Ep 83: The ROI of AI - Productivity, Creativity, and the Work That Actually Matters
Watch the YouTube video version above or listen to the podcast below!
Episode Summary
Alex and Dave explore one of the biggest challenges surrounding enterprise AI: how to determine its real return on investment. Starting with the cost of tools such as ChatGPT, Claude, Gemini, and Copilot, they examine why AI is harder to evaluate than traditional software. Subscription fees are only part of the equation; token usage, agentic workflows, repeated revisions, and human oversight can all add cost. At the same time, common ROI measurements such as hours saved can be misleading, especially when they turn productivity gains into arguments for reducing headcount rather than expanding what employees are capable of accomplishing.
The conversation then shifts to agentic AI and the limitations of today's systems. Alex and Dave discuss how agents can perform impressive tasks but still struggle with ordinary business processes, user interfaces, file handling, and unexpected workflow complications. Small errors or assumptions can also become amplified when they are carried through multiple AI-generated steps. For that reason, they argue that humans remain essential for reviewing work, applying judgment, correcting mistakes, and determining whether the final output is actually useful. AI is most valuable when it enhances an employee's area of responsibility rather than simply automating isolated deliverables like reports, presentations, or spreadsheets.
From there, the hosts examine the organizational side of AI adoption. Companies that treat AI like another SaaS purchase may miss the larger transformation required around leadership, training, incentives, and process design. If executives focus primarily on cost reduction, AI projects will naturally be judged by savings and efficiency rather than quality, innovation, or better decision-making. Using Formula 1 as an analogy, Dave explains how strong performance depends on constant feedback between engineers, mechanics, drivers, and real-world results. The lesson for businesses is that technology alone does not create better outcomes; organizations need clearly defined goals, frequent feedback, and coordination between the people designing systems and those using them in practice.
Alex and Dave also challenge the assumption that more output automatically means more productivity. AI can generate more blogs, social posts, campaign assets, reports, and presentations, but producing more material has little value if nobody reads it, uses it, or acts on it. In marketing especially, excessive AI-generated content can create sameness, dilute brand identity, and overwhelm the people responsible for reviewing and distributing it. They argue that creativity remains central to differentiation. AI can provide structure, prompts, and assistance, but the distinctive ideas, judgment, and understanding of an audience still need to come from people.
The episode closes by extending the discussion from software into physical AI and robotics. From guided journals and creative prompting to C-3PO, laundry-folding robots, robotic lawnmowers, and Roombas, Alex and Dave highlight the difference between something being technically possible and genuinely useful. Adoption depends on reliability, usability, cost, and whether a technology produces a meaningfully better result. That ultimately becomes the episode's central argument about AI ROI: the most important question is not simply whether AI can perform a task faster, but whether using it leads to better decisions, better work, stronger creativity, or improved business outcomes.
Ep 83: The ROI of AI - Productivity, Creativity, and the Work That Actually Matters Podcast and Video Transcript
AI ROI Debate Setup
Dave Dougherty: Hi, and welcome to the latest episode of Enterprising Minds. Alex and Dave here with you. Um, so in today's episode, after you like and subscribe and help, uh, help with the podcast, um We're going to talk about the ROI of AI and the business cases around that. There's been a lot of debates on this, and I know if you pay any attention to the markets and Wall Street, there's a lot of debate on the Wall Street side of things on what's actually the ROI on all of this infrastructure spending and, you know, the demand side of it.
So Alex, this was your idea, so why don't you- Yeah ... properly scope it for the rest of the audience and we'll go from there. Sure.
Building the AI Business Case
Alex Pokorny: Yeah, I kind of wanted to put together a kind of a imaginary situation. Let's say you're at a company, you're trying to get more AI or AI licenses or access to it. You basically now need to put together a business case to basically just pitch it.
And what I'm saying by AI is basically any of the LLMs out there. So let's say you're trying to get access to, Claude, Copilot, Gemini, ChatGPT, whatever your flavor is, um, and your company currently will not fund it. So there's a couple of difficulties that I want to throw into this. One is there is, generally speaking, a monthly fee or cost for all of these different kind of licenses or access levels.
Mm-hmm. And a higher cost that you pay usually just gives you more credits on it. Then the other side of it is once we especially hit, like, the agentic side where it's either doing things virtually as it's connected and you've given it permissions to access stuff, or it's doing things on your own computer, literally moving your mouse around like a ghost and clicking on things and doing stuff for you that burns tokens really fast.
So then you start getting outside of that, you know, straight monthly cost kind of thing, and now into how do you budget for the token cost piece of it. And then that kind of gets into this whole next piece of is the work that's being produced valuable? Can you actually make a business case out of this?
Maybe you're putting together a presentation and it's able to throw together a presentation while you do something else. That's cool. It's probably going to be a little bit generic, and you might go through five revisions of it, which is going to burn your tokens and credits like crazy, and then at the end of the day, you give this great presentation, and what was the ROI of that, right?
So it starts to get kind of tricky once you have these kind of non-tangible pieces, and then also speaking, Dave, about the kind of markets as well from, like, a job loss aspect.
Hours saved isn't always a great argument because then it sounds like, oh, so your coworker's expendable then, or you're expendable because you're saving 40 hours a week.
That sounds like- Well, the- ... a headcount reduction.
Dave Dougherty: Yeah. It's the, it's the bucket of work fallacy, right? So- Right ... you know, we have a certain amount of stuff that needs to get done each week, and if we can do it with two less people and AI, then that bucket of work is fine. But That bucket of work is- Right.
Sounds great ... has a bunch of holes in it. It is never full. It is always changing. It i-
Alex Pokorny: yeah, which then I, I try to like play a... I- it's is there a cost, offset kind of savings? So let's say instead of doing a marketing campaign that hits one segment, now you're able to do, you know, you're able to split that up to five different segments, and you have better landing pages, better ad copy, and maybe a better return on your campaign.
