Episode 2:
Operative vs. Performative AI with Garin Hobbs of InboxArmy

Scott Cohen (00:03.237)
Hello all, welcome to that Inbox Army podcast. I’m your host, Scott Cohen. Today we’re keeping it in-house for a discussion on AI. Joining me in the guest chair this week is Garen Hobbs, VP of Client Success and Strategy here at Inbox Army. Garen, welcome.

Garin Hobbs (00:18.978)
Thanks, Scott. Always a pleasure to be here.

Scott Cohen (00:21.837)
Absolutely. Before we get into our discussion, for those who are watching or listening to this podcast and don’t know you, why don’t you tell us a little bit about yourself and your background?

Garin Hobbs (00:30.762)
Yeah, that’s a great place, always a great place to start. Hello, good afternoon, good evening everybody, wherever and whenever you might be watching this. My name is Garen Hops. As Scott mentioned, I am the Vice President of Customer Success and Strategy here at Unbox Army. I’ve been in this industry as the Gray Hair will attest. This is now my 25th year, so starting back in 1999 when email marketing was just called email. Yeah, I did, indeed.

Scott Cohen (00:50.717)
Wow, the silver anniversary.

Scott Cohen (00:57.849)
Well, great. Well, Garrett, it seems like really any marketer, anyone can’t enter a room these days or read an article without getting hit over the head with AI. If you ask five different people what AI means for email, you’ll probably get five different answers or five variations or who knows variations on the theme. If I’m going back to my music background, you and I have talked a bit, quite a bit about AI and it seems like there’s a lot of

what I call the big science around AI, but most folks don’t have a full understanding of how AI really functions or will function. I mean, I was having a conversation earlier today even, we’re going, what’s everybody using? And everyone’s like, I don’t really know. So how it’ll function that day-to-day workings of an email or SMS program, I like to say, AI is just not small enough yet for folks to really fully understand. You know, when we talked previously, before we jumped on here, you talked about this concept of

Operative versus performative AI. Let’s start there. What do you mean by operative versus performative AI?

Garin Hobbs (01:59.934)
Yeah, that’s a really great question. So, you know, I am the furthest thing from an expert in AI. Like all of the rest of you, I’m on boarding, I’m learning. The art and science actually seems to be more science than art these days, but that seems to be evolving on a day-to-day basis. And as you mentioned, you know, you ask five different people, you get five different sort of perspectives on it. So what I won’t do today is get yet another perspective. What I’m not gonna tell you is how AI can be applied in your email programs or within your campaign universe.

One thing I am good at, however, is creating working structures, frameworks, and taxonomies that help marketers create greater efficiencies and efficacy. And that’s kind of what I propose to provide today with regards to AI. And I’m going to keep it really simple, a very simple T-chart, right? With two value categories to help you sort of decide where AI can be applied and how it can be applied within your program. The two value categories, you just mentioned them, are going to be operative. So think about putting operative on the left of the T-chart and the performative on the right.

With regards to what we mean by operative and performative, operative, let’s think about those as being those items strictly related to the building, composition, launching, and management of campaigns or the data that drives your messaging. Performative will be used to define those items that strictly drive the performance or otherwise impact the KPIs of your campaigns.

Scott Cohen (03:20.125)
All right, let’s, I mean, I know you don’t want to get too deep and get yourself in trouble, right? But let’s talk a bit more about operative AI. You talked about like more of the day to day, the functionality. What do you put more specifically under the banner of operative AI? And then sort of a sub question of that would be…

Because I like to think that a lot of what AI is, it’s been around for a while, or what we, it’s kind of a banner term, right? An umbrella term, but what’s net new that we should be thinking about, and then what’s sort of just getting more and more enhanced all the time.

Garin Hobbs (03:56.67)
Yeah, so let’s start with the first part of the question, some sort of examples of operative versus performative. Now, it can be really easy to blur the lines between the two and one could make an argument that most any activity could belong in either category or the other. But if we stick to the literal definitions I’ve just given, we can keep that kind of clear line of separation between the two. So in pursuit of your question here, let’s give some examples of each. So.

