Rethinking AI in Marketing: Some Practical Uses Beyond the HypeRethinking AI in Marketing: Some Practical Uses Beyond the Hype

Rethinking AI in Marketing: Some Practical Uses Beyond the Hype Transcript:

Jeff White: Welcome to The Kula Ring, a podcast for manufacturing marketers brought to you by Kula Partners. My name is Jeff White, and joining me today is Carman Pirie. Carman, how are you doing, sir? 

Carman Pirie: I’m not doing too bad. I’m joining you just a few days before a bit of summer holiday, so it’s pretty hard to be in a grumpy mood. I can be, anybody that knows me, knows that I can do grumpy pretty well, but… 

Jeff White: But at least your flight isn’t on Air Canada. 

Carman Pirie: Yeah. 

Jeff White: Yeah. So you’re good. Oh, okay. This is gonna be another one of those episodes, similar to one we did a little bit ago, where we’re just interviewing each other.

And in this case, I’ll be interviewing you to find out. Some of the things that you’ve been doing around artificial intelligence and leveraging AI, and not in the ways that I think a lot of people think agencies use AI. 

Carman Pirie: Yeah. I don’t even know if I’m gonna say it’s an agency. I think just maybe thoughts on how people in the world of marketing and sales can use the technology. Yeah. I’m very I’m I’m very aware that we’re recording this conversation just like a day or two after the, like the New York Mag article where they’re talking about Altman and Schmidt kind of cooling the jets on the AI hype.

Maybe it’s not all. It’s cracked up to be shifting the attention away from a GI. And stock markets are reacting even in the last 24, 48 hours negatively to that. So it, I guess it’s not, it’s interesting to be having the conversation and being very aware that it’s a changing thing.

I guess all that to say I’m probably going to disagree with myself just moments from now. Especially once this thing is live. Who knows if I’ll agree with any of it, but. I invite listeners to listen along and see if there’s something that they can pull from it that is maybe useful in how you use the tools.

Jeff White: I, I think, there, there’s no question that there’re going to be things that are useful. I think, some of that Altman hype, the Anthropic CEO is, he’s got about a month left till he had proclaimed that all software would actually be written by AI and no longer by.

Developers and engineers and humans, and we’re nowhere near that being the case quite yet. So I don’t think the bottom has fallen out of a lot of the marketing-related requirements for humans. But humans can certainly make good use of these tools. Yeah, absolutely. Yeah, absolutely. A lot of the way that, you know, and I think your use of these tools goes above and beyond my own.

I think you’ve really delved more deeply into this than I feel like I have, and I’ve learned a lot from watching how you are utilizing them. And a big part of that is, you’ve talked about AI and, related tools as your research intern. A bit, and I’d love to dive a bit deeper into what you mean by that and how you’re using it and explore maybe some of the things that you’ve learned and found along the way.

Carman Pirie: Sure. I think, look, using AI for research is not certainly not new. I think people that’s probably the first use that people use the tool for, is whether they’re researching something personal or for work. I, so when I say this, I guess what I mean is that, how to think about your use of the tool, how to how to think about your prompts.

And I find that it is helpful to. To think about it as like the world’s smartest research intern, or at least the world’s most keen in that, it will do precisely what you ask it to do. And you don’t have to tell it certain things twice. And. So I think it’s just a helpful lens to think through.

If you can imagine like the most keen junior research as assistant intern you’ve ever worked with or whatever times 10. If you think about your prompts that way I, you get some pretty interesting outcomes. And, but the out, but I guess the one little twist to that I might add, Jeff, is that, one thing I think people do when they’re mentoring junior team members or what have you, is that they sometimes, soften the language a little bit just to make it understandable. Like they, they take jargon down a little bit or they don’t maybe explore incredibly advanced concepts because, a, an intern or a junior isn’t there yet.

And that’s the one area where I think if you’re using that research intern kind of lens, they maybe need to flip it on it head a little bit, which is to say, I think in their prompts, you’re a lot better off if you’re. If you get incre, you do get incredibly detailed. If you do in some ways communicate to the ai what, how smart you are and what you know about the topic because it helps frame up where it goes from there.

Does that make sense? 

Jeff White: I think it does, and I think where, you can really explore the concept of what you’re trying to impart. This, but not this, and based on these other things that you already know about the ai, the assistant can build, come back with a research report or information that is more tightly integrated with what maybe you’re hoping to see.