Maybe, maybe not. Kind of depends on a whole lot of other factors, to be perfectly frank, like- Right ... your budget and your bidding strategy and a bunch of other things, or how you're maintaining your account. Competitors, markets, general economy. There's a lot of pieces there that kind of are going to put some wrinkles into that, that argument pretty fast, uh, that you can't control.
And then creating presentations and decks, like, okay. Is- are we creating value? Are we just throwing out slop? I, I, I have to admit, in the last couple of weeks, the first time it ever happened to me, I took something that was obviously AI generated from a colleague. I threw it into AI to summarize it so that it could then give me a concise version of the AI created thing in the first place.
Dave Dougherty: Yep.
Alex Pokorny: So we successfully avoided communicating directly with each other using AI. I have no idea if any thoughts were shared. So I don't know if it was actually a success.
Dave Dougherty: Well, and the biggest thing for me, I mean, when it comes to, if you look at a productivity side of things for me, I think-
Alex Pokorny: Yeah
Dave Dougherty: What work is valuable is an interesting question- Yeah
that people are now having to face because I think that depends company to company and what the culture is, right? Like- Sure ... uh, Amazon banning PowerPoints. I mean, how long has that been a case study on, "Yeah, you don't need to do this"? And yet how many companies do you know default to if it's not on a PowerPoint, it doesn't exist, right?
Alex Pokorny: Right.
Dave Dougherty: So yeah, cool. We could have already done this, but we haven't because change is hard, right?
Agentic AI Reality Check
Dave Dougherty: Um- Yeah ... the, the biggest thing for me is operations. I think, three years, four years into AI, like, you know, we were sold originally the creative part of it, but, It was getting better in ter- like, you could start convincing yourself like, okay, yeah, the chatbots are doing a lot, a lot better in terms of writing emails, doing higher context work.
Sure. And then you flip over to agentic and you're like, oh wait, you can't follow a process. Yeah ... like I think it's been interesting, New York Times had an article, I think I read it, it was reading this yesterday I think where they actually went and they did some of the examples that are used in the benchmarking studies.
Alex Pokorny: Mm-hmm.
Dave Dougherty: Uh, and so it was, okay, agents, what are you doing? And they tracked it the all the way through, and on each thing it failed. The agents failed in certain elements, but passed in other ones. And so it was- Sure ... so outcome driven that it didn't necess- it, it would like ignore certain things or, Yeah
it would like refuse to use the A- the UI, and instead would code itself its own thing so that it was easier for it to deal with. You know, and it, and it talked about like having a hard time with like basic Google Sheets work because it didn't want to deal with the user interface. It would rather- Sure
code than dealing with the onscreen visual stuff. So one of the takeaways was that, okay, for a lot of this stuff, you still need the person to go through and make sure it didn't screw up, and stuff that is fairly simple for a person, like uploading, downloading a PDF, can actually be very difficult for these systems depending on, you know, how they play with each other.
So that article is interesting. Sure. I'll put it in the show notes for people. You might have to log in or create an account, but it's worth, it's worth reading because it's a third party test of all the stuff that we've been hearing from the labs, you know?
Alex Pokorny: It will be interesting also to see... So AI companies, any of the frontier model kind of companies, the ones I, I basically rattled off before-
Dave Dougherty: Right
Alex Pokorny: have been really focused on solving coding for about a year and a half, a little bit more than that. And so Codex and a bunch of other kind of elements of it are definitely for software engineers, built by software engineers. That's kind of their bread and butter of the talent pool that those companies are made up as.
A lot of companies kind of fall into that default process of- The things that seem like the biggest problems in the world are the things really that are biggest problems for the employees.
As it moves more towards this for work GPT for work or, the Claude Cowork and as well as Copilot's Cowork, the naming is terrible.
Branding overall on all these things is just ridiculous. But as it moves more towards this agentic side of things, of like it's actually trying to do a day-to-day task which is Excel sheets and PowerPoint and email and communication and chats and everything else-
Dave Dougherty: Mm-hmm ...
Alex Pokorny: do you think maybe it's just not there yet?
Like I wonder from the benchmark aspect- I don't think so ... like maybe it's just not built for that quite yet.
Dave Dougherty: So this is a hot take, I might get in trouble. But I think if you look- if you look at what human beings do whenever they get put into a new job, one they might be unqualified for or unsure about-
Alex Pokorny: Sure ...
Dave Dougherty: they naturally default to what their experience is prior. Sure. So there's a lot of leaders where, all right, I've only ever done commercialization, but now I'm in this other thing, so I'm going to do this other thing with a view towards commercialization.
Alex Pokorny: Commercialization, right.
Dave Dougherty: Of course. Yeah. Because then at least something feels comfortable, right? Right. And so, okay, if you have a bunch of engineers and coders who are engineering and coding, the systems that they create to mimic human behavior will be to mimic the creators-
Alex Pokorny: Mm-hmm ...
Dave Dougherty: and what they prefer.
So the fact that Cowork is built on the coding platform, it's, it makes sense that it defaults to, no, I'm just going to write a Python script for this because that's-
Alex Pokorny: Yeah ...
Dave Dougherty: that's there. Now, it is easy to anthropomorphize the AIs in this example. Please don't. It is computer code, right?
Alex Pokorny: That's true.
Dave Dougherty: But it also, that's where the organizational change, it's a little eas- you know, it's a lot easier for smaller groups or, um, s- you know, solopreneurs or, you know, teams of five to adapt this and then, you know, just kind of look at each other and say, "You know what? I never liked reading your PowerPoints.
