Artificial Intelligence is Coming! Are You Prepared?

[00:00:00.000] - Dan Paulson

Hello and welcome to Books in the Biz. I am here with Rich Veltre. Rich, how are you doing? I'm good, Dan. How are you? I am doing wonderful. And we have another guest today, and this is going to be interesting discussion in AI, Mark Steuss. And your company is proof analytics, correct?

[00:00:24.420] - Rich Veltre

That is correct.

[00:00:25.380] - Dan Paulson

Awesome. Got it right. And working on these name things. Oh, Mark, We connected a little bit ago. And like I said, we've done our top 10 for 2023. And one of the things we did talk about was AI, and it was quite apropos that you popped up in my feed and would love to get an understanding of what you're doing in AI and how it's going to impact business. So why don't we start by, why don't you tell us what proof analytics is?

[00:00:54.220] - Mark Stouse

Proof analytics is essentially what we've done is we've taken classic multivariable regression and mixed it up with some AI, what's known today as causal AI, to enable users to create models quickly that assess cause and effect, mainly in the go to market area. But we have customers that do it more broadly than that. And so what's the effect It's the effect of all of my marketing spend on my sales performance, right? It's a classic example of cause and effect analysis. It includes things like time lag, which is super important because nothing happens right away. And if you don't have an assessment of time lag in there, you will never find the value because you won't know where in the calendar to look. So it's that thing. It is the fundamentals of any budgetary business case that's ever been written. For the last 15 years, business cases have faded because the cost and availability of cash was so low and easy, right? That everybody just said, who cares, right? And that has obviously changed a lot. And so the budgetary Derry business case has new vogue today in the corporations, and probably will stay that way for at least the balance of my career and probably a lot of your listeners as well.

[00:02:43.830] - Dan Paulson

Okay. So I know you mentioned go-to-market is the target for this. Walk me through an example of where this would be used or what type of products this would be used on.

[00:02:58.170] - Mark Stouse

It actually cuts across all industries, all vertical. It's not even about marketing, per se. This is the classic math that underlies the scientific method of inquiry. So it doesn't It could be matter what you're studying. It could be climate change. It could be epidemiology. It could be the impact of all the different components of go-to-market on each other. Net of all of the externalities that are either speeding you up or slowing you down. So if you want to understand... So the big word that everyone's tossing around right now is efficiency, cost efficiency, things like that, right? But before you can do that, you have to know what's effective, what's working, what's not working. In fact, the biggest contributions to making your overall operation more efficient is to find out what's not working and to stop doing that thing, right?

[00:04:02.910] - Dan Paulson

I talk a lot about that. Yeah.

[00:04:04.520] - Mark Stouse

Yeah. And so that's really what we do. I mean, we do some services around this, but we're primarily a SAS company. So this is a piece of software that is streamlined so that analysts on the back-end can model things very quickly to meet the needs of business users. And the business users can look at it and go, you know what? I totally understand what that screen is telling me, and I know now how to make a better decision this time than I did last time. The law of compounding is really at the heart of all this, right? As it always is. I mean, let's just say you have to make this decision 365 days a year, and your goal is that you're a half a point better each day, which is not a huge lift, right? Proof can help you do that. The math, forgetting the product for a second, the underlying math can help you do that. If you aggregate that over a whole year, you're talking about something like almost 2000 % improvement annualized. Anybody would say that That's a major triumph. If you cut it in half, it's still a triumph. If you cut it in half again, that's still good.

[00:05:39.030] - Mark Stouse

That's really the essence of it right there. But the only people who historically have thought about this issue in those terms are finance people or business people, business leaders who happen to have that bent. So the last 18 to 24 months has, to varying degrees, obviously, this is a bell curve, right, of people, convinced more and more and more people to have that point of view. And so not just for proof, but for the whole analytics business in general, right? We're in a hot zone right now.

[00:06:25.700] - Dan Paulson

So how did you get down this path? Is your background primarily No, actually, I'm a marketer, right?

[00:06:34.130] - Mark Stouse

Okay. So I came into it the other way, right? I got 20 years ago, I was at HP working for Mark Hurd, the CEO at that time, who was skeptical about the value of marketing and its impact on the business. And he he had a way of really getting in your face about it. And so things got rather existential. And so I was just sitting there saying to myself, I've got to do something to fix this issue, this question, answer this question. And I freely admit at that time, I was doing it purely ego reasons. I just couldn't stand the idea that I was not being thought of as significant and meaningful. Or I had to just go do something else. That was like, I'm so out of here if I can't fix this problem. I started reacquaining myself, by accident, actually, with math, because I'd been on it. This is actually one of the great, hilarious paradoxes of my career, because I was in total a math avoidance at this point. And now I'm the CEO of an analytics company. So the universe has a sense of humor for sure. So I started down this path, right?

[00:08:18.920] - Mark Stouse

And because I was highly incented, personally, even though it was ego, so not the best reason, I started to really master the and a half, and I started gradually scaling it. I got a lot of positive feedback across multiple companies about this. I mean, every CEO and CFO that I worked for, post-Mark heard, just thought it was just the cat's meow. And so by the time I was CMO of Honeywell Aerospaces under Dave Cody, who was my sponsor, my patron, in a sense, right? We had gotten it to a very high degree of maturity and success, and it was at that point that you started to realize what the real problems were. The real problems weren't solving the big questions about marketing's impact on sales. The real challenge was, how do we operationalize analytics at the clock speed of the business so that the output of the analytics is always relevant and becomes part of the warp and woof of the way that businesses make decisions? Because Because right now, even, data science teams are seen as a day late and a dollar short in many cases, right? They're asked a question. They don't have the answer.

