AI in the Enterprise: Strategy and Tools - Mike Kaput, CCO, Marketing AI Institute

Speaker A: At the end of the day, enabling your average person within your company to have a baseline of saying, oh, okay, wait a second, I could go play with this. Oh, I'm going to be really good at, ah, figuring out how to make this work for me. I think that's really just a critical first step, and it's not happening.

Speaker B: Welcome to the AI Edge Podcast for Enterprise Marketers. Uh, a show dedicated to sharing insights, strategies and experiences from a group of experts who have successfully implemented AI solutions in a large enterprise B2B software company. Specifically. Specifically within the context of global marketing and how that effort can connect to sales, IT product, and the rest of the business. I am Edine Porter De Leon, and I am joined, gleefully excited by my co hosts, Plural Jessica, Ria and Michelle Moore. Michelle, it's been a little while since you've been on the show as a co host. What have you been up to? What's in Michelle's world?

Speaker C: Oh, well, I have been consulting for a large enterprise company that makes, like, AI chips. And ironically, of course, internally, that company does not have any sanctioned generative AI tools. So my personal exploration continues. I consume a lot of content, including, of course, our guests podcast and Keeping up with the Trends. And I still definitely use and explore a lot of tools for my professional research and development. And that all continues.

Speaker B: Oh, it's. It's a journey. It never stops.

Speaker C: Exactly.

Speaker D: So good to have you on, Michelle. For our audience. It's been a while, perhaps, but Michelle was one of the original authors of our generative AI usage policies and really spearheaded that entire initiative. When we say it was updated weekly, that was all Michelle and her team's work creating the internal SharePoint style site with our guidelines linking to other tools. All of that really shepherded by Michelle and her team. So she is, in my mind, the resident AI policy expert. And really happy to have you back on this week.

Speaker C: Thank you.

Speaker B: Yes. Fabulous. The whole ethical AI workflow, too. That. That was really powerful. And Jessica, what is going on in your world?

Speaker D: Yeah. Good to see you again, Yadine. Excited. We're getting back on a regular Friday calendar, so for recording podcasts. So hopefully our audience, you'll guys see these pumping out more often. And super excited about our guest today.

Speaker B: Yes. And great segue. Thank you. Well done, Jessica.

Speaker D: You're welcome.

Speaker B: It's almost like we know what we're doing.

Speaker D: I'm getting there.

Speaker B: All right. So super excited to welcome Mike Aput. Mike is the chief, uh, Content Officer. I love that title. Chief Content Officer at Marketing AI Institute, an Education, event and media company that makes AI approachable, actionable for marketers and business leaders. As Chief Content Officer, Mike uses content marketing, marketing strategy and marketing technology to grow and scale traffic, leads and revenue for marketing AI, uh, Institute. Mike has published hundreds, hundreds people of articles on how to use AI in marketing to increase revenue and reduce costs. Mike is also the co author of Marketing, Artificial Intelligence, AI Marketing and the Future of Business. Mike, welcome to the show.

Speaker A: Thanks for having me.

Speaker B: Fabulous. All right, so today we're going to talk about really tools in the enterprise. And actually there's tons and millions and millions of tools. So there's really kind of, we're going to be taking a step back and tackling it from a, uh, strategic lens as well. But I think one of the great things you've done, Mike, is you've done a lot of shows, articles, webinars, enablement on here's this tool, here's what it does, here's how you can use it, which I think are absolutely fabulous and I definitely want to tap into a lot of that. And so the structure of the show, for those of our listeners who are familiar with this, we start with topics and objectives, then we go into strategy and tactics, teams and tools, then business impact, and then at the end we do a lightning round. So let's go in the first segment of the show, which is topic objective, people asked you a lot, I'm sure, Mike, what AI, uh, tools should I be using?

Speaker D: You're like, attend my 30 in 30 segment. Right?

Speaker A: Right. Yeah. How much time do you have? Right.

Speaker B: How do you answer that question?

Speaker A: Yeah, it's a good question. I think that there's no one answer here. I'd say that there's a single answer you should be thinking about.

Speaker B: Yes, exactly.

Speaker A: I'm going to actually kind of cop out a little bit and take a step back because I think you need to start thinking first and foremost about your use cases for AI. It's not about just shotgunning a bunch of technology at your team, at your staff, at your company without any conception of what you're trying to achieve.

Speaker D: Right.

Speaker A: That's not a new AI problem. We've been dealing with that for 30 years now with any type of marketing technology and Internet technology and um, software of any type.

Speaker B: Where are you saying, Mike, that marketing teams throw technologies at problems without actually really knowing what they're trying to do?

Speaker A: Those are your words, not mine. But yeah, yeah, yeah, yeah, 100%, yeah. So I genuinely do think you need to step back and say, what are my actual Top, say three to five use cases for AI. There's ways of getting to that, there's ways of figuring that out. And once you have some conception, then you're going to be able to fit the right tool in for the right job. Now Obviously there's the ChatGpts of the world, the frontier models out there, the big providers of these kind of cutting edge models. It's a wise move as a marketer, as a business professional to be familiar with at least what some of those can do. But really beyond that, you're really looking at. Okay, I need an actual understanding of what AI can do today, which sounds obvious but is not always. Things are changing really fast. So if you're a marketer or business person who hasn't spent the last couple months figuring out what's going on right now in AI, it would be a good idea to honestly like ground yourself in that. It doesn't take that much time. But understand what's possible today. Start understanding what you're actually trying to do and then fit the tool well.

