 Live from the FIA Barcelona Grand Villar Compensator in Barcelona, Spain, it's The Cube at HP Discover Barcelona 2014 brought to you by headline sponsor HP. Here are your hosts, John Furrier and Dave Vellante. Okay, welcome back everyone. We are live here inside The Cube in Barcelona, Spain for HP Discover 2014. This is The Cube where we go and extract the signal of noise. I'm John Furrier, Dave Vellante here and next guest is Colin Mahoney, SVP and General Manager of HP Software, Big Data. Welcome back, Cube alumni, good seeing you. Good seeing you guys. So, always great to talk to you on The Cube. Britica story continues to be the heart and soul of Big Data, Group and Software, HP Software. So what's new? We have Haven announcement, big innovations you guys took to the stage yesterday. A lot of the limelight for the most of the show is awesome. So give a quick update on the announcement. Thanks, yeah. So we announced Haven on Demand and the first two components of Haven on Demand are Idle on Demand and Vertica on Demand. So customers can come, try out the offerings. Idle on Demand is actually free right now, so encouraging anybody that wants to try it to jump on. And what's really nice about the on-demand offerings is the way that they're configured with RESTful APIs, making it really easy for developers. A big part of what we've been focused on is creating that platform that developers gravitate to because of how easy it is to quickly call upon a function, sentence analysis as an example, that they can bring into their own applications. And so that's been probably one of the most powerful parts of the on-demand offerings overall. But we've gotten a lot of requests for cloud offerings in our products and that's just a natural progression from where we've been on-prem. So we continue that. And I think also the way these work together is very different. HP, as you know, is really focused on the hybrid cloud. So public cloud, private cloud, on-prem, out in the cloud. There's a lot of good use cases where you can burst up part of your information into the cloud if you just want to get some more horsepower and then shut it down if you want to move on to something else or keep it up there. So we just saw Alan Nance come by. We interviewed at your event, the big data event in Boston, developer conference or user conference, which is awesome names. We always were so proud to be there because the names are awesome, big names. And it wasn't like a pimping kind of event. It was all a real deal. But he came by and he's like, excited to see the cube. And he said to me and Dave, we did our deal with Vertica. I couldn't talk about it. It's kind of cagey. So share. That's now public. So I see them here doing a little bit of speaking and so some tweets earlier. So he's on board, right? So tell us about it. We couldn't tell us on the cube. Yeah, I think hopefully he's going to jump on and tell you more now too today. But you know, great customer. When you think about data, you think about companies like Philips that have been around for a long time. They're doing a lot with Internet of Things, but they're just leveraging innovations around data and analytics to drive so many different opportunities that they have. But they're also putting a lot of stuff in the cloud and production, a significant amount of their business isn't the cloud. So where's the big data angle on that? Share at the post. We aren't following that deal. Where are you guys fitting into that? Well, I think so. So part of the announcement yesterday was Bortica and Idle on demand, Haven on demand, we're running on helium. And so those two things go hand in hand. I think you're going to see more and more that that's just a very natural combination, really. And so the cloud is a great place to generate data. The cloud is a great place to collect data, to aggregate data. And I think with every organization now, data is not just about what you have inside your firewall, inside your physical organization. It's about what's going on outside. It might be social sentiment data, it might be partner data, it might be customer data. But people want to bring that data in and to do that, naturally, you're going to have to go into some sort of cloud that is built beyond just your own on-prem. And some of those data sources are, you know, vast data, trickle data, you know, things to massive data sets that are real-time, like phone data, like it equals gesture data, massive computation. Yeah, I don't think people realize, but when you make a phone call on a 4G network, roughly 3,000 fields of information are collected about that call. So every time you make a call, there's 3,000 data points on that call that are being collected. They're used for making sure the network is behaving, making sure that you're getting the right optimizations. I think we as consumers just take for granted. I'm going to make a call. It's easy. There's a lot of things that happen. So what does that mean for the back office? You know, in the old days, you put some Oracle databases out there, you get the data. There's some guy in there going, oh my God, now, tsunami of data coming in. Yeah. Every single piece of data times all the users. Yeah. And well, first thing is, how do you process all that? Well, process is just ingesting the information is a huge challenge. It won't work in a traditional database. It won't work in a traditional platform. Then once you ingest it, you've got to make sense of it. You've got to process it to your point. So you've got to do things with the data. And that's, I think, where it gets really exciting. And I think a big change in 2014, 2014 is this is the year that people finally were starting to take a lot of the data that they're putting into these systems and do things with that data. So they stored the data. They started exploring the data. Now they're serving that information out in use cases that actually generate money or say costs or whatever it might be. So Colin, you're telling us there's a new big data group in software that you're running. Describe what that's all about. Yeah. So what we've done in software is we've taken Vertica, we've taken Autonomy, we've created this big data group, this organization and software. And, you know, it just makes sense. And I think the biggest driver for this is our customers are saying, we want