it more from a product physical perspective, but even in the software side, it is very important to have it done with the controls that it needs. And we need to have the transparency in how the AI works. We cannot be a black box where the AI is going to do its own stuff and then we don't know what exactly it is doing. We need to be able to understand its journey, where it is going, what it is doing, and that complete transparency and traceability needs to be there to be able to ensure the right outcomes. Indeed. And it was all about compliance and auditability. I was reading the critical things, particularly in the larger enterprises. Because the last thing you want is your agents to go rogue. Exactly. Those are two of the correct words, like the compliance. To be able to put your processes in place and make sure that it is compliant with your processes, because the processes change from one company to another. And auditability, traceability is very important. Just like you have a perception of what your team is executing, you need to be able to understand what the AI agent is executing as well. What about your own partners? You're talking about your team and how you're implementing it. What about the agencies that you use around the world? Do you expect them equally to be up on the use of AI or AGI? Exactly. They should be. They are embedded as part of the team, and they're doing something because they're augmenting the work that my team cannot do. And I expect them at the same level of my employees to be able to learn, adopt and use and provide the value that is actually needed and for them to be relevant. The equation for all of us, whether an employee or an agency, is the value that you deliver in the supply chain of whatever process that you're trying to do. The question you always need to ask yourself is, what value am I delivering here? Is it significant? If you're not, then you're out of that equation. Very interesting. Okay, so let's just stand up and look more across this whole initiative in terms of will this give organizations ultimately a competitive advantage? Who will be the winners and losers? So, yes, it is definitely going to give a competitive advantage. And it is the one who knows how exactly to use it and orchestrate it will be the winner. It's just like we're going for a sword fight. You have a sword, I have a sword. The one who knows to use it better is the one who is going to stay alive. The other one is dead. So literally you can think about AI in a similar way. The one who knows it better is able to exploit all the capabilities that it can bring, employ it within the business processes, get the results much more faster, much more accurate, while enabling the brand to be increasing the brand trust and the equity is the one who is going to win. The one who doesn't know how to do it and deploys it without guardrails, without any knowledge of it, is going to suffer. Interesting. Okay, but what about the creative aspect of all of this? And we talked about the trust of a brand, but also what about flattening, you know, brand distinctiveness, which obviously is a bit of a risk, isn't it? Because we, we have everyone using similar models and data in many respects. So tell me more about that. No, there is. I don't see that as a way. Like, even if you use the same AI, I use the same AI, it is going to come out with the same particular answer. At the end of the day, creativity rests with the humans. We are the creative ones. The AI and the agents are basically execution engines. So they can take our ideas and they can execute on them, but they can't create their own ideas. And that's the same is true with the computers. So to me, it is not, again, using the sword analogy. Like the one who knows how to use it better will provide the differentiation. So it all comes down to the human creativity ingenuity that they can employ in creating the distinctiveness, doing it much more better. And the AIs can help you get there much more faster. And that's the way you pair with them. You create the ideas. You are like the coach, but ultimately they are the ones who are playing in the field and they can win. Together you can win. Do you think though, that some companies are just treating it as a silver bullet, they're being lazy with it, they're, they are looking like everyone else. I mean, let's be honest, we just need to look across LinkedIn at the moment. There is a massive sea of mediocrity going on. And that cutting through the noise is a big challenge. Do you see that happening often and do you see that getting worse or better? Well, it is getting better. Mediocrity is always there. And I think it is important for anyone to be able to improve by themselves. And there are imposters, I would say, who are claiming themselves to be much more aware, knowledgeable and so on. But once you dig deeper, you will find that is not the case. But it is important, I would say, for people to get educated and be able to embrace it and use it. There is definitely a lot of push that is coming from the top down. The board, the CEOs, the CXOs, they're all pushing for this to be done much more faster. And they don't have much of an idea. They hear about it, as I mentioned, the hype is out there from the vendors and so on. They say, hey, I want to do this stuff. But still it is lagging. So you see that option being only 10% of it. They need to give it the time to make sure that the people are coming on board to be able to be very productive. Is there a naivety, would you say, about around agentic AI in particular across the C suite, beyond marketing? There is definitely. So they kind of think that here is something that I can bring on board and I can replace the humans and then they will just do by themselves. That is never the case. You read in the news how some of the companies have gone down the other side of the spectrum and now they are bringing back from, even from customer service agents to be able to use them for the human touch. And there is that unrealistic expectation because at any day, from a business perspective, you want to increase the profit, so that means you reduce the cost. Humans are a big part of the cost. And then with AI, you can do it much more faster, cheaper. And so On. But the reality, as I mentioned before, there is no AI without humans. So you need to have humans in the equation to be able to embrace it. And that unrealistic expectation is what actually creates a lot of the statistics that you mentioned. 