 Welcome to this Biz Ups Manifesto power panel, Data Lake or Data Land Field. We're going to be talking about that today. I've got three guests joining me. We're going to dive through that. Karen Taylor is here, the CMO of Broadcoms Enterprise Software Division. Karen, great to have you on the program. Thank you, Lisa. Kevin Serresa is here as well, Chairman and CTO of ATVAN. Hey, Kevin. Hey, Lisa. And Isaac Sokolik, author and CEO of Star IO. Isaac, welcome. Hi, Lisa. Thanks for having me. So we're going to spend the next 25 to 30 minutes talking about the challenges and the opportunities that data brings to organizations. You guys are going to share some of your best practices for how organizations can actually sort through all this data to make data-driven decisions. We're also going to be citing some statistics from the inaugural Biz Ups Industry Survey of the state of digital business in which 519 business and technology folks were surveyed across five nations. Let's go ahead and jump right in. And the first one in that survey that I just mentioned, 97% of organizations say we've got data-related challenges limiting the amount of information that we actually have available to the business. Big conundrum there. How do organizations get out of that conundrum? Karen, we're going to start with you. Thanks, Lisa. You know, I think, I don't know if it's so much limiting information as it is limiting answers. There's no real shortage of data, I don't think, being captured recently met with a unnamed auto manufacturer who's collecting petabytes of data from their connected cars. And they're doing that because they don't really yet know what questions they have of the data. So I think you get out of this data landfill conundrum by first understanding what questions to ask. It's not algorithms, it's not analytics, it's not math that's going to solve this problem. It's really, really understanding your customer's issues and what questions to ask of the data. Understanding what questions to ask of the data. Kevin, what are your thoughts? Yeah, look, I think it gets down to what questions you want to ask and what you want out of it, right? So there's questions you want to ask, but what are the business outcomes you're looking for? Which is the core of BizOps anyway, right? What are the business outcomes? And what business outcomes can I act upon? So there are so many business outcomes you can get from data and you go, well, I can't legally act upon that. I can't practically act upon that. I can't, whether it's layoff people or hire people or whatever it is, right? So what are the actionable items? There is plenty of data. We would argue too much data. Now, we could say is the data good, is the data bad? Is it poorly organized? Is it noisy? There's all other problems, right? There's plenty of data. What do I do with it? What can I do that's actionable? If I was an automaker and I had lots of sensors on the road, I had petabytes, as Karen says, and I'm probably bringing in petabytes potentially every day, well, I could make my self-driving systems better. That's an obvious place to start, right? That's what I would do. But I could also potentially use that to change people's insurance and say if you drive in a certain way, something we've never been able to do, if you drive in a certain way, based on the sensors, you get a lower insurance rate. And nobody's done that. But now there's interesting business opportunities for that data that you didn't have one minute ago and I just gave away. So it's all about the actionable items in the data. How do you drive something to the top line and the bottom line? Because in the end, that's how we're all measured. And Isaac, now you say data is the lifeblood. What are your thoughts on this conundrum? Well, I think they gave you the start and the end of the equation. Start with a question. What are you really trying to answer? What you don't understand that you want to learn about your business connected to an outcome that is valuable to you? And really what most organizations struggle with is a process that goes through discovery, learning what's in the data, addressing data quality issues, loading new data sources if required. And really doing that iteratively, and we're all agile people here at BizOps, right? So doing it iteratively, getting some answers and understanding what the issues are with the underlying data. And then going back and revisiting and reprioritizing what you want to do next. Do you want to go look at another question? Is the answer heading down a path that you can drive outcomes? Do you got to go cleanse some data? So it's really that, how do you put it together so that you can peel the onion back and start looking at data and getting insights out of it? Great advice. Another challenge though that the survey identified was that nearly 70% of the respondents, and again, 519 business and technology professionals from five countries said, we are struggling to create business metrics from our data with so much data, so much that we can't access. Can you guys share best practices for how organizations would sort through and identify the best data sources from which they can identify the ideal business metrics? Kieran, take it away. Sure thing. I guess I'll build on Isaac's statements. Every company has some gap in data, right? And so when you do that data gap analysis, I think you really, I don't know, it's like Alice in Wonderland, begin at the beginning, right? You start with that question like Isaac said. And I think the best questions are really born from an understanding of what your customers value. And if you dig into that and you understand what the customer's value, you build it off of actual customer feedback, market research, then you know what questions to ask. And then from that, hey, what inputs do I need to really understand how to solve that particular business issue or problem? Kevin, what are your thoughts? Yeah, I'm going to add to that, completely agree, but look at, let's start with sales data, right? So sales data is something everybody on this, everybody watching this understands, even if they're not in sales, they go, okay, I understand sales data. What's