unstructured data. We want structured data. We want it to just be seamless in how we work with you as a vendor. And so what we did was we aligned our organization around that. And so as you know, we've been working on Haven. This is the team, collectively, that's been driving it. And so having this as a single team, a single organization and software makes a lot of sense. And the other thing about it too is, on the Vertica side, we spend so much time talking about data as an asset. And we still do, right? How do you monetize data? How do you get a competitive advantage out of data? Same thing with idle against unstructured information. But there's a whole aspect of data that's not just the asset, it's the liability of data. How do you archive it? How do you secure it? How do you govern it? How do you make sure you're keeping in line with the regulatory requirements? That's a huge part of the traditional... I need to lead it defensively when you don't want it around anymore, right? Maybe, sometimes there's figuring out how long to keep data around from a regulatory or just from a cost standpoint. These are all policies that have to go into it. This is what we've been doing for a long time in HP software and with the autonomy portfolio, especially. So bringing that mix onto the side of the asset generation and similarly bringing some of the analytic capabilities of the asset generation aspect of data into the liability side make a lot of sense. And so having this portfolio combined already in the short period of time we've been doing it, there's just been a great cross-pollination of did you know you could take this analytic function and apply it here and do a knowledge graph? Did you know you can take this control point or supervisor technology and think about how you manage the information, this life cycle of information? So I'm really excited about it. I think the teams are really excited about it, especially on the engineering and the sales side. Customers are really excited about it. Partners are excited about it because it's a lot easier to understand how do they work with us and engage with us around big data. So the mission is to build big data solutions, is that right? We do. So we have, we already have several solutions, especially on the archiving side, on the digital save side, on the data protection side. And we will continue that. But, you know, there are going to be other solutions that come out. And as you know, Haven was created with this notion of nApps on top. We've got a lot of partners with internal to HP as well as external that are building the solutions on top. And we're going to continue to build that community, work with developers, work with application companies, services partners to build those solutions. So organizationally it's a P&L or is it an overlay? It is, yeah. It's part of software as a P&L. So you're giving a talk today? I gave it the support. Okay. So you talked about the balance sheet, right? We do. We talk, the notion that seems to resonate really well is when we talk about what we call the big data balance sheet. And like any balance sheet, you want to maximize your assets, minimize your liabilities. And information is no different. Information is perhaps one of the best competitive assets that any organization can have. But we go in and start talking to them about, let us talk to you in a complete portfolio sense of what we can do to really maximize your big data balance sheet. Help you get all those assets in the right place and help you in your organization share the data, leverage the data, make it work for you, and minimize all the liabilities around the information. And that's what we're hearing more and more. And more and more you're getting roles in organizations like Chief Data Officer, Chief Analytics Officer, Chief Compliance Officer. All these people are coming together saying we know what the tools and technologies are, but help us bring it all together. And Enterprise Services has a huge play in there as well. So before the big data theme hit, the General Counsel was really the sort of tail wagging the dog. If the head lawyer said, no, we can't do that, boom, everybody's hands were cuffed. That's changed, obviously. It's almost completely flipped, but there's still that General Counsel has a lot of influence. Where do you see that sort of balance of influence now between the business guys wanting to get insights out of data and pushing the envelope? And the General Counsel saying, well, hold on, we can't do that because it's a compliance or the legal reasons. Yeah, so I think there still are a lot of disconnects in these companies. And part of what's happening is you have a little bit of the Wild West with data where data pools are being created and everybody's going and analyzing the data. And that's not necessarily a good thing. You really need that conversation to happen between compliance, between legal, and the folks that are doing the data lake environments. And when it's done right, it's very powerful. The people that have access get access. The people that shouldn't don't get access. The right policies are in place to manage it. And that's what professional organizations need. They don't want to stymie innovation. They don't want to slow things down. But you've got to do it in a way that is legal and it makes sense with the overall sort of corporate rules. Data governance. Data governance. And that's something that no one likes to talk about it. It's the thing maybe that everybody sort of says, well, we know it's there. Let's not talk about it. You need to talk about it. Because when you're designing these environments, when you're talking about petabytes of data, if it's not done from the beginning, it's going to be really painful trying to unravel all that and make it work the right way. So it's a conversation that we're having very openly. We think it's a huge strength of ours. And it's just one more thing that can differentiate. You want to future proof your customers, right? You want to have a roadmap. So and bringing that back to the customer is the strategic asset concept that's been kicked around. Data is a strategic asset for customers, right? So how do now practitioners who really have to use it every day? That's the real value of big data is that it solves problems, everyday problems, like real-time stuff