[00:09:57.130] - Mark Stouse

The business team goes, well, we got to make We're going to make the decision, so we're going to make the decision. And then the data science team shows up three months later with a bunch of insights, and everybody said, man, that would have been great three months ago, but really don't care today, right? So we were And we solved it at Honeywell by brute force. The whole effort was seen as so valuable that we were able to spend $8, $9 million a year on on just that. And it was mainly to hire more and more and more data scientists, right? To get the throughput and get the latency down to where it needed to be. So we didn't have automation. And so we were just doing what has been the case since 1834 in the industrial revolution, right? Just stack more people into the equation, right? So So you didn't have to be a rocket scientist, though, to figure out that, A, this was really valuable, and B, there weren't a whole lot of companies who are going to be willing to spend eight or $9 million a year on it, and that coming out of the software industry, which is where I came out of, we just knew that automation was going to be a big part of this, and a certain aspect of AI was going to be a really big part of it.

[00:11:28.860] - Mark Stouse

And so that And that's what we ended up building proof on that basis, right? And it's really been an adventure. I mean, for all of the math and the automation and the software and the science and all this stuff, the journey into the human psyche and the human soul as it pertains to analytics and AI and things like that has been extraordinary, right? And not always pleasant, I might add.

[00:12:07.770] - Dan Paulson

And I'll let Rich jump in here because I'm sure he's got a ton of questions about the financial stuff of it. But I know you talked a lot about your marketing background. One of the biggest challenges that I think many, if all companies have is, okay, I'm spending a dollar on marketing, what is it actually producing? Really, how does AI help with that in real time versus, as you pointed out, it being a two to three-month lag of finding out information that by that point is obsolete.

[00:12:42.090] - Mark Stouse

Yeah, so I mean, the nut of the whole thing And the second one, is that cause and effect is understood today, and it has been the case for several hundred years, actually. It's understood through what's called multivariable variable regression. It's come a long way. There's been a lot of enhancements and all this stuff to the math, but essentially, that's it. It's been a slow, laborious calculation process. So the learnings from regression, historically, have always lagged the need for the insight, right? Need for the insight is right now. Got to know right now. Well, regression is historically anyway, not been able to meet that deadline. We were one of, if not the first to automate it, and also to add AI into the mix, specifically to help analysts. So our platform is built for two personas, right? The data scientists or the data analyst on one side, and the business user on the other. So the AI helps the analyst create models much, much faster. So what does that really mean? Well, to create a normal regression model is measured in days. To do it in proof with all the AI generated help and prompts and all this stuff, it's hours.

[00:14:35.540] - Mark Stouse

It's like single digit hours. So this is also very much of an agile approach Whereas historically, it's been more waterfall. So you're able to create a bunch of what we call minimum viable models. I mean, that's an extension of the agile methodology right there. So very focused model The business user says, Man, I've got this question. I don't need a giant mega model that's going to answer all my questions forever and ever and ever. I need this question answered with an appropriate amount of context. So it becomes very fast, again, the AI helps with this, for the analyst and the business user to collaborate within proof and get to a model that answers the question, at which point it goes into production in proof, and every time new data is presented to that model, it automatically recalculates it, right? So if you are presenting new data, to the model weekly or daily or whatever it is, it's going to recalculate. So from that point forward, proof actually operates very much, in fact, like virtually identical to a A GPS. So when you're using the GPS on your phone, it has the history, which is you're here, right?

[00:16:10.650] - Mark Stouse

You can specify where you need to go, what's your goal, what's your objective, right? It's going to give you usually three choices, three routes, right? These are forecasts. This is based upon current traffic patterns, weather patterns, all this externalities, right? And you're going to pick one, and you're going to be cruising. It's going to be tracking you as you move down that path. And it's not just going to be tracking you, it's going to be tracking all these factors that can either speed you up or slow you down. Let's just say that all of a sudden, there's a wreck ahead, and traffic starts piling up, and it says, Hey, Rich, guess what? This route that you're on, it was an awesome route. It was going to get you there like, no problem. It was going to be really fast. This is going to be great. And it's a beautiful route, too. So you're going to enjoy it. But the problem is things have changed. So what is the number one word that we're all living with right now, it's change. It's change. It's changed. And it's change we don't control. We are surfing the wave, not to mix a metaphor here, but we are surfing the wave that we do not control Control.

[00:17:30.530] - Mark Stouse

So things have changed. So guess what? You still need to get to this meeting, this dinner, whatever it is, right? So we're going to have to reroute you. And if you do the new route, You'll be eight minutes late. But if you stay where you are, you're going to be two hours late. So we don't really recommend that. That is almost exactly what proof does for the business on any business question. And if you stop and think about it, most of life's questions, most business questions, are rooted in causality, and they're all about... It's essentially a navigation question. Where am I? Where do I need to go? What do I need to do to get there? What's going to speed me up or slow me down? Do I have enough resources to make the journey? If all of a sudden the parameters of the journey change, how do I get more resources if I need them? All those kinds of questions you answer with something like proof.