Gary Saarenvirta on LEx [Podcast]

Listen to the LEx Member Spotlight Podcast

Daisy Founder and CEO Gary Saarenvirta joins The Leaders Expedition (LEx) HUB Podcast to discuss how artificial intelligence is helping high frequency retail companies make better decisions about the work that they do. Listen to "LEx MEMBER SPOTLIGHT - Gary Saarenvirta" on Spreaker.

Original Source: LEx MEMBER SPOTLIGHT - Gary Saarenvirta

Podcast Transcription

Steven: Hello, I'm Steven Hurley, host of LEx member spotlights. You know, from social activists to boardroom CEOs, from those just starting on life's journey to retirees, the remarkable people of leaders expedition have many remarkable stories to tell; stories about lives lived and passionate side stories that share a common DNA in our tenets and our values. And we believe that these values define the culture of leadership that will guarantee prosperity in a world of abundance. As you're hearing these stories, it's a culture defined by a drive for personal fulfillment enterprise success, environmental sustainability and social justice. Welcome to LEx member spotlights. So, what's so intelligent about artificial intelligence. That's just one of the many questions that I discussed with my guests today on the LEx member spotlight. Gary Saarenvirta is founder and CEO of Daisy intelligence, a company that uses AI to help high frequency retail companies make better decisions about the work that they do. He is also a longtime member of leaders expedition, and has a very strong affinity to the core values at the heart of Lex. So, there's more to unpack than just the groceries in this episode of the lex member spotlight on Wix radio. Gary, Welcome to the lex member spotlight. The spotlights on you this morning.

Gary: Thanks for having me. I'm really excited to have a chat with you this morning.

Steven: Well, it's, it's good to get people excited early in the morning. Artificial intelligence, I want to dig into that a bit, but maybe you could just give us an overview of what you do at Daisy Intelligence, what is that organization all about?

Gary: We're about helping our clients grow their net income. I've been working in this space for 25 years and, you know, learn that the, you know, the current traditional AI approach is really statistical analysis rebranded, and we're using methods from the aerospace industry that have been in play for 50 years. We help companies make smarter decisions, and smarter decisions lead to improved net income, and so for retail we help them decide what products to promote every week, what prices to charge every week, how much inventory to allocate. And those are weekly decisions, our system delivers the decision. Our vision is to kind of take the human out of the loop, let humans do what humans are good at, let machines do what machines are good at. And you know we elevate the human to be the pilot, you know, the analogy is like a fighter jet. Fighter jets are not actually flown by the pilot, the pilot inputs the instructions of what they want to do through the flight controls, and then the computer moves the flight surfaces, you know, 1000 times a second which is beyond human ability.

So, in a few minutes the pilot and the autonomous intelligence then executes the detailed instructions inn the same way our system operates by delivering the decisions. And we also work in insurance doing fraud detection which is again a beyond human ability task, and so those are the things we do, and when clients execute our recommended decisions we've been able to dramatically grow their financial results. You know, in grocery industry almost double or triple total company net income. So, anomic results are, you know, pretty staggering.

Steven: So, this is data based decision making, for, for those industries. What, what do you think the public misunderstands about that term artificial intelligence?