Operative right the easy one right at the top of the list is automation and to your point automation has been along for around for a long time And it was called just that automation It wasn’t given this fancy moniker of ai but examples of how ai can be used in automation automated email scheduling That’s already done, right? But what might be that new is taking an automatic programmatic view of things like email engagement opens clicks reopens Conversions to understand not just the best day in time again. We’ve had, you know, sort of time optimization

send time optimization around for a while. But, hey, I can take that a little bit more further and make it granular on a recipient by recipient basis. So your campaign newsletter, promotional message, it might go out to the entire list, might not go out to the entire list at the same time. Right, again, it looks at sort of those engagement factors for each individual and might optimize the sending of that email to the point in space and time where it’s likeliest for each individual to open, read, and otherwise engage with that particular email.

list management, even responses to customer inquiries, common customer inquiries, some of the most commonly used strategies, thanks to some of these AI driven tools. Let’s give some examples of automation, right? Extracting large sets of user data. So we can look at those, I can look at those sort of actionable patterns from data sets, right? And all of the data that all of us are forever gathering in all of that huge mountain, there’s maybe 5% of that data that represents an actionable insight.

Scott Cohen (05:28.605)
Thanks.

Garin Hobbs (05:47.286)
but it is a absolute Sisyphean task to have to parse through, hash through all of that data and find that needle in the haystack. And that is a perfect use case to apply practical AI. Looking for that needle in the haystack, what’s gonna move the needle? What are some of those patterns over time, either within a specific individual record or across a very broad audience that’s going to let us know when to send the right message to the right person at the right time to borrow a very overused phrase, right?

Scott Cohen (06:16.525)
Interesting, yeah, you know, you talk about, I mean, you’re kind of getting into the triggers and personalization piece a little bit here as well. You know, I think back to my, you know, five, six years ago in previous lives when we got into Next Best Channel, you know, or at least that’s what the idea behind CDP was, right? At least then, and it’s been a minute before I’ve been in the actual.

space and what the technologies can do now. But there are ESPs out there that have launched their kind of version of AI driven customer journeys, right? The idea being that the AI engine will not just give you those insights that 5%, which I love that term, but that sort of nugget of we collect a lot of data that is, let’s face it, useless. Or not useless, but unnecessary. Let’s put it that way. Maybe there’s some garbage out there.

You know, taking that data and deciding best message, best channel, all that stuff. I have my ideas of how I’d approach this, but how would you recommend approaching this technology?

Garin Hobbs (07:25.822)
Yeah, that’s a great question as well. Look at the points of friction within your customer journey. And then AI, obviously, can be one of those things that get applied to remove that friction. Do folks get stuck in the consideration phase? What are the things that are otherwise preventing more of that clean transition to a conversion? To your point, this could be firing a behavioral trigger at the right time, providing additional, amplifying educational information, helping folks build some of that veritas, some of that confidence that would otherwise

put, allow them to or compelled them to put fingers on mouse, hands on keyboard and complete that call to action. And channel is a big part of that, right? So when we think about audience segmentation, we always, we often think about groups of people, you know, that might receive this particular message with that, but, you know, segmentation to your point also includes things like proper channel. And again, segmentation now with the help of AI can be considered as segments of one, right? So for example, if I’m entering a welcome program,

I might have clicked or engaged with some elements in the email, and that might sort of tell the AI program that maybe the next best message for me is not necessarily the email in the sequence, but a very short alert delivered via SMS that might again sort of push me to that point of conversion. So when we think about how AI can be applied to segments, we want to think about including sort of more broad definitions with regard to segments or how we think about them. Is it…

segments of one, is it segments that identify audience based upon common shared characteristics or is it segments that might move people from one channel to another, whether that be mobile, whether that be social, etc.

