AI Is the Junior Collaborator. You’re Still the Leader.AI Is the Junior Collaborator. You’re Still the Leader.

AI Is the Junior Collaborator. You’re Still the Leader. 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 you doing, sir?

Carman Pirie: All is well. How are you making out?

Jeff White: Thanks. We’ll be talking about one of our favorite subjects today.

Carman Pirie: We always joke — today’s marketing and sales you almost can’t say the word marketing or sales or a lot of other things about business, without saying AI at the same time. It’s easy to develop a bit of an eye-rolly response to it, ’cause you’re like, “Oh my goodness, do I need to be talking about AI every day?”

But I guess what, folks? You kinda probably do. It’s moving that fast. If you’re not paying attention to it every day, then you’re not paying attention to it at all. And so we’re gonna add to that today. I think it’s it’s important for folks to lean into hearing more information about it versus less especially when we’re talking about the practical application of it in industrial marketing environments.

That’s that’s what a lot of folks need to figure out.

Jeff White: And I think the more that you know about what’s possible, what it’s capable of, how it can be used, and maybe where it shouldn’t be currently, I think the better off you’ll be at implementing it wisely.

Carman Pirie: Yeah. Yeah. And in so many cases, the more you know, the more you realize you don’t know much. I think is what a lot of people encounter, especially as they first start diving in. Yeah, important to be humble, I think, in that quest for knowledge around this as well.

Jeff White: So joining us today is Will Pierce. Will is the Vice President of Sales and Marketing at Flexaseal Engineered Seals and Systems. Welcome to The Kula Ring, Will.

Will Pierce: Thanks guys. I’m really glad to be here.

Carman Pirie: It’s awesome to have you on the show and thank you for joining us. It’s a fun topic to to kick off here. Maybe as we get started though introduce our listeners, if you would, Will, to yourself. Give us a bit of background about you and how you ended up at Flexaseal.

Will Pierce: So my background, I’m, I one of those unique individuals that got into sales starting down a technical path. So my background is I’m a chemical engineer by training and started off in a traditional engineering path and then got into the rotating equipment industry and the manufacturing side of it, about a little over 15 years ago, and started working at a pump company and got into application engineering and really leaned into the sales side of things and I had a passion for marketing and it was a small enough company to where if you raised your hand to do something, they said, “Yeah, that’s you. That’s your job now.” I basically got to take on whatever I wanted to and had a lot of fun growing and building that side of my career and then became a customer of Flexaseal’s actually at this pump company. And then when I left I did my own thing for a little while and Flexaseal asked me to join them in 2023, and so I’ve been here since then.

Jeff White: Very cool. I always found a lot of folks, manufacturing marketers and others within larger organizations like, “Yeah, no, that, that’s actually your job. You need to figure that out.” There’s just so much you have to learn and take on, and people just expect you to be a bit of an omnipresent person who can do it.

Carman Pirie: Or they’re certainly willing to recognize when you raise your hand to take something on that ha- isn’t being done. It’s like, “Hey, you can do it.”

Will Pierce: Oh, yeah. Nobody, no, no smart organization ever has a go-getting individual and says no,

Jeff White: Don’t worry about it.

Will Pierce: Yeah. You gonna do that thing that nobody else is doing? Nah, we’ll just leave that there.

Carman Pirie: A bit tangential I suppose, but it sounds as though you were more drawn on the sales side versus the marketing side initially. That’s what kind of pulled you not away from engineering, but pulled you to use your engineering background in a different way. What was it about the sales side that interested you or that got you heading down that, that side of things?

Just wanted more money from making commissions or what?

Will Pierce: That’s always a great motivator. I think that’s a strong motivator for most. But I definitely I’m a competitive individual and the thrill of the chase, the wanting to win, and from the engineering side, being able to see the problems that customers are encountering, almost no problem that any customer has is unique, right?

Very rarely. If you have enough experience with it, you see that problem and you’re like, “I know how to solve this problem, and I know it from a technical standpoint.” And the selling side of it just came naturally. “I’m confident in this solution and I can help you, and you can accept my help or not. Accepting it is giving me money,” and so that was fun. I’ve always been a creative or thought of myself as a creative individual and I would, I led with sales first and then was like I need these other things to help enable these sales or to pave the way for sales.”

