 Live from Boston, Massachusetts, it's theCUBE at the HP Vertica Big Data Conference 2014. Brought to you by HP. With your hosts, John Furrier and Dave Vellante. Okay, welcome back everyone. We are live now in Boston, Massachusetts. This is theCUBE, our flagship program. We go out to the events and sometimes go on stage, like we were here at HP Vertica. I'm John Furrier with Dave Vellante and Jeff Kelly kicking off day two of wall-to-wall coverage of HP Vertica's Big Data Conference. And Dave, Vellante and myself were the keynote stage act, if you will. theCUBE was on stage, technically with no desk, but with chairs. Dave, great session. We basically did up there, did a CUBE interaction kind of thing with crowd chat on the big screen. So they took theCUBE and brought it up on stage to entertain the crowd, which we did and we informed. Yeah, I really was proud to be up there with you and Jeff and having crowd chat in the background. So let's talk about some of the things that were said and we can get into sort of recapping yesterday. I'm really concerned, John and Jeff, of two things. One is the schism between IT and business. IT seems to think that, okay, this big data thing, it's working great, we've gotten the maximum value out of our project and business guys are saying, no, less than 20% of the business folks that we talked to say that their big data projects are paying off to full value. That's number one. There is a big shift coming. The cost, the expense, the ROI on traditional enterprise data warehousing and analytics initiatives and business intelligence initiatives which have not lived up to their potential promise, compared to the potential for emerging analytics projects is big disparity. I think that there's a share shift that is going on now and will go on for the next 10 to 15 years and I'm concerned that this audience doesn't see it. Before we go to Jeff Kelly on this, I know he wants his data on this. Rephrase the first one, the schism, what specifically, slow that down a little bit. Go in slow motion, please. So when you talk to people and ask them, are you doing big data projects and to what degree are your big data projects paying off a little bit, a lot, or none, none to a lot, that big spectrum? Over half of the IT practitioners say, our big data projects are successful. We're realizing the full value. I think about 54%. Less than 20%, about 18% of the business people that you talk to, by the way, the guys who are funding it, the CMOs who are going to spend more than the CIOs say that their projects are successful. So 54% of the IT guys say it's a success, 18% of the business guys say it's a success. Big schism, that to me says there's misalignment. The IT guys are measuring success based on, oh, the technology works. Did my job. Check the box. So check the box. Yeah, check the box. That's not how IT people should be measuring success. That's certainly not how the CIO is measured and is going to be measured. You should measure successes. Did it increase revenue? Did it cut costs? Did it save lives? I mean, those are the measures that matter. You agree? I agree with that. From, I think there's a few reasons for that schism we're seeing. One, as we talked about, the way you measure success is different if you're an IT person versus a business person. Get the technology stood up and operating success versus a business person wants value from the data. They want insights from the data and striving value, whether, as you said, cutting costs, finding new sources of revenue, whatever that metric is. And I'm saying warning, warning, warning, that's misalignment. Right, I agree with you. And so that something has to be done to bring those two together so that IT and the people on the technology side understand the impact that their technology's having on the business side. The other issue, I think, is also that business people have to change their mindset. We talk a lot about becoming a data-driven organization. It's not as simple as deploying some technology and throwing up some dashboards or some data visualizations on it. Business people have to buy into the value, buy into the hype, if you will, that data, big data analytics is going to change the way they do business, improve their business. If they don't, it doesn't matter how much the IT team screams and kicks, hey, you guys got to start changing the way you operate to take into account data. It's not going to matter. The business people have to under, there's a shift that business has to have. So IT has to change as well, but the business people have to understand this is a new world we're living in. So basically, if you notice in that room, Dave Vellante asked a strategic question, Jeff, and he basically said, I love it. The analyst did this perfectly. I love how, and Dave's a master at it. How many people have Hadoob in all the hands go up? Majority of the hands. Let's say maybe half, practitioners. So they're all big data savvy or learning about columnar data stores with Vertica. And then we said, how many people are paying for it? Almost every hand went down. There were three hands that stayed up. I mean, I think those were plants from probably Hortonworks. Those are MapR customers, probably, right? Those are either MapR customers or Cloudera customers. When you talk about Hortonworks, are you paying, meaning, you're not paying for the distribution? No, I said it couldn't be paying for the distribution. The point is the hands went down. The hands went down. What does that tell you? Nobody's paying for it. That means it's early stage POCs and figuring out where to integrate this thing. And so I didn't get the red hat for Hadoob question that I wanted to bring up, but in the reality is big data is a variety of use cases. We've talked in marketing, they call it omnichannel marketing is the new normal for with social business. With Clou, you have