interesting there is we know who our customers are. We could probably figure out if we have enough data why they buy. Are they buying because of a certain sales person? Are they buying because it's a certain region? Are they buying because of some demographic that we don't understand, but AI can pull out, right? So I would like to know who's buying and why they're buying. Because if I know that, I might make more of what more of those people want, whatever that is, certain fundamental sales changes or product changes or whatever it is. So if you could certainly start there, if you start nowhere else say, I sell X today, I'd like to sell X times 1.2 by next year. Okay, great. Can I learn from the last five years of sales, millions of units or a million or whatever it is, how to do that better? And the answer is for sure, yes. And yes, there's problems with the data and there's holes in the data as Karen said and there's missing data. It doesn't matter. There's a lot of data around sales. So you could just start there and probably drive some top line growth, just doing what you're already doing, but doing it better and learning how to do it better. Learning how to do it better. I say, talk to us about what your thoughts are here with respect to this challenge. Well, when you look at that percentage, 70% struggling with business metrics, what I see is some companies struggling when they have too few metrics, their KPIs, it really doesn't translate well to people doing work for a customer, for an application, responding to an issue. So when you have too few and they're too disconnected from the work, people don't understand how to use them. And then on the flip side, I see other organizations trying to create metrics around every single part of the operation, dozens of different ways of measuring user experience and so forth. And that doesn't work because now we don't know what to prioritize. So I think the art of this is management coming back and saying, what are the metrics? Do we want to see impact and changes over in a short amount of time, over the next quarter, over the next six months and to pick a company each category, certainly starting with the customer, certainly looking at sales, but then also looking at operations and looking at quality and looking at risk and say to the organization, these are the two or three we're going to focus on in the next six months. And then I think that's what simplifies it for organizations. Thanks, Isaac. So something that I found interesting is not surprising in that the survey found too much data is one of the biggest challenges that organizations have, followed by the limitations that we just talked about in terms of identifying what are the ideal business metrics, but a whopping 74% of survey respondents said, we failed to have key data available in real time, which is a big inhibitor for data-driven decision-making. Can you guys offer some advice to organizations? How can they harness this data and glean insights from it faster? Kieran, take it away. Yeah, I think there are probably five steps to establishing business KPIs. And Lisa, your first two questions and these gentlemen's answers laid out the first two. That is to find the questions that you want answers for and then identify what those data inputs would be. If you've got a formula in mind, what data inputs do you need? The remaining three steps, one is to evaluate the data you've got and identify what's missing. What do you need to then fetch? And then that fetching, you need to think about the measurement method, the frequency. I think Isaac mentioned this concept of tools sprawl. We have too many tools to collect data. So the measurement method and frequency is important, standardizing on tools and automating that collection wherever possible. And then the last step, this is really the people component of the formula. You need to identify stakeholders that will own those business KPIs and even communicate them within the organization. That human element is sometimes forgotten. It's really important. It is important. It's one of the challenges as well. Kevin, talk to us about your thoughts here. Yeah, again, for sure. You've got, in the end, you've got the human element. You can give people all kinds of KPIs. As Isaac said, often it's too many. You've now KPI'd the business to death and nobody can get out and do anything. That doesn't work. Obviously you can't improve things till you measure them. So you have to measure, we get that. But this question of live data is interesting. My personal view is only certain kinds of data are interesting, absolutely live in the moment. So I think people get in their mind, oh, well, if I could deploy IoT everywhere and get instantaneous access within one second to the amalgam of that data, I'm making up words too, that would be interesting. Are you sure that'd be interesting? I might rather analyze the last week of real, real data, really deep analysis, right? Build a real model around that and say, okay, for the next week, you ought to do the following. Now, I get that if you're in the high-frequency stock trading business, every millisecond counts. But most of our businesses do not run by the millisecond and we're not gonna make a business decision, especially humans involved, in a millisecond anyway. We make business decisions based on a fair bit of data, days and weeks. So this is just my own personal opinion. I think people get hung up on this, I gotta have all this live data. No, you want great data analysis using AI and machine learning to evaluate as much data as you can get over whatever period of time that is, a week, a month, a year, and start making some rational decisions off of that information. I think that is how you run a business that's gonna crush your competition. Good advice. Isaac, what are your thoughts on these comments? Yeah, I'm gonna pair off of Kevin's comments. How do you chip away at this problem at getting more real-time data? And I'll share two insights. First, from the top down, when STAR-CIL works with a group of a CEO and their executive group, how are they getting their data? Well, they're getting it in a boardroom with PowerPoints, with spreadsheets behind