Gary: Well, I think the vast majority of what people call artificial intelligence has nothing to do with artificial intelligence. It's just statistics. If you're learning from historical data, and you need historical data to build like supervised learning or regression models or deep learning models, that's really just statistics and curve fitting. There's no real intelligence in that per se, and the classic understanding with word intelligence means, I mean, it's really all math and science and computing, so I'd say it's a big leap to call that intelligence. I mean, what we do is optimal control, so, like, if you think of an autonomous race car, and driven by computer, It's goal is to determine what's the optimal sequence of steering, brake, and gas to get a fastest lap around the racetrack. And it's governed by the laws of physics, right? And so the laws of physics exist independent of the data. There's no, you know, that was invented, you know, Einstein came up with the laws of physics and Newton came up with the laws of physics with no data. There's no big data back in the Middle Ages, right? or in the early 1900s. And once you have the math, then you say, I want to learn about what are the properties of my universe in the case of the race car. It's the configuration of the track, it's the wind, speed, the density, the friction of the rubber on the road. And so, the data you collect, you learn about the environment. And so, a lap around a racetrack is like a year of business, right? And instead of instead of brake, gas, pedal, and steering for a retailer, it's what's the sequence of promotions, prices, and inventory. And in fraud insurance, it's what's the sequence of fraud underwriting, and auto adjudication that drives the company. And that way, and from the historical data, what you do you're not modeling that, you just learn properties that are fundamental elements of mathematical theory, and all this that I've described is the scientific method, which has been used for hundreds of years. Not a single thing in our life was built with a statistical model. And so, I think this was, you know, aerospace has been doing this since the 1950s, and the whole scientific method has been around since, you know, the 1600s, right? And so, that's what we're applying. This whole data first approach is, it doesn't work, and the world hasn't realized that, and once they do, they'll come back to the scientific method and do what we do.

Steven: So, it's beyond just collecting data. You bring some intelligence to it through, is it through algorithms that you that you do what you do? and we hear that word a lot now.

Gary: Well, it's a computer, computer software is part of the we design. Like the, you know, so, if you think of the laws of physics or a set of differential equations, we've designed differential equations for retail and differential equations for insurance, and we design a set of differential equations for every industry, and then we solve those equations. And those equations include things in retail, like customers don't buy products they buy solutions. That means every product has a halo effect, you know, if you're buying ground beef to make an Italian dinner, and ground beef is promoted, you'll say, Oh, I should make an Italian dinner, and you go buy pasta, tomato sauce, bread, cheese. You know, so every customer that buys ground beef buys other things, because you don't consume raw ground beef. So that's one fundamental, so from the data you learn what are the halo products that are associated with every single product. And then you think what are the, what are the cannibalization, because the customer said

I'm going to buy ground beef to make hamburgers, well, hot dog sales go down. So, there's a negative effect as well. Which products are cannibalized when customers buy product A. And then, because it's on sale, customers buy a two, three week supply. That pull forward, all those ripple effects, are like more than the number of molecules in the, in the universe, and so it's beyond human ability to do that. So from the data, we learn those properties. And then we plug them into our differential equations, and we use computers to solve the equations.

Steven: So, how do you figure in the psychology of all this? is that part of the equation as well?

Gary: Human psychology, I mean, the human psychology is present in the data. We learn that on the properties, that it's the human use cases it's, you know, that we see what people, the patterns and combinations of products that people buy, that’s there. We see price elasticity. When you lower the price, people buy more. When you promote a product, people buy more, you know. So that's the human psychology, it's captured in the effect in the data, and we see that we learn all those properties from the historical transactions. And humans are very, you know, their behaviors very up, when you look aggregated up to millions of people, it's very constant. And, you know, the recipe for an Italian dinner hasn't changed in 1000 years, what people do with Cheerios hasn't changed, and in decades, what people do with lettuce and tomatoes hasn't changed in decades, so there's, there's a lot of, you know, companies focus on the edge cases, these shiny little New Edge details, but the core things haven't changed in hundreds of years.

Steven: Well, it's interesting. As, as the shopper in the family, the grocery shopper in the family, I'm a little, not disturbed, but I am a little taken aback by what you're saying, because I tend to think that I make decisions, just personal creative imaginative decisions on what's going to be for dinner, but I know during this COVID-19 shutdown, my shopping habits have changed. It's, it's, there's a lot more tension in the grocery store. There's a lot more speed at which I want to move through the process, and I'm wondering, does that figure into what you're doing most recently?