Speaker D: And it's always like, uh, what tools do you have access to today? Going back to your company guidelines. Right. And do they even know?

Speaker A: Yeah, you may not have even control over all of this at your company. Your company may have already chosen approved tools for you. So again, that's where genuinely the use case methodology becomes even more important. You need to know what you're trying to achieve and what the tools you do or don't have access to can actually do to help you achieve it.

Speaker B: Yeah, and so I, uh, like the way that you phrase that too, because that challenges that idea of putting that sort of the, I guess the cart before the horse or everyone always talks about you should be people process technology, but it always seems to be reverse. Let's throw a, uh, piece of technology in there, then let's try and wrap a little bit of process around it and then let's give it to the people. Like you said, we've been dealing with that for the last 30 years, which is like, oh, let's implement Percolate. And then two years later everyone says Percolate's horrible. Well, it's because you just threw it at everyone and then it turned into a giant mess. Garbage in, garbage out. And that's pretty much the story of like 90% of implementations that I've seen gone wrong. And if they just took a step back and actually talked to everyone. And this is, I think the biggest thing is a recurring theme, is that you go and you do the process work and you find out how people actually do things, and you realize that there actually is no process and it's very dysfunctional. And what they'll do is, we need a consultant. The consultant comes in and says, oh, this is how the process should be. And then they create this wonderful, great deck. And Jessica, you and I have gone through one of those.

Speaker D: I'm like, oh, I feel it.

Speaker B: And I remember you all excited and like, I remember your printouts to the deck. This is going to be so great. And then it was like, okay, so when are we going to implement this?

Speaker D: And nothing gets done because they weren't ready.

Speaker B: Nothing gets done because. Yeah, because nobody wants to change the way they do things. So that's what you're banging your head against. So, Mike, you bring up a great point of is, what are you trying to accomplish and have that goal? And then once you created that goal, I think the big step, and you can give me your perspective on this, is you then have to go into the organization and change the way you do things. Because AI is something that's going to change the way you do things. So you actually have to change the way you do things. It's just crazy. So already kind of slipping into the strategies and tactics section of the show. So what are the key changes that you're seeing that need to happen within a marketing team to effectively integrate AI tools? And we're going to continue to push it until teams actually do what you tell them to do.

Speaker A: Oh, my gosh. I sigh because it's such a huge question too, and you hit the nail on the head. So much of this is actual change management, but I've stood on this soapbox forever. As a company, we stand on the soapbox, but we do it for a reason, which is step one that people still don't do is AI literacy. At a basic level, that's kind of fancy term, we give it. But every single person in your marketing team, in your company, the lowest of the highest level people, it does not matter what they do, how much experience they do or don't have, every single one of them needs to understand, just at a baseline, what AI is, what it can actually do, what it can do specifically for them. Too few people actually have it, are grounded in that kind of knowledge. In marketing, there's plenty of forward thinkers, uh, out there. Plenty of them, I'm sure are listening to this podcast. But the teams you're trying to bring along oftentimes are so much I don't even want to say further behind, they just haven't kind of groked the idea of, oh, wow, okay, this technology is going to fundamentally reshape what's possible and what I can do in my job. And you kind of have to open that door first, I think through a little bit of education. It doesn't have to be taking 200 different courses or watching a million different videos. You can get to it pretty quick. But you do have to ground people in saying, okay, this is different than what you've used in the past, than the software you're used to. Here's roughly how it works, here's why it's going to have an impact. Here are the expectations around this technology and what it can do for you. And then I want to enable you in some basic way to go experiment and play with it and figure out how to use it for your own role. Because you're the expert on your work, not someone else. There's a huge amount of leeway and space for like outside vendors and consultants to help with this. But at the end of the day, enabling your average person within your company to have a baseline of saying, oh, okay, wait a second, I could go play with this. Oh, I'm going to be really good at figuring out how to make this work for me. I think that's really just a critical first step and it's not happening. Mhm.

Speaker B: It's tough.

Speaker D: And even if you don't have access to it at work, having your own paid personal license, I always tell people even to bring it up at work meetings. I have custom GPTs now for CO parenting, for nutrition and fitness coach, by the way, that I'm obsessed with and I used to pay for that and all these things that give you an idea of how to use it. And even if you can't use it at work, it helps you build the business case of what could be done.

Speaker C: That's exactly what I wanted to explore next. Because for those who are in a situation where their company prohibits the use of that in house, and I'm sure people have approached you, Mike, with this question, where do you recommend they get started so that they could maybe simulate some of the work that they would like to be doing within their company, but do so in a, uh, safer, anonymous way outside of their company. So do you have any recommended, uh, starting points for that?

