 The Cube, news and analysis from Big Data SV 2014, is brought to you by headline sponsors Actian, accelerating Big Data 2.0, and WAN Disco, we make Hadoop invincible. Hi everybody, this is Dave Vellante, one of the co-founders of Wikibon.org, and this is the Cube SiliconANGLE's live production of Big Data SV. Big Data SV is running concurrently with Stratoconf, which is here at the Santa Clara Convention Center. We're across the street at the Hilton Santa Clara, and I'm here with Jeff Kelly, who is Wikibon's Chief Big Data Analyst. We've been giving you wall-to-wall coverage of this event, all the news and Big Data, and bringing in all the guests, the Cube alum, New Cube innovators, startups, people in Silicon Valley this evening at 6 p.m. We have a big reception here, a big Cube party. A number of people coming, we have a couple hundred people who have already RSVPed, they're still rolling in, so if you're in the area, stop by, I would love to see you. Jeff, good day yesterday, you were over there at the convention center, John and I were over there last night, looked pretty packed. Stratoconf really is becoming quite a big, big data show, a lot of suits this year, a lot of vendors. Remember when we originally came to Stratoconf, it was the T-shirt crowd, it was a lot of the hardcore practitioners, a lot of data scientists really is becoming sort of overwhelmed with the business people. What do you make of that change? Well, I think that's a good sign for the market generally, because business people write the checks, right? So it's a sign that the market's starting to mature, and that Big Data is being taken seriously inside the enterprise. But to your point, it's definitely the show, Strat itself, has evolved significantly over the last three years. We saw a lot of data scientists, hackers, developer types, maybe three years ago, and that percentage is doing a little bit. And now you're seeing, as you said, the more suits from a vendor perspective. The first couple of years, we're really focused on the Hadoop startup community. Last year, what struck me, walking into the show, remember the first big banner I saw was EMC? I thought that was pretty interesting. And of course now, if they've spun out Pivotal... Cloudbeats, big data. Yep, now they've spun out Pivotal, so there's Pivotal, and there's IBM, and so a lot of the big companies are taking notice. They know that this is a really important market for them. So it's definitely maturing of the market. That's a good sign. But really, when you look at the vendor landscape over there, it's quite a hodgepodge. It runs from the hardware up to through the database, through Hadoop, analytic applications, up to visualizations and services. So it's pretty much any kind of vendor you can think of is over there. Strat it today. So we're here in Silicon Valley. Of course, we're right next door to the new Levi Stadium that's going up. It's mostly complete. I think the Super Bowl is going to be here. What is it, 2016? Is that right, guys? 2015? Okay, yeah. Well, so build a new stadium, and you get a Super Bowl. So they're going to have to build some new hotels around here, I would think. So a lot of action going on here. Jeff, I want to talk about the big data study that you just released. I know we're going to talk about it a little bit later, but I want to tee up some of that. You released their first report. So you, we, it was a community effort, really, back two years ago. This is the third study. What prompted you to actually dig into this and release that first study, which is a market share, market sizing study? Well, I think a couple of things. One, we were getting requests and questions from our community about market share, which there's a lot of startups, obviously, in this community who's gaining traction, who's not. But just generally, I think we wanted the information as well. It's interesting to us as analysts. So we didn't see anybody else doing it, and we decided to take a stab at it. So I remember when the report first came out, you had a forecast. You were sort of targeting 50 billion by, I think it was 2016 to 2017, and it's roughly the sort of same shape of the curve right now. And it's got a little sort of slope to it, and it starts to pick up sort of mid-decade, 2014, 2015, really starts to rise. So last year, you said the business did almost 19 billion. And so that's a big number. Now when I look at who is driving that revenue, there's a lot of big guys in there. IBM's the leader, HP's number two, Dell, you don't necessarily think Dell is synonymous with big data, SAP, Teradata, Oracle, SAS Institute. So the first six or seven are large, established, traditional vendors. So what constitutes big data in your definitions? Well, a couple of different things. You can look at it from a technology standpoint. Things like Hadoop, obviously, things like NoSQL. But there's also a lot of technologies that are not necessarily new but are being applied to big data. That could be data visualization, data integration. You've got to get the data into these platforms. And then, of course, you've got services as well that help companies both from a technical perspective, architect systems, but also determine things like best use cases to start with and building the business case. Some of the other things that we include, of course, are the hardware. You've got to run these systems on some hardware and to scale out architecture. So you mentioned Dell, a lot of Dell boxes are used to constitute these Hadoop clusters and other things. So it's kind of a range of technology. But it's all those things like data visualization, et cetera, as applied to big data workloads. And the other part of the definition, the way I look at it, is big data is also a mindset. It's not just technology. It's just not the structure or