Going and meeting somebody and they’ve never heard of you before or your company that’s a failure of marketing. And I learned that kind of later down the road and I’ve actually leaned more into marketing later in my career as, ’cause I’ve found that to be more sales enablement than anything.

Carman Pirie: Yeah. Yeah, fascinating. Yeah. All right I could – I, I feel like I could do the whole podcast episode just on, on that nuance, frankly and the kind of the draw towards sales and marketing. So I’m gonna try to resist that.

Jeff White: I would like to know more about Flexaseal though, Will.

Will Pierce: Yeah – happy to tell you. So we are a manufacturer of mechanical seals typically go in pumps, essentially any sealed rotating equipment. So that could be pumps, vessels, mixers, agitators, you name it. If it’s rotating and it needs to be sealed in some way a mechanical seal can go in there and do that.

And we also make the support systems that go and support those mechanical seals. You’ve got two very flat faces rubbing against each other, creating a bunch of heat from friction. You gotta get rid of that heat somehow, and there’s a lot of support systems around the seal that do that.

And so we cover both sides of that equation.

Carman Pirie: Well, and so that fits the bill nicely for us talking about deeply technical categories and how you begin to think about integrating AI into the sales and marketing apparatus of a deeply technical firm. So let’s jump into it. What is the… can you give us a bit of the overview of kind of the current state of affairs of AI integration, what you’re currently navigating and and where you’ve come from in that?

Will Pierce: Yeah, I, that’s a great question. So I’m an early adopter of most things, and so as soon as AI hit the scene generative AI hit the scene, with OpenAI and ChatGPT I started using it immediately, just trying to figure out what it was. And I always looked at it as a tool, and not as a replacement for anything, but a lever for the different things that we’re already doing.

Because I don’t know what your experience is, but my experience is that you should already be an expert or well-versed in whatever you’re doing that you’re integrating AI in, because otherwise it’ll lead you down a deep dark path that you have no idea you’re at the bottom of a well before you’re down there just totally out of luck.

That has been how I’ve thought through integrating it into our business my workflow, the people that work for me’s workflow, is don’t have it do your tasks, have it help you with your tasks. So we’ve been trying to integrate it in that way. So on the manufacturing side, e-early on there were some early adopters with Esprit and other CAMs, CAD softwares where they started integrating AI and, there’s now engines out there that they’ll take a 2D CAD drawing, it could be in a PDF, and it’ll write the whole CAM program for a CNC machine. 30% of those programs might crash the machine, so again don’t trust it. But it’s getting better by orders of magnitude, I feel like it, it’s, we’re on a logarithmic curve. It’s not a linear growth path. That’s one area. On the marketing side, though, where we’ve been using it a lot is some content generation, a lot of ideation.

What are some different ways to attack this problem? Or, let’s come up with a theme for the year and you almost use it as a marketing collaborator. And when you’re on a small team, unless you’re in a organization that’s a billion-dollar organization with a marketing team of, 10, 15, 20 people, a lot of times mid-size manufacturers, the marketing team is one, and that’s usually their second or third job.

So having an AI generative, predictive, whatever, as a collaborator is helpful. You just have to look at it as a junior collaborator. It’s not the leader.

Carman Pirie: Yeah, that thought partner kind of usage of the of a proper LLM is I think quite impactful and But I wanna kind of zone in a bit on your previous comment around subject matter expertise basically being the killer app. It’s subject matter expertise which creates the best prompts, the best output. People talk about prompt engineering and whatnot, and I’m sure somebody listening to this is gonna say “ no, Carman, you’re wrong. We know prompt engineers that are dramatically better and they make everything better as a result of their work.”

And maybe that’s true, but I still think you can ignore most of that prompt engineering advice and substitute it with subject matter expertise, and you’ll end up in a better place as a result. And so it connects to your other comment about often you have a marketing department of one, and marketing isn’t job one for that person.

Not to be unkind to some manufacturing marketers writ large, but the subject matter expertise sometimes isn’t there on the marketing side — they may have the marketing title, but they… i’m Sometimes I wonder if the marketing expertise is there to be the subject matter expert in prompting the LLM.