Speaker A: That's a really good question. I would say as an initial caveat, like obviously don't go do anything that's going to violate your company's AI policies or NDAs or privacy whatever, but I would say To Jessica's point, you can probably invest 20 to 40 bucks to spend one to two months with a paid version of ChatGPT in your personal life. From there, you don't have to be putting in super sensitive company data to simulate, like you said, Michelle, what you could be doing in your job. Uh, plenty of marketers that I know have side projects doing marketing for themselves for side hustles, for kids, school or whatever. Also, we all have hobbies. Right. I love Jessica. What you mentioned of your personal stuff, that's actually a really good place to start, I would argue, is you get a real quick sense of what's possible when you're passionate about something. So if you can't use this stuff at your work, and if you're having even a hard time simulating marketing use cases at home, I think you could probably still figure out on your own, use it for something personal. What is the thing you like to do? Use it as an advisor or as a planner or as an assistant for that thing. Whether it's running, cooking, parenting, you'll really quickly get a sense, knowing some kind of domain of what's possible. So I think you're in a tough spot if your company is not letting you use any of this technology. But I do think there are ways to quickly say, okay, I at least have some expertise in a sense of what's possible with the technology off the shelf out there.

Speaker B: Yeah. And for those listening right now, hey, Wet, I thought this episode is about practical AI tools. Don't worry, it's not a bait and switch. We're just warming you up to say, hey, don't just jump in and use a bunch of tools. Okay, so we're going to get to some tools. And I think Michelle kind of put us into now the teams and tools section of the show. And I think the way we talk about this is not, hey, use this tool from this company. I think what you're talking about right now is like, let's look at categories of tools and things you can tackle that may be the easiest ways to do the quick wins, because quick wins is always a really great way if you're launching something new. Get a quick win, get a use case, talk about what the impact is, and then like you said, get a, uh, subscription to ChatGPT, of course, which then you'd have to choose. Is it 3.5? Is it 4? Is it 4? Turbo? Is it 4.0? Omni? Is it 4:1? Is it 4:1? Mini? I don't know which one. So I think getting into that really gets us too much in the weeds. So I do like the way that you're approaching it. What kind of categories of tools do you feel like is gone from a marketing perspective is going to make that first big win? And if it is a frontier model, then we talk about a frontier model and a use case. Say you have Gemini, you have ChatGPT for Teams for the Office Enterprise License. What's that first one? If someone's listening they should jump into this type of tool and this type of use case. I think it's probably good to pair those two together.

Speaker A: I'll try to hopefully break this down. How I would approach it if kind of I knew nothing about what was going on right now in the landscape. Escape.

Speaker B: It's got to be hard for you to pretend, Mike.

Speaker A: Everything changes so fast. You'd be surprised. Some days I wake up and I'm like oh, everything I knew before is like useless.

Speaker D: It's all gone.

Speaker B: Yeah, it's like we're recording this at 10 o' clock so if anything happens after that, it's not our fault. Okay.

Speaker A: Yeah, exactly right. Exactly right.

Speaker D: Yes.

Speaker C: You and Paul always caveat that like every podcast episode exactly when you recorded

Speaker A: it have to timestamp it. It's so true. But I would say first there are what we might call non Reasoning models like GPT4O is a good example, the latest, smartest non reasoning model that you have access to in something like ChatGPT. I would say if you have not experimented with the latest non reasoning models, go try to use them for general assistance, general advice, possibly producing copy content ideas from there. I would then look at reasoning models so something like OpenAI's O3 Gemini 2.5 Pro, the newest models from anthropic Claude 4 Opus I think has some reasoning capabilities. These are models that go one step further and actually take time to think through the steps they're going through. So you can do the same things with those types of models. I would encourage you to experiment with any and all use cases you might have, but you can also go a step further and say, okay, I'd actually like to use this as an ah, in depth strategic assistant or use it with a ton of data to really build out a very comprehensive plan or approach or strategy to some type of marketing tasks. So I'd encourage you to kind of figure out where did the fast, cheap and pretty intelligent non reasoning models work out? Like where are those enough for your quick hit stuff you're doing every day as a marketer? Where can the reasoning models then add a lot more value in terms of these in depth really strategic use cases that I think we're all kind of just starting to experiment with because people don't always realize, oh my gosh, like I could dump a, uh, transcript from a two hour meeting into one of these things and get like a full on marketing plan in like two minutes from an O3 or a Gemini 2.5 Pro or whatever. And then from there I'd also say just experiment with the category of what we call deep research tools. So OpenAI Google perplexity, they have these tools that they call deep research or capabilities within the current models where it will agentically go autonomously research topics for you and every knowledge worker, every marketer absolutely needs to go try this out for themselves. I think there's at least one free option with Google or ChatGPT I think to do this on your own. Go see what it does and what it produces on a complex topic. It produces research briefs that are dozens of pages, sometimes tens of thousands of words long that are pretty well cited. You still have to check them as a human. But my gosh, you're gonna really have a pretty eye opening moment seeing what these things can do. We can talk about image generation, video generation, all that, but that's kind of roughly those big categories right now that are really impactful, I would say for marketers.

Speaker D: That's what I was going to comment on. So it sounds like starting with language first, lowest hanging fruit.

Speaker A: Yep.