different implementations, different stacks and all this stuff. You're seeing different versions of solutions that are actually really tailor made for a specific use cases at scale. That means general purpose software is gone. So I think people are really trying to figure out where to go with Hadoop. How do you customize it? How do I play with it? So what's your take on that, Jeff? What did you get in the survey? Are people really still scratching their head or is it stuck in the mud spinning its wheels? I think people are, look, when you've got a technology that's still developing as quickly as something like Hadoop is, the application for Hadoop is changing from where it was two years ago we talked about on stage from pure batch map reduce style analytics to what is potential is today with things like yarn for interactive in real time. So based on the feedback we're getting from both the survey data but also talking to practitioners on theCUBE and elsewhere. We did dozens of in-depth interviews as part of our survey work. I think people are still trying to figure exactly where Hadoop's going to fit. There's a reluctance to pay for something where you don't yet quite know what the value is going to be. And increasingly I think people understand that it's Hadoop while important is just one aspect of this modern day architecture. And I think there is going to be a significant market for the Hadoop vendors for Hadoop technology and services but where the real value is going to be as further up the stack in the analytics and ultimately the applications that you're packaging that are delivering insights and triggering actions based on analytics. That's where a lot of the value is going to come. So it rains to be seen how much people are going to be willing to pay for what is increasingly becoming a commodity layer. A very important foundational layer but a commodity nonetheless. I still think that the IPA window for a public company is going to open up in 2015 and the truth will be told whether that happens or not and it's going to come down to Dave what we talked about yesterday is will the technology be in place where people actually deploy practitioners to deploy the software and infrastructure to do the big data and where is the killer app? Is it analytics? So it's kind of a red flag for me that Cloudera takes massive amounts of funding from Intel. It's kind of a red flag that you don't see the tsunami of apps hitting the table outside of analytics and visualization which by the way are in my opinion the killer apps. So without that present I'm skeptical right now that the IPA window may or may not be open because if there's no value creation for the vendors that are going public and it's going to shift to the practitioners it's going to be a whole different landscape than what people are looking at. So what's your thoughts on that Dave? Well I wonder, so I've been thinking about this a lot I wonder if in the boardroom of Cloudera somebody said hey guys you know we could do IPA we're almost ready to do that we're cleaning up our act but we're really not going to be ready until let's say mid 2015 and we're worried. Interest rates are going to go up and the IPO window could close so here's an alternative like Mark Andreessen says going public sucks why not give up 20% of the company sell 20% of the company to Intel and create an exit for some of our early investors and maybe some of our senior management I don't know John you might know the inside baseball there and that way we have our options open if the market gets super frothy we can go public if the valuation's there. It's a fact I've confirmed I've confirmed just for the record fact investors cleared cash off the table in some cases all their investment has gone. What about Mike Olson? That's been confirmed that employees I'm not going to confirm names but I know that employees did take off the table that would imply that management did take some off the table whether it's all or not. So there's an incentive and now they have the flexibility if the market gets really frothy interest rates don't go up the bull market keeps cranking maybe they go public but they have that option. So a question that came in that I didn't bring up on the stage because it was kind of long-winded it was really pro-vertica I didn't want to pimp on Vertica because they had pimping already out there but it came in from a practitioner and I got an email they wanted to remain anonymous I love the anonymous stuff it's like don't use my name but ask this question and it was not a bad question and it was Hadoop is moving into the production phase clearly from POC and cost to operationalize it as a data warehouse is becoming apparent period. Does the upfront cost to purchase Vertica versus open source worth it if you get a low cost data lake and cost versus cheap thoughts do you agree or disagree? Well so my take on that is Vertica is a very specific use cases Vertica to me is not your data data your one size fits all data lake it is for loading data fast MPP doing things that are conducive to that type of architecture so you can't do that with a traditional enterprise data warehouse you're not necessarily going to be able to do that with Hadoop but at the same time it's probably not going to be as cost-effective to do your data lake with Vertica so you're going to have Vertica compliments the data lake it's you called it the Ferrari I think you nailed that John Vertica is like the Ferrari that you put on you know you drive on the weekends or you drive when you need it you drive on the roads that are really fast that are designed for that Facebook needs a Ferrari I mean it's a high performance vehicle basically in this case a data movement vehicle it's not a general purpose platform but it's a flagship what they're doing with Haven that's why the strategy is brilliant from an HP's part is that they