AI's Impact on Client Engagement

Guy: [: 00:00:00

Relevant. Dot com. Today I'm interviewing Andy Wick of Hubbub and he has a fascinating story and he's doing some really cool stuff and, uh, looking forward to our conversation here. So he is the co-founder and CEO of Hubbub, which is an AI powered. Concierge platform, transforming client entertainment, a seasoned entrepreneur and an AI thought leader.

startups, in uh, in Chicago. [: 00:01:00

Andy: Thank you for having me.

Guy: Yeah, absolutely. So, uh, tell us, uh, how did you get involved in this AI and marketing opportunity and, and, uh, space here?

Andy: So at Hubbub, obviously we do a fair amount of marketing, content marketing. We're an early stage startup, but, um, I've also dealt with it other, at other companies in a, in a more labor intensive way.

And, uh, AI is really disrupting the space dramatically. So, um, it's a natural, anybody, anybody building a startup is resource constrained and time-limited in, uh, in, in many ways. And so, um, AI is a natural solution to a lot of that.

Guy: Absolutely. And um, it, uh, you know, and you're so right about a startup and, you know, being resource constrained.

Wouldn't it be great [: 00:02:00

Andy: No, you never have enough money for that or, um, other things, but definitely marketing and sales. You can always, you can always find more ways to spend it, but it's probably good to be resource constrained.

So you're, you know, it's like a diamond, you know, it gets polished by friction and, um, I think it's true of a lot of startups that you get polished through friction, and that friction is res, you know, constrained resources.

Guy: Yeah. Well, and it certainly makes you make the right decisions or, or try to as near Yeah.

Well that's a good point. At least the try to make the right decisions. Yeah, absolutely. Well, you know, one thing that's uh, happening though right now, and just to kind of change the subject a little bit is that. And I don't know whether AI is being overhyped or not, but, um, uh, and there's certainly some limitations to it.

t is AI being overhyped and, [: 00:03:00

Andy: So I think it is being overhyped and it's being overhyped by some very well known, famous people, famous entrepreneurs even, or people who would run very large public companies who you would think would have a better understanding.

I won't name any names, but I think thing talk of super intelligence or depending what that means, it's never really well defined, but I guess we can assume that it means something smarter than a, the smartest human right, or artificial general intelligence. You'll hear some people say a GI. It depends what you mean by that.

, I think is probably AC or, [: 00:04:00

Just as a shovel allows you to be more productive and dig holes faster and more. More holes. Ais are going to make us be able to be a great deal, more productive. They're gonna reduce the cost of a lot of things. Some jobs, I think, will get eliminated. They'll be automated out, which is not, you know, it's nothing new.

I mean, there's, there's nobody milling flower by hand anymore, and there's nobody, you know, shoeing horses unless they really want to. Right. Or doing blacksmithing unless they really want to. It's not a career anymore, so, but I don't think anybody sees that as a bad thing. Capital just flows into other more productive areas.

And creates even more jobs. There's a lot of jobs today that didn't exist 10 years ago from selling things on eBay and Amazon to Uber drivers, to social media influencers.

Guy: Yeah. You know, and it's, it's funny that, you know, you brought up Uber, you know, a lot of the taxi drivers in New York are now Uber drivers.

it was kind of a shift. But [: 00:05:00

And, and that's a, that's a big deal.

Andy: Yeah. It's staggering how fast it's snuck up on us. And I think there's still some room for improvement. If I had to guess, we're probably as far. I don't know better than half the way there, maybe even 75% of the way there as far as how good these LLMs can get with current computer architecture.

Quantum computing might change some of that, but that's at least minimum years out. But it's just been staggering how fast it's moved.

Guy: Yeah. Yeah. You know, the one, one of the, uh, objections to, uh, AI though is, um, uh, and a friend of mine, he brings them up. Of course, he's also, uh, you know, doing and starting to do quite a bit in ai.

things like MySpace and, uh, [: 00:06:00

I think, I think AI has proven itself in the short time it's been around that it is really a game changer and, and, you know, and reduce costs, you know, make things faster, better, and cheaper. So,

Andy: yeah. You know, MySpace is gone, but you know, digital networks are not gone. Social networks are not gone. They've just been, somebody did a better job than than MySpace ever did.

I don't know if you remember my, I don't think MySpace ever had a newsfeed. That was a brilliant innovation on the part of, uh, Zuckerberg. Or likes and stuff like that. So I think that's really, he just built a better mouse trap. When you think about AI in general, it's really mimicking. It's a brilliant pair.

