Ep 49: The Interplay of AI, Personalization, and Human Expertise in Business
Watch the YouTube video version above or listen to the podcast below!
Ep 49: The Interplay of AI, Personalization, and Human Expertise in Business Podcast and Video Transcript
[Disclaimer: This transcription was written by AI using a tool called Descript, and has not been edited for content.]
Dave Dougherty: [00:00:00] All right, and welcome to episode 49 of Enterprise Minds. We're gonna have to figure out what to do for 50, because that's, that's pretty big milestone. But yeah, we've got a couple of exciting exciting bits before we jump in.
Upcoming Speaking Engagements
Dave Dougherty: Ruthi is back from the Adobe Summit not just in the crowd, but also on the stage presenting out what she's been doing and kicking ass, taking names, and then, on the self-promotion side, I got a couple of speaking gigs coming up too. If you're in the Twin Cities area there's one at the University of St. Thomas coming up. So go over to LinkedIn. In the description you'll [00:01:00] find links to that and get more information as I'm able to say more on that.
So Alex, you owe us a speaking gig. Yeah, apparently I need todo list.
Alex Pokorny: I got my combine right here. I'll add it. Don't worry.
Dave Dougherty: So, yeah, with that, Alex did in fact bring topics because Ruthi and I were too busy with the speaking case to come up with any,
Alex Pokorny: Somebody has to do this, man.
Dave Dougherty: So why don't you why don't you queue up. What we're gonna talk about today.
Humanity's Value Proposition in AI
Alex Pokorny: Sure. It's come up a number of times over the last month and I keep seeing different kind of examples of it.
And I've been trying to like get a kind of my hands around it. I like analogies, but deem work well enough. And really what it is, is basically is humanity has a value prop. Basically there's a number of different situations. One of 'em was basically as we were talking [00:02:00] about AI agents and basically.
Where they're capable and where they're not capable. And really the limits on that would be greater context. So their knowledge of your life their knowledge of the situation that you might be asking about. For instance, I'm looking for a family vacation. I don't care where is very different than this is the flight book, me three tickets, two adults, two kids.
Here's all the information. Right. That's a very different ask. Like one of 'em requires much greater context and the other one doesn't. Even to the point of like knowing what kind of budgets I, I would typically have instead of just throwing out, you know, Fiji as well as someplace 10 minutes drive from me or something.
You know, it's like there's context here that's missing right. Which is very different than like a human conversation to cut you off
Dave Dougherty: because, or God forbid, you have a VPN and it thinks you're in like Istanbul and you do the, you know, restaurants near me and you're like, what? What? No, no. [00:03:00]
Alex Pokorny: Yeah, I get that pretty frequently.
I get run a V VPN on my phone and. It's never, never, never quite useful. No.
E-commerce and AI Limitations
Alex Pokorny: And then the other one was more of a, a business side of things of looking at e-commerce and complex B2B buys, as well as even individual consumer buys. I've seen recently, basically, of individuals who are buying particular products.
They bought them online, they sold themself basically online. The problem is that product actually requires a few other products to make it work. So now they're annoyed they didn't know the other things. I mean, yeah, you could put recommended products, you could try to like help some of the body copy or something like that, but honestly, a real quick conversation with a salesperson would've actually covered all that, which I realize is not an ideal scenario.
So it's like there's this weird kind of push and pull there. Like you don't want to have the sales call however. You don't know whether or not there's a level of complexity there or not, because you are new. [00:04:00] You're buying the product, you're the consumer, you're the customer. You don't know all the things that the, the business would know, right?
But at the same time, a human conversation would've covered that in five minutes or said, no, you don't actually need this thing that you're talking about. You need a cheaper version. Nobody needs to sell that one. I mean, you're, you're talking something else, or this product won't even work for you because you think it does based on these things.
But because of these other things that I know about from my expertise that I'll ask you about, oh no, you don't have those things, then you actually need something completely different. That is such a complex aspect. Mm-hmm. Of e-commerce kind of falls apart. So does AI play a part there? Is it just a failure of e-commerce needs to be better?
