AI Market Research Realities

Guy: [: 00:00:00

Dot Pro relevant.com. Today I am interviewing Peter Swim of Ville, and he's got a fascinating story on what, uh, and where he's trying to help his clients out in terms of their use of, uh, of AI and, and you know, and building it into the organization. But let me tell you a little bit first about Peter. He is an innovative force behind Toy Oilville, LLC.

e is collaborated with major [: 00:01:00

Peter: Well, thanks for having me on the show, guy.

Guy: Yeah, absolutely. So tell us, uh, how did you get involved in this?

What's your backstory on AI and marketing?

Peter: Sure. Um, so I, I've been in tech for my entire adult life and, um, I've started getting into conversational ai, which is basically the process of turning human interactions into terminal commands for like 10 to 15 years now. So, um. I worked on a open source, uh, chatbot framework called Botkit.

aunching it kind of exploded [: 00:02:00

You know, like I'd be talking to the d and c one day in the Grammys the next day, and Walmart the day after, and they all had. Omni universal needs, but they had very, a universal problem, which was just people wanna say their problem and then the solution comes out the end of the shoot. Right. And that kind of led me on this journey where, you know, for the past 10 years I've been, I.

Kind of seeing the desire and the implication and the, you know, messing around and finding out of all these companies trying to like, handle the huge scale of their communication needs, uh, through automations and then with ai and for all my tenure of building these products for, for people, I would talk to clients that you mentioned, Rabobank and all them, and they're like, we wish we can have a person like you come in and just like.

like, you know, an ounce of [: 00:03:00

So, uh. I think right now like there's a tense pressure for people to AI size everything, and they may not understand what their AI sizing and they may not understand like what is an improvement and what's a detriment to their existing processes.

Guy: Yeah. And, and I think you're right. And, um, uh, you know, there's the hype and, uh, and everything that, uh, you know, that wow, if we could do this or if we could do that, and then you get into it and you go, well, we can't quite do all that.

nd then, you know, work back [: 00:04:00

So, uh, makes a lot of sense. So now you put together a company called, uh, Ville. Mm-hmm. Uh, tell us about that and, uh, how you came up with the name.

Peter: Toyota has always been an idea bouncing around my head of the social contract of, of being in work. You know, I'm a creative in my off hours. Um, I have a family and, um, I like to work hard at work, but also I like to leave it on my desk and close up the office at five o'clock.

And, and I think, like you don't hear a lot of this in the, the pivot to ai, especially from tech leaders and uh, management, that like technology has the potential to. Improve the day-to-day life of the worker, right? To amplify their expertise and to like automate the things that are super annoying about their job and all that.

so. And I think that's kind [: 00:05:00

Guy: Yeah. Yeah, that makes sense. And um, and you know, one thing that, uh, is always interesting is, uh, you know, the boss or the client when they realize that, hey, you can do more, you know, with, uh, fewer people than they're always gonna say, well, how can I get more out of this guy or this team, uh, right now that they're using ai?

And so there's, uh, you know, hopefully, hopefully at least there's a short term respite for all of us. And then, um, you know, but I always see that, uh, you know, that we're gonna go back into the, you know, trying to stuff 10 pounds of stuff into a, into a one pound bag. I know I'm guilty of that. If I can find something that'll speed something up, then that just means I can do, you know, more of something else.

So, you know, maybe that's just the marketing mindset. I don't know.

times profitable of your [: 00:06:00

And, and it's just like things that grow uncontrollably isn't an organism. It's cancer. Right? Like it's, it, it's not like a. In the history of capitalism, it's not like historical way you've done a business historically, but in the past like 25 years, it's become the expectation that you're, oh yeah, we're gonna have year to year 10 x growth.

And, and that's gonna happen forever because there's infinite consumers and our product has infinite total market accessibility. And, and I think. That's kind of like the gold lust with AI right now is like, oh, I can just turn, lead into gold and, and uh, you know,

Guy: you mean you can't?