It [: 00:07:00

Both oral and written human speech, these ais are mimicking it, but there's no understanding. I, I remember a, a famous. Um, di uh, a real famous West Coast vc, uh, saying about a year ago, he said, you know, somebody, you know, we passed the Turing test that nobody seemed to notice. And, you know, but the truth is, it turns out the truth.

The Turing test turned out to not be a good test. Turing said, you know, if you, if a com you could talk to a synthetic, a computer, or something like that and not be able to tell it's a computer, you could assume it's understanding. Well, it turns out that it's not true. The computer doesn't understand, it's just mimicking.

know. Like for instance, we [: 00:08:00

There's just a probability matching up in the database layer. So,

Guy: yeah. Yeah. And that, that's kind of the scary part of it too, is that it is kind of like, well, what's the, uh, next most probable word? And, uh, you know, that's one thing. Uh, certainly it's advanced past that. But then there, that advancement though is kind of like a structure it seems like, that somebody has developed.

And so that's structure is really. Taking the intelligence of somebody else and kind of form forming it around or force feeding it around the answers so that they really look like they're very intelligent. Yet, yet in reality, they're just following some form that somebody else programmed into it.

Andy: Yeah, I mean, there's a lot of.

ed the Chinese room and the, [: 00:09:00

The person inside the room takes it, uses the filing cabinet, doesn't understand Chinese. But he is able to use the code in the filing cabinet to translate it into English and slide it back under the door. So the person that receives that on the outside assumes, well, whatever's inside that room understands Chinese.

It doesn't. And, and a good example of a Chinese room in real life is the Watson computer built by IBM that won Jeopardy and has, you know, beat Peter Gasper in chest. Things like that. It's a giant Chinese room. It doesn't understand anything. Yeah, it make, it just makes correlation. Computers are really, are, are decent and this includes the LLMs.

types of reasoning. There's [: 00:10:00

They've gotten better at realizing that, and, and so they do that reasonably well. Deduction is taking an own pattern or theorem or equation and deducing what you know, what the pattern is and, and making, you know, extrapolating and making predictions. Mm, they're okay at that, but they're not as good as they as they are with pattern matching.

and I don't know if they're [: 00:11:00

But the third type is abduction, and that that is basically an educated guess. And it's been shown that that's non-comp computable. And by non-comp computable, it means it cannot be solved algorithmically and computers only work algorithmically. And I hear, I hear objections like, well, it's a probabilistic instead of a deterministic machine.

Well, that's only the database layer. It's only, it's using proba probabilities to make associations of the database layer. It's still running an algorithm to produce, you know, to process the humans, you know, whatever it is, prompt and then produce output. It's still, you know, using an algorithm and some things cannot be solved algorithmically.

aper on computable numbers in: 1935

Guy: Yeah. Yeah. Yeah. Interesting. Um, and, and yeah, I think you're right. I think there's, uh, and I like the way you broke it down into the, uh, [00:12:00] uh, into those three classes there.

So, um, uh, but let's get back to, uh, more mundane things. So tell us about, uh, hubbub and, um, when you, uh, your, your position kind of as an AI driven concierge, what, what do you mean by that? And then how do you leverage AI to, uh, really drive the, drive the business and value for your, for your clients?

Andy: So we automate business client entertainment.

We're basically a SaaS platform for that, and that includes, you know, everything from customized recommendations that based on your purchase history through notifications, calendaring. Ride share payment all the way through populating a CRM tooling expense report. And the way to think about it is, is it's sort of your own personal assistant, but in digital form that handles.

consuming. And maybe they're [: 00:13:00

So we're, uh, making something available that was to the masses or to more people at least that was only available to the well to do, or maybe even the C-suite. And, and a correlate to this is. And when you into the broader AI world, you can think of, um, the coming cyber cab as being your own chauffeur, something only the rich could afford.

You can think of, uh, LLMs in the near term being your own personal tutor, something only the, you know, the rich could afford. And then with these robots from figure and, you know, optimists from Tesla that might be able to clean your house and cook your food. Now you can see a future where. Almost everybody can have their own personal chef and maid, right?

Mm. So

even beyond business client [: 00:14:00

Scheduling and stuff like that that nobody really likes to do and coordinating. If you're, if you're going in a group activity, then there's even more coordinated to do. Yeah. Calling up and following up and people dropping in and dropping out. All of that, we, that's how we fit into the broader AI picture, I think.

