A Chat with Randy Bean on His Book, Fail Fast, Learn Faster

00:00:00

Good afternoon and good morning, and even good evening to some of you. Thanks for joining our webinar today. My name is Beth Befferly, and I'm the VP of marketing at DataKitchen, and I'll be the host today. So we are very excited to have a very special guest, Randy Bean, join us today. He's the CEO of New Vantage Partners and an expert on the topic of data-driven business leadership. Many of you may be familiar with his work, such as his annual CDO survey and frequent columns in Forbes, Harvard Business Review, and The Wall Street Journal.

So in August, he'll be publishing his new book called "Fail Fast, Learn Faster: Lessons in Data-Driven Leadership in an Age of Disruption, Big Data, and AI." Today, we'll be talking with him about the book, and he'll share some of the lessons that he learned in writing it. He'll also be joined by Chris Bergh, DataKitchen's founder and CEO.

Chris is a leader of the DataOps movement. He has more than 30 years of research, software engineering, data analytics, and executive management experience. At various points in his career, he's been a COO, CTO, VP, and director of engineering. He's also the co-author of "The DataOps Cookbook" and "The DataOps Manifesto." So Randy will kick off the webinar by sharing some background about the book, and then we'll spend the majority of the time having a conversation about many of the main concepts in the book, which I'm sure will be a super stimulating discussion.

Before we jump right into that, though, just a few housekeeping items. So we hope to have lots of audience participation, so please enter your questions in the Q&A box on the control panel, and we'll make sure we get through as many of those as we can during the course of the webinar. The webinar is being recorded, so if you miss something, you'll have an opportunity to listen to it again later, and we'll send out that recording to you via email within 48 hours of the webinar. And then lastly, as mentioned on the registration page, a fun thing today is that we'll be giving away a copy of the book, "Fail Fast, Learn Faster," to 20 lucky webinar attendees. So we'll choose the winners at random after the webinar, and we'll notify you about that via email if you're a winner.

So at that time, we'll need to get your address so the publisher can send you that book as soon as it's published. So that's all the housekeeping. I think we're ready to go. So with that, I will hand it over to you, Randy, to kick us off. Great. Thank you, Beth. Very much appreciated, and nice to see you today, Chris. And what you didn't mention, I guess Chris is quite a chef as well from what I understand. At least from- Not for food, for data, but not food at all, except if it's grilling .

Well, data, that's what's important. So maybe we'll have some barbecued data after we . So thank you for your time today. I look forward to the conversation. Beth mentioned at the outset, she described me as an expert. I'm always leery of the word expert because I believe there's so much to learn, and that's one of the things that I try to convey in this book. This is the first book that I've written.

I've written many articles over the years for Wall Street Journal and Forbes and Harvard Business Review and MIT Sloan Review, and I've really resisted the notion of writing a book because I spend a lot of my time doing consulting. But with the pandemic of the past two years, wasn't able to travel, so had more time than usual, and as a consequence, agreed to write this book for Wiley and Company, Wiley & Sons. And it really shares my experiences over the past generation, 35-plus years now, working with data, seeing how organizations leverage data, seeing the challenges they face, as well as a look of what this means for the future in the years ahead.

So as Beth mentioned, the title of the book is "Fail Fast, Learn Faster: Lessons in Data-Driven Leadership in an Age of Disruption, Big Data, and AI." So that's a mouthful. So moving to the next slide by way of background.

Why this book and why now? First of all, as mentioned, over the past 35 years, I've had a front seat to the data revolution. My first professional job was with a predecessor bank to Bank of America, Bank of Boston, and I was interestingly enough trained as a COBOL and assembler programmer, but I was really more interested in the data that was being manipulated. And one day I said to one of my colleagues, "What do we do with all of this data that we're storing up?" And they said, "Oh, well, the regulators make us hold onto it for seven years, and then we're able to destroy it." And I thought, like, "Wow, isn't that kind of missing the

00:05:00

point? Here's a trove of information that can be analyzed to gain insights into your customers and the market." And that really precipitated my interest in data in the ensuing years. During much of that time, the past 20 years, I've served as an advisor to leading Fortune 1000 companies, particularly in financial services, life sciences, and healthcare, on how they can become data-driven and develop data cultures. When the term big data came into vogue and data really became something that was popularized with a more general audience as well as greater awareness within the c-suite, I was invited to write a monthly column on big data in The Wall Street Journal. This was in 2014, 2015. Have subsequently continued that with Forbes, and I'm a frequent contributor for Harvard Business Review and MIT Sloan Review and have a series of articles this year in both of those publications, an MIT article just last week and a HBR article a few weeks ago co-authored by former American Express president Ash Gupta So data is critically important as we all know, and I'm sure that's why you've joined in today. But now more than ever, and I think a lot of that was highlighted by the COVID epidemic and the need to look at data in terms of understanding the risk factors and later development of vaccines.

