 Live from Las Vegas, extracting the signal from the noise. It's theCUBE, covering IBM Insight 2015, brought to you by IBM. Welcome back to IBM Insight everybody, this is theCUBE. theCUBE goes out to the events, we extract the signal from the noise. It's really our pleasure to have two Gartner analysts here, the biggest analyst, the best firm in the industry, Gartner, Merv Adrian is here, Cube alum, and he's joined by Donald Feinberg, first time Cube guest, Donald, welcome Merv, good to see you again. A pleasure always. So Donald, since you're the newbie, we'll start with you. So you're a Gartner for a number of years, your 25 year anniversary, I can see on your analyst profile, database, watcher, expert, you're quoted as saying it's a pretty sticky business, the database business. So it's really interesting what's going on in that business. First of all, Insight, what's your take on what's going on here, IBM's big analytics show, they don't use the term big data, but what's your take on what's going on here? Well I think they're spending a bit more time on other products around the actual database. So yes, they're talking about DB2, I mean obviously, but they're talking more about Spark, they're talking more about DashDB, they're talking more about the surroundings products that fit with DB2, which is correct because DB2 is one of the strong, what I call, I guess, old, I hate the word older, but older databases, mature, that's a good answer. Mature databases along with the others that we know of who've been around for 30 years and stuff like that, and they're coming out with a lot of new things. So emphasizing the new, the no SQLs, the Hadoop, the Spark and that kind of thing is important because that's starting to have a real impact on the overall information management world. Yeah, there's at least, I think, at least five by my count major databases that they're selling now and there's probably more. We had a customer on yesterday from a South African bank extending IMS into- That's still strong. It's huge. Well, and you know, why convert? I mean it's just going to kill their business. Well we're talking about a company with a portfolio that actually treats its portfolio, like a portfolio, that is engineering its collection of products into a series of related engines that may sometimes overlap in functionality but often are far more suited to one workload or one type of data than another. And that's the challenge that their customers are facing, all this big data stuff. Donald and I have had a lot of fun saying there's no such thing, it's just data, but all of it is about new kinds of workloads with new kinds of data that require different kinds of processing. The idea that you're going to stuff it all into one database to rule them all has pretty much been proven relatively ineffective. So here they've got a portfolio, you have two ways to do it. You can do what some vendors in this space do and only talk about your flagship product and put the other guys at the little kids table in every conversation. Or you can really attempt to carve out an integrated story within some sort of a thematic infrastructure. And the story here was cognitive, right? The idea is how does all this stuff work together for a new style of computing that's about cognition? And well, I'd say both of us come away from this with the feeling that IBM is establishing a very differentiated approach to the market which has its challenges because it's going to be a transition that's going to take a little while to pay off. We all know they've had a couple of challenging quarters. This isn't going to turn their numbers around in the next quarter or the quarter after that but long-term or medium-term. It's a good differentiated strategy and it's beginning to get some traction in the market. It's clearly differentiated but you sound sanguine. Yeah, maybe Wall Street's a bit impatient but what are your guys' thoughts on? Wall Street's always impatient. Yeah, so what are your thoughts on the big bets that IBM is making? I mean, they've made some historic bets here. You mean like Watson? Yeah, like Watson. And is that going to pay off? But that plays to the cognitive that Mer's talking about. And, you know, it is a beachhead. None of the other vendors have taken that beachhead so they have a position that they can now defend regardless of how far along they are, they're ahead of everybody else. Well, and actually they are moving ahead. You know, a couple of years ago when Watson was on Jeopardy, okay, that's interesting. What can I do with that? Yeah, it came across as a gimmick but a good one. Today, I think it was Chef Watson. Yeah, right. What's in my refrigerator? What can I cook? But today there are 500,000 users and we dug a little deeper with them here and 5,000 companies using the software, not necessarily all of them paying for it but if you think about that ratio, that's 1%. You and I have talked to plenty of open source companies who've come on and said we've got a million downloads and we've got a thousand people already paying, isn't that great? One tenth of 1%, they've been happy. Well. Okay, so IBM's making progress and I think the key thing that we saw here and they were great pains to put on stage, in fact, to the exclusion of promoting their products which we can come back to, on the first day they put real customers with real use cases, applied cognitive computing and this is all about value delivery. It's not about whether I can do a better job of sentiment analysis if you give me the tools. I've got a sentiment analysis solution for you. You don't have to build it yourself, right? As one example first. You're right. I mean, the keynotes, we go a lot of these events and the keynotes are a lot of product stuff and almost too much. Not here. Not here. It's thought leading kind of use cases but real use cases that give you a glimpse of the future but I wanted to ask Donald about Comet that Merv's made about the open source guys at Zillion Downloads. Amazon's making a lot of noise, about a thousand downloads. I think of Aurora, it's a new database that's moving into that space. They got a million customers, thousand people downloaded it. Sounds good but out of a million customers. It seems like it's going to be very hard for the likes of Amazon to claw in to the cloud some of those hardcore transaction database workloads. What's your thought on that? You know what's interesting? We just finished our magic quadrant for transactions and this