Let's not do PowerPoints anymore." It's a lot easier for that team to do that than up in the big enterprise where, you know what, actually treating these tools as traditional software might actually be the problem instead of the solution that you're used to. You might actually have to treat these differently than just buying another soft- uh, SaaS, platform.
Yeah. So yeah, I don't think... The, the LLMs are definitely not there. I think the human in the loop has to be there. Like, you know, and
Alex Pokorny: d- Absolutely ...
Dave Dougherty: so-
Alex Pokorny: I mean, we're not even close to the point where it can do anything really solo of great value.
Dave Dougherty: Right, and for me, my biggest test has always been, did I at least hit 80%?
Yes or no? I don't... Do I ever feel super good about what I turn in by the deadline? No, not really ever. You know, because you can always think like, "Ah, I could've done, you know, the third paragraph better," or, you know, "These charts are wonky because I just didn't have time to, get the graphic design done well."
But I think there, there's little things too, especially with the agents and the chatbots that I've been discovering lately. As I try to do more with these things, if I don't address a small issue that's pretty easy to overlook in a document, at the beginning, it gets carried through all these other things.
And as it gets carried through these other things, because it's shown up multiple times, it gets amplified as we made this decision. It's like, yo, yo, yo, yo, we never made a decision on that. Yeah. Stop talking about that- Yeah ... in a definitive sense, so then that, that unwind is a lot harder to what you've talked about previously on how like it goes down a rabbit hole and it gets unstuck.
So how do you deal with that? Um- Right. So yeah, you have to have that editing and that person in the loop for whatever process it is to then say, "Hey, this is reasonable. This is okay. This is,"
Alex Pokorny: maybe this kind of runs back to that ROI discussion because-
Dave Dougherty: Mm-hmm ...
Alex Pokorny: instead of it doing something novel, we're really talking about that person's work amplified, upgraded, you know- Right
specialized, stuff like that. Like we're continuing to have the same person, same role, and that's the one thing I always get annoyed by whenever we talk about like layoffs in AI. It's not about that. It's about areas of responsibility.
Dave Dougherty: Mm-hmm.
Alex Pokorny: And AI is inert. It doesn't just care about your SEO so much that it just starts randomly sending you emails about SEO and random things, and going off on the latest thing that it found off of LinkedIn, et cetera, and all the rest.
No, that's an SEO. You hire somebody full-time, that's what they do. Yeah. Because they care. They get really obsessed about an area, and they- Right ... fight for it, and they make it happen because internal processes are terrible always. Right. And it might take two steps one day, and the other time it takes 14 and a business case and a sprint and a ticket later just because.
And a person handles that. An agent doesn't. So if you have an area of responsibility for something straightforward with an output driven thing, like let's say analytics, and you've got a monthly report that you're saving time on-
You can make it upgraded. And I think showing that you made a presentation from basic routine, one that comes out of GA or Adobe, like basic kind of monthly auto sent to you, maybe it's a SemRush one, whatever it is.
Right. Versus pulling out the insights on it and making it more actionable, which is actually like turning a report to being useful, and you're now able to do that, that additional analysis. I think that's worth a lot. And that, that's- It, I keep running back to this game that I did at Medtronic of, um, because we would be budgeting, uh, demand gen campaigns and-
Dave Dougherty: Right
Alex Pokorny: and what's the cost or the value of one additional webinar visitor? Right. And I- you have to ask that because you have to come up with a budget figure to put on a campaign. You're spending money, and everyone would always give me a blank face, and it'd be like, okay, what if... And I can't guarantee quality.
Let's say 500 bucks, and they'd say, "Uh, that's too much." And you're like, "Okay, 200." And they're like, "Well, that seems okay." I'm like, "250? Meh? Okay, 250. We're starting with 250." Right. That'll be my goal. And literally, that was the game because I couldn't figure out a better way to put a number on an intangible thing like a webinar viewer.
Like, I don't know what's that worth, and to be honest, I have no idea what it's worth to anybody, and it changed all the time because some people- Right ... would be like, "Oh, it's only worth 50 bucks." Other people would be like, "I'm willing to pay 1,000 because we're targeting surgeons, and, honestly, it's really hard to get a surgeon to attend a webinar, so 1,000 bucks is fine."
And that, that blows my mind, but an ROI discussion like that can quickly fail against other projects if you're in a budget meeting because you're saying, "Now I've got better insights." Cool. Well, this other guy says he can save us money with a logistics project, a black belt Six Sigma project. I guess I'm going to do that instead.
Leadership Strategy and F1 Feedback
Dave Dougherty: Well, and that's where there are some narratives and some environmental problems that I think we run into.
Alex Pokorny: Yeah.
Dave Dougherty: So this is why I think with AI, because it's more the general purpose thing, not only are we hit with, okay, a tool can do this, but a tool can do damn near everything better, right? So now it's hitting- Yeah
all aspects of the business, except nobody really knows because nobody's getting any training on anything, so everybody's just got their little pet projects, and then that's what they're doing, right? So that, okay, that's problem number one. Problem number two- Yeah ... because of all that, the leadership needs to lead on it.
So if you don't have leadership that's talking about AI, proposing things and, and encouraging people to do stuff with AI to figure it out, you're not going to have a good strategy, nor will you have the air cover to create a good strategy because- You're not... The system's not aligned to it. Yeah. So if you have- Right
a CFO and a CEO whose narrative to the outside world, I mean, assuming it's a public company, is one of streamlining costs so that the margins are better, which you will never be fired for that narrative. Let's just be honest. That you- Especially as the CFO, yeah ... will never be fired for that.
Alex Pokorny: Yeah.
Dave Dougherty: So if you can deliver on that, great How are you going to do that?
You're going to tackle cost of goods sold, whether that's people or inputs or processes or, you know, waste or whatever, right?