Or how do you, how do you meet in the middle between. The massive scale that is digital and the very personable, you know, context driven relationships that we have as humans. Mm-hmm. And [00:05:00] I'm just, I keep struggling with that and I just keep seeing examples of, we go really, really hard into the digital side.
Well, that's great, but that's really my topic. Like we just never seem to get the full context. And then we go too far on the humanity side and it's like, well that doesn't scale like. And it's not available 24 7, like, I don't know. So that's kinda where I've been. I'm struggling. Hmm.
Personalization Challenges
Dave Dougherty: Ruthie, do you have a hot take on this or, well, I don't know if it's a hot take, but the analogy that just popped in my mind towards the end there was this idea of, okay, so you're going trou.
Ruthi Corcoran: And I think of there being, I. Sort of three categories of ways that you can travel to a completely new place. The first one is, you know, you do your research ahead of time and you sort of book the things that you wanna do. Are they more likely to be the touristy thing? Sure. But you can do a lot of online research and you can sort of figure it out yourself.
That's sort of on, on one end. On the other end, you have, I know, a guy [00:06:00] who lives there. Who's gonna show me around, who's gonna show me the local restaurants, who's gonna show me the spots and mm-hmm. I'm really gonna get to know the place that I'm visiting. And then in the middle you have something like, you show up at the hotel and they have a concierge service and the concierge is there to, to guide you to different things.
But of course they have their own set of incentives and they're like less likely to show you the hole in the wall place where you can get that really good, you know, rice on muscles or whatever it is. And what I think about is. The, the co, the issue that you're describing is. Almost like you can, the first one's easy, right?
You can make that one better. You can use gen AI to make that concierge experience happen better wherever, right? Maybe you can have it in person, but now you can have it digitally, right? Marriott's doing this very well, and they're sort of bringing to light or bringing to digital a whole bunch of different exp.
Experiences that are more and more personalized and to the extent that they can, they're using gen AI to improve that. I think, Alex, what you're [00:07:00] describing is almost this other one, which is you're not going to easily replace that, you know, a guy experience because they're gonna. Give you something much richer, much deeper that I think there's, we've got a bit of a, we've got a gap before any sort of personalization in gen AI gets there.
And I toss in this word personalization because I think that's, that's where I. Dave mentioned I was just at Adobe Summit and that is one of the buzzwords mm-hmm. Of Adobe Summit is personalization. Personalization. Everybody's tried to crack that nut and everybody has been trying to crack that for the last 10 years because they know that the more focused you can be in your, in your messaging, your, your.
Targeting, the more effective you're gonna be. I think where personalization falls short is to, to what you just said, Alex, the amount of context you need to create a personalized experience that is helpful and not just insert your name here and throw in a couple of things. That's, that's the, that's the piece that's missing.[00:08:00]
It's how do you provide a helpful experience. You're not wasting somebody's time to, to give them what it is they're looking for. And it almost falls in an uncanny valley all the, a lot of the time where it's, it's personalized. They're really trying hard, but it's not actually helpful. Like you, you, you can almost see it trying.
Alex Pokorny: Mm-hmm. Yeah. I think, I mean, personalization, in other words is also know your customer. Like why are telemarketing calls annoying? Because I don't want what they're giving me. Yeah, because they don't know who I am. They're not giving me something that I, right. I want. Right. So it's about what they want. So,
Dave Dougherty: yeah,
Alex Pokorny: exactly.
It's about what they want. Exactly. So you have a business, you know, side looking for sales, but the customer side is wondering, why on earth are you bothering me? Like, but if you know your customer really well, there's trust there. There's a relationship with her, but also. It's helpful. And I, Ruthi I think that's [00:09:00] just, that's nail on the head.
I mean, that's the thing is like, without that context, it's not helpful with the context. It's helpful.
Dave Dougherty: Right.
The Role of Data in Personalization
Dave Dougherty: And I think too, I mean we've talked about this on, on a lot of other episodes where AI can be very good and very helpful for personalization if you're willing to give up that level of data. You know, and that's what's made Google and Google search so effective for so many years is because you just stay logged into your Google account, right?