Peter: No. Uh, you can turn lead into lead, you know, you can turn, uh, you can turn, make a nice old lead chair to set in when you're like paying your AI bills.

Um,

Guy: and you can paint it gold, at least with some gold paint.

ld paint. Um, and, but like, [: 00:07:00

Five x more for a person to answer the phone and do it 30% worse. You know, like, and there's like clear studies now. Like we, we, we have a lot of data to show that, like putting AI at the top of decision making is a huge risk and, and unsustainable at scale. And putting it at the bottom and then being cautious about it has a lot of ROI like, there's like.

You know, there's things that are effectively have been solved by this wave of ai, like the FAQ problem, right? And like the, what are the office hours problem and like, you know, when is my W2 coming? You know, like all those like things that are like, take up so much time in answering that are just like, just read the website, you know, and people don't read the website or whatever.

oblems of like, uh, my plane [: 00:08:00

You know, like, yeah, we should get that. We should get that person to our best support person and our highest tier kindness person as possible, because that's what makes or breaks loyalty to your product. Right. And you know, I think people have kind of lost sight of that, you know, and they're kind of like, oh man, I.

I can a hundred percent automate my contact center, you know, like absolutely not, you know, at your peril, you know?

Guy: Yeah, yeah, yeah. Well, and I think you're right, you know, the, uh, the simple stuff that is on the website that could be answered, uh, you know, even if it's on the website, you know, so many times, you know, I'll be looking for something and, uh, you know, I can't find it.

um, you ask the help and it [: 00:09:00

It's just incredible how good it is. And, um, so, but to your point though too is, uh, you know, quite often, you know, I know right away that I've got a complex answer that the, the simple person isn't gonna be able to answer or the, the chat bot isn't gonna be able to answer 'cause it's trying to structure some tree thing and can't get, get you down that path and you just gotta say, representative, representative, representative.

Mm-hmm. And finally you get somebody, and finally you can get to that answer. So I, I like your example that you, you mentioned about the, you know, the complex, edited, uh, you know, expedited change that needs to be done,

Peter: right? Yeah. I mean, like, I always like building things in my system. It's just like the minute a person says a lawyer, right, you need to get to a person, right?

erson says any of these food [: 00:10:00

Guy: like,

Peter: and it's like if you work in that domain, it's common sense.

But for the people building AI tools, they don't have that domain expertise, so it's not common sense. Right. So I think that's also a big hallmark of, of my consultancies. I really allocate strongly for domain experts to be involved in the transformation because they're the ones who are like. You know, aunt Sally, who runs the contact center for 30 years knows way more about, uh, banking than I can ever forget.

Right. And I could probably teach her about AI quicker than she could teach me about banking. Right? Yeah. And so that's, that's kind of like my pitch to the workers, because your employees are gonna dictate whether or not your things succeed. Right? Because if it's like some kind of like. Try hard, uh, proposal.

e like you come along every, [: 00:11:00

It, they fail every time, I think, because they've not invested their entire organization in the concept of the digitization. So,

Guy: well, you know, you bring up a point, and that was gonna be one of my later questions, but, uh, we'll bring it up now and that is, um, you know, if you, you know, in building these AI applications, you do need to have those people with domain expertise.

And, uh, one of the challenges though is those senior folks with, you know, 10 or 20 or 30 years of domain expertise. They're few and far between, but then it makes it, if the AI is gonna take over a lot of the, the things that that person is doing, it's gonna make it harder for somebody with no domain expertise to break in.

And, uh, have you thought about that at all? Any, any comments or?

use gen on their day-to-day [: 00:12:00

Uh, doesn't help them find the clever hacks that the 30 year veterans have because like, and so that kind of begs the question of like, AI is good on the front lines, but if you deeply integrate it into your system, are you kind of shitty your future in the foot? You know, I think that's a very valid concern to have because when people have hired, when you interview a person for a job, you, you, you're talking to 'em about their experience and what they've done in the past, and you try to get a sense of like.