Alex Pokorny: Yeah.
Dave Dougherty: But there's going to be some, some friction that's necessary in order to produce quality things. All right? You can't just keep cutting your way to greatness. There's nothing in the world that has ever worked by cutting and then being great at something. That's just not how it works.
Alex Pokorny: Yeah.
Dave Dougherty: And so yeah, it...
Like, you have to, especially if you're in an enterprise environment you know, outside of the, the small business stuff where you can be agile and adaptable, you need to really look and be like, "Who's actually helping with this? Are we set up to actually be successful with this?" If you care about it. And if not, you gotta find somewhere that does because, you know, otherwise you're, you're not going to get the training or the support or the whatever to be adapted to the, the next thing.
So I don't necessarily think that it's going to hit as fast as a lot of people have predicted, right? The financialization of damn near everything really bugs me. So I'll admit my bias there. But I do think that if you have the C- the CEOs and the CFOs right now are way too focused on financial things for AI to actually be implemented properly, because it's not just a financial thing, right?
Alex Pokorny: Yeah. No, that's a good wrap-up point to it Can help. It can make things easier. It can have a quality of life change. Aspect of, you know, taking away some of the, the annoyance work. I mean, stuff like that. Yeah, when it gets agentic solidly, I think that's when the job fear is going to get resurrected again of-
Dave Dougherty: Mm-hmm
Alex Pokorny: well, it can go create the report. What's the value of having an analyst? I mean-
Dave Dougherty: Well, what's the value of the report though? It's- You know, I mean, that's-
Alex Pokorny: I know. That's, that's the thing is like that, that gets to a really dumb question because it's like, well, you hired a person for a reason, and the reason was not to create reports.
It was to provide recommendations and insights and- Mm-hmm ... make the business better so it can become more profitable. It's not to-
Dave Dougherty: Right. One, I think one of- ... create
Alex Pokorny: reports ...
Dave Dougherty: you and I are both humanists, I think that would be safe to say.
Alex Pokorny: Yeah.
Dave Dougherty: Yeah. And that's not necessarily a, a good thing when it comes to, to business brutality.
But let me reframe it this way. So for a lot of people, uh, if you don't know I'm a fan of F1, and I'm... I don't just watch the races, I watch all the pre-races. I watch the, um, the special things that F1 TV does and puts on the app. I'm in it. And there was one... They have this really interesting show called Tech Talk.
Now, granted I'm not an engineer, I am not interested in building things. I get really frustrated as soon as I have to measure twice on something, right? So I, I am not set up for this world, but I find it really interesting because it's a lot of smart people going to the pinnacle of something, right? How do you try to get to that very specific outcome on the edge of what's possible, right?
That to me is the interesting thing. So this Tech Talk piece, it was the, um... Sam is the host who has an engineering background and is very interested in aerodynamics and cars and, you know, whatever. He was talking to a now retired engine mechanic, formerly of Red Bull, and they were talking about how do you actually put things together between the engineers and, you know, you mechanics on the ground level from the designs, the materials, the whatever, and the teams, and what are the processes with that, and how does all that work?
And the guy immediately started talking about how when you're looking at the designs from the engineers, and the engineers are using all of these new AI models and all of these machine learning techniques and all of the, you know, CAD designs and, you know, they're using- Sure ... all the top mathematic things- Top of it and whatever.
When they start the new season in January, they have to then look at all of those things the engineers came up with over the holidays, and then figure out how that's actually going to work in real life And that theoretical to real life thing is really where the team's success happens, because the mechanics have to say, "Hey, the...
I know this worked on your CAD design, but I c- I don't have space for a wrench to tighten that bolt, and if we don't tighten that bolt, we're going to have an engine vibration, and it's going to ruin the whole thing."
Alex Pokorny: Yeah.
Dave Dougherty: So they have to go redesign everything, right? So the key thing with that is just that back and forth between the theoretical and then the on-the-ground thing.
My brain, when I first heard that, because I'm a nerd, I immediately thought the difference between marketing and sales. You know, you get a lot of these marketing centers who are like, "Here is the best practice, and theoretically- Yeah ... we should totally do this and target this persona. Isn't that wonderful?"
And then the sales guys are like, "Y'all are in a tree house. I don't understand what you're saying, but that's not what anybody's saying on the ground." It's like, okay, well then tell me. Let's talk. Yeah. Let's work this out, right? But the thing between an F1 team versus regular businesses is there is a defined goal, and every weekend you get the feedback on whether or not you're getting close to that goal.
I want to be on the podium.
Anything below- Top three ... the first three point, the first three positions is a fail.
Alex Pokorny: Yep.
Dave Dougherty: So okay, if you're in sixth and you're consistently sixth, all right, you might be the best of the rest, but do you think sponsors want to work with the best of the rest? No.
Alex Pokorny: No.
You got limits. Yeah.
Dave Dougherty: You know? And how much you win depends on how much money you get the next year. I mean, it's just like all of these different things between the F1 system versus, you know, how business actually works is fascinating to me. But that's a different episode. Um, but it was that feedback loop, right?
And having that really defined goal of every weekend we want to be in the top three, or at least- Sure ... fighting for top three, right?
And if the engine system fails, well, that's an engineering problem. Or the carbon fiber rips to shreds because of a crosswind, well, okay, materials guy, go redefine this so you don't, ruin it for everybody else.
Alex Pokorny: Do you know what's not important in that story?
Dave Dougherty: Reporting?
Alex Pokorny: The AI system that they- ... used to make the thing in the first place. because what matters is the output. It's basically is-
That coordination that's far more advanced than chatbot telling you, "Hey, by the way, this would be, this is an interesting insight that could save you some money or do things a little bit differently."