those PowerPoints, with analysts doing a lot of number crunching, and behind all that are all the systems of record around CRM and the ERP and all the other systems that are telling them how they're performing. And I suggest to them for a month, leave the world of PowerPoint and Excel and bring your analysts in to show you the data live in the systems, ask questions and see what it's like to work with real-time data. That first changes the perspective in terms of all the manual work that goes into homogenizing that data for them, but then they start getting used to looking at the tools where the data is actually living. So that's an exercise from the top down. From the bottom up, when we talk to the IT groups, so much of our data technologies were built at a time when batch processing in our data centers was the only way to go. We ran these things overnight to move data from point A to point B. And with the cloud, with data streaming technologies, it's really a new game in town. And so it's really time for many organizations to modernize and thinking about how their streaming data doesn't necessarily have to be real-time. It's not really IOT, but it's really saying I need to have my data updated on a regular basis with an SLA against it so that my teams and my businesses can make good decisions around things. So let's talk now about digital transformation. We've been talking about that for years. We talked a lot about in 2020 the acceleration of digital transformation for obvious reasons. But when organizations are facing this data conundrum that we talked about the sort of data disconnect too much, can't get what we need right away. Do we need it right away? How do they flip the script on that so that it doesn't become an impediment to digital transformation, but it becomes an accelerant? Kiran? A lot of times you'll hear vendors talk about technology as being the answer, right? So MIML, my math is better than your math, et cetera. And technology is important, but it's only effective to the point which people can actually interpret, understand and use the data. And so I would put forth this notion of having data at all levels throughout an organization. Too often what you'll see is that I think Isaac mentioned it. The data is delivered to the C-suite via PowerPoint and it's been sanitized and scrubbed, et cetera. But heck, by the time it gets to the C-suite it's three weeks old. Data at all levels is making sure that throughout the organization, the right people have real-time access to data and can make actionable decisions based upon that. So I think that's a real vital ingredient to successful digital transformation. Kevin? Well, I like to think of digital transformation as looking at all of your relatively manual or paper-based or other processes, whatever they are throughout the organization and saying, is this something that can now be done for lack of a better word, by a machine, right? And that machine could be algorithms, it could be computers, it could be humans, it could be cloud, it could be AI, it could be IoT, it doesn't really matter. And so there's a reason to do that and of course the basis of that is the data. You've got to collect data to say this is how we've been performing, this is what we've been doing. So a simple example of digitization is people doing RPA around customer support. Now you collect a lot of data on how customer support has been supporting customers. You break that into tiers and you say, here's the easiest lowest tier. I had formed that out to probably some other country 20 years ago or 10 years ago. Can I, even with the systems in place, can I automate that with a set of processes, robotic process automation, that digitizes that process now? Now there still might be 20 different screens to click on and all different kinds of things, whatever it is, but can I do that? Can I do it with some chatbots? Can I do it with it? No, I'm not gonna do all the customer support that way, but I could probably do a fair bit. Can I digitize that process? Can I digitize the process? Great example we all know is insurance companies taking claims. Okay, I have a phone. Can I take a picture of my car that just got smashed, send it in, let AI analyze it and frankly do an ACH transfer within the hour? Because if it costs an insurance company on average $300 to $500, depending on who they are, to process a claim, it's cheaper to just send me the $500, than even question it. And if I did it two or three times, well then I'm trying to steal their money and I should go to jail, right? So these are just, I'm giving these as examples because they're examples that everyone who's watching this would go, oh, I understand you're digitizing a process. So now when we get to much more complex processes that we're digitizing in data or hiring or whatever, those are a little harder to understand, but I just try to give those as like everyone understands, yes, you should digitize those, those are obvious, right? Now those are great examples, you're right, they're relatable across the board here. Isaac, talk to me about what your thoughts are about, okay, this data conundrum, how do we flip the script and leverage data access to it insights to drive and facilitate digital transformation rather than impede it? Well, remember, digital transformation is really about changing the business model, changing how you're working with customers and what markets you're going after. You're being forced to do that because of the pace digital technologies are enabling competitors to outpace you. And so we really like starting digital transformations with a vision. What does this business need to do better, differently, more of what markets are we gonna go after, what types of technologies are important? And we're gonna create that vision but we know long-term planning doesn't work, we know multi-year doesn't plan and doesn't work. So we're gonna send our teams out on an agile journey over the next sprint, over the next quarter, and we're going to use data to give us information about whether we're heading in the right direction. Should we do more of something? Is this feature a higher priority? Is there a certain customer segment that we need to