lack of structure of the data or the volume of the data. It's the mindset, how you go about analyzing data, looking for new ways to do things, new ways to drive new lines of business, new types of revenue, become more efficient. So it's really a mindset as well as a set of technologies. So I'll pressure you a little bit because a lot of the things that you just said could be considered, you know, driving new types of revenue, et cetera. I mean, Larry Ellison could stand up and say, yeah, that's us. We do big data. But isn't there an element of the definition that suggests, just again from a technology perspective, that you're doing things differently than you would with traditional systems? In other words, the type of data that you're dealing with, the amount of information that you're dealing with, the speed at which you're ingesting data requires you to do things differently. Like, for example, ship function to the data as opposed to bringing data into a pipe. Do you make that distinction in your definition? Yeah, absolutely. There are different ways that you need to process this data. It's not so much that you couldn't necessarily do some of these workloads in the past, but it would be just too expensive or take too much time to make sense in a business context to do these workloads. So from that perspective, certainly it's the way you, as you said, you bring the compute to the data rather than moving the data around. You look at different types of data. You integrate multiple types of data from different sources outside and inside your enterprise. It's about doing things in real time, making real time decisions based on analytics. So absolutely, there's different components to it and from the oracle perspective, they have their definition of big data for sure. But one of the keys, as I mentioned, is the big data movement, as we constitute it, as we define it, a key component is that it is practical from an economic perspective. So while you could potentially buy a really expensive buck from Oracle to do some of these workloads, that's not really a viable option for a lot of enterprise. So I want to bring John Furrier into the discussion. The big data movement started here in Silicon Valley. Really it was Google, things like Bigtable and MapReduce and the activities that were going on at Yahoo and then you had subsequently startups like Cloudera. And John, you were here watching that, but before I get into the big data, the social media, the whole Web 2.0 trend, preceded big data. So I want to start there. That was to create a lot of excitement in this area. You're seeing some big IPOs, Facebook, Twitter, LinkedIn, et cetera. My first question is before we get into the big data pieces, are we in a social media bubble? Dave, good question. I think one of the things that's exciting to watch is if you look back at 2008, that was really the great recession point where you saw a kick up from there and that's been a growth cycle ever since. And if you look at the activities, especially on valuations on the market, you've seen significant growth in opportunity. So you saw Hadoop come out of that, the unstructured movement of databases, and then you saw Facebook platform, which was launched a year earlier, come on, Twitter, et cetera. And then you've seen since then a huge data-driven social media perspective, obviously the iPhone, propelled that with the smartphone. So that brought the consumerization trend in there, which really highlighted the value proposition that big data can offer. At the same time, that was the beginning of the cloud kind of Kool-Aid. People were drinking the Kool-Aid around clouds. So we're now good five years into that movement, and you're seeing huge growth. So the market is growing, it's been growing like a tsunami of opportunity. And valuations have been reflected on that. So if you look at the companies here that are now growing, the startups, per se, the valuations are very, very high. So there's a huge growth opportunity that's a premium on those startups. And the big guys are coming in with strategic changes with their plans. So you look at the data warehousing, business intelligence markets being retooled, you're seeing the notion of data science, which is being driven by machine learning, and some other technologies that's really built on top of open source and the database changes. So really you've seen since OA to today, a massive opportunity. And this year I think the valuations are reflecting on that trend, that mega trend. So it is a perfect storm. It's the confluence of those mega trends. And in Silicon Valley, we are in a very frothy bubble environment. But there's growth there, Dave. So I think the question is, who is going to tap out? Who's going to extend that lead? Who's essentially rearranging the deck chairs, as we say, versus really delivering values? The value conversation ultimately is what investors want. That's what consumers will vote with their dollars. And ultimately the valuations will level out and or grow based upon that valuation. So clearly guys like Facebook are delivering value. You made the call. When Facebook's IPO came out, it was not well received because they sort of oversold it. They did a great job actually of picking Wall Street's pocket for a change. But so you made the call. You were a bull on Facebook. And that's at least thus far turned out to be right. Twitter was sort of the opposite. Twitter IPO very successful. And then it took a hit after it announced its first earnings release. But I think you're still a bull on Twitter in terms of its value creation potential, obviously LinkedIn as well. So those are three strong examples of value creation. I mean, I think LinkedIn is one of those things where it's interesting. They're getting kind of good marks right now on the heels of a Twitter