How they'll approach problems yet unmet. And, um, AI cannot do anything with problems yet unmet. It's entirely at a met problem solution proposition. So you'll never, like, you'll never reach that one-to-one map of the world with ai, and it'll never get better than the world is today. Without, like someone hacking it or, or reinforcement learning.

bias and all that. So like, [: 00:13:00

But there has to be some sort of way of the human brain evaluating processes and making the decisions of whether things approved or not, and be able to measure that to see if that actually worked. And I think, you know, the a hundred percent unattended system is dream of entrepreneurs and CEOs around the world, but.

They have the CEO of Microsoft talks to 25 of the smartest human beings on the planet every day in meetings who get personal reports to them. And they think that's what agents are, you know, and actually you're gonna get the bottom 10% of humanity for most people, and the CEOs get the top 25 people in the world, you know?

me the only game in town for [: 00:14:00

Guy: Yeah. Yeah. Well, I liked your example, although I might wanna push back on it where you said, um, you know, the AI reduces the level of critical thinking and it doesn't, you know, somebody that's got 20, 30 years that's kind of like, you know, got all the scars and, and has come up with everything.

Whereas the younger folk, the younger person that's, you know, entering it in gets a fast answer and doesn't really. Critically think about what the real challenge is. You know, I think, I think you're right, but it did take that person. 30 years to get that far. Mm. And, um, you know, maybe some of the scars that the 30, you know, year experience person has, maybe those are then already answered in the, uh, you know, in the chat conversation.

do some critical thinking or [: 00:15:00

I, you know, I don't know, I'm a little bit torn on how, how that that process is actually gonna, you know, unfold. I. Because I agree with you that, you know, it is gonna be faster just to, Hey, let me ask ai. Oh yeah, there we go. And then move on. Right. But, um, uh, yeah, I don't know any, any thoughts.

Peter: Well, and I think like that also raises a good point about the arbitration of AI decisions.

Like I. You know, uh, we see a lot in like dystopian PHI movies where a person's trying to do something and the, and the answer's no. And can I talk to your managers? Like no. The, the entire system's the manager, and it's, and it's ruled this. And, you know, that's a very kind of fascist, kind of like view of the world.

up. There's people who bring [: 00:16:00

But it's the right thing to do, right? Like, you know, the the person who's just like, oh man, uh, you missed the flight for your grandmother's funeral. I'm gonna pull my strings that I know and make it happen, even though that's not policy, because I feel for you and that's the right thing to do. I'm a human being, and you're a human being.

Let's make this happen. Right? And I think like the, you know, you, you make a really co point of like, the risk of that. Is, um, how far back that 30 years of experience before you enter the system goes of the autonomous thinking, right? Like if people, you know, we got kids entering kindergarten now who have their whole life been dominated by chat GBT, and you know, are they gonna be in 15, 20 years?

u know, and I think the, you [: 00:17:00

Like, yeah, this is clearly not something this person wrote, and they can't understand the thing that they claimed to have wrote, and I have to fail them, you know? And it's like. You know, it is kind of like bewildering and sad problem.

Guy: Yeah.

Peter: For, for people. Like, you know, am I getting this person's authentic self, you know, when I'm talking to 'em or am I talking to like what they think is the work?

So that thinking smarter, harder thing has a double-edged sword where people are kind of like using it to like get outta work and get out of like their critical thinking and all that. So there's definitely, there's definitely concerns there.

Guy: No, I, and I agree with you, you know, the, um, for education, uh, the education is now, you know, very different or it has to, you know, evolve to be something very different.

of us that grew up without a [: 00:18:00

Finally, you know, we got that in a calculator and um, and then we didn't have to know that stuff anymore. But, uh, what ended up, I think coming out of that was. That, uh, we ended up learning how to define and specify the problem better. Mm-hmm. And then, you know, put the boundaries and the, the box around the problem.

And I think that is what applies now to, uh, what ai, uh, is able to do is. What needs to be taught to the young folks is how to define the problem and, uh, and then make sure that each and every one of the little things that goes into that definition are clearly stated. And then of course there's iteration.