Scott Cohen (09:02.045)
I’m always interested in the audience of one, the whole one-to-one, is it scalable? Is it not, well, I don’t think it’s scalable. Certainly not with current technology. I mean, is AI really gonna get us there? Is, how do you even quantify it, right? I mean, is it just overall? I mean, I come from kind of an old school testing control direct.

marketing environment, right? Where you test and then you have an aggregate and then there’s always this, how do you know if a test won when we always say you need enough people for statistical significance, but then we’re trying to get to an audience of one. And if it only works at the individual level, like a truly individual level, do we know, how do we know if it actually worked for, it’s sort of this weird balancing act that I’m.

I don’t even know what the question is here. Let’s be fair, but maybe just, you know, it’s, I always go, I think about the example I heard at Email Insider Summit back in December, where it’s like, hey, we ran 750 versions of this Black Friday message, because AI enabled us to do that and results went up. And it’s like, would results have gone up anyway? Right, like the, I always go back to incremental value. Like where do you see AI?

appeasing to the people like me that probably have a little bit of an old school view of how incrementality comes in.

Garin Hobbs (10:32.886)
Yeah, that’s a great question. And I think probably the best way I could answer that is by correlating an example we’re probably all familiar with. So I don’t know that there’s a person out there who likely doesn’t have an Amazon account or who doesn’t buy things frequently through, through Amazon prime. Think about the way that product recommendations are presented both within the app as well through email themselves. Every single message you get with product recommendations, you in fact are a segment of one, and this is a really, really great example of how AI can be used in the performative sense as well. But think about personalized product.

recommendations. I mean, right now, the state with a lot of sort of, you know, the ESP technology is more from the categorical point of view. People who bought shirts also buy pants, people who buy skirts also buy blouses, for example. But what that doesn’t necessarily do is hone that down to color, size, style, and all of those things based upon availability. Right. Amazon does a really, really great job of that. You know, just because I bought AAA batteries and maybe a 40 inch stovepipe, they somehow know that I also want grits. Right. Not something that I would have consciously thought of.

But they’re like, you probably also want grits. And I’m like, shoot, you’re right, I do want those grits. But that’s a really, really great example of sort of the power of the segment of one. Now, I don’t think anyone can argue with one, sort of the feeling of that personalized experience that we have when we’re within Amazon. You mean, they’re not always on the money, but more often than not, they pretty much are. But think about how successful Amazon has become. Now, there are a lot of different reasons for that, but one of the primary reasons for that is that very curated, that very personalized, that…

individually relevant experience. And that technology doesn’t just have to be reserved for the Amazon juggernauts of the world. It is now something that’s becoming largely democratized and becoming increasingly available, either as standalone technology or as an inherent part of email platforms or messaging or experience platforms out there.

Scott Cohen (12:18.453)
what I would say is walk up to the line of creepy then take one step back, right? Like that’s the ideal, right? Like you don’t wanna say, hey, I noticed that you, you know, said hi to your mom at 1223 today, but you know, it’s, there’s a level where people will appreciate it and then there’s one step over that level. And I think it could, yeah, it could get really interesting. Well, I mean, we…

Garin Hobbs (12:22.819)
Yes, exactly.

Garin Hobbs (12:41.714)
Exactly.

Scott Cohen (12:46.233)
Let’s talk about performative AR, not AR, wow, AI. What would, I mean, like you said, there’s a lot of gray area here, but what would you sort of classify as performative AI?

Garin Hobbs (13:00.982)
Yeah, so I think the things that probably pop to mind are things that maybe have already existed but haven’t necessarily be classified as AI. AI is adding that automation, that elegance and kind of taking it one step further. Things we’re all familiar with, predictive analytics, right? So looking at the likelihood of a subscriber making a particular purchase or unsubscribing and allowing marketers to send those campaigns at that particular moment. So think about things like re-engagement campaigns.