50 to 70% of the projects are not working because they have no idea how to make this work. So when you think about, and I know we talked about this when we last spoke, I was talking about the short tenure of CMOs. There's a report out not so long ago, particularly in the B2B space, and there are many headwinds facing the CMO in these sort of situations. But would you say AI or agentic AI or AGI? Would you say this is the primary reason why they're losing their jobs so quickly and that CMOs need to keep up? Or is it a different issue altogether? What do you think technology is just an excuse? Chris, as I mentioned, everyone needs to ask whether you're a CMO or whether you're just a marketer down the line, what is the value that you're adding to the equation? Okay. When I was working at Oracle, Larry will say, there are only three things that we test you on. Results, results, results. So you need to be able to produce the results. Ask yourself, are you producing something that is of significant value to the company and to the equipment and to this process? Sorry, are the KPIs changing though? Yes, yes, the KPIs are changing. In the sense from a timing perspective, you need to be able to done things much more faster. If I'm launching a campaign today without the AI, will probably give you two, three months, but with the AI, I'm going to expect that to come back within a month or so. What other headwinds do you think are CMOs facing beyond AI? The big challenges that everyone is facing, the biggest headwind I would say is basically the education. There is so much, as I mentioned, the innovation that is going on out there. There are so many vendors. You have to be able to differentiate what is truly AI and what is not. Everybody's whitewashing AI. Before we would go to a data conference, now it is called the Data and AI Conference. A vendor that was out there for the last five plus years, even before AI came into force. Now they call all their wares as AI, you know, native stuff, which is not the case. So CMOs need to be able to educate themselves in addition to the execution and be the leaders who can actually impart this knowledge to the team and bring them up to speed. You cannot Rely on your team to do the work for you. You need to be in the front, leading the crowd, understanding it and imparting that knowledge and bringing the team along with you. It shouldn't be the other way around. Excellent. And just one final, final question. I'm guessing when you were a little boy, Ravi, you didn't say, when I grow up, I want to be a cmo. What did you want to be and how did you end up here? So growing up, I always wanted to be an astronaut. So to me, like, the idea of exploration was very important for me and I developed a sense of curiosity and how to understand things. They kind of work. But reality hit when I got closer to the college. At that time, computer science was really fascinating. It was kind of coming into this modern world and I saw that as this modern version of the space exploration. So I went down the computer science, became a developer at first, but then it evolved from me. It was not just how the technology worked. I wanted to understand how people choose it, why they adopted it, and how the businesses scale it. So with that, I did my mba, switched over to the business side and became the marketing side. Now I kind of understand how this journey has kind of evolved for me. I did not quite make it a space, but I landed as the CMO doing the marketing and I still enjoy doing some of the innovative works in this frontier. Excellent. And you work with a technology company. Now just quickly describe who that is and what they do in a sentence. What's your North Star? So denodo is a leader in the data management space. Our North Star is called something data virtualization, which is the ability or logical data management, which is the ability to bring together the data from the different systems into a unified version without having to physically move the data into another repository. That increases the speed it does with the fewer resources and it is less expensive. Excellent. Thank you for that, Ravi. Thank you so much for your time today. I really enjoyed talking to you. And is there anything that we should have said that we haven't, that we've missed? No, I think you covered it all. I really enjoyed this conversation over here. And so my final thing is that really get educated on the AI side. Bring along the people in your team because as they say, humans are needed for the AI to basically work and keep it in control, don't get ahead of it. And that can tip you over. So make sure you put the guardrails in place, make sure that there is a good control of the AI before you can actually start letting it loose. Great advice. Thank you for wrapping that up, Ravi. So long. Thank you so much for your time. See you again soon. See you, Chris. I really enjoyed this conversation. That wraps up another episode of CMO Unplugged. Thank you, Ravi, for being so awesome. And thank you, listeners, for tuning in. You'd be doing me and the algorithm a huge favor by pressing the subscribe button. And if there's a talented or outspoken CMO you'd like on the show, please message me directly. My name is Chris Hewitt at CMO Unplugged. Thank you for listening.