So people who never paid attention to data were all of a sudden paying attention. And I just want to read briefly this opening quote from the book. "The world is in a race to become data-driven, now more than ever. The warp-speed effort to organize scientific and epidemiological data from across the globe in a heroic effort to find a COVID-19 vaccine has illustrated the urgency and existential nature of this quest.

We need data, science, facts, knowledge, and insight to make informed, wise, and critical decisions. Now more than ever, data matters, and having good data matters tremendously." So that's really the core message of the book. And moving to the next slide, I'll talk a little bit more about who the book is written for and how it's organized.

So basically, I wrote this book for three audiences: CEOs and senior business decision-makers, people who sit in the C-suite on corporate boards. And part of this was because these are people that care less about how data is manipulated, if you will. They care about what is the benefit, what is the business value. So I really want to translate some of the complexities of data management and data science into terms that a CEO could understand. And that's what I do in all the articles I write.

Sometimes I tell people that I'm writing at the third-grade level, but what I mean by that is I'm trying to simplify things so that any executive can understand. I've been in conversations where all of the complexities of data environments and movements of data have been explained to executives, and they kind of get a blank look in their eyes until at the end, they ask, "What is the ultimate value to me?" And often, it's as simple as being able to do something faster, being able to do something more cost-effectively, and have trust in the data that you have.

The second audience was really industry professionals. These are the practitioners. These are the individuals who work with data every day. And I've included about 25 case studies from leading Fortune 1000 companies and other firms to really show how these organizations are tackling data issues, and so practitioners can learn from the experience of others. And then finally, I wrote for general readers.

I run into a lot of people that ask me what I do, and basically, what I try to do in the book is to help educate and help general readers understand what is all the fuss about data, why does data matter, why is it important to all of us?

In terms of the title, the title actually comes from an Irish playwright, so it was a good opportunity to mix literature and business. And I came across this quote a number of years ago. I was actually watching Wimbledon, I believe, and I saw the Swiss tennis player, Stan Wawrinka, and he had a tattoo on his forearm, and it said, "Ever tried, ever failed, no matter. Try again, fail again, fail better." And I thought that was fabulous because it was really a metaphor for continuous improvement, learning from failure, how you can get better, how you should never be complacent, how good

00:10:00

is never good enough. And so that inspired me to look into the source of this quote. I learned it was from the Irish playwright, Samuel Beckett. And I was looking for something bold and provocative. Facebook has had the mantra, "Move fast and break things." But I want to imply something about learning from experience.

Nothing ventured, nothing gained. An iterative process, reducing cycle times, trial and error, test and learn, which is very popular in the data science field. So hence the title, and this is really also about the lessons that can be taken away that everybody can learn from.

So moving to the key themes of the book and how the book's organized. I organized it into 10 chapters. There's a forward by my colleague, Tom Davenport, and I start off with a little history of big data and data initiatives, really over the past 35 years. Sometimes people ask me what's different and what's similar, and one of the things that I say is that 35 years ago, people were saying, "How can we learn from the data that we have?" And people are still saying this.

But the tools and technologies and techniques and capabilities to learn from and analyze data have vastly improved. Data has also proliferated at the rate that there's more and more data to deal with. A second theme of the book is thinking different. Organizations are often challenged in terms of becoming data-driven, and they need to think about new ways and have new mindset in terms of how they proceed.

I have a few chapters that are interludes, so I have one on insight and knowledge, and I take a look at examples of how data science and facts are used in decision-making. In chapter four, I refer to the annual survey that we've been conducting for 10 years of C executives from leading Fortune 1000 companies and provide a snapshot, a glimpse at the state of data in the corporate world today.

And then a deep discussion around one of the major challenges that all companies face, and that's establishing a data culture. For most organizations, the vast majority in our survey, it's roughly 95%, they indicate that the biggest challenge to success is not really around technology, it's really around the cultural issues that are impediments to organizations fully embracing and deriving business value. I also speak about the rise of the chief data officer role. In 2012, only a handful of organizations had established this role, and now it's become pretty much an industry standard for major organizations. I also speak briefly on the issue of data ethics, and some of the challenges and ways that data can be misused as well.

A couple of people called this the most terrifying chapter of the book, and sometimes people ask me what's my favorite chapter, and I point to this chapter because it's different and it's cautionary. A little bit then about data disruption, innovation, and how data can be used to change industries. And then I wrap up with two chapters, one, a glimpse of the future and data-driven AI, and how AI has been enabled by greater volumes of data and faster access to data.

And then finally, a closing chapter on data-driven leadership and one company's odyssey over the past decade in becoming data-driven. And then in conclusion, speaking about data as a journey, not just as a destination that you get to, but as something that you need to continually strive for excellence in. As mentioned, I include a series of real-world case studies on the next slide.