While you search, while you watch YouTube, while you get your calendar together, you know, it knows a lot about you. Whereas, you know, new AI startups like, well, I don't know. Do I trust them with that? I don't know that I want to trust them with that, you know? 'cause what happens when you get bought out or what happens when your product goes bust and you're no longer [00:10:00] supporting it?
But I still have the, you know, hardware device that was the, the key to using your, your AI thing, right? Like there's some downstream effects you have to put together. And then the good, you know, the, the data that goes into it is really. The outputs are only gonna be as good as that. Right. So that, you know, with your, your, you know, B2B sort of system solution kind of thing.
Yeah. I mean, you could put that in the, in the copy. The AI would know how to do that, but then it also has to run against whatever database that, hey, these products are related to these other ones, you know, and you can do that, but it's. That technology has been targeted to B2C stuff for so long that it's just not even in the wheelhouse of a lot of B2B marketers right now.
Because it's a different way of thinking of, you know, let's take on the entire experience, you know? 'cause I was thinking about that the other day. I was using the target drive up and, you know, it's like, why does. You know, [00:11:00] target and Amazon and Walmart, all, all in the same sort of area where they know a lot about whatever the purchaser is because they know, okay, they like these drinks, they're Coke versus Pepsi.
They, you know, the less fat cream cheese, but that doesn't make sense 'cause they're also buying donuts. Like, you know, like it's. You can personalize around that stuff. And it's fairly easy for like drive up orders where it's okay, I'm just gonna get the stuff that I buy all the time anyway. And then for the, hey, surprise me kind of things, I'll go into store, you know?
But again, I'm thinking about Depends on what industry you're in too, right? I mean that, it's, that's why marketers are always yelled at for saying, it depends 'cause it really does.
Ruthi Corcoran: Thinking about your, your comment about the complexity of the B2B world and something Alex had said about how a lot of [00:12:00] these personalized or recommended for you type of experiences that are becoming more and more powered by gen ai. A lot of where they fall short is you, you're, you might as a customer be missing something in that picture.
You know, it's saying, Hey, recommended with these products. But we've all worked in, in the B2B space where you're not just buying a single product. You may be buying into a system, or you may be buying a product that works with several other products and in order for it to work well, I think you, you need the collection.
And the, some of the thoughts that come to mind when you're, when I'm thinking about this is digital does play a huge role in your ability to sort of explore and understand and, and learn on your own, whether it's via videos or looking at webpages, et cetera.
Future of B2B Sales and AI
Ruthi Corcoran: But then the value of the in-person sales com conversation changes to an extent.
You've gotta have a much higher value sales conversation or else it's not [00:13:00] worth it. It's not helpful. To the customer if the sales person is simply repeating what they already read online, right. To, to Alex's point, it's providing the additional context of, did you know this about this? Did you know this?
You may not have been thinking about it in this system or getting to know the system. Which then changes the role of the sales person. They've gotta be they've, they're not just repeating back what's on brochures now. They've gotta be the expert on this product set and on potentially what the customer is trying to do with that product set.
Mm-hmm.
Alex Pokorny: That's a really interesting point about the complexity of the purchase or the interaction as I, I, you know, kinda going back to Target example too, like sometimes they make use of that data, but sometimes they really don't.
Dave Dougherty: Right?
Alex Pokorny: I mean, I may have bought one particular product the last time I was here, but that doesn't mean that's a repeat thing that I actually want a second time, like Amazon keeps trying to sell me.
Subscription, [00:14:00] like re like automatic shipping to me various products that are single use products. Like I, I, I'll never use it again. Like it was a one time thing. I used it and it's done. Like, or, you know, there's a quantity in it. Like, oh yeah, every few weeks you could have it again. It's like.
Yeah, I, I don't buy like that. What do you think I'm doing? I don't need that
Dave Dougherty: quantity of super glue,
Alex Pokorny: like Exactly. Just don't need that much or frequency. Yeah. So there, there's like. A large amount of data. I mean, heck, I'm struggling with this at work too, because talking about like average sale volumes and you know, particular industries, what kind of products might they want?