Speaker D: Because then I think it's really easy to get into complexities or access issues once you get into image and video as well. But I feel like a lot of questions that we get around like what's the best tool for PowerPoint, what's the best tool for video, what's the best tool for imagery?

Speaker B: I just wanted to say, have you found anyone who's fixed spreadsheets? Is there a tool out there I won't have to do spreadsheets again?

Speaker A: Please.

Speaker D: Yeah, I've been doing a lot with Google Gemini in sheets because it's a Google shop and yidding you are too in terms of data analysis being right there next to it and going through math without having to do formulas. Like I know you can do the equals AI, but even just asking Gemini, I'm talking survey data or multiple columns be used, different metrics of time and things like that. Gemini's been really helpful. I still had to create the spreadsheet, but I kind of like doing that part, I guess. I don't know.

Speaker A: Right.

Speaker B: I Guess that's the part I was talking about.

Speaker D: It makes me feel organized. I feel it.

Speaker C: I'm sure an agent is coming your way, Yadin.

Speaker B: Yes.

Speaker A: Yeah, I'm at the stage where I think that we've gone very far down the road of great data analysis with AI. Tools like Gemini are awesome. But yeah, I want to get to the point of I don't want to see the spreadsheet ever again. I just want it to tell me what's in it or tell me what to do with it.

Speaker D: I love that breakdown by the way, of how to approach the different language models and use cases. So to recap, number one, a non reasoning model, number two, reasoning model and then number three, deep research before we move on, because I do want your take in video and audio or any other category I guess is do you have an example first Deep research project because it's still blank page syndrome if people don't know what to do within the tool. Like is something out of the box that comes to m mind. That's just an easy way to show capabilities.

Speaker A: So for deep research, the advice I would give is don't overthink your prompt. To start, just type in a sentence of like research something and that's something. Make it something that you know pretty well. I would say actually, because this is where these tools can go a little wrong. If you could get some amazing research on a topic you know nothing about and it looks great, but an expert looks at it and says, well okay, that's pretty interesting, but there's a, B and C things wrong with this. I would actually say go just as a test with something you know pretty well or that you're interested in because you'll quickly see how powerful it can be. And you can also really quickly start understanding, okay, here's where it went right, here's where it went wrong and kind of learn from that. Moving forward research X. Yeah, yeah.

Speaker B: And one of the things I found super, super helpful and I think this is something that you and Paul have suggested as well where you don't know what amazing prompt you should do. So you have a large learning model who can create a really amazing prompt for you. So don't get the prompting out. Nobody has to be a prompt engineer. You just have to say, hey, I want this really cool thing and can you give me a prompt so that I can use in Google's deep research tool to do this kind of stuff? And then boom, it will put out like a three page prompt with details and sub bullet points and you're like, oh, yes. And I wanted that and that and that. And you just go ahead and check that. Yeah, that looks good. And then you dump it into Google Deep Research. And that's what last I did. And that just created like a 50 page brief of something that was just utterly amazing. And the hardest part was really just cutting out the stuff that I didn't need.

Speaker D: Yes. I was going to say quick plug, uh, for Mike's prompt GPT, because I save it and actually use it a lot with that wiser framework.

Speaker C: I have used it too.

Speaker B: Yeah, yeah. Yes.

Speaker A: Yeah, it's super helpful. Like honestly. Oh, of course. Yeah. I mean it's been like a huge unlock for me because I actually, you know, as a writer by trade, I kind of took to prompting and really enjoyed actually doing it. But then I was like, this needs way more time and attention. Anytime I'm trying to do something with AI than I have to give. So we all default towards like a sentence or two prompt. And I think eventually we'll get to where the models are just smart enough that it's like, okay, you won't need to do all this stuff, but for the time being, even like I still find it really valuable like that. GPT will use a framework out there that's helpful to like assign it a role, come up with some tasks and subtasks or instructions for it to follow, and you just type in a quick sentence like, hey, I want to do this thing and it writes it for you. Copy and paste that in, you're good to go. You could ask any model you need to do this for you as well. And honestly, that's huge. That's probably step one I would do. For anyone listening. If you're struggling with this is just ask AI and use the prompts it gives you and see what the output looks like.

Speaker C: Yeah. And quick side note, Claude has a feature that I think I just set up in the settings and it will ask me clarifying questions. So I can start with just a really generic prompt. Then it'll ask me like five or six clarifying questions and I'll answer them and then it'll give me the intelligent response.

Speaker A: I love that.

Speaker D: And there's some tips that answer the clarifying questions and then go back and edit your original prompt so that it's in there from the beginning because it'll get a better output than the continual back and forth. But that came up on the AMA Mastery members AMA today with Paul too Michelle, about actually going through and answering the questions that the, uh, chatbot is giving you about whatever your strategy and whatever you're doing. And even Paul said he used to not answer all of them all or like rush through them. And now he actually enjoys the process of going through all the questions. And my light bulb moment was there is even all of the AI assisted enabled marketers. There's still levers to pull to differentiate yourself as an individual. Because not everyone is going to go through and take the time to do that.

Speaker C: Yeah.