to ingestion large logs, files for your phone or whatever your app is. This is a lot of real-world examples of big data. It's a strategic asset. What do companies are doing now to understand that, value it? ROI is a huge issue that data now is a big part of, and it's not talked about much. You guys do talk about it, but in the industry, there's not a lot of discussion around the ROI piece. Like, okay, data is measuring things. So now we have the systems and the speed of Vertica and Haven to explore the data and govern the data, protect the data, all that life cycle stuff you guys talk about. But into the day, it's like, okay, my investment is the data. What's the return? How are people thinking about this and what do you see as solutions that they're justifying their other parts of the business? Yeah, so we did, we recently published a report from an AT&T customer and they published 657% ROI on a, this is a Vertica project that they did, but they had all the metrics from the beginning of this project through the sort of delivery of it, that they could back up and show exactly what they got out of it. And I think they did it right. They knew going in, this is the return on investment that we're looking for and what they were referring to as well as return on information. And I think when it's done, when you go in thinking about, I'm going to invest this much money and this is what I'm going to get out of it. In their case, it was revenue generating, reduced costs, reduced time, then there's a very clear ROI. I think that analytic engagements have some of the highest ROI in terms of IT deployments right now over almost anything else. And because they're so data focused, it should be pretty easy to get the ROI data. But too often there's still not enough thought going into the financial side of it. I think great IT organizations do this very well. Every single thing they do, they have an ROI attached to it. But we're seeing incredible ROI's from organizations that are doing this properly. Do you see value chains being measured in silos or under activities or you see companies looking at the data measurement aspects or asset aspect of the data across the value chains their whole business? Or is it you seeing it on a project by project basis? I mean the AT&T is a great example. But that's a vertical project, that's not their whole business. But as a customer, how am I supposed to be thinking about not just projects but big picture? Yeah, I think I think the CIOs, their offices are thinking about the big picture. I don't think it's rolling up perfectly today where across the entire chain they know exactly where things are at. But on a project by project basis, I think that we've come a long way and understand it. And honestly, I think these projects that if you if you say I want to buy a big data solution, and it's going to do everything. Go to Walmart, pick up big data on the shelf. I mean, what do you? But if you try to just purchase this thinking it's going to boil the ocean and solve a big problem, you're going to fail. The best big data projects are the ones where a line of business says, I think there's something here. I want to go learn about this. And and it's narrow enough where they really can say, yeah, it took us this many resources. And in this amount of time, we delivered in that amount of time should be days, not months and years. So you see an acceleration. First of all, we're in basically first generation big data. I mean, would you agree? I mean, we're like general, genuine. Yeah, like the pioneers, you're the pioneer. So complexity has been an issue, right? So, you know, we know we live our big data product. Dashboard has got to be simpler. It's doing a lot of stuff to look at. No one wants another dashboard, right? I mean, so but they want the data. So what are you guys doing to make it less complex? So a big, a big part of what we've been doing is we talked about serving the data. And when we talk about serving, it's about showing people the right dashboard or sending them the right alert at the right time. I think to your point, John, people are inundated by dashboards and high charts and, you know, the traditional bi just tell me the insight. Give me the answer. And more than the insight in the answer, I think what everybody's most interested is tell me what's going to happen so I can change the outcome before it actually happens. And that's where a lot of the predicted elements are going to work on. And that's what's incredibly exciting to me. Being able to look at something and say what's going to happen. We have one customer, you might have heard the story of the BDC, they talked about it, but it's a gaming company and they just decided to track every time somebody's finger hits the screen, they can track that. And they didn't know what they were going to use it for. It turns out they have a 600 millisecond timing after the game ends before a bonus button comes up. And there was one person, one of their games was able to hit the screen 13 times and 600 milliseconds, which is amazing in and of itself. They still didn't know what the data was going to do. What they ended up learning was that when somebody does it 13 times and then only nine times and then only seven times, that diminishing number of how many times they try to hit the screen before the bonus button was actually the perfect prediction of when they were going to leave the game about a month before they churned out of the game. So they just randomly collected this data. They didn't know what the data was going to do. Turns out though that this is the best correlation to when someone's going to turn. So what they did was when they start to see that activity happening and in focus groups, they would ask the people that people didn't even realize that they were losing interest in the game. They were they were acting on it, but they didn't even know in their minds. So here's a here's a data machine data in a way that you can collect. They will predict so they did two things. They would try to make the game more interesting at that point in time to keep them in the game. And if it's still went down, they'd introduce them to another game. And I think there's so many examples like that in this world where you can get data like that, maybe on how you're driving a car or your health care or whatever it is. But