Jeff White: You mean, Carman, more on the marketing skill side or the understanding of the, of what that company manufactures and that side?

Carman Pirie: Oh, no, they’ll understand that certainly, but it just, it struck me when Will said “Oh, we have a lot of people in smaller mid-size industrials. It’s a marketing department of one, and sometimes it’s their third job,” and it struck me. We, and we’ve done a lot of episodes in this show, Jeff, about the marketing department of one, but it wasn’t until Will said, “ And usually sometimes it’s their third job,” where I was like, it just smacked me in the face.

It’s like, ’cause then the thought partner work that you’re doing with the LLM, maybe there’s only one partner there that’s thinking, and it’s not the one that should be. You know what I mean?

To be unkind to the category, but I wish I had a question in all that, Will, but I think it was more of a reflection and get your response to it.

Will Pierce: It’s a great point. So I have an agency, right? That I’ll make it – I can only speak from my personal experience and maybe what I’ve seen. But from my personal experience I am a marketing department of one. Even though we’ve got a company that’s 300 people large, a large, robust every other department our marketing department is me.

So I have an agency that I work with that does, I delegate that. So in a real sense, it’s obviously much larger than me. But there’s one person in Flexaseal making those decisions. Now, but that bottlenecks growth or evolution. So I have to program time for myself to sit down and I’m going to be a marketer today, and I’m going to focus on strategies, campaigns, this, that, and put all this together.

But what if I had an AI agent that I said, “Every week I want you to review our inbound leads our quote volume and assess the state of the market. Where are trends leading? What’s excelling? What’s not excelling? And help strategize with me on next steps for things that we can either improve upon or pull the company into a different direction.”

And you could just have that be automatic, and then you go and you consult with it and now you’re the leader with the junior persons coming to you with a report saying, “All right this month, this quarter, let’s look at these things.” And as the leader, you can say, “No, I don’t wanna do any of that.”

But it’s probably gonna jog you into things that you hadn’t thought of doing.

Carman Pirie: Absolutely. That makes sense to me. Since you mentioned working with an agency, I should disclose to our listeners that yes Jeff and I happen to own an agency, but in no way are we working with Will. So this is not a, not trying to tee up anything here on…

Will Pierce: We’ve got an inside job.

Carman Pirie: No, exactly.

I’m curious about what your… So I’m being very selfish here too, but I’m curious about what your attitude is towards AI usage on the agency side as you think about the partners that you work with as a marketing department of one. I’m assuming you’re eager for those partners to be leveraging AI as well.

Can you just reflect on that relationship between a client and the agency and the AI usage that sits in the middle of it?

Will Pierce: So that’s interesting. To me, it’s a mixed bag in that regard. When I… And we recently changed agencies last year, and when I was interviewing different ones, I wanted to know what their AI strategy was how they’re implementing it. Because it’s a siren song to a certain degree.

You can let it do all of your copy and, God help you. But and so that was a point I made with them is like, “Look, this, we’re in a very technical field. Our brand voice is technical, authoritative, trustworthy. We can’t go out there with loose copy. It has to be tight and on point and factually correct.”

And so I just I made the point with our agency, you cannot use AI to do your main copy generation. You need a true technical copywriter. Don’t have them go out on a limb. I’m still the bottleneck for copy, right? So at the end of the day, I have to review every word that gets through there and make sure it has the check mark. And also start avoiding those same things that we’re starting to see across the industry, that sameness. All of us have gotten emails where we’re like okay, did Gemini write this? Did Claude? Did Copilot?” And you start to learn the different AIs’ voices, and it’s this isn’t the same person I’ve been emailing for the past five years.

And I’ve actually called out some people on it. I’m like, “Look, just write me an email. Your emails went from two sentences to seven paragraphs. I’m not reading it.” I saw a standup comedian do a joke. He’s like, “I used to tell my employees to get their stuff done, but then I realized that was too blunt.

And so I had Copilot write me a very flowery message, and they received the flowery message, and they had Copilot tell them what I was actually trying to say, and Copilot told them to get their stuff done. So I think that’s better.”