Great, but if you're presenting crap leads to your sales team and your sales team's like, "This is terrible, and your messaging makes absolutely no sense for our customers, and I have no idea what on earth you guys are doing, and please stop talking about discounting because we are trying to work on quality and, you know, in justifying cost right now."
Right. Like, there could be so many disconnects there that then the tool doesn't matter so much. And you want to get better, and this tool can help you get a little bit better, cool. Grab the tool. So let's keep going.
It's not important.
Dave Dougherty: Right.
Alex Pokorny: And maybe that's it. Maybe it's just, it's getting the attention of some brand-new super expensive thing where really, to be honest, most of these tools are pretty darn cheap, and the work is probably far more valuable, especially if you're talking about outputs of your work.
Dave Dougherty: Right. And that's where- Once you get that
Alex Pokorny: coordination, once you get that leadership. Once
Dave Dougherty: you get that point. Right, and that's, and that's one of those things, right? I mean, what do you... When y- if you were to sit through a race weekend, which is a large investment of time, but if you were to do that, you would see the first practice of every weekend is a, all 22 cars going around the track at least 15 times to make sure that the car is set up properly for the environment they're about to race in for the rest of the weekend.
Right? So they will do up to, on a normal race weekend, they will do three practices Two on the first day, one on the second day, then there's a qualifying-
Race for- Mm-hmm ... what position you start in for the race to... on Sunday, right? So those first three one... Those first three practices are all about tuning your car and collecting enough data on the car to make sure, A, any upgrades you brought to the track are working the way you thought they would.
Alex Pokorny: Yeah.
Dave Dougherty: And B, the car is actually set up to be successful for the rest of the weekend, and then you address any of those issues before you do the qualifying race. because there are certain tracks- Yeah ... where you if you qualify poorly, you will not be able to move up- Yeah ... because the tracks aren't set up for overtaking or anything, right?
Yeah. That's a good- And with each practice you have all the racing drivers in their interviews going, "Yeah, I don't know. I haven't seen the data. I need to go talk to the, talk to the people about what race data we have and what changes we should make." So you get that constant feedback loop, and the reporting is handled a lot by some AI systems and whatever, but it's to facilitate a conversation.
Alex Pokorny: Yeah.
Dave Dougherty: And that's where my... I guess what I'm saying with this example really is my problem with the productivity debate really is why do you need that output? If it's to create another spreadsheet, if it's to create another report, who cares? If, if all you're using AI is to, a CYA strategy, cover your ass strategy, why, why?
That's not really- Yeah ... a good use of the tool.
Alex Pokorny: Right? I'm with you on that. And that makes sense also the, the preview and practice rounds, you can imagine that also between a marketing and sales team saying like, "I'm not setting expectations that we're going to nail it in one."
Dave Dougherty: Right.
Alex Pokorny: We're going to look at things for three months.
We're going to meet, separately, and then- Right ... we're going to talk about basically what we have to go on and work on separately. And then we're going to come back, try it again, and say like, "Okay, if this was a real campaign month, leads would be coming in here. This would be what we would be getting from you.
You'd be getting feedback on this. What kind of things are missing here? Okay, what, why was this bumpy? Why was this hard to pull together?" Mm-hmm. "Why was this feeling like a lot of extra work?" I can see that too, and especially from the leadership perspective too, of just like, look, this is going to take a while until we ba- basically find this rhythm where we have something that's race worthy.
Mm-hmm. And you can do it hypothetically even for three months, or you could do it for real for three months and make changes to a live campaign and the like.
I could definitely see that working too, of just like focus on the result instead of focusing on all these inputs.
Dave Dougherty: Mm-hmm. '
Alex Pokorny: Cause the inputs might get cut at some point too.
You provide some report, but we found out that seemed like a great idea, and the moment you met with our sales, they're like, "Yep, we never actually used that. That was actually just a waste. Moving on." Mm-hmm. "Month two we won't do it again."
Dave Dougherty: Yeah, you know how you just- That changes the business
spent $30,000 on campaign assets? Yeah, we only launched about 1,800 of them, so thanks.
Alex Pokorny: Yeah. I mean, stuff like that happens. And like, you know, oh yeah, a creative agency, they sent this to the five zip files, and we were like, "Wait, what?" They're like, "Oh yeah, they created a thousand cuts of the same image." Okay.
Our budget- Right ... is $500 a month. Oh. So you're going to run like two of them? Yeah, like maybe.
Dave Dougherty: Yeah, I don't know why I'm so- But this is what you asked for. This is what the scope of work was.
Alex Pokorny: Oh,
Dave Dougherty: yes. Contractually I lived up to exactly what you said. Yeah.
Alex Pokorny: So it spent a long time resizing.
Dave Dougherty: Yeah.
Alex Pokorny: Oh.
Dave Dougherty: Yeah, that poor schmuck.
Alex Pokorny: Being billed at a rate that they will never take home.
AI Productivity Burnout
Dave Dougherty: Well, and that's- Yeah, I think, I mean, we've, we've tackled this topic a number of ways where, okay, we're operating under an assumption that any increase in productivity is good.
Alex Pokorny: Yeah.
Dave Dougherty: But then you have a lot of people reporting complete burnout because I have so many outputs now that I'm not- Yeah
understanding what's going on. I was talking to a good friend of mine, and they were talking about how in their new role they have to... They're in charge of content, but the expectation from the company leadership is that because of AI, you can now write three blogs a week, plus social media, plus promoting- Oh my gosh anything else, plus whatever, whatever, right? And so she's like, "Okay where do I apply AI to this? Is it the writing maybe?" Editing has to go through me
Alex Pokorny: Yeah. We want decent quality.