the most exciting thing to me is predicting and being able to figure it out before it happens, save a life or save the environment or save a customer. That's where it gets really exciting. And that's why I think you guys are right in very early days. We're seeing ROI and I want to comment this is very bifurcated. You got guys like AT&T that are getting off the charts ROI. And frankly, I have to say I would observe at the Vertica user conference. Many of your users are in that sort of early adopter with high ROI phase, but there's a lot of experimentation going on, especially in June. And people, I don't know where the ROI is. So it's really split with the guys that are getting telephone numbers on ROI. The other guys like the IT guys say, Hey, project work, the business guys say, Well, yeah, but what do we get out of it? And so I wonder if you can comment on that. And you sort of did before on the future of the potential of big data, not just to affect, you know, how many clicks, you know, you could do retail even but other industries. What are you saying? Yeah, I think it is absolutely moved from the early adopter phase. If you look at our customers, say, currently versus a year ago versus two years ago, there is a progression where in the very early days, we had a lot of companies that were just very innovative, very progressive. And then larger organizations that you might consider more conservative, just started coming in as customers as well over the last, say 18 to 24 months. I think that will continue to happen. But one of your other points in terms of the business and the analytics people, every project I see that's successful, there's very good communication between those people. The statisticians, IT organization and line of business are all in complete sync. The projects, a lot of them that fail, I'll hear IT say, look at this amazing thing we built. And they're very proud of it. And it is probably amazing. But maybe somebody on the business side doesn't understand it. Or in some other areas, there's somebody on the line of business that takes a relatively small piece of data that they can get access to. And they do a great analysis or a great visualization. But it wasn't on enough of the data because they couldn't get access to it. You still hear a lot of that. And that's partly probably why there are chief data officers and chief analytics officers are trying to bring those groups together. But I think you're going to continue to see this in a much more mass market adoption manner over or certainly FY 15 and beyond. But I think it's getting there. Hadoop is out there. Hadoop is part of what we do with the platform. It's a great place to ingest data to store data. I think the world's still trying to figure out exactly how much analytics and processing they should do there versus other places. Hadoop as we talked about before just continues to evolve. But I think customers care less about the technologies. They just want to solve their problem. And I think the other thing that we know in our research is, I think you're right, a lot of experimentation going on and they haven't figured it all out yet. But one thing we see is that they are baselining their investments in traditional legacy data warehouses. They're saying, OK, well, that stuff really didn't live up to all those wonderful promises. We're not sure there's a dupe stuff. Well, so we're going to experiment there. So they might be baselining what they might have spent a dollar on in the old stuff. They might only spend 30 cents on the new stuff. Now, the interesting thing about Vertica and Shilpa makes this point is you guys sort of fit in both camps with that bridge point. And so it's obviously reflected in your business and makes it that you guys were tripled high double digit growth rates. I don't know if there's other metrics that you can share with us. But so what are you seeing there in terms of the shift from the traditional EDW into this new big data? Yeah. So two things. So traditional EDW world, relatively small amount of data, relatively to all the machine and other data. And secondly, very expensive, very rigid, very expensive. So customers are frustrated for both those reasons. That doesn't mean that their investments are bad, right? They should still leverage everything they've done there because there's been great work done in traditional enterprise data warehouses. As you said, what we like to do is say, don't throw out what you have. We can be that bridge. We can we can take all this new type of information having to do. You can have it straight in our engine. We don't care. And also you can leverage your existing EDW and other traditional legacy environments and will help you with that bridge. And then over time, what these customers do is they realize I can do a lot more on this new platform for a lot less money. I'm comfortable with it. I see that my applications are working on it. And then over time, they probably spend less on that core EDW or with maybe the core EDW traditional vendor and they move it over. So I think it's it's absolutely price related. But it's also just the volumes right now. These traditional legacy database products were never built for them. And so we are able to take advantage of that. And that's been a great model for us sitting right in between. Yeah. And the market potential is enormous. A lot of people are thinking, oh, wow, it's a trade of dollars and 30 cents. The market's bad. But no, the volume is huge. It's going to it's going to far surpass my opinion anyway. I want to be great where we were with the traditional EDW market opportunity. I completely agree. I mean, it's the same thing as storage. If you look at the store in the market, it's elastic. Drop the price. Boom. Price goes down. People just store more. Colin, thanks for coming on the side. I really appreciate it. Big data is driving new creating new leaders, making business change. You guys are a big part of it. Gen one of big data. Just getting started. Congratulations for all your success. I'll see Haven on demand. Good move there. Free products for people to use idle on demand as well. Good stuff. It's theCUBE. We're getting all the data sharing it with you. We'll be right back at the strip break. I'm John Furrier with Dave Vellante. Live in person alone is theCUBE. We'll be right back. Thank you.