Jeff White: It’s almost like everybody who is creating anything with AI, the number one task that they could learn other than being an expert in what they’re doing is just the edit button. Just understand how to fine-tune that or start from your own content rather than letting it provide.

Will Pierce: Yeah, so I mean, I don’t know. I think I walked around your question a lot, but to be frank, I don’t know exactly how our agency uses AI on a day-to-day basis. I know for a fact they have to use it for copy generation because the amount that they’re coming up with, they don’t have degreed engineers on their staff writing this copy who are in the mechanical seal industry, but it’s pretty accurate for what they’re writing.

So yeah, I’m just cognizant that it’s there and I have to monitor it.

Carman Pirie: Yeah, when you mentioned that you said, “I’m the bottleneck here. I need to review it all.” And man, I feel that. And it, the, it pointed me in the direction of, frankly, one thing that I’ve noticed on the agency side is probably one of the most malicious uses of AI that I’ve encountered, is the using AI to review copy that somebody else generated.

Because, of course, if you just plug that into the LLM and say, “Please review this,” in no way is Claude or ChatGPT or whatever, doesn’t matter what LLM you use, they’re not built to say, “This looks good. I have nothing to add.” That’s never going to come out. It wouldn’t matter what you did, it wouldn’t matter who wrote it, how good it was, there will be a suggestion.

And so it just can put you into this endless loop of in the worst-case scenario, you have AI generating the copy, and then the client side is reviewing it via AI, and

then it’s just this endless Endless AI loop with a couple humans in the middle just for giggles.

Will Pierce: The dead internet theory, right? No, it’s really funny you bring that up. And it’s funny that you called it malicious ’cause I did that recently. I took a blog post and I was slammed, and I was like, “I wanna get this out. Let me just dump this into…” I’ve been using Gemini a lot. “Let me just dump this into Gemini and have it give me notes.”

And it was all grammatical stuff. And I was like, “Okay fine.” And I gave them the grammatical notes and moved on. I forgot that I did that. Two weeks later, I go back and somehow that blog post copy came back up in my queue, and I actually read it top to bottom, and it was, several thousand words.

And I had tons of technical edits and all types of… I probably edited 10 to 20% of the overall blog. And I submitted it. I said, “Hey, I’m finished reviewing it.” They were like, “You reviewed this two weeks ago and said it was fine.” I said, “Oh

Jeff White: Claude said it was fine. Sorry.

Will Pierce: Exactly. And so I learned from that, okay, can’t do that anymore. One, it was mean. Two, it didn’t do a good job.

Jeff White: The poor copywriter is just, “Oh.”

Carman Pirie: I love this. We’re doing like the agency client side kind of disclosures here or something. And we’re gonna have to rename the whole episode.

Will Pierce: Yeah. I’m a transparent person. I, I’m gonna tell, let you know what I’ve done, good, bad, or indifferent.

Carman Pirie: Yeah. I think the, on the brief side is the other area where I would say the agency-client relationship has suffered some malicious AI use. If there’s ever a document that really deserves to have human attention poured over it, it’s a brief. And briefs ought to be brief for a reason.

AI’s not great at making them brief. And so yeah, it’s interesting. I think these are the types of dialogue that both client and the agency side, m-m they need to be using AI enough to be able to have this conversation in an intelligent way. Because you have to be able to encounter the problems and understand what you’re encountering, just like you just said, before you can move on to its useful use.

Jeff White: Will, do you think that there is… Like I, I feel like I’m leading the witness a bit because I feel like I have an answer to this just in terms of how we’ve conducted some of these things. But do you feel like there is a path forward for utilizing AI to develop highly technical content that can achieve your goals as a manufacturer?

Will Pierce: Yeah. That’s the inevitable end. It’s – I’ve seen such growth in the couple years that I started using it. It’s gone from being a middle schooler to a high schooler to a college graduate to it, it will get there. The one thing that I talk about with my colleagues is that right now my impression is that it’s limited, it is limited to its knowledge base.

It’s not extrapolative. It’s not going to look at a problem and come up with a novel solution. It’s knowledge base is essentially the internet and whatever we feed into it. And so it’s gonna work within those bounds, but it’s not extrapolating and going, “Okay, I see this. Let’s take it here.” And that’s where the human in the loop… And I don’t know when – And that’s probably AGI, right? Or is that, is AGI artificial general intelligence? We’ve truly got something that’s thinking.