Maybe taking a little bit more precedence over win back campaigns to have more opportunity now to retain the customers before they’ve turned Rather than having to chase them after they’ve already exited AB testing as well. So again something that’s been around for decades at this point, but AI powered AB testing tools I mean, I’m thinking about things like Persado I’m thinking about things like Frazee for example that really helped marketers optimize email subject lines Content and then again the cooperation with the ESPs

this send times to maximize the engagement and conversions. Again, rather than looking at larger segments, so one to some, again, we’re looking at this now from a perspective of one to one. So it’s that segment of one theme coming up again. We talked about the product recommendations, again, probably one of the stronger examples, but even in understanding the customer lifecycle, right? So we can, it’s an exhaustive exercise, as you know, having been on the brand side to look at your customer journeys from a very large scale audience and understand.

Where do most people enter? Where do most people exit? What are the points of friction in between? And how do I otherwise optimize their path from entry to conversion, to retention, to repurchase? And AI is getting one of those things that really can be applied at scale to help us understand the gaps. Are there any particular missing biomarkers or experiences in my path through a customer journey? What are the elements that are most needed within content experiences or emails or messages to get people from point A to point B, to get them from welcome to…

through again that consideration phase up to the point where they’re ready for conversion. They can identify those or analyze, identify, anticipate, and then deliver that messaging or that experience at that point in place and time where it’s again, it’s most likely to lead to a conversion. But however you define your conversion event, or again, that might help increase the engagement value of the content itself. Even if folks don’t purchase from this message, let’s make sure that they open the next one and have that continuity of experience.

Scott Cohen (15:19.109)
Well, it’s the goal setting piece, right? Like the, you know, not every message is designed to sell. You know, if you just bought something, maybe it’s not time to buy, right? Or I think about when I worked at 1-800-CONTACTS and it was the conversion at that point was to get a new prescription in place, right? Cause your prescription, you know, I wear glasses, I, you know, for work, I don’t.

Garin Hobbs (15:21.453)
It is.

Scott Cohen (15:44.077)
really need them that much otherwise. And I never have worn contacts. I don’t really wanna stick things in my eyes. All power to you people who do that. But I learned, and if you wear contacts, you know this for real, but it’s not just the parameters you get a prescription for, it’s the specific product, right? So it was a true single skew business, if you think about it. And…

So it was, okay, now we need to make sure your prescription’s up to date. Well, now we need to personalize and you get put in the bucket to talk about prescriptions. You get put in the bucket for if you’re wearing, and the big challenge they were trying to, we were trying to figure out then was wear patterns, right? If you were supposed, if you wear daily contacts and you’re supposed to throw them away when you wear them after a day. Some people would stretch them out, some people wouldn’t wear them every day, they would wear glasses some days, and you had to figure out a model.

that was, well, what, what fate you talked about customer journey, what phase are they in? You know, how, how close are they to actually needing to buy again? Are they, or did they go somewhere else? We have turned and everything else. So it’s, there’s a lot that I think AI, the people are looking to AI to do, but I, you know, I’m really much more of a brass tacks kind of guy, right? So when we talk about day to day adoption, what

Can people who are listening or watching this now, where can they start today?

Garin Hobbs (17:13.986)
That’s a great question. And that’s one of the reasons I really wanted to produce this framework is one, to give folks an understanding of how and where it can be applied, how to think about AI in terms of its areas of application and the impact in those areas of application. And then I see a lot of folks really scrambling to figure out where AI should be applied or could be applied. And I’m almost encouraging them to flip the script there a little bit. If you’re just figuring out…

you know, where can I use AI? Should I be using AI? It’s almost sort of the solution looking for a problem type of aspect, right? So I would encourage folks to flip that and start with their program itself. And again, just to kind of borrow some language we already use, what are the points of friction in your production process? What is the most resource intensive portion of your campaign production efforts? Where are you spending the most time? Where are you needing the help of others? Where are you getting stuck, right? And then the same thing in the actual customer jury itself.