Dave Dougherty: But now I'm reading-
Alex Pokorny: It doesn't look like
Dave Dougherty: slop ... a bunch of AI blogs for my job. That's not why I got
Alex Pokorny: into this. You're posting them.
Right. That's your job, is just the process piece, upload. You get to click the buttons. Not do the creative, right? Not create content, not create a strategy even, it sounds like. I mean, you're just pumping and dumping at that point. Like, that's- Right,
Dave Dougherty: and at what point is that, are you creating differentiated content?
Probably not.
Alex Pokorny: No.
Dave Dougherty: Eventually, no, maybe. So then when they say, "Hey, why is, why is marketing not working?" because you outsourced everything to AI, and AI's only good at- That wasn't very good stuff ... what has worked in the past. It has nothing to do with novel ideas or creating things. Like, do you think- Yeah
AI would ever recommend anything that Liquid Death has done for their marketing? No. No. It wouldn't even come to its i- you know, idea space. Like-
Alex Pokorny: It's actually built against that. There's a P value, which is a creative lever, and that's always set very low with any of the public models.
And it's really, really hard to get any of the public models these days to change that value- to get it so that it actually starts pumping out creative stuff instead of just in the same vein. I can't tell you how many times I have tried, sorry listeners, this is true though, I have tried using AI to come up with podcast topic ideas. And my gosh, is it awful. I've been doing this for, like, two years, and I've been trying to be like, "Okay, I don't have any ideas.
Like, what, what do you got?" And it's like, oh, those are terrible.
Dave Dougherty: Yeah.
Alex Pokorny: Just dry, terrible comments. And, like, why would you ever want to listen to any of these? Just a terrible list, and it's... And I can't get it to, like, ever create, like, the kind of creative things that we come up with for every episode, that we just come up with from life experience, you know, stuff we're exposed to and everything else.
Like the- Right ... the human aspect of creativity. And man, I, I mean, our transcripts are out there. We've got, what, 60, 70 episodes out there? You think you'd have enough. No. Right. My gosh, no.
Dave Dougherty: Or, yeah I was looking at that on YouTube, where it's like, hey, YouTube Studio now has a- Oh,
Alex Pokorny: yeah ...
Dave Dougherty: a generator.
You should
Alex Pokorny: create a topic on this.
Dave Dougherty: Yeah. And it's like, okay, well, you know, do I want to do this? Do I not? Da, da, da, right?
Content Mills Kill Trust
Dave Dougherty: But I mean, this, this again comes back to productivity for what? So, like, great- Right ... I can pump out a bunch of social media posts, but why? Like-
Alex Pokorny: Right ...
Dave Dougherty: I know I need to adapt into doing more video on Instagram and TikTok, and that's the way of the future, whatever, but then I look at, like, what video has done to LinkedIn, and I'm like...
The biggest problem for me with short videos is it's a whole bunch of people assuming that I give a crap what they think about me And I know I'm guilty of that with this whole podcast medium too, but like, I just don't. S- you know, I'm sorry, like, I'm kind of in a mood today. So like, I'm sorry, but if you're just out of college, don't tell me operations advice.
You haven't lived long enough to give me operations advice. I'm sorry I can't prevent you from doing that, and that's fine. Maybe you have a good idea in there. I'm open if th- if there is a good idea, I'm open to hearing good ideas, but at the same time, you got a lot of hills to climb before there's any of that kind of trust there, right?
Um-
Alex Pokorny: And if you're pumping out content like crazy, you erode whatever trust you've already built with your reputation. Right. And that's, I think, something that companies don't think about with the whole AI content mill. You know, mass con- content generation thing is, you're going to pump out how many LinkedIn posts per week, and your idea is that the more output is better, and who's actually liking any of these?
Your own employees, really. If you look at the data, it's almost always your own employees and nobody else. Right. Or some random people that who have absolutely not inside your customer set.
Dave Dougherty: The people who- That's it ... want to be a vendor to you are liking your stuff too.
Alex Pokorny: Oh, yeah.
Dave Dougherty: It's not,
Alex Pokorny: yeah. Exactly. Future employee kind of thing, or people who just randomly like random stuff, and it doesn't- Right
mean they're going to buy.
Bad KPIs and MarTech Waste
Alex Pokorny: No, the metrics, I mean, so many of that stuff gets gamified. I mean, you talk about like the, the over-financialization of business. You hit it nail on head, especially with once you start looking at output driven roles. That's bad leadership. It's not bad roles. Mm-hmm.
It's not bad people. It's bad leadership- Right ... because that is a terrible KPI.
It's not posts per week.
Dave Dougherty: Right.
Alex Pokorny: Come on.
Dave Dougherty: But if you could have that conversation of, "Hey, we know that if anybody reads three of our blogs per week, or interacts with 20 pieces of content, they are six times more likely to purchase something," then great.
Alex Pokorny: Yeah. That's
Dave Dougherty: at least strategic.
Alex Pokorny: The main base intent based stuff, you can absolutely pull that if you use the tool set to pull that stuff. You totally can.
Dave Dougherty: Right. You could. How many companies- Yeah ... in your experience ha- are doing anywhere near that level of data work?
Alex Pokorny: One that I've ever met, or two. That's it.
And I've met, like, hundreds.
Dave Dougherty: Right.
Alex Pokorny: You know?
Dave Dougherty: Even the people who have the budget for all the tools to be able to tell you that, they just like to pay for all the tools.
Alex Pokorny: Oh,
Dave Dougherty: yeah. You know? They...
Alex Pokorny: Oh my gosh, yeah. The MarTech stack of, like, the graveyard. There's so many belt payments being made. Like, it, it's crazy.
Because also, like, that kind of intent data is super valuable and actionable, but that's the key piece. Action. Mm-hmm. Who is going to care enough to take the action, and are they going to do it consistently month after month? And if the answer is no, you probably don't need the tool. Right. I mean, sadly, that's an internal process problem or an internal hiring and leadership problem.