Jeff White: Mhm

Will Pierce: I don’t think these engines are thinking yet.

Jeff White: I don’t think so either. I think to your point, what’s out there, it can’t necessarily develop a cogent hypothesis and go out and solve it and bring it back to you. But I do think that you can feed into it, either through a conversation with an engineer or, and generating a transcript and using that to kind of power things may be able to get you that little bit closer, that coupled with web search, that the agent is using in order to get there.

Have you experimented with that approach?

Will Pierce: I have a little bit. One thing that I’ve found it to be extremely useful in is, so – And this is actually fairly new, but all right. Every, every company goes through stages of brand growth, old logos, old voices, all of that. And so we, and we all have collateral that was made 10 years ago or 20 years ago.

We’re still selling those products but we need that collateral to be updated. And updating collateral has gotten so easy. I take an old document that I’m happy with 90% of the copy, I’m happy with the images. Maybe I need to do some image updates certainly logos, whatever. And then you take a current document that you’ve got and you tell it, “Hey, take this old one, insert these new images or replace these images get it on the new brand format templates,” and we’re changing how we’re doing it now.

We want to start with – We want it, we want this to be like a story. Start with the solution and then get to the product. Don’t start product first. And it does it. It actually like it… and this is – I’m using Claude for this and I was shocked. Something that I would’ve had to pay $1,000 for and wait two to three weeks with several revisions, I did in about 15 minutes.

Carman Pirie: Really nice tight context there, right? You’ve got this is the previous piece of collateral. This is the new information I want to inject in this. I want to do it th-this way. The room for the AI to mess up is less, right?

It could really… Yeah, it’s a great use case.

Will Pierce: I, and I actually wanna call that back to what you had said before about prompt engineering. I think for a period of time, everybody thought that prompt engineering was gonna be so vitally important. We have to understand how to talk to these things. I feel like that season has passed.

I talk to it just like I’m talking to you right now, and it gets me exactly what I want typically.

Carman Pirie: And that, that is what’s exactly my point, is that your subject matter expertise, when you’re just talking to it the way you would talk to me, just right now, it your subject matter expertise can’t be faked. It’s just gonna come through. Right? You will ask the question in a different way than I would because you’re an engineer and I’m not.

And in depending on the question, that’s gonna be really useful for the output of that LLM, right? yeah. I guess the thing that people need to lean into more is like the… people say treat it like an intern. We’ve mentioned a couple of times in this conversation how it was in high school and now it’s university student or what have you.

Sometimes I think that leads people into a thought that when they’re prompting, they need to, in some ways, dumb down what they’re asking. And I think you almost need to do the reverse. If you’re a subject matter expert and you’re prompting a tool, you should be trying to appear very smart to that tool.

’Cause it will meet you at your level, right?

Jeff White: Yeah, and then it has the context to give you what you need too.

Carman Pirie: Yeah

Yeah, it’s the one, it’s the one part of treat it like a very smart intern, that model is the one part of where that falls down to me, ’cause you would still never… you would really risk talking over the head of that intern, right? But you don’t have to risk that with Claude.

Just go ahead and talk over its head if you can thing, right?

Will Pierce: No I agree completely

Carman Pirie: I was just wondering more on the pure sales side, I guess, yes, it’s, it makes scaling outreach possible. Yes, it makes in-depth research possible, and it’s just like you just said at the start of this conversation. And my goodness, we all get those emails, don’t we? And we know exactly what they are. And part, email, cold outreach, things of that nature, it’s getting tougher by the day anyway, long before AI showed up. I guess when you look into your crystal ball, Will, is it just going away entirely as a tactic because of what’s happening with the slop?

Will Pierce: So going away as a tactic as far as using AI for outreach, managing direct…

Carman Pirie: Just email cold outreach to prospects with any hope of getting anything in return? Is it just donezo because AI’s muddied it so much?

Will Pierce: It’s a fair question. I s- I still get cold outreach and granted, I feel like marketers and salespeople are the easiest people to market and sell to.

Jeff White: A great LinkedIn kind of circle. Yeah.