Here we go. So, companies such as American Express, Charles Schwab, MasterCard, Bloomberg, among others, Citizens Bank, Capital One. A number of these firms are recognized leaders in terms of being data-driven, and telling their story helps other firms learn from their experience and helps practitioners understand some of the approaches and methodologies these companies have taken. And then lastly, just in terms of the book, I have two slides in closing on the next slide.

There we go. Some advanced praise. So, sent early copies of the book out to a number of folks in the field. So one of those was David Edelman, former Chief Marketing Officer at Aetna, who led

00:15:00

McKinsey's digital practice for many years, and he called it, "Inspiring and terrifying and should be required reading across C-suites and boards." Allison Sygraves, who's Chief Data Officer at M&T Bank and a global Data Powerwoman 2020 leader, as designated by CDO Magazine. She adds her comments. And Geoffrey Moore, who's well known for writing "Crossing the Chasm," he says, "Big data is indeed crossing the chasm, and there's a wonderful hoard of business anecdotes for readers to learn from." And then lastly, one additional sets of comments from Don Peppers and Martha Rogers, who a couple of decades ago wrote the classic book on one-to-one marketing, and they point to the many case studies.

Cindy Hausen, who's very active in the contemporary world with her Chief Data podcast, talks again about, "A must-read for CEOs, CDOs, and all data and analytics leaders." And Cameron Kerry from the Brookings Institution, also a brother of John Kerry, who says, "Data has become an essential element of business strategy in the 21st century." So that's a little background on the book and the process and what it's all about, and I look forward to now engaging in a discussion with Chris.

Yes, great. Thank you so much, Randy, for giving us that overview of the book. There's certainly a lot to unwrap here. So I'm going to, let's see, start. I have a bunch of questions for you both, and then also encourage, again, anyone in the audience to submit your questions, and we'll get those into the discussion.

So just to start out, Randy, you talked about how companies need to have a different mindset to become data-driven. Can you talk a little more about that? Why is that? Yeah. I think so much of this relates to change, and that to become data-driven or to use data effectively in your business, you have to

change your technologies in many ways. And sometimes you have to change your people and skill sets, but you also have to change your processes and how you approach things. Data is typically something that flows across an entire organization in many organizations and are organized on that basis. So it's often fragmented. It exists in silos. And so for an organization to get their arms around data is really a different sort of problem and something that many organizations have not thought about. And change, even when people pay lip service to it and say, "We embrace change, we love change," usually they like it when it's change that impacts somebody else and less when it's change that impacts them and means they have to learn new things, do their jobs differently, have some of their responsibilities shifted to others or have new responsibilities shifted to them.

So there's many components of the mindset issue, but really I think they all emanate from change and what that implies for an organization. And Chris, would you agree with that in your experience? Yeah. And sort of reading Randy's book before the meeting, it caused me to reflect on my own career. And I think what's great about Randy's book is it talks about,

in a lot of ways, how an executive can help get their company to be data-driven. Right? See the business value of data, see the value in data literacy and a data culture, and see the value in learning and a fail fast culture. And I think all those things are incredibly important. And in my career, I've always kind of been the guy that a person like a CDO would go and say, "I convinced the organization to do this. Now, Chris, help me figure out exactly how we're going to do it," and sit down with the people who are actually going to do the work.

And there's a lot of aspects to that once you've decided to be data-driven and work in a more iterative failure way that is the culture of the team that does the work with the tools that they have. And I think that's reflected in the process view that Randy has in his book. It's very little about a faster database or a better tool.

It's like, how do you get a team of people to really embody failing and learning? Because we're all professionals, and we all have positional authority and reputations in our organizations, and failing's hard for people. And the shame that's associated with failure, the risk that's associated with failure, is hard for people. And putting that culture from shame and blame to a learning

00:20:00

culture, from trying to do things perfectly to getting something out there that's 70% right, and then focusing on the infrastructure to do that, that's been my journey with running data teams

and trying to actually live the data-driven, fail fast, learn faster life. By the way, some people have told me that their favorite chapter of the book is the closing chapter, the one on data-driven leadership. The one company's odyssey is actually American Express, who I had interviewed and profiled in articles, probably about six articles over the course of the decade.

And what's nice is that in the later articles that I profiled, American Express was able to be very honest and say, "We tried this and we failed, and we tried this and we failed, and we tried this and we failed, and then we learned from that experience and succeeded." They probably weren't so anxious, or most people aren't so anxious to talk about it when they're in the midst of that failure, but when they're on the other side of that failure and they can offer that perspective and openly and honestly, it's interesting to read that because that's really a good metaphor and illustration of what learning is often about.

Well, my next question I think you both somewhat touched on is, is becoming data-driven a destination or a perpetual journey? Randy, do you want to start on that? Yeah. For me, it's a journey. It's not a destination. It's not like being king of the hill, that you get there and then you can relax. It's kind of the opposite.

And the story that I often tell is that people ask me, "What are the most successful data-driven organizations that you've interacted with?" And I point to a couple of organizations, among them American Express and Capital One, and say that, however, when I speak with these organizations, they're never complacent, they're never satisfied. They're always nervous. They're always, "How can we be doing better?" They're always looking over their shoulder.