Gary: Absolutely. I mean, so, we learn the patterns from the data literally every day. And so, we can see the patterns changing dramatically. So during COVID, we observed, you know, all these things I talked about, Halo, pull forward, elasticity, all of that, you know. We see those dramatic changes, so if you're doing, you know, statistical modeling, this recent event, COVID pandemic pantry loading, screws up all the historical data, so your statistical models will no longer work, whereas our approach, because we have a mathematical theory, we can learn the new properties day by day, and we can see the change in the patterns. Basket sizes going up, certain categories being bought, the use case, the solutions that people are buying dramatically changing, and we see that there has been kind of three phases that we've observed in retail, you know, it's going to give us kind of panic buying. Then some kind of somewhat return to normalcy. But there's new categories emerging as people are at home and they're looking to do more, more than they normally do, and they're trying new things, or entertainment type activities, hair products, you know, we see all of that. And there'll be some new normal emerging which hasn't started to happen yet. I think we're kind of in phase two, and the new normal will be something new that won't look anything like it looked like before, so if your technology is based on historical data only, I think all those approaches will break, or if you're a company that just says let's just do what we did same week last year, that's not going to work anymore because the new normal will be different. And I think, I'm hoping, that our approach can really help.

Steven: Well I, for the past 10 years, I have been making my own pizza on Friday nights, and I am almost out of yeast, and I don't know what's happening in the yeast market, but I have a feeling it has to do with people wanting to feel that they are more at home and making more.

Gary: Yeah, we've seen baking products are going up, and we've now, we've seen, you know hair products were going up, and entertainment, fitness equipment is like sold out. You know, it's people are ordering more online, so the whole, people's patterns have completely been thrown up in the air, and it's a big fight for customer service and survival for a lot of mid market retailers. They don't realize that they're seeing increased sales through increased baskets, but what they're seeing is that maybe they could be losing customers, because now, you know, there's 20 million new customers, and, trying eCommerce for the first time, and some of them may never come back to bricks and mortar. There's, you're not going to two or three stores anymore, and most people are picking one store. So, and to your point, your patterns of change, you're, you're doing different things getting through the store, so it's a real fight for customers at the current moment for retail, and the more insight they have into what's going on, the more they can service their customers better, and be in better shape after this new normal occurs.

Steven: So, your insights come from constant flow of current data. And so, where I guess that, that whole ecosystem has changed over the past number of years with, with more data mining technology out there, I mean, we probably as consumers are unaware of all of the places our data goes.

Gary: I think, I mean, the transactions are what we primarily do. We're not, we don't, we don't have to have personal identifying information on the transactions and retail. Certainly in the, in the health insurance side we. We do have personally identifying information, and we're looking for fraud. But, you know, we can see the patterns occurring in transaction receipts. Having personal information is helpful, and I think, you know, companies don't look at an individual, and they, you know, spy on you. I think, you know, they're looking at aggregate patterns to try to service their customers better, so if you buy more from one store, that store is meeting your needs and it's good for you. And I think as long as we observe the government regulatory rules around use of that data, securing that data, or protecting that data, I think companies are just trying to service their customers better, and the data we collect is from that perspective. And certainly as long as we put in the safety protocols to protect that actors, I think, for the vast majority, our usage is very positive and targeted at improving the quality of customer service, adding value.

Steven: So, this is obviously something that has captured your imagination for many years. What is it about this? Is it the, is it the the logic behind it? Or the, the, the impact that you're making? What is it about, about what you're doing that gets you up early in the morning and maybe even keeps you awake at night?

Gary: Well, my goal was to, you know, the goal is to reduce poverty, you know. And I think I thought my, you know, and I went came out of school with a graduate degree in aerospace engineering, and I ran into this opportunity to say wow I can bring more science, math and science, to businesses, and it works like this: if I make if I make a grocery store more profitable, what they're going to do is lower prices, because that's what's gonna happen. There's gonna be another price holy war and the new, this new normal, because of a recession. People have less disposable income. So, it's going to be more promotion intensive than ever. But, when we, when customers are, when companies are profitable, smart companies reinvest in price and innovation, and that means lower prices for you and me. And if we do that in every industry, lower the price of goods and services in every industry, that means the cost of living goes down, and if the cost of living goes down for all of us, then poverty goes down. And that's really my

motivation, that's what I want to change in the world, and you'll help everybody have a fair shot at life and living a good life, and we'd be efficient. So, making the world smarter I think leads to leads to reduction in poverty. And I can't control what our clients do, but I know that the smartest companies do what I say, and retail has been being a race to the bottom in price for a long time, and will continue to be so.