Guy: Yeah. Well, and that is, uh, you know, and you're so right too, the amount of time it takes, especially when you have a group, uh, and you don't, they're not on a group calendar or whatever, and you're trying to figure something out and everybody's got an opinion and, oh, I can't make it that day, but I can make it this day.

And, yep. And, uh, yeah, that's definitely something I think that ai, you know, if you have an AI agent working for you, that you could talk to that AI agent, so to speak, and then the other AI agents could talk to themselves and, uh, figure out, you know, something that really makes sense,

Andy: right? So only the rich in the celebrity's gonna afford that right now, but everybody will have it in the near term.

Guy: [: 00:15:00

Andy: The, the first thing I would say is think about something that affects you personally. A problem that affects you personally, or a friend or a family member, and what that's called is founder market fit. And it's important to have that because you're, whatever solution you're gonna come up with is probably gonna change at least two or three times.

Maybe even more and maybe, and also maybe your go to market strategy and other things beyond the solution will change. And, uh, to be able to persevere and stick that out. And you really have to feel fairly strongly about solving that problem. Otherwise, you're just gonna give up. It doesn't make sense that to persevere with something, you just don't care that much about.

The other advice I would say [: 00:16:00

You'd be surprised for a lot of these guys, these serial real entrepreneurs, it's like they're golf, you know, instead of golfing, they like helping founders.

Guy: Well, yeah. And then, uh, they've definitely got, you know, the, the, uh, the college of hard knocks behind them, so, right. You know, and I think, I think that's a, you know, a big thing.

If you can weed out 50 or 70% of the mistakes that you might otherwise make, that leaves so much more time and money and opportunity to make things Right.

Andy: There's one more thing I would add that I think is a really key ingredient, and that is there's a notion of a vitamin versus a pill. In the startup world and vitamins are okay.

ngly about it. But a vitamin [: 00:17:00

Or,

or

Andy: you know, you've got a migraine and you need a Tylenol, that's a pain. And it's, there's something called switching costs whenever, whatever product or service you're building. People are, you know there, Daniel Kahneman wrote that book, thinking Fast and Slow, and he talks about system one and system two and people are, system one is pretty much autopilot.

that do succeed and, um, so [: 00:18:00

What pain can I fo solve for people? And it doesn't even have to be. It can be just something that gets the initial flywheel going, and it doesn't have to be a long-term thing. And I'll give you a couple of examples. Uber, when they first started, they were, they're targeting delivery drivers in San Francisco with downtime and businessmen or business people who didn't got out of a meeting, didn't have time to look for a hotel, to find a cab.

All the cabs are by the hotels, right? So now they can push a button on a phone delivery. Drivers who's got some downtime between Rise can earn some extra money. They're solving a pain, a serious pain point for both sides. Another example is Amazon. When they first started, they were selling books of course, but their target customer was antiquarian, booksellers.

hey could sell more books to [: 00:19:00

Sellers are buying more books obviously than the average person. You know, the average person's buying me what, four or five a year at the most? These guys are buying hundreds a month, right? Mm-hmm. So they got, that's how they got the flywheel on the third example I always give, and these are kind of B2B examples.

go around and they'd pick up: 2000

That got the flywheel going. And what pain point were they solving for those users? Those people wanted to know whether so and so was dating somebody else without having to ask them and get embarrassed so they could just go, look. Now I'm a, that's a, that's the biggest problem for that age group, right?

Is so and so available. So there were, and that got the flywheel going now. I mean, that's almost a minuscule amount of the users using Facebook for that.

Guy: Yeah, yeah, yeah. Funny. So think

about that when you're belt, [: 00:20:00

Guy: Yeah, absolutely. Well, it's certainly, you know, when you think about where a company starts and that initial problem that they've got and, and how it evolves, it's, uh, it is pretty in incredible.

I mean, it was pretty surprising when Amazon, you know, moved outta just books and into everything, and now they're into everything. Yep. And then, you know, and then they bought, uh, stores, they bought Whole Foods and now they've got, you know, brick and mortar instead of just only online.

Andy: Yeah. And Amazon go.

Guy: Yeah. Yeah, absolutely. So, uh, what are some of the, uh, big, early mistakes that, uh, founders make? I.

Andy: Those not thinking about your go to market. A assuming that if you, first of all, not talking to your customers, not thinking that if you build it, they'll just come. And not only building it, but not building it, uh, in a, there's something called a lean startup method.

. You want to see if you can [: 00:21:00

So build something first. Talk to customers before you build anything. And if you can do something, you know, using what they call the wizard of Asthma that do it and don't build anything. And you saw that with like cameo. Groupon, for instance, like cameo, they didn't, or even Airbnb didn't provide payment transactions when they first started.