They're always looking at new and emerging competitors in fintech or insurtech, depending upon the industry. And so I actually find that refreshing, that they're forever vigilant. And I've also been asked what's the worst case scenario, and I say I walk into some organizations, frankly, and not infrequently, and I ask them about their data initiatives, and they say, "We've got it all figured out. Everything's under control." And when I hear that, there's really not much you can say.

You can just say, "Well, congratulations, and let's speak again if at some point you have any issues or challenges." And ultimately they do. But when I go into an organization and they feel that they've got it all figured out, I worry a little bit about them.

And Chris, would you agree with that? Is it a destination or a journey? Yeah. And I think, in my career, I've been 30 years doing tech and the last 16 focused on data and analytics. And I think when I started, it was about sort of construction. Like you constructed something and you walked away, and it was seen as I'm building something that's going to live forever, and I'm done. And I think what's changed is that it's more like kayaking than construction.

You're going down a river and constantly pivoting back and forth. And I think that it means that you are on a perpetual journey, and I think Randy touched on it. You're never done. It's never finished because there's always the follow-up question. And in my career, I've just seen too many data and analytic teams be depressed after getting 10 follow-up questions as opposed to being excited because they know how much work it took them to get there, and they know that.

And that's just really has made me sad to see because if you've got 10 follow-up questions, that means you are doing awesome. And you should continue to paddle down the river and pivot to those questions. And I think that's really an important mindset, and a lot of organizations are kind of getting it. It's never done.

It's always improving. There's always more data. There's always more algorithm. There's always more ideas. And the competitors are always at your heels. And so I think it really is a perpetual journey Yeah, and I like the kayaking metaphor. If I had thought of it, I would've included it in the book, maybe in the sequel.

And to that point, sometimes when people ask me about what I do in the industry, and I've talked about one particular client and over the course of the decade, working with them three times on developing their data strategy. And sometimes people have said to me, "Wow, that's really horrible." And I've said,

00:25:00

"No, that's actually really good because it means that they're continuing to focus on what's important. They're rethinking because the business environment changes, the technology, the data environment changes. So they're continuing to revisit, look at this from a fresh perspective, as opposed to coming up with a data strategy, putting it on a shelf, just thinking that it will guide their future over the course of the next decade." So, the process of continuous improvement is always critical to successful outcomes.

So why is it so difficult for organizations to evolve and change and become data-driven? Yeah, let me share some data from the annual survey that we can conduct. So let's see how the-- There. Maybe you can see that. We can see it. So we've been conducting this survey for a decade now, and it was really originated out of a series of C-executive roundtables that were conducting quarterly.

And at the time, it was a CIO from JP Morgan who said, "Hey, the board's asking me about big data. What's our strategy about big data? I never even heard of the term until two months ago. I don't know what it means. It'd be great to hear what other organizations are doing." So, we started this survey, and it began as a big data survey, and it evolved into the big data and AI survey over the years. We conduct it each year of leading Fortune 1000 companies.

This year, there were 85 companies that participated. And of those 85 companies,

I forget the exact number, roughly 90% of the participants are C-suite executives, so they're chief data officers, chief information officers, chief digital officers, in some cases, CEOs themselves, or line of business presidents. And I want to read some data to you because it's actually startling, but it's illuminating as well. And though it may sound pessimistic, there's a silver lining in this. So we've been asking this question for the past several years.

Have you created a data-driven organization, yes or no? In 2021, 24% said yes. Okay? And by the way, that's down from 31% in 2019. Have you forged a data culture, yes or no? 24.4% said yes, down from 28.3% in 2019. And I'll give you one more here. Are you managing data as a business asset?

39.3% said yes, down from 46.9% two years ago. So what does this tell us? It tells us that organizations are facing a lot of challenges. It also tells us that there's a lot of opportunity for growth and development and maturity and an opportunity for organizations to be more successful. I've also been asked why the decline in recent years of people getting worse, and it's hard to pinpoint exactly why the decline, but my answer is I think that organizations are becoming more realistic, that in the past, people were asked if they were data-driven, they had a data culture, and they said, "Yes, of course, we do." And then when they got deeper into their jobs and became more honest with themselves and actually became more confident in their roles and that they could be honest, they had to acknowledge that there was still a lot of work to be done. So the great news is that there's so much work to be done that people entering the profession, people coming out of school, midterm career professionals have decades and decades to go secure and high-demand employment.

And Chris, do you have anything to add to that? Yeah.