pay attention to more? Is there a set of defects happening in our technology that we have to address? Is there a new competitor stealing market share? All that kind of data is what the organization needs to be looking at on a very regular basis to say, do we need to pivot what we're doing? Do we need to accelerate something? Are we heading in the right direction? Should we give ourselves high fives and celebrate a quick win because we've accomplished something because so much of transformation is what we're doing today, we're gonna change what we're doing over the next three years. And then guess what? There's gonna be a new set of technologies, there's gonna be another disruption that we can't anticipate. And we want our teams sitting on their toes waiting to look at data and saying, what shall we do next? That's a great segue, Isaac, into our last question, which is around culture. That's always one of those elephants in the room, right? Because so much cultural transformation is necessary, but it's incredibly difficult. So question for you guys, Karen, we'll start with you, is do you advise leadership should really create a culture, a company-wide culture around data? What do you think? Absolutely, I mean, this reminds me of DevOps in many ways. And the data has to be shared at all levels and it has to empower people to make decisions at their respective levels so that we're not kind of siloed in our knowledge or our decision-making. It's through that collective intelligence that I think organizations can move forward more quickly, but they do have to change the culture. They've got to have everyone in the room. Everyone's got a stake in driving business success from the C-suite down to the individual contributor. Right, Kevin, your thoughts. You know, Karen's right, data silos, one of the biggest brick walls in all of our way, all the time, you know, SecOps says, there is no way I'm going to share that database because it's got PII. Okay, well, how about if we stripped the PII? Well, then that won't be good for something else. And you're getting these huge arguments. And if you're not driving it from the top, certainly the CIO, maybe the CFO, maybe the CEO, I would argue the CEO drives it from the top because the CEO drives company culture. And, you know, we talk BizOps and the first word of that is biz. It's the business, right? It's ops being driven by business goals. And the CEO has to set the business goals. It's not really up to the CIO to set business goals. They're setting operational goals. It's up to the CEO. So when the CEO comes out and says, our business goals are to drive up sales by this, drive non-cost by this, drive up speed of product development, whatever it is. And we're going to digitize all of our processes. To do that, we're going to set in KPIs. We're going to measure everything that we do and everybody's going to work around this table. By the way, just like we did with DevOps a decade ago, right? And said, Deb, you actually have to work with ops now. And they go, those dangerous guys way over in that other building, we don't even know who they are. But in time, people realize that we're all on the same team. And that if developers develop something that operations can't host and support and keep alive, it's junk, right? And we used to do that. And now we're much better at it. And whether it's DevSecOps or Dev2AOps, whatever, all those teams working together, now we're going to spread that out and make it a bigger pie around the company. And it starts with the CEO. And when the CEO makes it a director for the company, I think we're all going to be successful. Isaac, what are your thoughts? I think we're really talking about a culture of transformation and a culture of collaboration. I mean, again, everything that we're doing now, we're going to build, we're going to learn, we're going to use data to pivot what we're doing, we're going to release a product to customers, we're going to get feedback, we're going to continue to iterate over those things. Same thing when it comes to sales, same things that, you know, the experiments that we do for marketing. What we're doing today, we're constantly learning, we're constantly challenging our assumptions, we're trying to throw out the sacred cows and what status quo, because we know there's going to be another island that we have to go after. And that's the transformation part. The collaboration part is really, you know what, what you're hearing. Multiple teams, not just dev and ops, not just data and dev, but really the spectrum of business, of product, of stakeholders, of marketing and sales, working with technologists and saying, look, this is the things that we need to go after over these time periods and work collaboratively and iteratively around them. And again, the data is the foundation for this, right? And we talk about a learning culture as part of that. The data is a big part of that learning, learning new skills and what new skills to learn is part of that. But when I think about culture, you know, the things that slow down organizations is when they're not transforming fast enough or they're going in five or six different directions, they're not collaborative enough. And the data is the element in there that is an equalizer. It's what you show everybody to say, look, what we're doing today is not going to make us survive over the next three years. The data equalizer, that sounds like a, that could be a movie coming out in 2021. Gentlemen, thank you for walking us through some of those interesting metrics coming out of the BISOP's inaugural survey. Yes, there are challenges with data. Many of them aren't surprising, but there's a lot, also a lot of tremendous opportunity. And I liked how you kind of brought it around to from a cultural perspective, it's got to start from that C-suite and to Kieran's point all the way down. I know we could keep talking, we're out of time, but we'll have to keep following. This is a very interesting topic, one that is certainly pervasive across industries. Thanks guys for sharing your insights. Thank you, Lisa. Thank you, Lisa. For Kieran, Taylor, Kevin Serace, and Isaac Sokolik, I'm Lisa Martin. Thanks for watching.