 Hi, I'm Peter Burris and welcome once again to a CUBE conversation from our studio in Palo Alto, California. With all CUBE conversations, we pick a topic and we get down to the meat of it. We're going to do that today as well. The topic today is the role that data is playing in an organization, but even more importantly, the changes that the organization has to make to take advantage of data assets. And to have that conversation, we've got Harry Glazer who's the CEO and co-founder of Periscope Data with us today. Harry, welcome to the CUBE. Hey, it's great to be here, thank you Peter. So introduce yourself, who are you? Yeah, I'm Harry. I'm the co-founder and CEO of Periscope Data. So we started this company, gosh, six years ago in my CTO's second bedroom. And we've scaled it to about 1,000 customers, about 150 employees all in downtown San Francisco. And Periscope Data does? Yeah, we make a platform for data teams. So data teams play this increasingly important and powerful role in organizations where they drive the way the company makes decisions with data. And we make their system of record, their source of truth, a platform that they use to do all their work within those organizations at places like HBO and Uber and Harvard. Okay, so let's talk about data teams. Because it starts there. A lot of organizations are trying to adopt practices associated with better utilization of data and are failing partly because it's catch is catch can. It's everybody's responsibility to figure out what data is, where it came from, what the value is and how they're going to use it. That's right. That sounds to me like the notion of a data team that says, no, we have to bring some degree of, at least centralization in thinking to make sure that we're exploiting data properly. Have I got that right? Yeah, that's exactly right, Peter. So often an organization will start working with data because someone somewhere in the organization wants to, maybe it's marketing. They hire a business analyst into marketing, then sales decide to do it and they hire someone into sales. But to your point, it's catch is catch can. So marketing is looking at one view of data from the marketing world. Sales is looking at another view of data from the sales world. They get into big fights about what the truth is. There's no one to orbit. It goes all the way up to the CEO who has no fucking idea what's going on with this fight. And that ultimately gets solved when you hire a centralized data team, ideally reporting to a chief data officer who can form a source of truth and form best practices in the organization for how they work with and make decisions with data. And presumably take some responsibility for diffusing those practices and diffusing the data about the data to the rest of the organization so you get more common utilization. That's right. So if you fast forward a few years, now all the data is centralized and the data team is centralized and they have formed a source of truth. Maybe they got into a fight about how many leads marketing delivered to sales. Marketing analyst says 10, sales analyst says we only got five. Now you have one source of truth, one piece of data that tells you how many leads. Analysts farmed out to the rest of the organization but farmed out from that central team where they have best practices and sources of truth. Okay, so presumably there is some degree of maturity or questions of maturity associated with these teams. Let's start with day one. I'm going to do this. What's the difference between that and someone who's a little bit further along? What's the first thing that a corporation needs to do? Day one. Yeah, day one is you, if you're doing it right, day one people do all kinds of things that turn out to be wrong. But day one, if you do it right, you hire that head of data first and you empower them first to build that organization and to build that sort of center of excellence. A big mistake that you'll see is either hiring data people to fuse with the organization or hiring a data team but stuffing it somewhere like finance or IT. Those are service organizations. They're not driving their own sort of source of truth and business practice through the organization. You want your data person reporting to a COO or a CEO and you want them to be empowered throughout the organization. So the way I thought about chief and you tell me if this corresponds. A chief is an individual within a business who has responsibility for generating a return on the assets under their control. So the chief financial officer is responsible for generating returns on assets, returns and capital. The COO returns on people and the operations of the firm. Chief data officer presumably then would be responsible for generating a differential return on data assets. Absolutely, so the CDO will take control of all the data being generated by all the various systems in the firm. And then they'll be like to your point, they'll be responsible for generating a return which is the returns from the improved decision-making at the company. If we spend all this time hiring this data department and spend all this time logging and storing this data and we don't actually make better decisions at the company, what is the point? The whole point is that everyone else at the company now has an ability to make much better decisions and that those decisions lead to profits which are the return on the data. So I'm the CEO board, I don't have this today. My first job is to hire a CDO and give them responsibility for increasing the returns on the data assets within my business. So let's talk about one year later, we've hired a bunch of people, how is a more mature data team operating? Yeah, so there's a number of things and they all happen in lockstep. You will have data people who are much more mature and advanced in their careers and their skill sets. That's one, so the people are advancing in their maturity. Your first analyst might be really good with Excel pivot tables. Few years later, you're gonna have people under the chief data officer who are data scientists who work with machine learning and AI capabilities miles beyond your Excel pivot tables. So that's one thing. The other thing is how are decisions made throughout the company, right? So on day zero, your chief revenue officer maybe comes into a meeting and goes, you know, we're going to sell this way