Will Pierce: It is. It is. My wife won’t let me answer the door when someone’s coming to sell something because I’m like, “Wow, they had a really good pitch. I liked it.” And, and so

Carman Pirie: Sales people love to be sold to if it’s done properly, yeah.

Will Pierce: So with that said, I may not be the best judge of it, right? ’Cause, I got an email yesterday from a Berkeley researcher who is – they’re doing a startup and they’re building agents for, entering quotes and doing all this. And I took that meeting ’cause it’s oh, that sounds really cool and I wanna hear about it.

And, I can give you 15 to 30 minutes of my time to hear about it. Do you guys ever see the cartoon of it’s like in, it’s in medieval times and there’s this, this king fighting a war on this battlefield, and he’s got his he’s got his wagons on square wheels, and this guy’s up behind him with the round wheels.

“Look, I’ve got…” He’s like, “I can’t. We’re busy. We’re busy fighting this battle.”

It’s like I don’t wanna be that king fighting the battle, like with this, somebody’s coming up here telling me about this tool, and I, I – so I don’t know. I don’t… I think I’m biased. I don’t know if I can answer that question effectively.

Carman Pirie: It’s like my FOMO makes it so I can’t answer this question.

Will Pierce: Yeah, that’s exactly what it is. ’Cause I probably take more meetings than I… Which, all right if this thing, if people watch this, my inbox is probably about to explode. We found a mark. We got

Jeff White: He’s the guy. That’s the guy. I had one of those meetings yesterday. It was interesting. It was not for me.

Carman Pirie: But are you adjusting your outreach tactics or how you think about it from the sales side as a result of what’s happening on the AI front? Are you doubling down? Are you using AI for those sales messages? Are you getting away from cold outreach or any, any changes?

Will Pierce: So interestingly we’ve grown really well without doing a lot of outbound. Almost all of our marketing is currently inbound and our business development group is almost handling completely inbound leads. And our outbound is our outside sales team going out there, doing regular old gumshoe work, right?

Getting in front of people, trying to go up to the gate, get past the guard shack, that type of thing. So we as an organization have never really done that. But that’s not to say… I think the AI will help it. I think it’s gonna flood it with crap for a long time, and then eventually it’s going to improve.

Carman Pirie: All right. I love having a prediction. ’Cause that way when we bring you back on the show in a year or whatever, Will, we can see if the prediction has come true. So

Will Pierce: Well, let’s take for example, and I don’t know if it’s just the LinkedIn algorithms have improved, but I felt like for the past 12 months, LinkedIn was just flooded. Major billion-dollar organizations were putting out crap AI-generated graphics that all looked the same, all were riddled with errors, and it was just like, what?

Who authorized putting this out there? And I stopped looking at it for a long time, and I’ve recently engaged more and I found that all that, a lot of that crap is gone. So I wonder if the novelty – people are past the novelty of it and that’s when we’re gonna hit kind of the pay dirt with it

Jeff White: I wonder if we’re just past, or not past, but if there’s been so much pushback outside of the marketing and sales industry on slop that people are just so concerned about being called out that the bigger organizations have realized they need to actually still have a quality filter on this stuff.

Will Pierce: I think that would ring true from my experience.

Carman Pirie: I don’t think they’ve given up using AI to generate it. It’s just a stronger slop filter.

Will Pierce: A lot stronger.

Carman Pirie: This has been a good meandering conversation. As is the danger when talking about a topic like this, you can just, before you know it, a half hour is up.

I guess as we part ways today, any parting advice for those that maybe aren’t as advanced in AI as you are?

Will Pierce: Oh, I, there’s absolutely no harm in diving in, and there’s so much information out there. If you haven’t tried it, you have to or you will be passed by. Just jump in. Jump in, and there’s plenty of places like this and others to get more information and yeah, I would say you need to dive in.

Carman Pirie: Yeah, I think we’re past that point if you’re not in these waters, then you’re done now in so many categories. Look this has been lovely to have you on the show, been a fun conversation. Thanks for joining us.

Will Pierce: Yeah. Thank you guys very much. I appreciate it.

Jeff White: Yeah, it’s been wonderful. Thanks.