Sometimes my job is to shine a light on the ugliness, right? In that people are regressing in being data-driven, that projects fail. There's a lot of sort of BS in data and analytics projects because they're not actually meeting what they set out to do. And as a result, people are frustrated. People who've come into the career, people who've been in the career for a long time get this sort of hair shirt of, "I'm going to do whatever it takes." And it's not a fun place to be in a lot of organizations. And to me, the way I touch the elephant is it's not that you shouldn't be data-driven, it's not that you shouldn't have a data culture. It's that part of that culture is a management style of the teams and people who do the work that is less reflective of political parties and more reflective of how

00:30:00

you do good manufacturing and how you build a good car. And that sort of management style that focuses on cycle time, iterations, lean, just in time, went through the manufacturing industry, right? And I grew up in Wisconsin, and there was a company nearby called American Motors that produced really crappy cars and got beat by Toyota.

Why? Because Toyota had a better manufacturing system, the Toyota Production System. And we've seen that set of managerial ideas go through software with Agile and DevOps. And to me, I think the same, "I've got to run my factory in a better way," is coming into data and analytics. And it's not about the workstation or the robot that makes your factory better. It's about getting a group of people working in a technically complicated environment to hum.

And there's a set of principles that go on it. And I think, to me, that's why data-driven is hard, because leadership is kind of focused on the wrong thing. They have a sort of a piece manufacturing mindset as opposed to a mass and even a lean manufacturing mindset. And once leadership gets that, they start to think the factory is more important than what comes out of the factory.

And that's all, actually, as a leader, you can control, is how good your team and the factory work. And by the way, I drove one of those American Motors cars for-- Well, not my first car. A Rambler was an American Motors car, right? Yeah. And how long did it last? Well, it lasted a long time, but it was a

step up from the horse and buggy.

Yeah. My dad was a union guy in Wisconsin, and he drove a Toyota Corolla back in the late '70s and '80s and took a lot of crap for it, and he's like, "It's cheaper. It's going to last longer." And he's right.

So, Randy, you said earlier that culture was one of the biggest challenges or one of the biggest impediments to becoming data-driven. Do you have any recommendations for addressing these cultural issues?

Yeah.

Just to underscore the issue around culture, let me give this example. I walk into many organizations and meet with the data organizations and the data team, and they talk to me about the robust capabilities that they've created, and they're rightly very proud of those. And then I go into the technology organization and meet with the CIO, and they'll share with me what they've created and the technologies and the processes, and they're rightly very proud of what they've created.

And then I go meet with the line of business executive, the president or CEO of the particular business line, whether it's consumer insurance or commercial banking, and they'll shake their head and frown and they'll say, "I don't have confidence in the information that I'm receiving. I'm not receiving the information that I really need to make decisions in a timely fashion.

I'm not receiving the critical pieces of information that are most important to me." So you have all of these people that have good intentions, but yet there's a disconnect with the ultimate business value. And when I started this business, Novantage Partners, 20 years ago, with a fellow, Paul Bartha, who was an MIT PhD in high-performance computing, we figured we would be spending 95% of our time on technology issues and 5% on the other stuff, and it's really been the reverse of that.

There's so many great technologies out there that can help organizations and enable them, but it's really the people issues. Everything from breaking down the silos so that people work together, collaboration at an enterprise level, collaboration across skill sets. So, the challenges can just go on and on. What's the optimal organizational structure to be effective in managing data?

How do people understand data from a production through a consumption process? We have done work over the years in helping organizations understand the lineage and all of the individuals and parts of an organization that touch data or create new data, and that's been informative for organizations because up until having that level of tangible understanding, data was often abstract for many executives. So it's really an educational challenge, a literacy challenge, an organizational challenge, breaking down the silos, and these are things that, again, relate back to fundamental changes in how people operate. So these are the things that take more

00:35:00

time, years, and sometimes decades for organizations to get comfortable with. Chris, do you have any recommendations for addressing the cultural issues?

Yeah, I find that too. There's a lot of data teams who are kind of delusional that they're being successful. Right? They think everything's great. And they have intellectual frameworks that they're putting on and words that sound really great, but you go off to and talk to the people who are the recipients of their value, and they're shaking their heads going, "These guys aren't giving me what I want. I don't trust the data. They're too slow.

They're always talking nonsense. Just give me..." And they do a lot of work around it. They hire consultants, they do the data themselves. And so there's this disconnect between sort of, I think of as my people, right, the data people working in the back, and the business ones. And I find that really quite frustrating and a little depressing sometimes that people want to live in delusion. And I think there's just a benefit of really facing the truth, and I actually learned that many years ago when I worked at a startup that raised a lot of money and didn't go anywhere.

And the CEO was always talking about how awesome we are and how we're going to win. And then I met with the CEO of the acquiring company, and he didn't talk about how awesome it was. He talked about all the problems they had and all the areas they needed to improve. And I was sort of shocked by that because he was very focused on what we could do to do better as opposed to what we do to do great. And as a leader, you need to do both, right?

But I think having this cognitive dissonance of being brutally honest about what's not working, but entirely optimistic about What the team can do is a great combination. And I think that's partly what organizations need to do to be successful. And by the way, I open each chapter of the book with, I guess they call it an epigraph. I call it a quote, but I guess if you're a literary person, you call it an epigraph.