year for the first time we allowed cloud only on there. Amazon just bolted to the top. Really. Now not because of Aurora but more because of DynamoDB which is so widely used but Aurora has, first of all, it's MySQL compatible. So if you're running MySQL already on Amazon it's literally a no-brainer to move over and it's pushing one button, literally. If you're not running it on Amazon let's say you're running on-premises it's 100% compatible. It has scalability that's never been seen before except in DynamoDB. It has stability with 11 nines, they claim, which 11 nines is 100%. I'm happy with six or seven myself. So we believe that you're going to start seeing some workloads moving from on-premises in MySQL initially because it's 100% compatible to Aurora. My, our prediction really is that over the next three years MySQL is not going to exist, okay? Because you have alternatives, MariaDB and Aurora now that are going to be able to do it better, faster, cheaper and really not just say that, do it. And Oracle is really not, you know, incented to at least 100 million dollars with the revenue. Yeah, it's not even worth it but let's take it to another part of the market for a minute. Think about the open-source non-relational DBMSs that are out there. One of the great stories on stage here was a gaming company that was using CouchDB, Amazon CouchDB or Apache CouchDB. They had a nice little run and then they decided they really needed to be scalable and they really needed to be reliable. That was this morning. They turned to Cloudant, right? This is, here's an example of the transition IBM's undergoing. It's fascinating to see a guy who's part of a company they only acquired a little while ago getting one of the key spots on stage here. It's a reflection of how well IBM integrates technology that's why. Hothead games, hothead games. Yeah, so we talked about the portfolio. That's a new piece of the portfolio right there. And you and I have talked about open-source in general is tough to make money by definition, right? IBM, we kept saying they're going to put all this money in Spark. How are they going to make it work out for revenue? And they figured it out. They're embedding it. How many products now? 37? Beth said, I think 15 already. This is a couple months worth of work. So that's pretty fast. When you consider that some of these products are stable, mature products that have been in the market for years, they see the value. They're investing in code contribution which is a historical thing IBM's done in the open-source community with Java, with Hadoop. They've trained hundreds of thousands of people already. You know, this is mind boggling. This is a major commitment in a very hot space of the market. And again, I think they have a very differentiated position to Donald's point. Nobody else is talking about embedded Spark. But IBM's going to have Spark in their products. You're not going to have to go acquire it and figure out how to connect it to the other stuff you've got. It's just going to be built in. But you have to acquire the product which is revenue. And that's how they're using it, using open-source literally to drive real revenue. And they're charging for that integration and that capability to deliver it, essentially as a service. It will vary by product. In some cases, you're just going to get the next upgrade of the product and it's going to be there. And they're not going to monetize it. But you've already paid for the product. Right. Right, so they, now you've talked about a couple of tentpole conclusions coming out of Strata Hadoop world, one of which was complexity. Yes. And simplifying that complexity. Yes. Now we talked about Spark. I think you guys just did a big survey and I think Hadoop complexity came up. Does Spark sort of solve that or is it just less complex? No, in fact, Spark is another new set of problems to consider. It increases the complexity. In fact, if you've walked around the floor at Strata, we didn't get to this very much in our conversation there. There were seven or eight companies on the floor I'd never seen before whose only reason for being there was we have an environment for you to develop Spark apps and get the thing up and running, make it work. And it'll be simple. Well, they wouldn't be there with a business proposition and getting funding if it wasn't hard and complex. But then if you're an IBM customer and it's not just them, other mega vendors too, if your vendor says I got this covered in case of IBM, they're making it really easy. They're embedding it. You may not even have to touch it at all. But if you're IBM or you're Microsoft and you sell people DevTools and you make some of those drop downs where you pick data type or processing that you're going to do or what library you need, some of those are Spark, that's relatively simple. At least it's something you already understand. So that marketplace may be relatively short-lived if these guys figure out how to just have their regular tools work with this environment. It's hard to imagine they won't do that. They haven't said so yet. Well, and of course, database, again at the center of that conversation with Cloudera and Kudu. So it all comes back to the database. Is everybody sort of going hard after it? That's what we think. Well, it all comes back to information bottom line. Yeah, and so I want to ask you about Oracle. Oracle Open World's going on this week. It seems like Oracle is sort of milking the maintenance from its database business and pouring it into applications, pouring it into Fusion, because it sees, hey, that's where the margins are going to be. What's your take on what's going on there? What's my take on what's going on at Oracle? Certainly they are pushing applications and moving to Fusion, which is in the cloud. They recognize that the future of any of this is going to be cloud. We have our own idea about that, by the way. People want open source pricing because there's no upfront license fee. Oracle can't stop charging upfront license fees, but when they go to the cloud, there's no upfront license fee. The yearly price is a little bit more, but there's no upfront license fee. It's their answer to open source pricing. And so they want to move the model more and more to the cloud. It's much easier to move at SaaS applications than it is to do DB Pass, or it is to do even infrastructure as a service. I mean, Oracle wants to compete with Amazon and Google. Whoever won a race to the bottom. Simple question. It's