Steven: So, as CEO of Daisy intelligence, you're the one that's ultimately charged with looking around and seeing those signals of change. And I remember being at Palo Alto a few years ago taking some course in strategic foresight, and there were, I think the majority of people there were from the grocery retail industry, and we were talking about signals of change. And those emergent questions. What are you seeing on the horizon, not only given the current COVID situation, but other factors in the in the economy and the environment that that are going to impact the work that you do?

Gary: I think there'll be more and more need for us. I think as the world, you know, continues in a challenge, like in a pandemic, it should hopefully wake companies up. I think to, you know, get efficient, use technology for efficiency. Imagine if we didn't have Zoom or GoToMeeting, we wouldn't be able to work. Like so, there's a really great use case for look at this technology that we just had around, really enables us to be way more productive through this pandemic than we otherwise would have been. In the same way, the artificial intelligence the technology that we use can really help businesses operate smarter. And so, they're, you know, in some industries, there will be some displacement, because they're very inefficient and companies will look to reduce cost. And Intelligent Automation is going to be a key, key way forward. If you look at the stock market, that's a good proxy, what's, what's going to infer what's going to happen, you know. In the day, I think 60 to 70% of all stock market transactions are happened on autonomous computers. And so, that's what's going to happen in every industry, and that's good for humanity. There'll be, will be some displacement in some industries, not as much as people think, but there'll be more new work created. I think we just need to support the people who are displaced, and can't be retrained for the new worlds, and as long as we do that then I think this progress is important and this technology will help the world be more efficient. Resources are getting more scarce, there's massive universal problems that we need to solve like climate and other and, you know, this political climate we're in as well, I think all these things are big problems that technology can really help with, and I think we need to listen more to the scientists, you know. I saw a great LinkedIn meme that said every disaster movie starts with politicians not listening to scientists, and I think that's a great, great thing to be looking at what's going on in the world today. If we just listen to science, and fact, and reason, we could have gotten way out in front of this pandemic three months earlier, and it wouldn't be nearly as bad as it is right now.

Steven: Well, that political science dynamic is something that, it's a bit of a juggernaut, but it'll be interesting to see how this plays out. Because there are, you know, there are front end finances or economies that work here, and all of that is Uber political in these times. And so, we can hope for the best, but I'm, I'm interested to find out about your relationship with LEx leaders expedition. You were an early adopter, I met you at one of the first LEx lunches a number of years ago. What drew you into the community initially and what keeps you here?

Gary: I think it's the, I think there's, I stand for what what LEx is all about, which is this improving the quality of leadership and bringing more humanity to leadership, and I think that, I think we need more responsible leaders to take us through the changes that are happening in the world today. And, there's really a dearth of leadership, so that, that really is one part. And then the charter really connects with me, I think some of the values that are in the charter, like honesty and transparency, equality of opportunity for everybody, those are, you know, perseverance and surviving through anything, you know, those are things that are my core, personal core values. So, I think probably half of the charter values are on my core values list, and that, that's why, you know, I love, I have great love and empathy for people and humanity and, you know, I think it's important for us to take responsibility. People who have the opportunity, I think my gift that I've been given is to try to help the world and, you know, and I think LEx is, is doing a great, a great thing, and anything I can do to help LEx grow and propagate the charter and the vision I think is important.