What kind of, you know, cart or basket, are they looking for that sort of thing or you know, what's their common reorder cycle or related products given their. Customer types, industry, science, whole bunch of other factors. You can kind of think about all that stuff and say, oh yeah, if I got a lead, it would be average value of [00:15:00] $187, but that's wrong.
Every single lead is different. I. Some leads, they're worth a million dollars. Some most are worth zero or they're a waste of time. So they're actually negative amounts because they're actually mm-hmm. Holding your team back because you're getting junk. So that's the thing is like you start to throw some of these figures and it starts to fall apart and, and practice in usage.
So that, that's the other piece of like, we collect all this data, but it doesn't necessarily mean that we can take this random person who just showed up and say, I know what you need. I. Hmm. You just don't have that quite yet. Or maybe that's just like, I mean, getting to a well quoted thing from Ruthi of a chat bot instead of a website in the future where maybe that is the purpose with a B B2B site is it starts with questions.
It figures out what you need. It gets you down to where you're going, but also gives you some. [00:16:00] Sense of control or enough options to feel like you've explored all the options and you're correctly, you know, get, get to the bottom of it. Maybe that's it. Maybe it's just the e-commerce functionality for certain complexity just doesn't work.
Dave Dougherty: Ooh,
Alex Pokorny: well, yeah, that's, I can't click on the value. That's where you get
Dave Dougherty: into the whole regulation sets and you know, whatever else like it. Yeah. Sorry, Ruthie, I cut you
Ruthi Corcoran: off. No, you're good. Hot take on the value of a website. I still think that proprietary chat bots. I have some potential. I, I'd put some, I'd put some percentage on that one.
I think that's a, that's a cool idea. And the di the idea that you could be speaking with the AI of a particular company about their products that and that they could tune it and train it to be better and better about answering, answering the run of the mill questions like that seems like there's a lot of potential there.
But. The other piece about the value of a website, and [00:17:00] this was, this was something that one of us speakers mentioned from a company called Grad Deal in one of the sessions, just that the value of the website increases as we have more and more gen AI come online because it becomes the source of truth mm-hmm.
For a given company, like, and I think about. The, the three of us have all encountered or at least heard case studies about counterfeit in in China. As being a particularly difficult issue and how oftentimes customers in those countries wanna buy direct from the company rather than from a distributor or something.
And the reason is because it's really important that they're not getting a counterfeit product. I think of something similar to that in the information world, where the value of going directly to a company goes up because you're, you're getting the, the correct information that company's on the hook for providing you the correct information versus what you might be getting from.
One of the many, many LLMs out there providing information. Sure they have the [00:18:00] source, but you gotta get it linked back somehow
Alex Pokorny: on. Correct. Go ahead, Dave. Well, I'm take this in a tech bend, so go ahead.
Dave Dougherty: Okay. I'm going a different way.
Innovative Solutions and Future Vision
Dave Dougherty: In preparation for one of my upcoming presentations, I've been revisiting a lot of our favorite presentations from Clayton Christensen and Rory Sutherland.
Right. The sort of classic thing that Rory brings up, right, is if if you change your perspective, you might find the solution that is actually the solution, even though it's unexpected, right? So if you wanna make the trains faster, you have to spend 6 billion pounds and a whole lot of time and years to do so.
Or for a fraction of that, you can add free wifi [00:19:00] to the trains and people don't care. 'cause now they're, they can do their work, they can do their homework, they can watch a movie, like whatever it is. It doesn't necessarily matter that they're there faster. So, you know, it's, you have to use the different lenses too.
Like if you get stuck using a particular engineering lens, well then, okay, let's go try to find a more creative thing for that. Or, you know what if we came at it as like an architect instead of a, an engineer? What does that look like? You know, just as like little thought experiments to break you outta your rut.
And I think for me, with the B2B side of things and the struggle for a lot of incumbent B2B organizations, the problem is going to be getting, I. The innate knowledge inside of the organization, onto the website in a way that actually communicates without without feeling fluffy. You know what I [00:20:00] mean?
And that's, that's gonna be really hard, especially, you know, if you have a lot of science involved in what you're selling, like convincing scientists to step outta the lab to do a blog or a podcast or something, that, that can be a really heavy lift, you know? But if it becomes part of the culture, then you know you have a shot.