And the one that I begin the book with, which I think carries through the entire book, and that is, "Perfect is the enemy of good." I see too many organizations often striving for the perfect answer, the perfect capability, the perfect solutions. And I actually meet with a number of technology or architecture teams, and they talk about how they're designing something for the long term that will create capabilities that really enable the organization.

And that's a noble goal, but at the same time, the business has to produce quarter-to-quarter results. And if you don't produce the results, you don't get the funding, you don't get the business, you don't grow, you don't stay competitive, you don't continue to serve your customers. So you have to play in the short term, as well as trying to develop the right platform for the long term. So there's that connection, or that intersection between the hard, brutal day-to-day reality of making your numbers and the perfect design of designing the academically perfect engineered solution for the long term, and there has to be that collaborative middle ground.

So that is a perfect segue to the next question, which is, Randy, can you speak a little bit more about the title of the book, "Fail Fast, Learn Faster," and why it's so important? And maybe you just did.

Yeah. It's funny, and I didn't mention it in the opening, but that is

learning from experience. Organizations and individuals can't expect to be successful or to win all the time. That's not the nature of life. What do they say in baseball, if you bat 300, they put you in the Hall of Fame, which means that three times you got a hit and seven times you made an out. So, seven times you failed and three times you were successful, and that makes you great in your profession.

So maybe they should develop stats for data people, too, so that when people succeed three out of 10 times, they have a comparative measure. And I think sometimes what's lacking is that perspective in the data field. It is still a relatively nascent field and some would argue that in Silicon Valley, they've charged ahead in many ways, both for good or for ill, depending upon your point of view, or maybe there's a lot of ill.

Well, it's both good and ill, but a lot of that comes from that hard-charging, the Facebook metaphor about "Move fast and break things," and not being afraid to fail.

And Chris, do you have anything to add to that?

00:40:00

Yeah. I keep saying love your errors. And I think it's really important that errors are part of the flip coin of learning. And the more errors you make, the more you learn. And so having a team that can really love it, but the other part of loving your errors means don't make the same one over and over again, and learn from them.

And so what organizations need to do is love their errors, but when they have an error, when something didn't work, what's the learning? And then is that learning some new approach that we need to do, or is it some automation that we need to put in so it never happens again? And that saying, "Do I need to change my approach or do I need to automate?" And turn the failure into something actionable and love it so that it makes you up forward.

And I think just looking at your problems and trying to bucket them as to, is it our approach or is it some automation that we need to wrap our system in so we can make a better factory? I think that's a really good discussion for people to have. And it takes failure and depersonalizes it and puts it more into a process of iterations and learning.

And I think whether it's Silicon Valley or Toyota or other companies, they've taken that idea that things that go wrong are really fertilizer and not the other name for fertilizer.

Absolutely. Couldn't agree more. So what do you think characterizes great data-driven companies from everyone else? Randy, I know you talk about that in the book. Yeah. It's that relentless drive to do things better, a relentless drive to think about things differently. If you look at Capital One, for example, they're an interesting company because they started as basically a data and analytics group, a team of people that said, "Where can we apply our data and analytics learnings and our algorithms?" And they said, "Oh, we can apply it to credit cards to target underserved segments or to serve those segments more effectively." And they've grown into one of the leading banks in the country, and certainly a leader from the credit card perspective, and they have their Capital One Cafes. So they've done things entirely differently.

They didn't follow the Citigroup, Bank of America, traditional banking model. They've been innovators in the industry.

And I try to write about firms of this kind as often as possible, and there are different stages. I wrote a piece in Forbes last week about a company called Traffk, T-R-A-F-F-K, who's an insurtech company who's taking snapshots of large volumes of data that Insurance companies traditionally haven't taken a look at it. They've been data-intensive, but more a limited set of data, which has been actuarial data.

And TraffiQ has developed a platform to bring in new sets of information. So, often I speak to really large companies and they say, "We've existed for generations, over 100 years, and we believe in change, but doing it very evolutionary." And that can cut both ways because obviously if you have a loyal client base that's very committed to you and you've served them well and continue to serve them well over many years, they'll retain that loyalty.

But at the same time, if new competitors can come along that can offer a convenience in a way that you can't, that's where you can get eroded, and that's where traditional companies have been eroded. You look at companies like Sears and other leaders from past decades and generations, and when industries change, they change pretty quickly.

I use a metaphor in the book, and I used to read, and still do, a lot of literature, and this was from Ernest Hemingway, the book, "The Sun Also Rises," and they have two characters speaking. And one of the characters has gone bankrupt and the other character goes, "How did you go bankrupt?" And this one responds, "Gradually and then suddenly." And I think that is the perfect metaphor for how disruption happens.

You often don't see it sneaking up upon you, but when it happens, it happens like that, instantaneously.

Chris, anything to add to that?