disastrous first earnings call from the Wall Street's perspective. LinkedIn's showing some pretty good numbers. But there's some nuances here that I'd like to explain. I'm bullish on Twitter. I'll tell you why. And LinkedIn I'm not too sure on. I think LinkedIn has some characteristics of a viable subscription market with their target audience, which is headhunters and people looking for jobs. And I think there's a nice revenue model there that's steady and cool. The problem with LinkedIn I see is that their data is not open. And you're hearing a lot at the Stratoconference about closed data. And that ultimately may or may not be an Achilles' heel for LinkedIn. On Twitter, I think what's interesting about Twitter is they kind of took it in the shorts on their last earnings call because their active uniques aren't growing as fast in terms of new net new users. And I think what you heard from Dick Costello at Twitter was essentially saying, look it, we're making some tweaks to our algorithm on timeline impressions. And timeline impressions is the metric that everyone loves because it's a big number. And that's really not the preferred user experience to see with Twitter to make that growth number work. Now Twitter to me is absolutely viable, great utility. The user experience issues will be addressed by the company, but they're biting the bullet. Now biting the bullet means they're shifting the metric to not timeline impressions but interactions and engagement, which is the right direction. So in a way they're biting the bullet. And I wrote a blog post on that and saying, hey, I'm long on Twitter for that reason. However, Frank Slutman at Service now pointed out on my Facebook thread, my private Facebook thread is that's roadshows are for. They should have taken care of this before the road show. So the thing that's the wild card in the Twitter equation is they all made money. They're all rich on the IPO. And should they have taken care of this business prior to the IPO? Unlike Facebook and Google, for instance, they took care of their revenue model business per se prior to the IPO. Google more than Facebook. Facebook had some great revenue coming in just on sheer volume numbers of interactions and engagement. They then tweaked their model to be much more eccentric in the newsfeed. So I think that's a little bit of an exception. But Twitter certainly is viable. I think it's a good call by management, smart move to shift the metric from Wall Street from timeline impressions to something more specific that actually highlights the flywheel of user engagement. So I want to get your take on one of the things before we break here and bring on our first guest. So the social media movement, the whole web 2.0 thing, it was all pure place, all startups, all new companies. You didn't have sort of traditional vendors, quote unquote, get in there. Of course it's not necessarily an enterprise play. With the big data movement, you're seeing all these companies, IBM, HP, Oracle, name any big company, and they're sort of co-opting the big data themes. So is this a case where the big guys are actually innovating or is this all marketing? Are the startups getting overwhelmed because they're part of the distribution channel? What's your take on the innovation cycle with big data and how it differences or similarities with the web 2.0 internet? Well, I think as you know, Dave, I started a venture-backed company in 2004, a five-time frame around the podcast. It was called PodTech. And my contemporaries, Evan Williams, started a company called Odeo. There's another show, one I'll call PodShow, all venture-backed, all to democratize media. And it was a great vision. The bottom line is that podcasting really never made money because of iTunes and the freeness of podcasting. But we were all part of that first generation social media movement. And what happened there was a couple things. A business model just never materialized. And I think we all kind of wanted this to happen. We felt good about the democratization of media and user experience. And then the 08 recession hit, and that really kind of dampered up the whole web 2.0, the media market from podcasting, blogging, and web 2.0. But web 2.0 just never had a business model that could hang their hat on, saying this is a scalable, recurring revenue model that drives growth. Enter big data. Big data with the smartphone highlights the benefits of the value you can get out of big data. And I think the deliverables you're seeing from companies that are bolting on a revenue model and a business model with big data is value. And people will pay for it. We heard from MapR yesterday they take an approach with distribution when they charge subscriptions. Bankers love that. The other open source models are banking on support, both viable, both business models. And I think the key with the big data movement is that you have the maturization of the platform and cloud. You have the smartphone, a mobile-only mobile first. You hear that buzz. And more fundamentally, you have a business model behind it that delivers value that people will pay for. So to me, a business model, good market timing, and that ultimately is really what Web 2.0 didn't have. So Web 2.0 kind of transitioned into the mobile market. So to me, all the Web 2.0 stuff hits right now with mobile and the scalability of cloud, getting in, low cost, a lot of leverage. And those are nice factors in building a business. John, great perspectives. Jeff, thank you as well for sharing your thoughts on the big data market. We're going to hear more from Jeff Kelly later. And of course, John and I will be back throughout the day, as will Jeff Kelly, interviewing guests from Big Data SV, from Stratoconf. We're here live in Silicon Valley at the Santa Clara Hilton. We'll be right back after this word.