You know,
Alex Pokorny: I'm gonna give you a vision of the future of the internet. Based on this, I think we're, I think we got it. Okay. Because Dave, you're pointing out the exact problems and Ruthie, you're picking out kind of what the user experience solution is. Mm-hmm. I think in between is basically reasoning agents crawlers with content, and.
Or transcripts or chat bots like fed data sources that chat bots are working off of. Mm-hmm. And you would supply that like a, a closed [00:21:00] captioning transcript like you would on a YouTube video, which Google still sees those basically as images and then polls from the transcript, whole bunch of content, and now it knows how to rank the video.
Without those transcripts, they're, they're pretty lost, to be honest. They're still pretty lost on it. Mm-hmm. Link signals aren't. That useful. So I think that's it. Basically. You have to, I mean, that, that is the, the key piece there, Dave, I mean you hit it perfectly, was how do you extract that information out of that company?
You got a million PDFs, great.
Challenges with Current Web Pages and Patents
Alex Pokorny: You got a ton of web pages that are fluffy, not great. You got a bunch of patents and some references to 'em, not fantastic. What do you need to do to basically get your arms around that? Collect that information, reduce it down to something that's useful, and then respond back with it.
The Role of SEO in Data Parsing
Alex Pokorny: So if I'm interacting with a Gen AI product, let's say it's Gemini, let's say I ask Google about this particular B2B situation or [00:22:00] whatever upcoming conferences, you know, whatever that information that's gonna. It's gonna be from the company, and it might be the new SEO job is much more database like focused, more, more kind of tech.
SEO focused, where you're trying to make sure that all the information is indexable, including FAQs, interviews, you know, maybe even transcripts from customer service calls. I don't know. Everything is basically parsed in some way, and that's your job is to make sure that it's, you know, this crazy PDF is actually could be parsed out so that.
These agents can basically collect that information, reduce it down, and then feed it. And whether or not that's on the company side to do the feeding, or if it's gonna be really heavy lift on the search engine side to scrape through all this stuff, reason through it, filter through it, basically kind of how they are now, but done way better with the reasoning agents.
That would basically be re-looking at this information because you, you'd have to look at it from, [00:23:00] like you said, multiple angles, like different lenses. You'd have to look at a science company with a very specialized bot that understands the scientific implications of these particular patents.
Because a patent describes what's going on, but it doesn't describe the problem that it's solving. You know, the situation. That's much more of a case study sort of thing, right. That's marrying in the marketing copy with the patent, and now you have something that looks like a complete. Helpful solution.
The Problem with Big Data
Dave Dougherty: That's always been the problem with big data. Big data is always looking at the past. It's never telling you what's happening right now. It's always, you know, or what, what could be happening, where you're going. It's only ever what worked previously, which is why I think I. The amount of creative ads that we've seen since the era big data came around has plummeted.
Yeah, because it's no longer fun because it's, you know, oh, we know this segment will react positively to this. If we show, you know, puppies for 2.5 seconds and then we do a jump cut to, you know, a toddler laughing on the driveway, [00:24:00] like, shut up. Just come up with good creative. You know, like some of the stuff you need to tip on its head, you know, I know I've talked about it a lot recently, but, you know, liquid Death, a water company playing off of the fact that, you know, all the branding for water is about life and essence and birds and meadows and, you know, all that kind of stuff.
No, we're gonna flip it. We're gonna do a ton of collaborations and we're gonna do something creative and fun. You know, they're killing it. So, you know, I think the, like ai, if you do that version of the internet, you will get to a very nice law state, right? It'll be nice and average, but then you're gonna have.
Your customers go, yeah, okay. I know what you said, but like, I'm different. My situation's different. Help me. [00:25:00] And you go, yeah, but actually you represent 80% of our customers. This is what you want. No, no, no, no, no, no. I'm special. You know, which has always been the problem.
Alex Pokorny: Yeah. And maybe actually it's gonna get a lot worse because basically now that you're pulling more reasoned, larger data sets.