, maybe the teaching and the [: 00:19:00

Uh, you know? Mm-hmm. Because if if you don't understand the answer, then that's not right either. You know, the, the, the education system has to tell you how to get to the, make sure your answer is right and that you understand it, and then, you know, then of course it's all of the, the problem definition that has to go around it.

Peter: Yeah. And, you know, I feel like a part of my consultancy that gives me my competitive advantage is I am kind of like a cusp millennial, and I say this as an advantage because like, I. I remember the day my computer lab got the internet, right? So there was a point before the internet and a point after internet.

Um, I went to college for audio production and they taught us how to cut tape, you know? And the year after I left, they brought in computers to do that. And so like my whole life has been this, like I. You know, the bridge from the old guard to the new guard. And I think that's a per that, that's really a pertinent point.

laugh every time I'm using a [: 00:20:00

Like, you know, how do kids even know with a dial?

Guy: With a rotary dial? Yeah. With a dial.

Peter: Yeah. And, and it's just like, you know, what is the reference point for that when phones haven't looked like that for almost 30, 40 years now? You know? So.

Guy: Yeah, yeah, yeah. Uh, that is, that is so true. And yet, uh, those words are still in our vocabulary.

al strategies that, uh, sure.[: 00:21:00

Uh, you know, that you can talk about and, uh, and, and go from there.

Peter: Yeah. Uh, we we're developing this framework at Toy that I call spell work, and that's basically just, uh, systemically, um. Mapping out what are the wishes that come to your big microphone in your organization? Right? And so like I think a lot of companies, like they broadly understand their mandate of what they have to do.

And then there's the yada, yada, yada, which people handle, right? Well say we sell telephones, right? And so like, oh yeah, people call us up and they'll wanna order the telephone. And then you're like, well, do they need one for them or for their company, or one for their family, or what are their needs? And that's where you kind of break down into like the minutiae of your line of business.

comes back to bite you. And [: 00:22:00

'cause all you're trying to do is make a pizza bot. And now your bot is like on CNN being, uh, showing, supporting, you know, racism and stuff around the world, and you're like in a huge nightmare. And so you'll just blame a hack or whatever and fire everyone and move on. But you know, I think that is kind of like the, I.

You gotta know what you're gonna build, where and when and be very crisp about it. And so our framework is, and once you can ship one thing and know and do it well, then you can horizontally expand across your things. And some things may never be adequate. Like, you know, the sensitivity of handling someone in a personal crisis, like the flight example I gave earlier.

mpany that was so large that [: 00:23:00

And they found out that the average call time was two and a half minutes and a minute, and 30 of it was user verification. So I'm calling and I, I'm, I'm proven that I'm guy and I live here and this is who I am. And they were so intent on automating the whole thing that they didn't realize that if they automated that.

Verification process, then you would cut the call time in half and double the capacity of the call center. And it, it was really hard discovery, you know, like, and you know, they were like, oh yeah, this, uh, well she's gotta be very personable chat bot. And she's gotta tell jokes. I'm like, whoa, whoa, let's back up.

phone tree. So why don't we [: 00:24:00

And it was just such a ion to them because they're so intent on the the end goal that they missed the low hanging fruit, you know? And I think a lot of organizations have tons of low hanging fruit that are just like, yeah, like I. There's no reason you should wait on the phone for 10 minutes to get office hours or business hours or a phone number.

You know, anything that's like, like you said on the website, but also it's like an accessibility thing. Like not everyone can use a website, not everyone can use a phone. Not everyone can like speak. So like having automations that like answer things in people's preferred modality, you know, is also like accessibility issues.

So there's things that you could do that are like warm and fuzzy and win-wins and. Get that out there and learn to cut your teeth on that. And then you can like, you know, approach these like stronger, longer term problems of like, how do you automate, uh, insurance quote, which is like 142 steps, you know, like, so

Guy: Yeah.