have the flagship high performance Ferrari that's the core product and they're going down into the product positioning having an entry level in the mid-range it's really smart because as you saw in the room Jeff the practitioners are just now kicking the tires and one question is still people think it's a big data bubble which I'm shocked I was surprised by that as well you know I would say an HP is not positioning Vertica as your data lake HDFS is your data lake right but that's what the question was I think right was essentially should I just use you know Hadoop or should I use Vertica for my for everything or is it too expensive that's what I understood the question then if that is the question as we understand clearly that the answer is no HDFS is your data lake I mean you can decouple compute with storage when you're using Hadoop so you can I mean you can go to a Hortonworks get the distribution for free and store as much data as you want and basically you're paying for the hardware Pivotal gives away their Hadoop distribution same thing and they've changed their pricing structure to basically you're only paying for compute you're not paying for storage like I said nobody's gonna become a trillionaire selling Hadoop distribution absolutely which is another back to your other question about you know when when's this window for IPOs I think it's it's not gonna be a big window because I think investors as they start to figure out this is not a huge opportunity here that the opportunity is for the practitioners that's where the billion dollar companies are gonna come from not necessarily I mean look there may be one company maybe Hortonworks is the billion dollar company maybe Cloudera gets there but this is not a huge market in terms of Hadoop distribution revenue guys let's shift gears to the insights we gleaned from our esteemed panel Q on stage Dave I think we hit some really interesting points from the crowd and the crowd chat and I think two things that jumped out at me this conversation that we just introduced to the industry called born in big data is a really huge deal and what I mean by that is that being mobile first is great cloud first is great data first is which we coined in the cube a few cubes ago but born in big data means that you are born with no baggage you are born with you know a glean in your eye you got a twinkle in your eye for creating value you have no legacy to clutter your mind up you are either a new cultural force like DevOps was in the cloud and that interesting is a cultural shift I think that is a really important new trend that is very clear you either see it or you don't and you know there's a question that said oh there's every tool for the job but if you're a hammer everything looks like a nail so if you're old school IT and you think you're successful with big data projects you're going to hammer that nail and that's what you're going to look for whereas the business people don't see it as success with big data so I'm seeing this culture shift with born in big data what's your thoughts? I think you're right on I think there's a bifurcated market out there you're either born data driven born in big data and in that case you know you've probably got much more alignment to your IT guys and your business guys are probably much more aligned with regard to the objectives of the project and the success of the project and then the other end of the spectrum is and I think many people in this crowd surprised me when I asked are you shifting resources I 100% guarantee that there's going to be more spending and more spending growth going on in the new emerging big data space than there is in the traditional enterprise data warehouse space if you ask business people how's your enterprise data warehouse working out for you? You got self service BI you liked having to wait a week two weeks, a month, two months for data consistently business people tell you now you know we use it we rely on it it's the best that we have but it's not what we envisioned and now the big question I still have is will Hadoop and big data live up to that promise? What would you say? People process and technology so that's if you're born in big data break that down people process technology okay so your people are are new generate younger skilled people in process younger skilled in dev ops or whatever we're calling them software engineering they're data hackers and they're much more aligned from a process standpoint there's no such thing as an IT project in companies that are born data driven it's a business project with a revenue target yes, yes it's not going to get the list and get the form filled out in the technology the guy from Saber actually I think nailed it the definition we brought in Hadoop because we had too much data and we had to handle it differently that's big data integrated stack so integrated stack so the process is simple there's no process it's you know, move fast, deploying process technology is stacks Hadoop I would add to the process part and this was an interesting session yesterday while you guys were doing some great cube interviews I was able to go to a couple of sessions and the lead data scientist from Nimble Storage had a session about really the business value of big data and he talked about how you know, Nimble Storage has been around since I think around 2008 so they're you know fairly new companies so they have the I guess the luxury of doing this but they you know they instrumented their storage arrays and their products with the ability to create and create lots of data so when I say so the process part of this is understanding beforehand as you develop new products that data is going to be key to developing and servicing those products so not kind of going back and saying okay we're going to slap some sensors on our products which is one way to go about it and it's important for