Speaker D: And it's going to improve your outcome. I think just overall, like improve how you use the tool. But it'll be more strategic from the beginning. Taking the time to actually think through everything.

Speaker C: Exactly. I just love being prompted to do so.

Speaker D: Yes.

Speaker C: Psy. Keep me honest here.

Speaker B: That's fabulous. And so one of the things that uh, I think came up too is I think you guys are already pulling on that thread is like what does it really mean to be a marketer in the future with these tools? Whereas people used to say, well, I'm the HubSpot person or I'm the salesforce person or uh, know that I do the ABM tool. And it's almost like they people there were functions of people that needed to be like, hey, we need to put a body in front of this tool or in front of this workflow when now you can start to automate those things in ways that we couldn't before. And marketers aren't just people who are super experts in pivot tables or other things where you're seeing that evolution of what it means to be marketer and what kind of skills are really going to be. What sets you apart from other people.

Speaker A: This is like the trillion dollar question.

Speaker B: You guys are writing a book about this, right?

Speaker A: Yeah.

Speaker B: Right.

Speaker A: We're asking AI to write us a book about this. Probably I would say at this stage.

Speaker B: Oh, uh, there you go.

Speaker A: And it's written my first draft answer. Because it's a question I think about a lot and just in general, what does it mean to even do knowledge work in general? I don't have a perfect answer at all here, but I would say first draft thinking for marketing would be whatever it is you do, you're probably going to be. And I didn't make up this analogy, some multiple other people have written about this. You're going to be the conductor of some type of orchestra of AI tools, models, capabilities, agents, whatever terms we want to end up using there. So maybe the HubSpot person or the copywriter or the E Comm person, they might steal whatever that secret sauce is. That makes them really good at that job. I think parts of that will still really be relevant, but you're going to have to move into that conductor role with these tools to have any hope, I would say, of continuing the relevance of that role. And you could also probably argue. And Jessica, I think we talked about this a little bit yesterday when we recorded our B2B summit panel. When things are always changing, look to what doesn't change. Human psychology is kind of changing a little bit with AI tools, but it's always going to be super relevant. So the whole consumer psychology piece of marketing and building relationships with customers, I would imagine is going to be super relevant as well. But I think all marketers become conductors of the orchestra, not the individual contributors to it.

Speaker D: Yeah. And I think it's an interesting opportunity even to say, like, there are people who don't ever want to become poopoo managers, but maybe you get management experience still, because now you're managing the agents and then at some point you have to decide maybe that's a path and you still have to prioritize, communicate, manage up, even what's going on. But Mike, what we were talking about yesterday, I think too is the human traits of it all. I really started thinking about all of the things that we look for in high performers. It is a lot of times execution oriented, but executioners are executioners, I think, to some degree, no matter what type of role they're in. It's the willingness to get stuff done, to work through problems, to see it through. It's that tenacity, um, I think to problem solve, which is also the same tenacity, you need to work through that initial AI process that is then going to scale across the organization as well as everyone always talks about curiosity, but I think it's even questioning programs and strategies and the infrastructure that you have already set as an organization or as a team. What do you value in high performers outside of their work? And it's all of these other characteristics that I think make people leaders in their role, even if they're not in a hierarchical leader position.

Speaker B: Yeah. So ChatGPT says strategic thinking is number one and AI literacy is number two. And then following that is creative direction, data interpretation. Um, I don't think that's maybe not going to need that one.

Speaker D: And the Future of Work report had a lot of these stats in it too.

Speaker A: Yeah, yeah.

Speaker B: Uh, experimentation mindset. I think I was kind of one of the threads you were pulling on Jessica. Experimentation mindset.

Speaker D: Is there a different way to do this?

Speaker B: I'm here using Gemini and ChatGPT and a bunch of other stuff in the background constantly, just because I like to pull on threads and see what are the tools saying. And adaptability, though, is one of the key ones. I think that's was kind of to the beginning of our conversation. Why can't we get people to change? What they're doing is because people hate change. And, uh, that's constantly what we're running into with absolutely everything, hey, I have a new idea. Or we want to do a workflow or want to do this event differently, or we want to try something exciting and people are like, whoa, we've already done it this way for many years and don't change anything. But I think the listeners to the show are listening because they want to do things differently, because they want to be adaptable, they want to have AI literacy. And I think you've given them some good pointers on where to get started, because I think that's probably that biggest hurdle. Is there just some website? Is there something I should listen to? So people are coming to the show, I think to be part of that, of what should I be trying? What should I be doing next? From a tools perspective and a company perspective, it's tough because we keep running into some of the same themes. And like you said, Jessica, uh, you and Mike were just talking about this in another recording. Where do you see that inflection point? Or are you seeing that inflection point in any companies right now who are starting to become successful, who are becoming the least. Not AI first, but maybe AI Forward?