Carman Pirie: Yeah. Yeah. And I maybe a really interesting example would be like I guess I think every marketer, my assumption is that every marketer, if they’re tackling a problem particular challenge, whatever, they can imagine information that would be good to know. But the juice just isn’t worth the squeeze to get it right.

There are things you could be sitting there thinking, it would really nice if we do this, and this. And if we could explore that and do that, how that intersected with this over here, that might help shape our decision a bit. That’s a five week project then $80,000 and we’re not going to do that.

There’s.

AI can shortcut some of those things now. So as an example, we were into this conversation with a a marketer some time ago who was marketing for an organization that’s it’s basically a PE-backed industrially roll up without getting too specific. And and they were really struggling with some kind of odd brand challenges, and they just hadn’t really thought through the impact of brand strategy on exit valuations and how that, as a. PE backed entity that was going to be doing a bunch of m and a over the next three to five years before that next flip. How does a brand strategy attach to that?

So you’re having that conversation. You’re thinking geez, it would be really nice to. To know within related or tangential categories, how house of brand strategies and branded house strategies implemented in a roll up. How those two brand valuations have shifted. Like when eventual exits have happened, is there any data out there to let us know that indeed a branded house is more successful than a house of brands or vice versa?

Jeff White: Before you go too much further, and I know we don’t always do this, but can you explain the difference between House of Brands and branded House just 

Carman Pirie: I, I could, but actually I would say if you’re listening to this and you don’t know, then go to AI and search a difference between House of Brands and branded house because then I will prove the use of the tool, but in, in that moment you’re sitting there thinking yeah, that, you could do research, you could find out in related categories who have conducted roll-ups, what their brand strategy was what they, what multiple they exited for, and do some comparisons. You can imagine that research in your mind if you’re a somewhat of a marketer.

Good luck executing it. But man, now, like you’re one prompt away from getting a bit of a taste, a bit of a snapshot, some signal to noise. And I think that’s just it’s just a really powerful use of the tool for research. If you think about I guess again, letting it know a little bit about what you already know, you’ll get a little bit more out.

Yeah. I didn’t mean to side step your branded house versus house of brands question. It’s just, man, that’s a, we’re gonna go down 10 minutes. 

Jeff White: I, and I wanted examples, all of this, so I know where you were going. That’s why I stalked that.

I think part of what makes. That’s so powerful too, just in terms of the, yes, you could do this, you could give an intern a, an eight week thing or a huge budget to go and try and find, and how many rabbit holes are you going to go down that don’t yield fruit? And, that’s mixing metaphors a bit, but the rabbit holes that don’t yield fruit, but part of what makes it so powerful.

And able to come back with results that you simply wouldn’t be able to find easily on your own. Is it’s rather dispassionate in terms of how it takes in what you give it, and do you think that makes. I don’t know, because, any, anytime, even if you’re speaking with a very intelligent person who is a, a category expert or whatever, they’re, they’re obviously going to have a a propensity to believe certain things.

And, they’re less dispassionate. So you may not be getting a more objective survey of whatever. Query you’re looking into. 

Carman Pirie: Yeah. A bit of a different type of research, and maybe even we talk about even if you are as a marketer, maybe you’re doing customer interviews or maybe you are having conversations with the sales team to better understand what they’re dealing with.

Even if you’re out on a ride along or something like that, you could imagine recording the conversations that you’re having. We can there’s. Pretty ubiquitous ways now to ensure that con, especially if it’s video chat or something like that, it’s incredibly easy to get transcripts of that kind of conversation.

And that’s where I think the dispassionate side of it really comes into your point. Like I think it’s ’cause it’s a better, in some ways AI is a better listener or a less biased listener in some respects. ‘Cause I think a marketer, it’s easy to begin to, hour long conversation with the sales person about the buying journey that they that they take people through and how it works.

You’re trying to understand it and so you can help them improve it, but you’ll come to conclusions on their own pretty quickly. You may only be into that conversation for 10 minutes and you think you’ve already gotten the nugget. Whereas if you were to upload the entire transcript of that conversation to AI and ask it to give the top five marketing takeaways or what have you, I think you’d probably be pretty surprised at some of the other elements beyond the one that you’ve identified that it surfaces.

’cause it’s just, again, it’s a little bit more dispassionate. I think another cool use Jeff, is I’ve known people that have used this for sales coaching. Oh yeah. Where they’ve taken basically, many sales organizations have a model or a bit of a coach or whatever that they subscribe to or a co, a sales process that they try to adhere to.