They were just displaying pictures. Groupon, you had to go to a web form, uh, and submit, uh, submit a web form, and they were just testing whether or not people would actually buy things this way. There's that famous, uh, story about that shoe store. It's in Vegas now and they were bought by Amazon. I forgot the name of it.

High-end shoe store though.

Guy: Oh, it's um, is a PO is a soda. A ZI think.

Andy: Zappos. Yeah.

Guy: Zappos. That's it.

ent to a Nordstrom's, took a [: 00:22:00

But it proved that, yeah, people will buy, particularly people will buy shoes online. Groupon showed that. And so that's called the wizard advise method. If you can do that even better, if you have to build more, then so be it. Then build more, but build the minimum test it. And don't, don't be wedded to it.

You know, get feedback, be willing to pivot. And then I would say the third biggest mistake is people will get early, early, maybe some early sales, and they'll say, think they heard it hit a home run, but they haven't really thought about the go to market. They haven't thought about their, yeah, cost of ambition.

nd of, what kind of language [: 00:23:00

What kind of, what converts them? Right? And, and that's probably the next biggest mistake, getting some initial success. And not really thinking about your go-to-market.

Guy: Yeah, no, I, I, I agree with that. 'cause it, you know, the first one or two customers, you know, typically they're your friends, they're your very close contacts.

Oh yeah. Hey, I got them. You know, they we're helping 'em out. We're doing all this great stuff.

Andy: Yep.

it like the ten one hundred,: 1000

That process is, uh, is very difficult, is right? I mean, you can get lucky. Some, some, some products, you know, it's just they, you get lucky. But for most startups, it's very hard to go from that, you know, one, 10, a hundred thousand.

Andy: A lot of people make the simple mistake thinking, well, I'll just do some social media advertising, or I'll do some SEM.

of free stuff away at those [: 00:24:00

Guy: I'm just here for the free lunch. That's all. Yeah. I mean,

Andy: so you gotta, you've gotta really understand, and that comes down to empathizing it.

It really, the, in the bottom line, you know, sales is empathizing, sales and marketing is empathizing with your customer. Right. And, and understanding the problem, and then, uh, listening to what they tell you. I get all kinds of people selling, trying to sell me stuff through LinkedIn or whatever, and I tell 'em no, and they never, almost never asked for why or feedback or they, you know, I responded to you.

I might give you some answers, you know, but almost never. That's a big mistake. Yeah, always be learning. Always try to empathize and understand your customer.

Guy: Yeah. Yeah. So how do you think, uh, AI and, um, certainly automation, but AI can help, um, you know, mid-market companies, uh, compete with, uh, larger companies?

otential to punch above your [: 00:25:00

There's very large companies who, you know, they, they employ a, a, an army of executive assistants who do this work for their Salesforce or their partners, or, you know, whoever's doing the selling. Right? Whether it's a Big four accounting firm or a law firm, or a, a Google or a Facebook. The companies that are smaller and they, they can't afford that.

Now they can punch above their weight with this automation. And not just with marketing sales, but you know, software development, you know, development that used to cost is probably dropped by half. Cost of software development is probably dropped by half with these LLMs and it's probably gonna drop another half.

Yeah. And, and probably in: 2026

So smaller companies can punch above their weight now because of that.

Guy: Well, you know, one thing I've always found, well, I don't know if always, always is a, is a loaded term, but um, yeah, the smaller companies are always afraid of. Uh, what's that?

Andy: In marriage and in business? Don't, don't say always to your spouse.

You always knew this. Don't say that. Yeah, yeah,

Guy: exactly. That one I gotta write down. But, um, it's, it's funny because, you know, the small and midsize companies are always worried about the big companies, you know, coming in and, and you know, and destroying the market. And, uh, the big companies are sitting there going.

u know, there's a, there's a [: 00:27:00

They certainly can. They have the money and the resources and they can throw stuff at it, and, uh, but they just can't be agile. They're just, it is just not in their nature to be, to be agile and pivot. Around, you know, around whatever the new, new technology or new capabilities are.

Andy: Some, a lot of times I think it's philosophy, it's, it's the culture of the company, you know, and you see that with Apple computer, you see that with Tesla.