existing products but as you develop new products instrumenting them in a way that they create valuable data that you can then use to both improve the product and service your clients that's another process aspect that I think people have to wrap their heads around well I think the process is business process right it's again the alignment between business and IT and so my argument is that the guy it's like it's like cloud first it's like mobile first I mean that those are disciplines that every single company is going to be adopting and I think data first is also born and big data born and big data driven great insight here in the cube again another first original piece of content coming out of the the greatness of the crowd sourcing second observation Dave you asked about the bubble right the question and this is interesting we were talking about this on the panel in prior and I want to bring this up if you look at the big data I truly believe that looking at the data and talking to all the experts and looking at the customers we are not in a big data bubble there's just demand for big data action and born on big data the kind of personnel things we talked about so there's a rationale around big data that's justified you don't see people crazy big data stuff it's really pretty rational in markets that are growing projects are in play but the bubble shifting to other areas and this is where the storage connection comes in storage and cloud are very bubbly because of big data I believe so I think that the thesis here is don't get your thoughts Jeff Dave is that if big data is truly real does that justify the shift to the enterprise with cloud and storage so it's what I was saying the practitioner is going to create more value than the vendors having said that you know what practitioner is getting an obvious valuation like cloud era so I want to go back to Netscape you remember the Netscape bubble you remember well because it jacked up prices in Palo Alto that was sort of the and Marc Antriesen was the co-founder who's my Twitter buddy Twitter friend so what happened in the next day everybody thought Netscape was going to that the browser was going to be the next essential operating control point it didn't happen and then Netscape Jim Barks said no problem if Microsoft bundles that's okay because we're going to go up the stack and we're going to be a software vendor and that all died why did Netscape why was there a bubble because people thought that Netscape was going to create a lot more value than Netscape actually did they were looking for the next Microsoft and the next Intel and then everybody was looking for the next Microsoft and the next Intel it turns out it merged in the face of Google well no it was called the web Dave so Netscape actually had the browser had an opportunity to create some cohesiveness and somewhat lock-in spec if they wanted to with the public free web but what happened was combination of Netscape kind of getting too arrogant which well documented certainly and then making a move into software enterprise software and all this other stuff and then three Microsoft's monopoly chokehold with IE3 just crushed them so that was just a systematic dismantling of Netscape by Microsoft and their own blunders and so that's well documented but they missed it but the web was the value the web came in through practitioners so here's my point but here's my point there was no Microsoft essentially that emerged out of the web like Google right but Google kind of came out of it as a surprise so here's the question related to Cloudera somebody thinks investors think that Cloudera is going to be sort of the big next value play or maybe they just no no your web analogy is actually a good one and I think it's a more stethesis of Jeff but maybe they think it's greater fool than I there you know how VCs think hey if I can make money it doesn't matter but so the question is is Cloudera's valuation justified will somebody come out of the big data world with a valuation such that it is justified and will that be Cloudera or somebody else? I want to answer that one question this is really key so this is what's Bill Gurley from Benchmark who was an analyst before he became a VC was very ProBubble.com bubble so was Mary Meeker you know all the people who promoted it and they all got egg on their face but what Bill Gurley came back and said if you look at all the stuff that they were promoting actually happened so selling pet food the hype was justified was early but the market lived up to the hype how they were allocating the value to the players was incorrect but actually everything actually happened the tooling was slower all the automation came in put moving digital to digital from paper, email all that stuff happened and this is where the value shifted it shifted to the practitioners the people putting the websites up the software, the web, ad tech software so that is exactly the same thing that we are seeing with Jeff Kelly's report and what Jeff's promoting which is the value creation of practitioners the people using the technology that's the web so your investment now certainly Google and others made money and they were worth billions and that was the vendors of but I would argue they're a practitioner yes I would say the same thing Google's a practitioner Facebook's a practitioner Amazon was a practitioner before they became a vendor well then the question is is cloud era a practitioner and I would say no no definitely not and so this makes investing really hard in big data you have to find the companies that are going to transform and the industries that are going to be transformed and make bets based on that so that's retail that's healthcare that's the things we talked about there was a question today about you know when are we going to tackle the really big problems and impact society and you know those are where