You're saying, I have a particular problem. Well, here's some things you might never have thought of. And now suddenly the customer base is very different.
Unexpected Customer Bases
Alex Pokorny: One of my family members owns a company. It's long history to it, but dealing with mining equipment and jackhammers and pneumatic tools, right?
Mm-hmm. Standard minors, construction. You know that kind of group, that's their core customers, right? Then they started getting some calls from surgeons, orthopedic surgeons. I don't that if
Dave Dougherty: you're using a jackhammer, I don't want that.
Alex Pokorny: I'm not kidding. It's very hard to move bones during a surgery and there is [00:26:00] a lot of chipping away and.
Work that has to be done with hip plant implants and construction work. And what they were looking for was basically is how can we basically bag this thing so it's hygienic and use it on a human being and what kind of size like compressors would I need to chip away bone? And they're like not, not an expected customer set.
Turns out, still a customer set been going on for a long time now, well over a decade that. Surgeons are buying from them for their pneumatic tools. Are also sold to construction workers. Like that's a really, sometimes the less you know is better. Oh man. There's a whole lot there. You don't wanna
just be thankful that it's a nice surgeon who just, you wake up and, oh look, it's, it's better now. Yep. Yep. You don't wanna know. But this is basically an [00:27:00] individual really branching out. They're not looking at the regular medical device companies, they're, they've looked well beyond that to finally find this supplier that might fit this need.
Right? So it's very inventive, which may be the crux of the problem that basically is in the future. That's the better. You have a reasoning agent, the more options there are. I mean, think of like all the crappy websites that are out there that are, you know, two pages and one of them's a contact us page, and that's it.
Right? The more that can be pulled out of that, well now they're a competitor. And there might be some off the wall competitor for you that you just don't know about yet. So I think as well, those creativity would've to break through all of that or inspire those searches and inspire those people to kinda start from those different starting places, which would allow them to discover these things, right?
Allow them to discover your company.
The Importance of Understanding Pain Points
Ruthi Corcoran: This goes back to something we've talked about in the past, and then also I'll tie it back into an article that Dave recently shared about [00:28:00] knowing pain points is the key, right? Mm-hmm. Those surgeons knew exactly what they were looking for and what wasn't solving their problem.
And therefore they took it upon themselves to go find a different solution. Right? And so better understanding either your market's pain points or maybe it's not your markets, understanding the, the pain points that different customer bases and different marketers are, are facing. Opens up your ability to solve for a, a number of different problems.
Or perhaps it's on the flip. Maybe this the. The increasing ability to get information from anywhere allows more surgeons or insert audience here to go and find novel solutions, so the problems they have so that they can better solve their own solutions. And to your point, Alex, that just increases your competitive set because your customers can go and find novel solutions that aren't yours.
The article that it reminds me about [00:29:00] is this idea that. Information and therefore, you know human's IQ is now less of a value, right? You being intelligent or having a lot of knowledge, like the value of that has gone down because you can have access to information more than you could even with search.
And therefore the important thing now that. Like the currency that matters is agency somebody's ability to just go out and do a thing, to go solve a problem, to just, to just go do it or to run with some, an idea. And that's, there's, there's something to this of that surgeon who reached out to your to your family member to say, Hey, that you guys are solving a problem.
Like it's, it's that that is the thing that's gonna drive going forward.
Dave Dougherty: Well, I think the, I put a larger currency on critical thinking than anything else. You know, like I was talking to a, a lawyer friend and they said that somebody called them and [00:30:00] said why are you guys saying we have to do this in this particular state?
The AI tool is telling me that we don't have to be compliant with that law. It is like trust the lawyers, don't trust the AI bot.
You know, it's like
there are certain things like that where I would just take it for granted that you wouldn't trust something as important as you know how compliant your company is with the laws to a chat bot. I. You know, like that. Yeah, yeah. Anyway, brought the mood down. Sorry.
Go
Alex Pokorny: ahead.
Ruthi Corcoran: Where my mind goes is sort of a couple different spots. The first is I, I think about this, this tension with marketers working [00:31:00] with legal or compliance and it, you know, over the last decade, I think that relationship has shifted a bit where. It was what the lawyer says go, and it's just, this is how it is where it's been shifting more to, it's a bit more subjective.