Speaker A: Yeah, it's hard to have a one size fits all answer because I've seen so many companies do this in so many different ways. But I would say that coming back to this point, one recent example, I can't name the company, but a large enterprise that we worked with, what really, really worked, even though they had all sorts of technology investment and education, was really at the ground level, taking in this case at least a small team and helping them understand what they could actually do with something as simple as GPTs. In ChatGPT, it was the simplest thing ever. I'm sure everyone listening to this, like, okay, great, I've heard of that, I've used that, uh, whatever. For them, it wasn't. And they didn't know a lot about AI. They were all domain experts in their own areas and they had previously had courses, workshops, they'd gone to events, they were given literally thousands of licenses to certain AI tools. It had kind of stuck in certain patches. But really what moved the needle in my mind was sitting down a small group and saying, look, here's one really narrowly focused thing you should focus on for the next. Call it 1 month, 2 months, 3 months, however long you need, here's how to actually make it work for you and then literally turn you loose and let me know how it goes. And with regular check ins, with regular support, not a huge amount of hand holding, but that kind of process, it was like night and day because a bunch of people that previously had been, uh, on the fence about a lot of AI stuff came back and were pretty over the moon about what they'd been able to do. And it's half the stuff they did is stuff I wouldn't even have thought of as someone who is helping them because I don't work in their role. Like I understand their role, but I'm not in the day to day. So that kind of process really at the very granular root level, showing individually, people, no here, focus on this one thing. Let's niche down whether it's GPTs, whether it's, hey, we're onboarding Jasper, whether it's, hey, we have a new tool, whatever it is, maybe get a little more focus on doing one thing really well and understanding it and then use those lessons. Because you can extrapolate so many lessons to other tools from that kind of initial pilot or initial project.

Speaker D: People have to know what to do within the tool. So there's a difference between the experimentation because you're learning through experimentation. When I log in in the morning and I log into my said AI tool, this is what I'm using it for. That is directly related to my day to day job. And so that's why we're always trying to find those really specific use cases. And that's how you're rolling it out to new teams.

Speaker B: Yeah, and I know we dig deep into it in one of the previous episodes, stay focused when adopting AI. And that was super key. And I like where you're going, Mike, because if people are listening to this, what tools do I need to use? You pick something that you want to tackle. You get people in a room or on a call and you walk them through it and say, this is the tool, this is the impact you can make. And like you said, Mike, it just can be like light bulbs going off everywhere because that's kind of how we ingest everything. We're forced a meeting on our calendar. I am now committed to this hour of my life on whatever's being talked about, even if sometimes we go to meetings, we're like, what? What did we just do for that hour? We don't know. Do something with AI and walk them through it and say, this is what you do. And then actually get them to do it. If they have a license, get them to do it and then like you said, have a check in. Because if there's no accountability, then people are gonna put it at the very bottom of the stack because there's 20 billion other things that they're trying to do. So this is competing with all the things that they normally do every day in the way that they've learned to do them and, um, done them for maybe years or even decades. And now you're asking them to step outside of that and do m more work on top of what they're doing. But the more work will end up yielding a massive difference in time saving or doing things that just simply were never possible before.

Speaker D: And even before you pick a tool, pick a problem, the side. I just want to overemphasize here.

Speaker A: I was even going to say, I mean, that's exactly kind of what we try to get at is I get doing talks, doing classes, whatever. I get so many people that will come up to me and very well meaning and very interested, which is great. And I'll be like, oh, these three different models just came out. Which one's better? And it's like, okay, it's a valuable question, but. But which one's better for, let's say, copywriting? My next question would be, what have you done with copywriting with AI that has failed, that you're asking this question? If you haven't done the use case with a tool and you're asking about different tools, you're probably asking the wrong question or asking the question out of order. Because I want you to come to me and say, here's my use case. I'm really trying to write killer landing page copy. So many people get bogged down in that analysis paralysis before even trying the tool in the first place. Like, go fail first. And I know that's not a clear cut answer. It's like, takes time. But I want you to come ask me about tools when you have a specific use case exactly like you mentioned, Jessica, that you are trying to solve for that you've already experimented and tried a bit with another tool.

Speaker B: So I think we're floating into that. The impact section of the show. Michelle, I think you actually had a question to kick it off.

Speaker C: Yeah, I love that we're all AI optimists here. More or less. And this conversation has been really focused on what's possible.

Speaker D: Are you, Mike?

Speaker A: Yeah, well, it depends on the day, honestly.

Speaker C: So we've been talking a lot about what's possible and great starting points and all that, but the flip side of the coin really is the need to introduce some cautionary thinking and critical, uh, thinking as you approach what's out there and the available tools to marketers. So for marketing leaders who are listening, what capabilities exist today that you think that they are just not prepared for and should maybe prepare themselves for?

Speaker A: That's a really good question. So I guess I would kind of break this up into a couple of them. I would say they're probably not fully appreciative of how good reasoning models are and what that means for their workforce, their teams and their plans moving forward. Because I think we're going to see in a very positive way some very AI forward, AI savvy marketers come into the workforce that are able to do things with these models in like five minutes. That would have taken you 20 hours to do if you're in the traditional marketing mindset. So if you're any type of marketing leader and you have not really pushed the limits on something like an, uh, O3 Gemini 2.5 Pro, whatever reasoning models are out there, I would say that's really, really important to do. And I think related to that, the deep research tools. Because while it's still really early, there are so many things that those tools can start doing for you that you frankly should not be spending time on ever again. I would argue. Yeah, so I would say those two, it's like, you know, these tools exist. As a marketing leader, you could say, okay, of course, I've heard of reasoning models or deep research. Go really kick the tires. Do it for stuff, you know, well, go give oh3 the hardest marketing problem you've solved in your career and see what it does.