What are the sticking points so going switching from the operative to performative now? What are the sticking points in? In my email program, what are the friction where people getting hung up? Where are they exiting? How can I retain more people? How can I drive more conversions? So I would say look for the problem areas and then look for a solution to fill it rather than starting with the solution And try to figure out where to apply it It’s it should be practical just like anything else that you know exists within our world if you before not applying it practically

then it becomes more of the nice to have, the need to have. And when you lose the necessity factor of something, certainly the value of it starts to plummet, as well as the practical application.

Scott Cohen (18:45.541)
And I think a good add-on to that would be when you approach it from the solution looking for a problem, you tend to look for band-aids, right? And so you may end up.

Scott Cohen (19:00.061)
having five different solutions that when you look at it holistically might compete with each other. You’d have like if you have five, if you go out and get five models or five AI engines, they’re all, I guarantee you they’re going to learn things just a little bit different.

Garin Hobbs (19:16.237)
Yes.

Scott Cohen (19:17.465)
And then you get that consistency problem. So I love this holistic approach because it’s, these are the problems overall. Can we, I mean, you know, there’s always the proponent of, do you go and get industry best solutions in every channel, for example, right? Like you get the best ESP, you get the best this, or do you go for like the person, the platform that can do everything, right? And there’s arguments to be made for both.

But I think when it comes to AI, the danger is it has a mind of its own in a certain case that gets trained. What are you training it on? Are all the inputs going to be the same? Is it going to be the same garbage in garbage out? Like there’s a lot of things here. So absolutely love that. Let’s let’s talk about upsides. So like a lot of times with any new technology or anything, it’s not just a personal usage thing. It’s getting buy in from leadership, both in concept. There may be some costs involved. I mean,

you know, chat GPT basic level is free, but the premium version has a cost and whenever you have to ask for even $15 a month for something or whatever it might be, finance goes, I don’t know. So, you know, how, what are the upsides that marketers can take to leadership to get by it?

Garin Hobbs (20:31.626)
Yeah, it’s just like anything else when you’re looking to make an important purchase or otherwise acquire new technologies about building that value case. So you start with the practical application. Here’s specifically how I’m going to use it. Here’s specifically what it will address. And here’s specifically what I expect those incomes or that impact to be. But as far as the areas of impact, obviously decreasing costs. So some of these marketing solution engines, they make it or AI solution engines, they make it a lot easier to design campaigns, lowering the amount of money spent on…

on creative, on design, on layout, on iterative rounds of testing, or even on advertising, right? According to the 2022 Statista report, 51% of respondents saw the cost of their firm’s service operations reduced by more than 20% after incorporating AI into email marketing. That makes sense, right? One of the other big sort of upsides is it really frees us up as marketers from the day-to-day blocking and tackling that takes up 90% of our regular day, and otherwise sort of…

prevents us or limits us from reaching and working on some of the more transformative aspects of our campaign strategy. So you’re getting a bit of double value there with AI. Not only are you getting more efficiency in the production, but I’m getting a lot more value out of my program because my application of AI is freeing me up from these repetitive mundane tasks and allowing me to focus on things that only I, as a thinking experience marketer, can truly address, which obviously is going to advance the state and value of the email program overall.

And all of those things add up to faster time to value, right? The faster we can get to optimization, the faster we can get close to perfection, the faster we can get things out the door. Obviously, the more opportunity we have to earn. You take a look at your email program. I would imagine that if you did some very quick and simple math, you would say, look, for every email I send, I earn on average of X dollars. So it only makes more sense, or it only makes sense that if I could send Y number of additional campaigns, I can then make X more additional dollars.

So it is a bit more of a force multiplier in that effect. So putting those things together should build a pretty easy value case for approval.

Scott Cohen (22:34.401)
All right, flip side, there’s pros and cons. What are the drawbacks to AI that we marketers should be aware of?

Garin Hobbs (22:41.326)
Global domination by computers. No, I’m teasing. I mean, everyone always has the whole kind of Skynet fear in the back of their heads. And I’m sure we’re very… Yeah. Nothing’s perfect. That’s one of the downsides. Nothing is perfect. As I mentioned, there’s a state of science and art for everything that we do. And again, I think AI is now at the science where it’s more about the development, the understanding of what it can do.