enormous value is going to be created so I want to bring up the third point so we're born in big data the bubble shifting the enabling of big data rationalization to other areas like storage and also the practitioners but let's talk about the practitioners I think the HP putting up the United States Postal Service was a genius move and I want to highlight that because I think that is representative in today's marketplace globally in the U.S. at least of a legacy vendor would you agree their legacy I think we can they're slow with their government but they have to compete now with privatized companies and they have all the elements of big data internet of things so this is going to be a great presentation this is actually a working case study of the modernization in action so if you look at what they do they have people process technology issues that are legacy and now trying to be born in the cloud born in mobile so to me what they're doing is very interesting they have all kinds of instrumentation issues but that to me is a great case study and I think that's worth watching look I think people naturally people are impatient that comes to especially when you've got a movement of an area of technology like big data that's got so much hype around and people are impatient to see the success stories where is it and then they're quick just as quick to jump on it and say well it's not living up to the hype these things take time it's not going to happen overnight the tooling itself is still developing but the impact that not just big data but the intersection of big data in cloud and mobile is going to have is going to be enormous where the winners though are not going to be in my opinion the vendors it's going to be the practitioners well but so who create value create valuable services that people will pay for that impact their everyday lives that help people run businesses better that's where the value is going to be so when vendors will win I'm not saying vendors won't win but the investing community saw what happened with nobody wanted to own Microsoft and Intel in 1990 when George Bush the elder was president you could buy Microsoft Intel at the time compact for a song if you invested you know fifty thousand dollars back then it would you know you'd be multi-millionaire now so nobody wanted to touch him so when the internet came around with Netscape we said we're not going to make that mistake again well guess what they got it wrong so now vendors will make money now where are they going to make money the woman from Singapore said it what about integration this stuff is too hard I'm a business person this stuff is too hard so developing solutions the question is is it going to be HP IBM EMC the big whales that are going to do it and which startups and pure plays are going to emerge let's use the last minute to address that so Bert also Bert Lattimore was watching has a question is the Red Hat next Microsoft how will the cloud era significantly larger using the same business model or will be bought by somebody basically it brings up the M&A question this is the interesting thing if you say that the practitioner will be the next Google someone will the next Google in this new world company will be a practitioner that has to emerge so the question is people these startups are getting bought left and right so when to sell will they even make it will the big guys actually buy out the next one so they don't want to make that same mistake twice so the question is entrepreneurially are we in a marketplace that's going to be set up for it's going to consolidate guarantee that who's going to get bought first M&A M&A if I were a betting man I'd say map are would be my first bet second would be Intel finishes their acquisition of cloud era Hortonworks I'd probably put last on that list mainly because a large part of their value proposition is being independent so they go public I would if there if I if there was only one company at those three that's going to go public and be an independent company it's going to be Hortonworks well the Hortonworks thing is interesting and I think the Red Hat for Hadoop never is never going to happen but I think they play that card I think Rob Bearden doesn't play it as much or even at all because it's not a comparable or Red Hat had other dynamics and so I just say it but what's different about Hortonworks now is that their business model still isn't playing my opinion because of the the critical mass goals that they want similar to to the old way but the global market place is a very interesting dynamic we brought the bubble up on stage here and we thought about China the global market is a huge issue so if you look at the distribution opportunity TAM and you factor in China and other countries Hortonworks has a run to be massively huge under their current business model by the way that's what I love about what Tableau is doing Tableau is nation building and I love that there's this huge demand for simpler visualization out there the global market is very important and a lot of times these smaller companies don't spend enough time building international because it's so hard and they're putting investments in there okay so we have a big day coming up here day two exciting got our juices flowing really early with the keynotes exciting Dave, Jeff great job for our first real cube we did want EMC world was a little bit different more scripted but this one was awesome HP did a good job we got Chris Zealand coming up we got Zynga we got the SVP of technology at Royal Philips autonomy SVPs we got the West Fox from Tallend Tom Davenport's coming in he's a Babson guy so look at the Babbo thing going for me I've got my where I got my MBA we got Tibco Les Bonnie CO Quilk Larry Schwartz great lineup day two here live coverage in Boston, Massachusetts the cube we'll be right back after this break stay with us