Like- Right ... sorry, you need a better strategy first. Then you can And you know
Dave Dougherty: who doesn't want to hear their leadership is bad?
Alex Pokorny: Yeah.
Dave Dougherty: Leaders.
Alex Pokorny: Leaders.
AI as Creativity Scaffold
Alex Pokorny: Okay, on a slightly different note, I'm going to throw this in at the end. Just to brighten the mood a little bit. Uh, this is a very random story, but I was helping my daughter shop for a birthday gift for one of her friends. Mm-hmm. And we came past a K-pop Demon Hunters whole slew of toys. Right. And one of them was a guided journal, and I flipped through this thing while she was debating Lilo & Stitch versus whatever other toy it was.
I mean- ... different licensed IPs, it's one versus- Right ... another, let's be honest. But I was flipping through this thing and it's actually really good for a guided journal. I, I like journaling personally, so I was like, "This is kind of interesting, actually," because it was like, uh, the Demon Hunters, after they play a show they like to relax.
What's six, six things you like to do to relax?
I was like, "Okay, I could probably name three, but I don't know if I could name six." That's actually a decent writing prompt. Or they like a lot- a bunch of different Korean foods. Circle the ones that you think you'd like, and stuff like that. Right.
And it was like... And they're short, little, tiny journals. I mean, you can probably fit 100 words on the little, tiny piece of paper that they give you, but decent prompts. Like, actually- Right ... pretty decent. And I have seen half the movie, so I don't even really know the, the show that well, but I do know Firefly, Star Wars, Star Trek, those one sci-fi stuff, definitely my jam.
Dave Dougherty: Mm-hmm.
Alex Pokorny: So I asked Claude, I was like, "Hey, I came across this toy. I think it was kind of cool. Create for me a bunch of a journal-based kind of writing prompt, guided journal thing based on these things, around the idea of, like, new AI ideas and concepts, kind of pushing towards that area."
And it knocked it out, and it was great.
I mean, it even created, like, a little, like, cover art for the thing. And going through them, some of them were awkward and stilted and didn't quite work, but the majority did. I mean, you can kind of work around it. And one was kind of cool. Like, it got to a point of, like, what would be, like, some things that you'd create, you know, just five minutes, no bad answer, make a list.
Dave Dougherty: Right.
Alex Pokorny: So it kind of came up with different ideas and stuff like that. So just- The creativity was not the system- ... or the AI tool. Yeah. The creativity was me. This was a guided way to basically get that creativity out of me into a journal of something that I could use. Back to AI again, kind of a little bit of a loop there, talking about human loop.
I think that's kind of where some of the key is, is basically is still the human creativity. Maybe it can help structure it. Maybe it can help prompt you. I've talked before about a, a book project. A lot of that was helpful because the AI tool kept prompting me, "Okay, what's next?" You know, "Do you want to work on the chap- next chapter now?"
And easy to say yes to that versus just having a blank Word doc that just ends when you s- the end.
Um, I don't know. It was cool. I created a little random game. I don't know, fun stuff.
Dave Dougherty: Right.
Alex Pokorny: Try something fun.
Dave Dougherty: The, that to me, I mean, you talk about the MarTech graveyard and, and all of that.
Throughout my entire career thus far, it has always been, how do I free up- menial, stupid tasks, like templated tasks.
Alex Pokorny: Yeah.
Dave Dougherty: So that I can be more strategic, more creative.
Alex Pokorny: Yeah. Creativity is it, man.
Dave Dougherty: That, that is what sells products. Strategy is really creativity
Alex Pokorny: at a core. Yeah. Right.
Dave Dougherty: Yeah. You don't get brand differentiation without creativity.
You don't get brand differentiation- Sure ... without a unique voice and understanding who you're for, who you're not for, how you're going to talk to them, right?
Alex Pokorny: Like- It's all the way down. Like, you don't get a better click-through rate, you don't get a better conversion rate, you don't get more people stopping at your booth without creativity.
Right. Like, all of those.
Dave Dougherty: Right.
Alex Pokorny: It all is the same answer.
Dave Dougherty: And one of the main problems of having a bunch of engineers and a bunch of coders putting together things to do work, and to do creative work, is the fact that they see that creativity is a problem to solve. Creativity is not a problem. No. It is the method.
Alex Pokorny: Yeah. Um- No, that's a really good point, and maybe that's really kind of where these, some of these agentic tools and all the rest are kind of failing because they do give really generic... Like I had one create a PowerPoint recently and ah, man, if I was an entry-level employee, solid.
Dave Dougherty: Right.
Alex Pokorny: For me, not good.
Right. If you compared it to my prior decks, and I've even taken classes on creating decks and stuff, got into
Dave Dougherty: it. Of course you have. You're the only person I know that would have said that.
Alex Pokorny: It was required by a job. I, I, I kid you not, everyone at the company in a certain role all had to take the same class.
Oh my God. And it was about presentation and communication and clarity, and it got down to basically creating really good decks.
Dave Dougherty: Wow.
Alex Pokorny: But it got down to the point of basically it's like, it's kind of like billboards. You can present very little information per slide, and you need to kind of keep it moving, and you need to keep the story very strong, and that's-
Dave Dougherty: Right
Alex Pokorny: it, really. Create the argument, nail that argument to the wall with short, very clear aspects. And you can tear apart a graph, I always do, so it's a build, so it's three different slides. So you kind of talk about each element as it comes together. You don't just throw up a giant graphic and try to help people's eyes kind of go everywhere.
Like that kind of stuff. And this thing, man, it just threw out like... I was like, "Okay, give me an executive summary." And it was an executive summary, I swear, it put 350 words inside a slide. And I was like, "No."