So you take some of that curriculum. Upload it to AI and then basically as you’re having sales conversations that are recorded and transcribed, you can upload those transcripts. And ask the AI to be a sales coach and to analyze your sales calls or what have you and how well they align with the prescribed curriculum, et cetera, that the organization has been rolling out.

Is it perfect? No, but. It’s shockingly good and incredibly inexpensive. Most people aren’t making the decision between okay coaching and amazing coaching. They’re making the decision between coaching and no coaching. So this at least gets you to, you, you can get there pretty cost effectively now.

Jeff White: Yeah. And I, just to go back to your point about taking a a discovery transcript or something like that and trying to find out what are the, tell me what you tell, get the AI to tell you what it thinks the key takeaways are from that to make sure that you’re not locking in that thing that you just heard.

That aligns with your sensibilities in the matter. But the other part too is that you can build on these, not, from the sales coaching side, from the, manufacturer understanding side. If you’re doing multiple discovery interviews with multiple salespeople, trying to parse all of that information yourself while also being a good host and extracting as much as you can, means that you’re not necessarily able to objectively look at all the different bits and pieces, but you can upload, a dozen transcripts and and whittle those down.

To the key points to make sure that you fully understand it. Now, this is not to say like that this should replace the hard work of an expert, in terms of harvesting the useful information, but it can help guide that and shorten it and ensure that you maybe find things that you didn’t even think of.

Carman Pirie: Yeah, I was going back to that intern comment from earlier. You wouldn’t give an intern a project and then never look at the output again. Just put it into the wild as though it was gospel. You wouldn’t do it. You’re gonna check it and help it, refine it and make it better. So yeah, you should be thinking about doing the same here.

Jeff White: Yeah. For sure the subject of that intern or, may just more junior marketers in general are often reporting up to people, they’re, they have to convince their superiors who aren’t necessarily marketers or salespeople or, any of the roles that, that we typically help out with, but they have to convince those people to either open their pocketbooks or make a decision on something.

How can AI help those people understand how to communicate things? Better to their superiors. 

Carman Pirie: Yeah, to your point, it’s not just the juniors, it could be a VP of marketing and you need to you’re pitching a campaign strategy and significant investment to the CFO and the CEO and a few board members or something.

Jeff White: Yeah. There’s always some higher up the ladder than you. 

Carman Pirie: Yeah, exactly. And I think the, in some ways, it can be very. One thing I’ve noticed, and a lot of marketers I guess, that I’ve met over the years is they it’s almost like the CFOs are from Mars, marketers are from Venus to, did you write 

Jeff White: a blog post called that in like 2006?

Carman Pirie: I don’t think so. But are you 

Jeff White: sure? 

Carman Pirie: If I did, then it’s really interesting that it came back and I don’t remember. But you can. It can be really hard to put yourself in that position. Using AI to say, through the lens of a hard nosed CFO who has repeatedly said no to requests for marketing investments over the last five years and tends to be incredibly skeptical about the value of marketing and industrial categories.

How would you improve on this strategy to make it most palatable to that kind? You can imagine what kind of prompt you could begin to get out of it to be, make to just improve the quality of the output and how you’re thinking about addressing objections, particularly from people who are going to be looking at whatever it is you’re proposing through a different lens than you are bringing.

Jeff White: Again, this is about, there’s a certain amount of. Introspection required here to even be willing to, it’s weird because you’re only exposing your, yourself to the computer. It’s, it’s not even like you’re dealing with what other people might say if you ask them how could I make this better so that Jimmy and the C-Suite are more likely to buy.

Most people aren’t really willing to take that kind of coaching; you have to be thinking about it in a different way. 

Carman Pirie: Yeah, but look, I encourage folks to just even so much as just this is this is the task that I was doing. This is the output that I’ve created. Ask me 50 hard questions about this to help me make it better.

And then buckle up. It’s an incredibly effective way of using AI to just make what it is you’re doing better. It’s not about replacing humans, it’s about making the human output just frankly more thorough 

Jeff White: and considered. Yeah. This is not about using it to craft hundreds of blog posts.

No. But to that end. In providing your output, your work, your strategy, any, anything, any piece of content, AI is really good at also finding out what you don’t know, but it’s potentially just as good at finding out what it doesn’t know and looking for gaps in that way. 