because this is how we sold this way at my last company and I know it's the right way to sell based on my years of experience. Fast forward five years, they're going to go, the data shows me that we need to sell in a different way that we've been selling in a couple different ways and this is the most profitable way and I can see that in the data. And so the CEO should expect as a return on hiring this data organization, the people who are coming to him or her with their decisions are backing it up with data as a result of the CDO and their organization's work. Yeah, and that doesn't diminish the value that chief revenue officer experienced, but it just gives them an opportunity to test a proposition, see if it worked, test a proposition and improve things over time, right? A good test is the CDO and their people should be the most popular people in the organization. Everyone loves them because they bring free value all the time. The CRO is probably almost certainly comped on the revenue they generate for the company, right? So if they've got data scientists helping them generate more revenue, that's awesome, that's money for the CRO, that's great, so they should be very popular. If you're the CDO who's getting in everyone's way and causing friction, you probably don't have a good one or something is wrong. Okay, so the CDO is now installed, their power and their influence and their authority is accepted by the organization, practices are changing. Now, the next level of maturity, what are they focusing on? Let me give you a little bit of a background as I think about this because I look back at history and you see over and over and over functions that have processes, some that come from the outside, some that might be developed inside and they try to instantiate, they try to manifest those processes in software because it helps improve the productivity of their people, the certainty of the operation, the certainty of the execution. So I'm into this process, but it's taken me some time. What do I do to accelerate my maturity? Yeah, so I think there's a number of things that's driven from the people, right? But if you start, you know, day zero, maybe you can't even get the data that you want or you don't know that you want the data, the CDO helps you get relatively, gets the data and helps show you what it is and you at least understand the data and can start making some decisions. Then they start joining the data together, right? So maybe I was like, okay, now I can see what I'm spending on marketing and what the return is, whereas I couldn't before. But you still can't say, okay, what's my total spend to get to acquire a customer until you merge the marketing and the sales data? So now you merge into a single source of truth, you resolve all the conflicts and differences between the organizations, that's good. Then you start predicting the future, right? And this is where the CDO kind of takes control of the discussion because previously we're going, maybe we started from a place of sales and marketing, one thing and can't have it. Then the CDO staffs up, builds the technology and answers the questions and this is where they get popular. But then they start driving the discussion. Well, hang on, I can hire some data scientists and I can build some machine learning and I can actually predict based on all the inputs, run all the scenarios for the future of the business and go, this scenario is best. So let's actually invest this way. And so now they're proactively bringing differentiated value based on technology that the company and capabilities of the company did not even know about until they started hiring this team. And so a very mature organization, the data team is actually driving the business towards what decisions they should be making and it's kind of in a much more powerful position even than some of the other chiefs. But I want to talk a little bit about that notion of the future because as someone who has been something of a student of the way that business uses data historically, it's interesting that a lot of the OLTP generation was recording what happens. So it's really using technology to better understand the past. And then personal productivity in many respects was how do I build a consensus amongst different thinkers about what's going to happen a little bit further into the future. So the Excel pivot table often is used to forecast two, three years out. Get people to agree that that's where we want to go. But you're talking about a more immediate notion of the future. The future that's relevant to the chief revenue officer. Like in the next quarter or the next couple of quarters. Have I got that right? It's both. I would say, yes, the Excel pivot table is used to forecast the future, typically in a relatively straight line fashion from what's happened in the past. And that's great. But you can, when you really have a mature data team and you really have a strong source of truth, you might say, actually, the thing that drives revenue more than anything is not the historical revenue trend, but it's the number of active users of your product. Let's say, for example, or does the viewership of your video get past the halfway mark? Those are your best customers, and if we can drive more of those customers, we get sort of differentiated value. And so that requires a more sophisticated technical approach than the simple Excel pivot table. Right, but it's still, at the end of the day, what you're doing is you're helping to, you're using, you're allowing data to drive your next action. Yes, that's right. And that's different from a historical process orientation where you let the process drive your next action. That's exactly right. And so if you do, and to your point, you end up with a, or you end up requiring a more agile organization because you're going to be getting more and more insights over time and changing direction based on those insights, as opposed to saying, here's my process, let's just run the process. Okay, so one of the, you've mentioned a couple of times the notion of system of record for the CDO, and ERP was kind of the CFO's software platform for running the finance and the business. What role does Periscope data play in the world of the CDO? Yeah, I mean, I think your analogy is exactly right. So all of these chiefs, all of these departments will have their systems of record, ERP