Speaker D: Multitouch attribution.

Speaker B: Yes.

Speaker A: Right, right, right.

Speaker C: Well, and the irony is that the conversations, the narrative is still that AI will automate those repetitive, boring tasks and free people up to do more strategic work. What is that strategic work? We've been talking about these reasoning models who can be strategic partners. So when do you see that narrative shifting? I think those who are really in this space are already thinking differently, but I feel like the general population, the conversation is still around automating repetitive tasks.

Speaker D: But I think the deep research is the strategic work sometimes. Right?

Speaker A: Yeah. I think in our circles, Michelle, kind of to your point, I think the narrative is starting to get to the point where people are starting to be like, oh, uh, wait a second, what's going to happen next? I don't think that's permeated broadly to the general public yet, but it could happen very quickly and all at once. I wouldn't be remotely surprised if the next few years are pretty rocky in a uh, public discourse perspective with some of this stuff. I mean we're already kind of starting to see it at the edges as something takes off like that goes viral or becomes like a hot button issue in the general discourse with AI and people freak out. And I think we're going to start seeing that actually quite a bit more among people that aren't following this stuff regularly and don't realize things are capable. Yeah, VO3 by Google is a great example. They dropped their new VO3 video model which is jaw dropping. It's incredible. I look at that and I say wow, we're getting very close to a future that I knew was coming. Your average person that doesn't follow this stuff looks at a hyper realistic video is like wait a second, two years ago I saw some crappy video that AI video made. Wait a second, you're telling me it can do this? And then they're like, but what about misinformation online? What about knowing what's real? And then suddenly it's like okay, good luck. You're gonna have many months of sleepless nights like we all have. So you know.

Speaker D: Mm, mhm. Yeah, I think you mentioned like the bee buzzing video. Was that yours Mike?

Speaker A: Yeah, I think Paul had posted about that, which is an incredible example.

Speaker D: Yeah, I was gonna say a video that's we can talk now, what are we gonna say now? But I think the other one is even more impactful and I think there lot of managing up that still needs to happen here. So my plea for marketers listening to you who may not be at the top of the organization is there's still a lot of opportunity to manage up, uh, showing the reasoning model, showing the deep research, getting your marketing leaders to get beyond travel planning, which I'm still hearing from some leaders m Travel. Yes, it's do the work for them to show them what's possible but also make it meaningful. So maybe there's a project that you're paying a consultant to do or a question that came up that someone's like yeah, it would be amazing if we knew that and move on on the side, like do it for them and then show them what that looks like and then walk them through how you did it. Because I think that's also a huge eye opening thing for leaders is they don't want to ask or say that they don't know how to do it. But getting them in a small group where it's not in front of the whole organization to show them step by step by step and they'll say like, wait, can you go back three steps? Because they're in a safe environment. But now you're the one showing them and it gives you a lot of leadership capital in that sense too.

Speaker A: Also, we all know this. We're marketers. You have to do a lot of internal marketing.

Speaker D: Yes.

Speaker A: So if you're a marketer, don't describe to me in 80 slides the history of AI. Go show them two minutes of what this can do. Yeah, yeah, Go video a deep research report running on the last strategic thing you guys were researching. Show the executive and then trust me, you're going to get interesting questions about like wait a second, you can do this? How do you do this? How do we do this?

Speaker D: Uh huh.

Speaker B: Yeah. Knock the head back. Yeah, Like I just did like to my CMO, I just dumped all of our customer stories into NotebookLM and say, look at, you can just say which customers are doing this product from this vertical in this Geo and then create a spreadsheet for all the customer quotes. And she was like, make a demo, share it with the entire organization. And that's the knock your head. And then I was talking to Jessica about it. She's like, you should share that on LinkedIn.

Speaker D: So it's on LinkedIn now because I didn't think about doing that. Every use case shared inspires another use case. One of my favorite quotes ever. We all need to share more.

Speaker B: We do. All right, so I should get to the lightning round part of the show. We've just got a few more minutes left. And this is where each person on the show shares one thing that's happened recently about Genai that is meaningful, insightful. It doesn't have to be something that's news. It could be just something general. And wanted to just share a quick roundtable of lightning rounds. I'm going to do one that's about Google Veo because while you were talking about it, I was watching the demos and I'm like, it is insane multitasking. It's crazy. Google Veo so V E O for those who've not experimented with the Google image. And I'm just like blown away just with some of the basic stuff and I'm sure it's super basic and it's like, wow. So this is something I think no leader's prepared for. How do we wrap our heads around AI generated hyper realistic video and our use in marketing? I'm still wrapping my head like, well, is it copyright issues or.

Speaker A: We have.

Speaker B: Anyways, all that's my lightning round thing. So I'm just looking at it right now.

Speaker D: It's your homework lesson for everyone.