Steven: Talk to me a little bit about that, and I think it's a false dichotomy, but sometimes there's a dichotomy set up that its profit and corporations against social ideals, and those values that you talk about. And I know that in your life that's, that's not true, but can you say something about that?

Gary: Well, I think, I mean, profit is a part of it. I mean, they, so, if you're, if you're a socially responsible business, you have shareholders, you have to deliver profit, because you need resources to achieve social change. You can't do social change with just desire and the will of one and the will of people only. There needs to be resources, and that resources as money and capabilities. So yeah, you can be socially responsible and focus on profit. So, it's just what the, what's the company mission. So, we want to build a highly profitable, highly financially successful business, because that means I'll have more resources to do good things, you know. And as our company has grown, we've been able to do more and more positive social things, support kind of charities in our local community, and that's only because we've been able to be profitable. So, it's, responsible leadership and profitability are not at odds with each other. It's, it's defining a mission that has something that, that is more than just profits. And, I think if you look at, there's a study by a u of t, Professor that said since the 1970s, the return of, the return on investment of companies has been lower than prior to that because the world changed, and in the 70s where companies became more profit focused, whereas before that they were more mission focused, and I think we need to get back to what's your mission, focus on the mission, and then, then the profits will also come. So, I think the two go hand in hand. You can't be a business and make more profits, that's not responsible.

Steven: So yeah, Gary, does that require different leadership capacity than maybe it did 30 or 40 years ago?

Gary: Um, well I think, I think it was there in the pre 70s, we were more mission focused, I think it's been very financially focused I think. It does require some type of leadership, that you have social awareness of the impact that the decisions you make on your employees, on your customers, on, on humanity at large. So, I think it takes them, you know, like the kind of leaders that LEx is, saying that we need more of that. You know, leaders who are both business savvy and socially aware, and I think that's a different type of leader than, you know, I think a lot of companies were, you know, hiring ruthless, business only oriented leadership, and there's a lot of companies that are doing the right things. I'm not saying that there's nobody who's being socially responsible, and, you know, we're all human beings, we try our best, and I certainly am not perfect, but, you know, we try to be, trying to, try to do both as best as you can.

Steven: So, let me ask you one one final question, and this is kind of a selfish question, because I'm always on the lookout for resources, good information, and good books. Is there a, is there a book on your bookshelf, or maybe on your bedside table, that, that kind of captures some of the imagination that we're going to need moving into the future? something you're reading now that's particularly motivating for you.

Gary: I mean, for me lately I've been reading books on leadership, right? Working on my personal human to human skills because I know my very technical background. And so, I spent, you know, two decades focused on technical capabilities, so I think the books I'm reading today are more about leadership and humanity, and that, reading books like Lencioni books on how to build great leadership teams, reading those types of books. So, it's not really what the future of technology i, that's my expertise, and I, you know, I read a little bit about that. But so, I'm not sure if the books I'm reading would get you what you're looking for, right?

Steven: Well, it could be, could be, and I think that, that resonates. What was that? what was the Len?

Gary: He was Lencioni , Patrick Lencioni, and it's l-e-n-c-i-o-n-i. And he's got a number of books like The Five, The Five Habits of Successful Leaders, and i think there's about six or seven books, really great books, easy reads, great common sense. I'll, I'll email you some of the names of the books. Off the top of my head, I don't remember the titles, but it's a whole series of them. Yeah, if you go on Amazon you'll they'll pop up right away. Very popular books.

Steven: Fantastic. Well, Gary Saarenvirta, thank you so much for the time, and thank you for your work in this leaders expedition community.

Gary: oh, thanks very much, I enjoyed having a chat this morning. Have a great day.

Steven: Find out more about Gary's work by visiting his website daisyintelligence.com, and to find out more about leaders expedition, and the LEx organization, visit our website leadersx.org. I'm Steven Hurley for next radio. Thanks for listening.

Subscribe to our newsletter and learn how AI will provide a competitive edge and elevate your people to do the tasks only they do best

Sign me up