It's interesting because when we started this company, we spent a bunch of months on, I think the top of the pyramid on trying to help companies be data-driven, which is like how do you not use analytics as just plastering of an executive intuition, right? And pick and choosing a chart for what you believe already, because truly being

00:45:00

data-driven means the data, not always, but at least sometimes actually tells you and it goes against your intuition. And so we couldn't figure out a software product to do it, although we tried very hard. But I still think that's a problem, is one is that tip of the pyramid, how to get executives and leaders to actually adopt it and trust it and do it. And then, I think the other part is just the factory part that's really hard for organizations.

They're in the sort of building cars by hand and trying to get to manufacturing and working towards lean and agile and DataOps. So I think both those things are challenges. By the way, that's one of the sub-themes in the book and the chapter, data science and facts, and later about data ethics. And that is, to your point, Chris, you can really take any data and tell whatever story you want by selectively presenting it or cherry-picking it.

So, there are objective truths, but as one knows from social media and so forth, you can choose which data you present, and then that creates a different picture for the audience. Yeah. And I think that's a lot of ways, at least 10, 15 years ago, I saw executives doing it. It's like, let's pick the report to cherry-pick what I already think, and it looks good, but it doesn't really... It's not meaningful.

And that's a very different metaphor and how people go from that data has power to I should actually do what the data tells, but we've got to understand the data, its limitations. Yeah, maybe it's generational. I liked your quote on "Moneyball" and the attitude towards data is taking, it's-- So I think all those things go into helping the desire of organizations to be truly data-driven, not just sort of pretending data-driven.

So a question came in that's related and is, what are your experiences in using data to drive fact-based decision making that yields real, consistent, and predictable results? Randy, you want to give that one a shot? I was hoping Chris would.

Well, it kind of goes to what I said before, right? Is because you can do everything right and then your leadership won't pick it up, right, or won't do it. And that's very frustrating for people. And in some ways, if you take that data is more kayaking and you're in a constant dialogue with people, over time, because I think you really are an advocate for the data-driven, fact-based view of the world, right? As opposed to intuition.

And if you've ever made a business decision, you know that it's not always data. There is intuition, there's social pressure, there's your own personality, all this stuff, and decisions are complicated things, right? And so I think having a bit of humbleness around all the attributes or all the things that go into a business decision, including data, right? And accurately representing that data can inform, and sometimes it's so completely obvious that all those other things should be, but in other times, it's more complicated. So I think it's really a journey in a decisioning process to help organizations more tend towards data-driven decisions and help data teams under-- That's what's great about data is it really is an entry into a big world of how organizations operate and market and sell and their decisions.

And there's very few careers where you can actually jump industries, and go from pharma to finances to manufacturing and have the same set of skills and learn all about it. And that's why I actually really like it, because it just gives you a unique window to understanding these complicated decisions. Great. Thanks, Chris. Yeah, I think that's a really good point.

I like your point about data as an entry point because- To your point, I like to have the data and see what the data is. You may make a decision that says, "From my experience in these situations, the outcome's going to be a little bit different. So I know the data, but I'm not necessarily going to follow it strictly on this particular decision." But you should always have the data because it informs you, and then in some cases, you follow it literally, and in other cases, you may overrule it, but because other experiences and factors come into play.

Or it gives you an opportunity to say, "Okay, let's get some more data on this, because right now what we have isn't good enough." So, because it's about building data sources that you can use to make predictions later on. Yeah. More data. All right.

Great. Thank you both. So another question that came in, it's two parts, one part for each of you. So this is about data sharing among organizations. So for Randy, what do you think are the most fruitful areas for data sharing?

00:50:00

Most fruitful areas for data sharing.

It really depends upon the context of that question, but I think certainly, one of the things that's been learned over the past year or two is in the life sciences industry, traditionally, you had these various drug discovery groups that operated completely autonomously. There was no sharing of data across lines of business, across epidemiological lines. And as a consequence of the pandemic, there was much greater data sharing, not only within organizations, but across nations, across international boundaries. And there was

factors or patterns that were discerned that wouldn't have been discerned otherwise. So, I'm not sure if it answers the question precisely, but I think certainly in scientific fields, there's enormous benefit to sharing data. And the second part of that question is for Chris. What are your thoughts on data pipelines running across different organizations?

Well, that's the way the world works. And so, what's interesting about data and analytics, I gave at the beginning saying you can imagine a leader in front of a team of data scientists and data engineers and people doing bids in one room, and they all work for him or her. But that's not the reality, right?

It's that there's a business customer and part of the work is done in one organization, and part is done in a different organization, and maybe their bosses meet six layers up. And so how do you get a distributed organization to have both consistency in a central way, but local freedom, sort of federal government versus states rights?

So I think that's really one of the key challenges in data and analytics. It actually makes it more interesting than software, because in software, you just got front end, back end, and your operations team. Here, you got a bunch of teams. It's sort of a complex many to many problem.