Scott Cohen (22:48.581)
I for one welcome our robot overlords, you know?

Garin Hobbs (23:08.278)
the artist more in the application and how it can be combined with other technologies, again, to serve as that true force multiplier. So it’s not perfect. And it certainly isn’t a Ronco oven where you just sort of set it and forget it. Use of AI is still going to require your review and oversight. It’s one of those trust, but verify situations. Right. And here is one that maybe does seem a little sort of Skynetti, but that’s, you know, too much reliance on AI can potentially degrade a marketer’s own expertise in certain areas or disciplines over time.

It’s one of those sort of muscle memory type of things. The more we do something, the more we become adept at it, the more skilled we become in it, the more understanding we become of that particular task or effort. So if we’re sort of relegating all of those things for AI to take care of, there is a risk of our own sort of skill set becoming slightly degraded or becoming, you know, over time narrowed on those things that we do spend our time doing.

Let’s see, a big use of AI today is NLP, sort of natural language processing. Again, to give those examples of things like Phrasee, Persato, even Moveable Ink to a great extent is using a lot of this natural language processing. Now the EQ on some of this automatically or programmatically produced content can be a little low. Be the emotional quotient, right? I mean, thinking of sort of the, a bit of the carrot and stick, there’s always the intrinsic.

motivations that compel people to do the things that they do. And I have yet to see a very, very deep understanding or representation of that in AI. What I mean by that is, you know, we often want to talk about the benefits, but we also talk about doing something. But we also talk about the risks of not doing something to make sure we kind of drive, or capture both of those emotive drivers. And I haven’t seen a whole lot of really great EQ built into those things to really understand sort of what is going to move the needle from one campaign to the next.

It’s coming and I’m seeing it evolving, but we’re not quite at the point where we can put full reliance upon it. And then lastly, kind of in the same theme, there’s no programmatic replacement for a marketer’s own mind, knowledge and experience, right? I mean, we know from a marketer’s point of view, we know what we’re selling, we know why it’s important, why it’s different, why that should matter to the customer and what the benefit to them will be. And those are certainly things that we can program into AI.

Scott Cohen (25:09.806)
Yeah.

Garin Hobbs (25:25.974)
but the extensibility of that currently is rather limited, especially with how quickly business has evolved, especially with how quickly the market and the market conditions and the feelings, or let’s call it the dispositions of the audience can change over time as well. Marketers, as human beings, we’re connected to everything going on around us from an environmental and atmospheric perspective, and that’s a very human understanding that we bring into our work every single day.

and the current state of AI, I just don’t, with regard to how it’s being applied to email marketing or marketing at large in any way, I don’t think it’s at that state where it’s taking in just this, all of this macro information and then boiling that down to today’s understanding of the human condition and what’s gonna drive best in my marketing programs.

Scott Cohen (26:10.893)
Yeah, if your corporate AI engine passes a Turing test, I think we’re in trouble. So, you know, it’s… But you mentioned Ronco. I think we’re showing our age there with that reference for sure. The kids today don’t… Yeah, I know. I’m just saying like Ronco. Wow, set it and forget it. Yeah, the kids today don’t know. Look it up. If you don’t know what Ronco is, it’s probably on YouTube somewhere. Yeah.

Garin Hobbs (26:17.422)
Hahaha

Garin Hobbs (26:27.427)
The gray hair shows my itch daily.

Garin Hobbs (26:36.61)
Google it.

Scott Cohen (26:39.613)
All right, well, you touched on, you know, time savings, cost savings as a benefit. So final question, I’m going to get a bit provocative here. And I saw an article, I think it was in a book that, you know, underreported quote from Sam Altman, CEO of OpenAI, said that in about five years, he expects AGI, artificial general intelligence to become a thing. And ultimately from a marketing perspective, he thinks it will be able to do.