Carman Pirie: Yeah, it was I’ve seen a couple of really cool uses of AI for that where people are basically prompting to try to understand the areas of like within a certain, area of academic research or thought leadership or what have you. Rather than thinking about it as a like research assistant, go out and find me what’s out there. You’re asking, find me what isn’t out there? What are the key topics that tend to be undercovered. By, you can imagine how you can begin to use the tool and prompt the tool to find gaps if you’re building out a content strategy or a thought leadership pr program for your CEO.

Maybe you’re it is a CEO of a packaging manufacturer or something of the sort. So you’re like, okay, one of the key themes for the organization is sustainability. You’re gonna find hard to stand out on that topic, right? But. I would bet if you leveraged AI in that example, we could begin to find where are the areas of sustainability intersecting with the packaging industry that tend to be not talked about, tend to be under, less understood, less widely discussed and even begin to, and find ways that some of those less discussing intersect with emerging regulations or other kind of impacts that are gonna make them more important in the future. It’s just a different way of thinking about using the tool rather than saying, what’s out there, tell me what isn’t. 

Jeff White: Yeah.

And this can be used a couple of different ways. You can do that and then publish content about those things with your own thoughts and research and become the answer for that and start to show up, in, in other ways there, you can use that as a strategic advantage for your organization.

Or I have a good friend of mine who is also an agency owner in Pennsylvania and used this. Similar tactic to write a book about things that nobody else had really talked about in the higher education space. So he was actually prompting it and, when it wasn’t able to come back with answers, that’s what he wrote about and that was how he structured his book.

A, it’s a pretty interesting way of thinking about things. 

Carman Pirie: Yeah. I suppose it does beg the question. It’s is there no information out there? Because nobody wants to know about this. 

Jeff White: Yeah.

Maybe nobody cares. 

Carman Pirie: Yeah. And look I do think it’s a, an interesting use of the tool and I think just given the act, it is a bit of a, I like the kind of contrarian approach. Do you like the contrarian thing? Because if you can start thinking, if you start thinking about that you’ll also think another different ways about the use of the technology, if that makes sense.

Jeff White: No I, that makes a lot of sense. As we. Kind of wind down this episode, and this has been really fascinating to explore these areas with you. Some of these concepts are, probably not the first things people think of when they think of in interacting with A GPT but there are some other kind of more tactical ways of using this so that marketers can begin to learn how to interact with the, with an AI.

Tool, especially seen some great uses around building persona bots and things like that. How does that, how do you think that should come to life? 

Carman Pirie: That’s been one of the early earliest uses I think that we had as an agency for leveraging AI, was to try to get into the head of buyer personas a little bit more, and round out the profile of a buyer persona in AI, and then use that to query for areas that might feel content strategies or what have you.

It was I do think that you can. It’s a nice tactical way to, to use the tool to get inside the head of who you sell to a little bit more and understand the information that they’re ingesting in a deeper way, potentially. Look, even everything that we’ve talked about on this show, Jeff, we’re not at the absolute bleeding edges of AI advancement here.

We’re not vibe coding anything that’s true. We’re not setting up agents that are building other agents, et cetera. But so I, but I do think that. It’s it’s surprising to me how little use the technology is even in the areas that we’ve discussed. Yeah. Like the, there’s, I talk to marketing organizations every day that’s, haven’t really got behind, beyond using it just as a bit of a glorified Google.

Jeff White: Yeah. 

Carman Pirie: And so yeah, I think there’s lots of advantages yet to be harvested and and folks ought to dive in. Headfirst. 

Jeff White: And I think once you do and you begin to see the power and it begins to shape your thinking it’s it’s pretty compelling. Yeah. 

Carman Pirie: It could, I think it just helps people understand there’s not something to be scared of, which is absolutely a weird thing to say, but it’s surprising to me how prevalent that is.

Even some research that we have coming out in a month or so here at Cruella Partners looks into the use of AI by industrial buyers. And examines that a little bit by organizational size and kind of interesting to see the small and medium size.

Industrials, maybe leading the way in AI experimentation usage. I think that there’s, there’s something to be said about a bit of a lack of resources fueling the use of a tool that can help be a bit more efficient and make, take some shortcuts. Anyway, I think we live in interesting times, and I hope people get some use out of this episode.

Jeff White: Yeah. I certainly have enjoyed it. It’s been a great conversation. Thanks so much. 

Carman Pirie: Thank you.