Dave Dougherty: Right. That's
Alex Pokorny: not a summary. That's a mini book. Uh, that, that won't work. Right. That just doesn't work.
And that's the thing is it solved the problem. It created the deck. Technically yes. It discussed the notes that I put in.
But man, it was not a good one, and it was not a good enough pervas- persuasive argument, which does crea- take a little creativity knowing your audience, and nailing that audience who this is a less technical audience with a technical topic, which means I got to cut down the word count.
I need to make the argument stronger and clearer. Like-
Dave Dougherty: I mean, that's... Okay, since you're a sci-fi nerd, I was thinking- Yeah ... the other day, assuming AI continues to go the way that people think it's going to go, and then all right- Yes ... it's not enough just to have AI agents or chatbots, but now we have to bring the AI into the physical space.
Alex Pokorny: Yes.
Dave Dougherty: Okay. What do you think the adoption rates would be if AI came out like a C-3PO?
Alex Pokorny: It depends on usability. There's one that just... So robotics and soft robotics and stuff I've started following really closely because it is- Okay ... advancing super fast due to AI.
There's a robot that just came out recently that had a 99.7% success rate with folding random clothes that were basically thrown at it, except it took two minutes and 51 seconds per article of clothing.
So awesome. It's, uh, laundry robots have been a big idea for a long time. It's very, very difficult because of randomness. So-
Dave Dougherty: Right ...
Alex Pokorny: how do you have a random, you know- Big
Dave Dougherty: shirts, little shirts, pink shirts, brown shirts.
Alex Pokorny: Oh. I mean, just a tumbled mass that comes out of a dryer, and now fold it all perfectly.
That's a lot of different objects to recognize and things to manipulate and all those, that kind of stuff. So it's, it's, it's difficult, and I get that. There are always... I think it's a great example though, because it, internal into a house, if you look at cooking methods across the last 100 years, you'll see the rapid change of appliances, and how- Mm-hmm
those appliances have changed the way and the amount of time that we spend cooking. Microwaves are amazing in terms of that. The time savings from that versus maintaining a fire is astoundingly great. Right. You're not chopping wood anymore. You're zapping the thing for a minute, and you're good to go.
Dave Dougherty: What's a better meal, though? Like,
Alex Pokorny: so I know, I know. It's, again, it's questionable on the end result, but I mean, yeah, you got something. It's, it's technically food. Um- Uh-huh. But looking at all the... I mean, let's include the, you know, better pans and pots and pans and everything else that kind of goes along with that kind of stuff, that adoption curve happens slowly.
I mean- Right ... look at air fryers, which was a big fad, as a mini convection oven. Those have been around for a long time. Insta Pots, that is a pressure cooker, also have been around for a really long time.
To get it to popularize can take 30 years.
Dave Dougherty: Mm-hmm.
Alex Pokorny: And it can take, new brands and tons of brands trying it and failing at it, and tons of cookbooks finally including recipes using the new-fangled device, and all the rest.
I mean- Right ... it's slow. So I mean, look at robotic lawnmowers. They have not replaced regular lawnmowers yet. They're getting
Dave Dougherty: to a point- I just experienced one for the first time the other day, and it was fascinating. It was a lot- I bet ... bigger than I thought it would be. But I was like- Oh, yeah
"Wow, if I had a huge lawn, I think I might go for this," because- Oh,
Alex Pokorny: yeah ...
Dave Dougherty: it's just kind of fun to watch.
Alex Pokorny: Yeah, it's a Roomba, man. Which also vacuum cleaners still around, Roombas and robotic vacuum cleaners also still
Dave Dougherty: around. Mm-hmm. Yeah, I use my stick vacuum way more than my Roomba because it actually cleans.
Alex Pokorny: Yeah. Yeah, that's the problem. There's limitations- Yeah ... to what it can handle, and having a big giant robot that's only purpose is folding clothes-
No. Can't.
Dave Dougherty: Maybe in the industrial scale if you can bring up the speed.
Alex Pokorny: Oh, sure.
Dave Dougherty: But-
Alex Pokorny: Yeah, if you're a dry cleaning business or something, that sounds great.
Dave Dougherty: Right.
Alex Pokorny: In a home-
Dave Dougherty: But- ...
Alex Pokorny: no ...
Dave Dougherty: so you went really useful with this. In my mind
Alex Pokorny: Dude, it's going to be pr- people, let, let's be honest, they're probably going to be dating the robots before that thing starts folding clothes. They're going to be
Dave Dougherty: dating
Alex Pokorny: the robots before they, yeah ... and those ones will probably be, uh, you know, people at homes pretty fast.
Dave Dougherty: But also- ... look at CPO and how he, uh, how just obnoxious he is.
Alex Pokorny: Yeah.
Dave Dougherty: Do you really want the know-it-all, like, "Hey, that's a great point, but if you did it this way, ba da, da, da, da." No. Shut up. Never tell me the odds. Right?
Alex Pokorny: Yeah. The C3PO, I never really saw the, the value. I
Dave Dougherty: Yeah.
Alex Pokorny: Translation, that is awesome, but it seems like it could be handled in better ways, but-
Yeah.
Dave Dougherty: Yeah.
Alex Pokorny: Diplomatic service-
Dave Dougherty: Anyway ...
Alex Pokorny: I don't know.
Dave Dougherty: An interesting discussion. We have to end for, unfortunately for other things. Subscribe, share.
Thank you for sticking around this long. Please tell us what you think. What... Do you think the F1 world or the, uh, Star Wars robot universe is a better proxy for this discussion? Thank you for your time. Appreciate it. Have a wonderful next set of weeks, and we will see you in the next episode of Enterprise Mind.
Take care. Take care.