90, his words, 95% of what marketers use agency strategies and creative professionals do today. I don’t think that’s gonna happen in five years. And to your point, I, you know, not to put words in your mouth, but that the human element will always be required, but five years isn’t a long time. What are your thoughts on what Altman’s saying?

Garin Hobbs (27:31.466)
As a 25 year veteran of this industry, there’s always a sky is falling, impending doom. How many times have you seen, you know, is 2020X the year of the death of email, for example? You know, I’ve seen many, many other endless predictions, but, you know, none of them have really come true. Email is still very, very strong, you know, decades and decades later. And I think it’s gonna take those human marketers driving it.

Scott Cohen (27:45.553)
Hehehehe

Garin Hobbs (28:00.818)
even far into the future for all the reasons that I just outlined. I know the speed of AI development is accelerating daily. And we may at one point get to the point where it can process everything that’s going on around us and build that understanding into email. But again, to be able to balance things like data, emotion, temporal elements again, such as conditions, feelings, time.

I just don’t see, and again, this could be my own sort of intellectual limitation, but I just don’t see AI getting to the point where it’s thinking, behaving, reacting, responding, and speaking in the same way that you and I do on a day-to-day basis. It can understand what you might want. It can understand what you might be likely to buy. But in terms of positioning that to you in the most effective manner.

I still think there’s a lot of ground to be met up before we get to that point. Maybe it’ll happen in five years. I remain skeptical personally.

Scott Cohen (29:01.877)
Yeah, I think whenever I think about AI taking over, let’s keep it simple here, taking over or taking over the processes. I think about there’s going to, you know, I think for the small businesses out there, there’s kind of a level, there’s kind of like acceptable level of mediocrity, right? And that might get consistent, like a consistent level of, okay, I mean, I always say, why does spam work?

Why is spam still around? Because it works, right? Because you only need one sucker to make spam work. You know, and there’s basic level, I mean, heck, plain text emails can’t perform really well, right? So it’s not like we’re asking AI to do a ton of stuff in some cases. But there’s a, to your point, there’s a level of what’s good enough because there is value just being there. And then what gets you over the hump.

to be great. And can AI get you to that basic level by itself? Maybe. I’m with you, I’m not sure five years is enough time. Now, I joined the industry in 2009 and literally my first month, email’s dead. Like going to your point, like email’s dead. I’m going, okay. Then why did two years later, I think it was 2013, 2013 when Facebook tried to do email themselves, right? And that lasted.

I don’t know, about as long as Google Wave did. If that. Who remembers? Callback, who remembers Google Wave? Or Clubhouse. Or there’s, you see these things rise and fall. It’s like, why is 90s style back? I already lived that decade.

Garin Hobbs (30:35.726)
Yeah.

Garin Hobbs (30:47.954)
high-waisted genes should have stayed in extinction.

Scott Cohen (30:50.137)
Yeah.

But I think there’s gonna be, I think about like Star Trek. Even in Star Trek, the humans were integral, even though all the technology existed. And they could have, and like, there’s a little bit of like, hey, this is all run by AI and everything else and it turned evil. We could get into conspiracy theories here. I think there’s a different, there’s also probably a key difference between

is the technology capable and are the companies ready for it? Right? So there’s like technology capacity versus readiness for it. I mean, you think about, I mean, do I see big old insurance companies, big old banks, big old, you know, all these sort of highly regulated industries, are they going to just go, oh, hey, I can handle our stuff? They’re right. The number of, you know, review cycles for things that were written by people in house are stunning in those industries. Right?

there’s going to be a control element that you can have the best technology in the world, but if nobody uses it, it doesn’t matter. So it’s, yep.