You know, Steve Jobs is always famous for saying, don't worry about disruption, you know, dis uh, cannibalizing our own products. You saw that with Kodak, they put, uh, digital photography on a shelf because they were worried about cannibalizing. Yeah, yeah. Right. That's, that's the most famous example. Yep. And, and he was saying, don't worry about it because if we don't cannibalize ourselves, somebody else will.

have to reinforce that they [: 00:28:00

And then there's also a question of focus that comes from the C-Suite. I, I remember seeing an interview not too long ago. A replay of an old interview of Elon Musk. I think he was on, maybe he might've been on 60 Minutes or 2020. It was a big program, and this was probably like 10 or 15 years ago when Tesla was, you know, making very few cars.

They were making the model of roads. Maybe they were on the s or something by that point, but they were, they were not at all threat to Ford or uh, right. Jim or anybody else. And the interviewer, whoever it was, I don't remember, asked him, he said, you know, are you a competing with. Is Ford, a competitor is G And he said, yes, they are.

You know, he said, it's not [: 00:29:00

And if you're. You are minting money with F1 fifties or you know, Cadillacs or whatever, and you're content with that and you don't look over the horizon and see where things might go and put some effort and focus on that. You know, clay Christensen wrote some famous books about this. He's, he's dead now, but dead about 10 years ago.

But he was a academic and, uh, he wrote The Innovator's Dilemma, which is big on this. And uh, and then he came up with the, uh, jobs to Be Done Method, if you've ever heard of that. Very influential guy. He influenced, uh, Steve Jobs quite a bit.

Guy: Yeah. Yeah. Well, you know, it's funny too, just your example there, you know, you think if Ford, you know, they could build an electric or a, you know, gm and of course, you know, Elon's uh, was perfect.

s and he said, you know, the [: 00:30:00

Started up a whole new thing, you know, from one day to the next. Yeah. And, um, and, and that's, and that is the way that, you know, I if, if you want to be agile is a big company, you've gotta take that, you gotta take that chance like that.

Andy: He created, he created a skunk works and that's. There's a lot of famous examples of that, you know, from Henry Ford with, if you read about him in the, you know, around the turn of the century, he was making Model Ts, but he would, you know, block off an area in the factory and that nobody could see it and only restricted to certain personnel, and that was their skunk work.

s and they made, they made a [: 00:31:00

Back in the mini computer days, they had a, uh, a 32 bit machine that was super successful and, and they got caught with their pants down. People were building 64 bit machines, which take three years to build, and they didn't have that much time. They're gonna outta business. So they built a skunkworks and they hired all these young guys who didn't know how long it took to build a new machine.

And they got it done them like nine months. So yeah, they the same, same solution. They built a skunk works. Yeah. Yeah. I mean, em did that a little bit with their PC kind of when they, you know, they were making mainframes. Yeah, yeah. And they built a skunkworks for their first pc.

Guy: Well, but then they hired, uh, uh, where we're trying to hire folks out of Microsoft to, uh, build their operating system.

at's for sure. No. Well, um, [: 00:32:00

Andy: Empathy. Think about your customer.

Think about their problem and empathize with 'em. And if they're not right for you, your, your product or service is not right for them. Empathize with 'em and tell them that, that builds trust and probably in the long run creates a better company, a greater environment and more, and that that customer right, might come back later.

And that, that, that really comes from empathy. Feel their problems.

Guy: Yeah. Yeah, well definitely that. And I guess, uh, even, you know, if you're coming right outta college is, uh, how do you empathize with the interviewer? Because you know what's funny? When you think about the most important product that you're ever gonna sell, it's yourself.

t. You know, you can somehow [: 00:33:00

Andy: That's right. That's right. Think about it from their point of view. Yeah, that's true. So many things, you know, even negotiations, buying a car or whatever. Yeah. It's true. Many things. Understand who you're buying from.

Guy: Yeah. Yeah. Absolutely. Well anyway, Andy, thank you, uh, so much for, uh, participating today and, um, where can folks, uh, reach you to learn more about you and your company?

Hub Hub.

Andy: So you'll find me on LinkedIn. I'm pretty active on LinkedIn. Uh, search under Andy Wick, CWIK. You'll also find me, we have a website, hubbub.me HU. It's just spelled like it is behind me. HUBU b.me. We're not a.com. We're a.me.

Guy: Ah, fantastic. So HUBU b.me. That's right. I love it. Fantastic. Well, Andy, thank you so much again, and for the audience, please stay tuned for many other videos in this series of the backstory on marketing and ai.

And if you'd like to find [: 00:34:00

Andy: Thanks for having me. I enjoyed it.

Guy: Absolutely. Great talking to

you.