Yeah. No, that makes [: 00:25:00

And then you go, oh, well I can't solve that and let me pass you over to somebody else and they gotta reverify you. I mean, just pass the token over, you know? And uh, so I really like your point that because that, that verification now is, uh, you know, it's painful and, uh, and I wish too, I mean, this. Two factor authentication.

You know, as much as I like it because I know that my bank accounts are then you know, that much more secure. Mm-hmm. Uh, you know, it is a pain when you have to, you know, go through that every time and, uh, you know, it's just, uh, you, you wish there was a way to kind of get around that

interested in solving those [: 00:26:00

Yet no one's made. Like I have to wait and hold for 20 minutes to talk to my bank. Why can't I just tell Siri on my phone that I need to call the bank and have this done? Mm-hmm. And they'll arbitrate. And then, uh, when two people need to talk, they'll get, it's like, Hey, pick up the phone. And you talk like, you know, the fact that anyone has to wait in a queue when everyone holds a phone in their pocket 10 times more powerful than put a man on the moon is, is is silly.

You know, like, and so. I think there's a lot of like tunnel vision, you know, of people, like, they're so obsessed with making all the money that they leave all the little bit of the money on the table, you know? So.

Guy: Yeah. Yeah, yeah. Absolutely. And that, that makes a lot of sense. Well, I could go on, uh, here forever.

[: 00:27:00

Last question though, before we, uh, close. Uh, so what advice would you give an up and coming new marketer? Uh, that's trying to, you know, I don't know, break into the marketing, uh, function. Yeah. Coming outta school or, or, or wherever they're coming from.

Peter: I think right now it's a battle of chum, chum versus chum and everyone's like, I get tons of AI marketing already and I'm trying to get out to people and I'm surrounded by AI market and it's very hard to stick out because it's all kind of like a gray morass of like sameness.

e. A way to match the output [: 00:28:00

And the ingenuity and decision making to like match their output, but with good, good decisions and, and coaching information. And I think that's like a lot of working with, like I, I find a lot of things, you know, I've worked with Fortune five hundreds and fortune fives and there's the Fortune five millions out there that are kind of like.

Don't know what to do and they need help. And there's plenty of like, opportunities to learn from people there. And there's people like who've been in the industry for 30 years who are like winding down their careers that have a lot of stuff to give. So, you know, really embrace those kind of situations and don't just kind of like, you know, attach yourself to the hot thing and lose out on the, you know, advertising has existed for almost a century now in the modern form.

I have a relative who worked [: 00:29:00

Soak in the culture. Right? And, you know, being able to like, have access to culture and new ideas is essential for the lifespan of anyone's career. So while you're young, you have that finger on the pulse and you should exploit it and be promiscuous with companies and, and get as much experience as you can while using technology to amplify your reach and, and use it to kind of like.

Attain your dreams, you know?

Guy: Yeah, yeah. No, that makes a lot of sense. So, well, thank you for that. Well, Peter, uh, you know, thank you so much. Really appreciate it. And, um, uh, that's been, you know, you're, you're, a lot of the stuff you're talking about really, you know, brings to fore where kind of the human side and the AI side can fit together.

d out more information about [: 00:30:00

Peter: Sure. Um, you can reach us on the web. Uh, we're at it's to.com. We hold office hours, uh, at least once a month. So, uh, you can come to forum and just talk about your problems. Uh, we also do, uh, consulting and one hour consult.

I'd love to get on the phone and talk to you about your problems and. See if we can find a way to help you out.

Guy: Yeah, fantastic. So it's the website is, it's tole V-V-I-L-L e.com? Yes, it's tole vil.com. Wonderful. That's right. Well, yeah. Fantastic. Well, Peter, thank you so much. I really appreciate it. And to the audience.

Please stay tuned for many other videos in this series of the backstory on marketing and ai. And if you'd like, please go to marketing machine.pro relevant.com and download some valuable materials about my upcoming book, the AI Marketing Machine. Uh, Peter, thank you so much. Really appreciate it.

Peter: No problem.

Thanks for having [: 00:31:00