You as the business are sort of assuming the risk, so to the extent that the chat bot is there to sort of lay out here, here's sort of the, the range of risk that you might be facing, and then you get to assume the risk. There is some value again to the conversation we had earlier though. There's, there's a context and there's a complexity that matters for which having a lawyer might still be helpful.
Like you might still wanna talk to somebody because they might see the nuances, or they might, you know, see how judges might interpret it in a different way than a chatbot at this point in time. The other thought that comes to mind though, is on the other side of that. Where, where critical thinking is not involved [00:32:00] in the digital space.
The Rise of Gen AI in Marketing
Ruthi Corcoran: Gen AI is taking over, I think, much faster than I had originally anticipated. So if your job involves button clicking based on documentation, and that is it, I do not think you'll have a job for very long either, because I. The, the rate of outsourcing that type of activity has increased. But now it's real gen AI agents are doing this sort of activity.
And, and me as a user of a platform would much rather just use a gen AI than work even with a, with an outsourced developer. On doing those sort of button quick activities because A, it happens faster. B, I have more rapid control over the situation. I can say, Hey, set up checkpoints here, check checkpoints here.
And I, I think I was a little bit blown away by this last [00:33:00] week of just how quickly the ability for gen AI to be basically software robots as. Become a reality. Yeah. I was running an experiment with a friend about a month or so ago, and just to see, 'cause it was, you know, the, the deep research stuff had just come out.
Dave Dougherty: Right. And we did basically a, a UX audit of their company's website in about 10 minutes. Instead of the three months, it would be to, you know, get an agency to do it, tell 'em about the project, have 'em go through each of the pages to do a project. You know, and then we threw in some optional extra, like SEO stuff about, tell us about the links that you find on these pages, and how many of 'em are, you know, this, that, or the other.
And I was like, oh, damn, this is actually farther than. I thought, like, is it a perfect [00:34:00] UX research thing? No, it's not. But does it at least give you some quick hit ideas on, okay, here are some areas we can improve. Yeah. Yeah, it does.
Alex Pokorny: Let, let me ask you a question on this one. Does it cheapen the output? Yes.
No. So do you. Yeah. So here, here's your agency. Critical thinking like Crux right now. So if you were doing this manually mm-hmm. You go through every page manually writing down on your word doc, da da da da da. You spend all week and you come up with this thing that you are now probably pretty passionate about and you are very knowledgeable about all the details.
'cause you manually put that thing together. Yeah. Flip that. Ask. Gen AI to do it. It does it. I get this output right away. Do I care about that output the same way I did before? Will that project actually get [00:35:00] fixed? Which one will fix the project or fix the problem faster? So I.
Dave Dougherty: Here's
Alex Pokorny: one. Saves
Dave Dougherty: time, but it cheapens it.
Here's my analogy onto why I say it does cheapen it. If you think back to your agency days, think back to when you had to go to a client presentation with your manager and your manager was talking about all the work you did. And of course they messed up because they said, yeah, we saw this, we saw that we da, da, da.
Yeah. And you're sitting there going, no, I did the work. I know the site. I know what's wrong. Right. There's a level of knowledge that cannot happen unless you work through it. Right. Yes ma'am. Go
Ruthi Corcoran: guys. But guys, and I'm on the client side through and through. I've never been on the agency side. Sure. As a client, [00:36:00] I don't have to trust an agency.
Now I can just go do it myself. And I know my system. I can just go, I can just go fix it. And that the reason I said no was less about that. It was more that, so Adobe just showcased Jen's studio. Mm-hmm. Which is their way of creating a whole bunch of variations on contact. Then you can just then push out to social media, et cetera.
So it's mostly focused on campaign marketing, but the, the transfer over to web experiences is pretty easy. Right. And what they have built in, and Adobe's not the only one with this technology. Got it. A couple other places is you in upload all of your brand documentation. And your brand guidelines, you upload all of your legal and your compliance documentation.