Great. Yeah. Thank you, Chris. So, Randy, as we head into the final minutes here, what is your favorite chapter of the book, and why? Yeah, I actually kind of anticipated that question in my opening remarks, and I referred to the chapter on data ethics because that's important. And one of our executive roundtable breakfasts a few years ago, rather than focusing on the usual issues, we decided we're going to focus on data ethics.

And I said to my colleagues, I said, "This will probably be the least attended breakfast ever." And it turned out to be the opposite. It was the most attended, it was the most passionate in terms of the discussion. So I think data ethics is essential. Basically, the chapter in the book talks about all of the horrible things that are done with data.

There's anecdotes from the Cambridge Analytica experience, and there's shots from various parties at the Googles and the Facebooks and so forth. I quote a lot from Cathy O'Neil, who wrote the book, "Weapons of Math Destruction." I interviewed her a few years ago. She's at the extreme end in her beliefs about the horror of algorithms and algorithmic bias. But there's, in terms of privacy, in terms of people trading off access or sharing data for the sake of convenience, but without realizing the potential consequence.

I tell people when people hear that I'm in the data field, they ask me questions like, "Oh, do people know much about me?" I said, "Everybody knows everything about you. The government, large corporations, they know where you are. You can't hide. You can't disappear. There is cameras on the street that track your appearances every time you do a transaction." So data can be used-- It's like any kind of tool. It can be used for good, or it can be used for ill.

And in the case of treating the pandemic, it was treated for good, largely, but there's also other things that pose inherent dangers if data falls into the wrong hands.

And Chris, I know we didn't get the whole book, but do you have any favorite concepts from the book?

Well, I actually really liked his stories. And so I think people really respond to stories of people's change and look for models of how people have affected

00:55:00

a way to actually sort of fail fast and learn faster. And I think there's just a bunch of great examples that different companies have of how they've gone about it and thought about it. And so I think that sort of mindset change of like, "Hey, if I look at my customers, they're really not happy.

I don't want to live in a delusion. And you know what? Our data's crappy and I don't even know if our customers even know that it's crappy because we're giving it to them without checking it." And then, how do I solve this, right? How do I actually become a fail fast, learn faster work? And those sort of trio of things I think are really, it's important to learn from other people's stories.

And I think that's why it's an important book because

I think those are really true stories for a lot of organizations, and if you sort of start grokking that, you start to think, I think, in a more structural way about how to change your data and analytic organization for the better. Yeah. I love to tell stories, and that's one of the things I thought I sought to do in the book. And one of my sons, my oldest son, read the book because I'd always tell these stories like, "You wouldn't believe what happened at this meeting today," and so forth.

And he said, "Well, how come you didn't tell the stories about the time you spilled pizza on your shirt, and you had to go and meet with the CEO?"

I said, "Well, yeah, some of those stories. Maybe that's in the sequel."

The sequel. So we have three minutes left. So a concluding question is, are you optimistic or pessimistic about the future of data, Randy? Yeah, I'm hugely optimistic in terms of the profession. I think data's only going to become more important. It's going to proliferate. It's going to become more relevant. It's going to become more central to all of our decision-making, regardless of what industry that you're in.

I think that it's the place to be for the next generation in terms of work. But I think you also have to be vigilant about misuses of data as well because data can be used in autocratic fashion as well. Great. And Chris?

Short term, pessimistic, long term, optimistic. I still think there's a lot of delusion going on where like if I just buy the new database or the new tool, everything's going to happen, or the new two-letter acronym, and I think that's leading a lot of people and companies to fail and be frustrated. And so I think if we can solve that and really start focusing on how you get your team successful, and I saw that in software, I experienced it in manufacturing, I think that long-term change is happening.

So that's why I'm optimistic because I think we will solve it, and I think people, books like Randy's and spreading ideas, I think is very helpful for people to learn and understand. Well said. Great. Well, unfortunately, we are out of time. We could probably keep talking about this for another hour. I want to thank all the attendees for joining us today.

Thank you, Chris, as always, and an extra big thanks to you, Randy, for joining us today and speaking and sharing your insight. It was really great to have you. And we're really looking forward to the book being released in August. Now, is it August 31st, the date? Is that right? August 31st, yes. Okay, great. So we'll be on the lookout for that.

Hopefully all the attendees will too. We'll be sending out the recording of this webinar in the next 24 to 48 hours, so everyone should be on the lookout for that in their email. And as I mentioned earlier, we'll also be separately notifying you if you'll be receiving a copy of the book, and we'll need to get some details from you to get that sent out to you.

So if anyone has any additional questions about the book, about DataKitchen, or DataOps, please don't hesitate to reach out to me or Chris at DataKitchen. We can certainly point you in the right direction, and we can point you to Randy if necessary. So thanks again, everyone, for this great webinar, and I hope everyone has a great afternoon and evening.

Thank you for the opportunity. Bye. It was a pleasure. Yeah. Thank you, Randy. Bye. Bye-bye.