Garin Hobbs (31:58.126)
Yeah, that’s where that trust but verify comes in. You raise an interesting point. And so not to get too philosophical here, but I’m thinking about to the development of the space program, right? When NASA was working through the 50s and 60s to put humans on the moon, one of the biggest limiting factors was having to develop life support systems for the astronauts themselves. They said, look, it would be so much easier if we could simply remove the human element, in this case, the astronaut from the equation.

and simply develop the technology that would go to the moon. And the answer was a resounding no. And the reason why is super interesting. If we don’t send a human to the moon, who cares that we went to the moon in the first place? And it’s kind of the same thing when it comes to email or marketing at large or your life at large with the application of AI is nobody wants to know that they’re dealing with what equates to a rumba at the end of the day, right? You want to know that there is a thinking, feeling, responsive human.

who is not too different from yourself on the other end, marketing to you, understanding who you are, what you want, why it’s important, and when it’s important to you, and responding in like manner, right? Now again, that might be a generational thing, or it might be something that over a very short amount of time we all get used to, or that we sort of forget. I tend to not think so, right? There’s reasons we all live in communities and we seek out each other’s companies is we want to feel that connection. We want to know that we are associated however tightly or however loosely.

with the people around us. And I think that extends to marketing as well. I mean, think about some of the most engaging or satisfying or interesting moments you’ve had as a consumer. I can guarantee you 100% of the time there was a human behind that. And I just don’t think that’s a quality that any of us are really ready to let go. I don’t think the technology is at the state where it’s ready to assume that level of deep sophistication. And so for all those reasons, I would say, I don’t think this is something we need to worry about with regards to our own job security.

you know, at least for, well, beyond retirement age for me, that’s for certain.

Scott Cohen (33:57.805)
I mean, I would say you’re not going to lose your job to AI, but you are going to lose your job to somebody who uses AI if you don’t, right? If you don’t embrace it to a certain extent, five, six years from now, your job could be in jeopardy. Just because as we get more automated, as technology increases and gets… I mean, even the rate of acceleration is up, right? I mean, I saw some article.

Garin Hobbs (34:04.898)
That’s a cool, yeah.

Scott Cohen (34:26.397)
I’m gonna butcher the actual number, but it’s like the technology improvement that we’ve seen or that our kids, like I have a nine year old and they were saying something like, if you have like a 10 year old now, the amount of technology innovation they’ve seen in their lifetime to date by the time they’re 40 will happen every like 11 minutes or something. And I go.

That’s something, that’s a big pill to swallow. You go like, that’s just how fast we’ve come. I mean, my kids are younger than the iPhone. And think about the computing power in the iPhone. You just mentioned the moon landing. I mean, if they had this computing power in the Apollo missions, Lord knows what they could have done. So it’s, but to your point, it doesn’t matter if the humans aren’t there. Now, I will say, if I have to call somebody,

Garin Hobbs (35:12.797)
Yeah.

Scott Cohen (35:20.369)
to get something done that could be done online, it pisses me off.

Garin Hobbs (35:24.145)
Same.

Scott Cohen (35:26.225)
But for the big things, you want people involved, right? Like I don’t think I’d ever go out and buy a house without human involvement. I would not go out and buy a house. Now granted, was able to do that. I wasn’t trying to buy a house during COVID, stuff like that. But, you know, it’s like the big decisions. You’re not just gonna go, I’m just gonna go buy it online, right? Like it’s just, now even then that happens. I mean, I worked at Purple Mattress. We go, who the hell will buy a mattress online? A lot of people buy a mattress online.

You bought a mattress on it. There you go, see? It’s, yeah, exactly. And thank you for that, you know, one little conversion on my chart. You know, it’s, yeah, it’s an interesting world. Well, I think that’s a good place to stop. Garrett, where can people connect with you?

Garin Hobbs (35:55.758)
I bought one from you, Scott. Yeah, I bought one from you.

Garin Hobbs (36:12.346)
As always, folks can find me on LinkedIn, Garen Hobbs. You can connect with me also at Garen at Inbox Army if there’s anything that we can do for you here.

Scott Cohen (36:21.725)
Fantastic, well thanks for joining. Thanks to you, our listeners and watchers, for tuning in. As Garen just mentioned, if you’d like to learn more about Inbox Army, check us out, inboxarmy.com. Until next time, be safe and be well.

Garin Hobbs (36:34.926)
Cheers, all.