Speaker B: Yes. Who's ready with the lightning round one too? I don't want to put Mike on the spot next. Unless you're ready, Mike.

Speaker A: I could go next, I think. I actually have been thinking about similar related things quite recently, both about what's real and how we tell what's real online due to things like VEO, but also just overall AI's impact. Like Michelle was just saying and um, kind of alluding to like, I think we're hitting. We're getting close to a tipping point about the general public being aware, at least more aware. And some really vivid examples of what's possible. And I don't think we're ready for the ripple effects of what that will be. I don't know if that means anything like legislation, public outcry. We're going to probably see some interesting conversation around jobs and work. I just think we're about to. If you thought AI was already buzzworthy, I think it hasn't even broken containment yet, really.

Speaker B: For those with kids, you're already seeing next level cyberbullying and that's the first hot button topic where parents and um, people are really getting confronted with the AI.

Speaker C: Oh, we could do a whole episode maybe about that. Yeah. Like an AI for Parents episode.

Speaker B: Oh my God. Yeah. AI for Parents. AI for marketing. Parents.

Speaker C: We're all parents here. Yeah.

Speaker D: There's a new series for you, Paul. Mike. There's the Parent GPT or Kidsafe. Kidsafe GPT.

Speaker B: Give GPT. I love it.

Speaker A: Yeah. Kids SafeGPT. It's. Yeah.

Speaker D: Which you can also get to write your summer screen policies, it sounded like from the podcast this week. So Paul was chatting with it to help summer rules and kicking off summer here. So definitely top of mind.

Speaker B: It's fabulous.

Speaker D: I talk a lot about communication, culture change, leadership, and I want to share something that's been really cool. At Jasper, we have a women's erg, which I'm sure all enterprises have, and there's different flavors of these, but we've been really incorporating AI in this and not me. The leader has been doing different things. So every Friday there's like a Fun Friday prompt that they give out that everyone goes and uses their tool of choice to put in. So for example, today the prompt was around dropping in your first name and current mood in a prompt that gives you your women in Tech weekend horoscope. And it's like expect sass wisdom, tech jokes, a nudge. And it's just I don't care if anyone else reads mine. That was amazing and super fun. Like about my weekend plans and lift you up on the week. There's different stuff, like write a slack message that sounds helpful but says absolutely nothing one day. It was like a bunch of things in a session. What oddly specific hill would I die on at work? Which is where like the memory comes in, which is really fun based on what you know about me, what would my beige flags be? And I'm sure you can prompt for all of these ideas, but it's a cool way to get to know each other as a team, to push people towards the tools and to also have fun. Because I talk about getting maniacally focused on your KPIs goals a lot of. But you gotta do the fun stuff too to get people engaged and really getting into the tools.

Speaker B: I like Fun Friday prompt.

Speaker C: Thanks. So I would love to end on a humanistic note, since we've been talking about tools and there's a lot of concern floating around out there about people getting too attached to their tools, developing relationships with their chatbots, yada yada. Again, that could be a whole podcast

Speaker B: series, a whole thing, whole series.

Speaker C: But I read something just this morning. It popped up in my Quora feed and there was this whole thread around Genai being the best therapist ever. And I was really curious about people's comments and just wanted to quote from a couple of the comments. And one person wrote, I don't want to replace people, but people are so difficult lately. And then the response, yeah, we can all relate to that. But then the response that I just loved is it's in the difficult that we grow and learn. True, it's messy with humans, and my chatgpt gives me perfect answers, but our humanity is in the mistakes and the getting up and trying again.

Speaker B: Yeah, I like that like that because it's the messiness that creates the challenges. Yeah, I've talked to LLMs. There's no challenging. They're not like, oh, let me challenge you, let me push back on you. Which is one of the things I think from a strategic standpoint really is missing. And I have to put that in the prompts. Like, no, be critical Be harsh even. And then it was like, well, you might want to. I'm like, no, really, Come on, push back. That's what AI needs to get tougher, Mike, Relationships.

Speaker A: I think maybe that's one of those human skills that we should be thinking about, is the human thing moving forward that we should all cultivate as being a pain in the butt.

Speaker B: There you go.

Speaker A: When we're talking to people being counterfactual,

Speaker D: the anti sycophant relationships are built in shared suffering or shared moments. Right. Like we talk about our marketing, AI Council from our previous company is so strong because of what we went through together. And so if you think about it in that way, too, that relationship, you're not going to have that struggle with AI. This is why communities are so big right now, is because we're all trying to figure this out together and learning together, and that's what's creating bonds and relationships that will stand the test of time.

Speaker C: Yeah, Embrace the messiness.

Speaker B: There you go. Embrace the messiness. Instead of trying to be more of a pain in the butt. I don't know.

Speaker D: There's something there.

Speaker B: We're at the end of the show, Mike. It's been fabulous. We could have talked for hours more, but I know we all have lives to go and live. Outside of talking about AI, do we?

Speaker C: Speak for yourself, Yideen.

Speaker A: Yeah, right.

Speaker B: All right, well, thank you so much, Jessica. Um, Michelle, you thank. Thank you, Mike, for joining the AI Edge podcast.

Speaker A: Of course. Thanks for having me.