So not only are you having bots take a look at like qa, does it have broken links? Is it like created correctly? You also have scores. How close does this hit your brand? [00:37:00] How? How close does this hit your compliance guidelines? Mm-hmm. And so I think about those things and I go, well, shoot, I'm the client. I already know what it is I want.
I know my systems pretty well. I just need people to go do the things. I've already got an idea going back to the agency idea, I, I just need, it's just I need somebody to go and do it. And now we have these a AI agents who can just go and do it. Oh, and bonus points, they're gonna get rated on how well they did it compared to your businesses guidelines for brand, for compliance.
And I'm sure there's gonna be more that come out.
Alex Pokorny: Yeah, there's, there's a difference there though. You're talking about an approval step and the kind of resistance piece is where I was going with the UX question, right? So the project that you would have to lead to get it done,
[00:38:00] if you created the document or the findings manually versus having gen AI create it, do you push as hard, do you get it done? And which one maybe saves you time? Which one gets it done faster? I mean, one of 'em, you're gonna get sunk cost. I mean, cheap in the work. Fine, but sunk cost 40 hours already. Just to create the thing like that.
That's a lot of time to get the document together to then champion it and get it to happen.
Dave Dougherty: And if you're already on the client side, right? Ruth, for your own example, how long did it take you to understand. Your systems to understand what your brand guidelines are. Yeah. Because you lived and breathed that for a very long time and had to have lots of stupid meetings where people said lots of stupid things where you're just like, shut off.
Right. That's part of the learning. In order to be able to do something more quickly, you have to know [00:39:00] the ins and outs of it. And you know, we've said this about all of the outputs since, you know, 2023. When we first started talking about Jedi AI was okay. I. This is good enough if you don't know what you're talking about.
But if you know what you're talking about, you can easily see where this falls apart, you know? And I think the campaign side of things, the reason that's getting so much love is a, there's a ton of marketing spend there, but then b, too, like the main thing is, okay, there is a lot that you can automate there, right?
With your, your brand standards. I'm only ever gonna choose this color green. Right. And then you're only gonna have one or two sentence worth of text for a display ad or social posts or whatever else, right? You're not gonna have some sort of long convince me kind of thing that might come through and you might actually want.
Some, you know, agency things like it,
The Future of Entry-Level Jobs
Ruthi Corcoran: I think we're our, our viewpoints very [00:40:00] much meet is what shocked me is the piece about one, if you have button pushing where the documentation is already in place,
Dave Dougherty: right?
Ruthi Corcoran: Gone. You've got robots for 'em, right? Mm-hmm. In the form of, so entry level jobs, agents, or in the form of, I'm outsourcing a whole bunch of just the, the hands-on keyboard work.
It just needs to be done. I think that I. There is just it's, I don't think it's staying around very long. Like the idea that you would hire a bunch of cws or agencies in order to just do a bunch of what is very tedious manual work, but to. You have to have the documentation, which means that you have had the strategy upfront where you figured out, here's what we're trying to do, here's why we're trying to do it now, just go and do it.
And I think you guys are talking more about that upfront thing of like you've gotta get to the point of knowing what it is that you wanna do. And that is still very much [00:41:00] outside of the gen AI world.
Dave Dougherty: Mm-hmm. Well, and again, that goes with the use cases, right? I mean, we're talking big sort of incumbent organizations that.
Personal Projects and Gen AI
Dave Dougherty: Do a particular thing where I've seen the most benefit with Gen AI has been all of my personal projects, right? Because I can just say, Hey, I need to accomplish this thing right now. Get me 90% of the way there. I can look at it and say, yep, no, yep, yep. I'm gonna change this. Like that's really, really, really useful, right?
Because again, I know what I'm trying to do. I'm trying to get it done. In a way that allows me to still relax before I go to bed instead of like just shutting off and, you know, trying to fall asleep. Yeah. Yeah. Always interesting. Clearly, we will circle back to this topic.[00:42:00]
Dave Dougherty: Thank you for making it this far in this episode. Hopefully you have liked and subscribed, if not. Do so it helps us understand. Thank you to the person who gave us a five star review on Spotify recently. And I. We will see you in the next episode, so.