for the finance team, CRM for the sales team, marketing automation system for the marketing team, et cetera. And we provide that system of record and that source of truth for the data team. And that looks like a lot of different things. Tooling around integrating the data so that you can build a single source of truth with data, storage options for, in fact, multiple storage options for the data itself so that you can run the analyses. The actual system that runs the analysis, so you might be writing SQL code or Python code in the product to perform the analyses, integrations with machine learning systems so that you can predict the future, and all the different ways that you want to share and publish the data out in the organization. All that happens together in Periscope data, Chief Data Officer is managing all of those workflows so that they can manage the whole flow of data through the organization within the product. So as a CEO, pretend CEO right now, I hire my CDO, I empower them to generate a return on data. I give them the authority to do so, and at some point in time I have a team that's being diffused into the organization, but all this can be accelerated if I get the software that will help my people be more successful. Yes, in fact, I would say you probably can't get past a certain level of maturity without differentiated software like Periscope data because it simply breaks. Like the volume of data you want to be working with in that top end of the maturity curve is so large and the sophistication is so large that you really do need differentiated tooling at that point. Okay, so how is this going to change industries? So I've got all this stuff organized because I have a thought and I want to run by you. But from your perspective, how is this going to change the notion of industry? Yeah, so I think that in every industry, you at this point have sort of digital disruptors and you have the old guard. And the old guard is not necessarily dead. And I think you can see, we were talking moments ago about Walmart and the transformation they've made to digital and how that's become a real focus of the company, great example of a company from the old guard that is by no means dead, right? But you do have to embrace the idea that the way you made decisions 10 years ago is not the way you're going to make decisions now. And by hiring this organization and empowering them with differentiated tooling, what you can do is have a much more data driven culture as a result. So you will watch them as they get more mature with data, transform the way your company makes decisions. And it is a cultural change, right? The company becomes much more nimble and agile, probably has less management hierarchy and fewer layers, all of that kind of stuff. It enables you to survive and thrive in a world where you are constantly being challenged by new digital disruptors. And I'll tell you, here's my observation on the whole concept of industry. Industry is a general way of describing how assets are organized. So a financial services firm has certain classes of assets, say airplane manufacturer, certain classes of assets or a bottling company. And you can look at each of these different industries and say, oh, they have this common approach to thinking about what is valuable and how the assets get work, perform work. Data reduces asset specificity. Specificity, asset specificity is the rate to which an asset can be applied to a limited number of purposes. Data reduces that, makes assets more programmable, gives us a better job of monitoring. If we think about that, so that the industry is a function of assets, therefore asset specificity, as more people do more data, it increases, or reduces the barriers. It takes certain respects, it limits the impact of industry and you end up with new types of competitors, new types of disruptors that you didn't know about, what do you think about that? I think that makes sense. I mean, I think we were talking also moments ago about the return on these assets, right? And so the CFO of a major public company will be primarily responsible for investing the company's financial assets across the globe in a way that maximizes the return on those financial assets or minimizes the loss of those financial assets. And similarly, with data, you will start to think about data as an asset. It will be the CDO's primary asset and the return on that investment in that asset will be the profits from the better decisions across the company that you wouldn't have had if you hadn't had a CDO to steward those assets. And the options that are created. Absolutely. So it's a profit now, but also the additional options that are created. And that's where this note, that's where the industry notion starts to get very fuzzy. And like all assets, the return on those assets will compound over time. We'll get the increased optionality, we'll make better decisions. Because of the increased optionality and the better decision, there's now even more optionality, we make a good decision again, right? And it starts to build on itself and you end up in a much better position, relatively quickly. Okay, so Harry, one last question. What's 2019 hold for Periscope Data? A lot of growth, first of all. So it's nice to be a high growth technology startup. And lots of good things happen, but it is a little sort of mind boggling how much the company changes and how much the team changes, the software changes every sort of six months. And so we will almost certainly double or more our business again. We will move into, I've mentioned some of our customers, Uber and HBO and Harvard. That is indicative of a trend where we are starting to work with larger and larger customers and real true enterprise customers for the first time. And so I expect that trend to accelerate. And I will say the conversations for us are getting easier. When we started six years ago and we were talking about platform for data teams, people were like data teams. And now I think you kind of, everybody understands that there's a big wave happening and that's been sort of propelling the company forward. So that's been a lot of fun. All right, Harry Glazer, CEO and co-founder of Periscope Data. Thanks very much for being on theCUBE. Thank you, Peter. I appreciate it. You bet. And once again, I'm Peter Burris and this has been another CUBE Conversation. Until next time.