a lost leader for Oracle. Exactly, and I think their numbers are starting to show it. They're starting to feel that pressure from the lower margins, from the lower revenues that they're getting from the cloud, hence we saw their last financial statement was not great. It wasn't bad, but it wasn't great. And I do think that can they compete in the software side of it, DB Pass, putting Oracle as a real managed past service, which they don't have yet, or the SaaS applications like Fusion and stuff, yes they can. They definitely can compete there, but they need to start selling even more than they currently are to drive those revenues up and therefore drive up the margins. Yeah, so we've talked about this. You're seeing a slow motion collapse in infrastructure software pricing. Oracle trying to move up the stack, but ultimately don't they make more money in the cloud? If they're jacking up their annual price in over a four year period, won't they make more? It depends. Does the calculus work? Well, it can if it's as sticky there as it is when it's on-premises, and that's going to be one of the challenges. Can they keep the base paying every quarter, every year? That will depend on how they perform in ways where today it doesn't. If you're invested for a large on-premises implementation and you have to think about moving out of it, there are more costs than the moving of the data and the rewriting of the applications. There's the configuration and installation and operation of an entire new environment to which you're going to move. Some of that goes in the way in the cloud. And I think one of the more intriguing dimensions of what's happening right now is that we've always talked in the tech industry about the innovators dilemma as being about engineering. This is the innovators dilemma about pricing models. Go to market stories. Where is my monetization? Where do I monetize what I do? We're going to have to rethink some of these things. How can I sustain the stickiness I've typically had in the DBMS market when people are moving things to the cloud, when somebody like Amazon can come along and literally, as Donald pointed out, say, if you want to migrate this, point to it and click. Don't go buy a cluster, set it up, configure it, plan a migration, move this stuff over and then test it. Just click. And then say, if it doesn't work out, you haven't spent any money. You've hardly spent any time. But if it does work out, then you can sign up for the subscription. It's a proof of concept with really zero labor. It's a frictionless opportunity for migration and it's likely to partition the market from the bottom up in a very dramatic way where people are over provisioned today paying for stuff that's much more sophisticated than they need it to be. And now there's a simpler, less expensive alternative. The big move by Amazon, obviously, putting a lot of emphasis in the database. What happens in the open source world when all the funding dries up? That's what I want to know. Who achieves escape velocity? Right, okay. And who's going to pick up the remnants? So there's some good deals that are going to be out there. I mean, right now you've got companies that are spending two to make one. Yeah, we'll make it up in volume. Yeah. So what's the customer angle? So you guys, the big theme of Symposia this year was digitization of businesses. What are you hearing from CIOs, from your customers in the Gartner community? They care about outcomes. That's what we're hearing. And we're talking about the algorithm economy because it focuses on how do I get the outcome I want. It's remarkable how many conversations we have that used to be about platforms that are not about platforms. And a lot of that has to do with the fact that 50% of the spend on information technology is no longer being done by IT people. It's being done by lines of business. We had an interesting breakfast conversation with a very large financial institution where these folks had a five-bullet checkpoint list for how to qualify a new DBMS technology. And they had a bake-off among several large players. And one of them was last on all five of the bullets, but that was the one the CEO had already decided he was going to buy. Guess what they bought? Six bullets. Yeah, the CEO wants it. Other. That's what I told him. You left out bullet number six. The CEO wants it. I'll tell you, the big message though, I think, is that CIOs have finally realized that waiting for the economy to get worse and then cutting costs is not the way to do it. So they're all starting to pay attention much more than in the past to TCO, to how to control costs, which is a better way of saying it. And we're seeing it in everything. They're asking far more questions about everything's about the value to my business. You hear it when SAP started this with HANA. HANA is expensive, but we can show the value. If you can't show the value, people aren't going to buy. And they're now starting to ask, what's the value of cognitive? What's the value of Oracle Cloud? What's the value of all these things the vendors are asking me to do? And if you can't show me serious value and return on investment in a shorter period of time, I'm not interested anymore. So you guys, we talked about this, Merv. You have this concept of hybrid IT. Do you see the structure of the IT business, the technology business, I don't even want to call it the IT business anymore. Do you see the structure of the technology business sort of lining up along those hybrid lines? It has to. That same conversation I talked about a minute ago involved a project that cost $1.5 billion for this bank. And I asked him point blank, how much of that cost was the software and the physical hardware platform you ran it on? Out of that $1.5 billion, $50 million. Yeah. What was all the other money spent on? A handful. So let's start thinking about where we're making the investment and not just on what check we write at the beginning of the cycle. It's a completely different conversation. And so when you talk to an organization that has a defined outcome and you can convince them that to Donald's point, I can deliver that value you're asking for. And oh, by the way, you don't have to have real estate power, computers, people to operate the data center, run the backups, because I can do that for you. It's a very different conversation and it speaks to why in the last two or three years Amazon has really started to make dramatic progress and why it's very obvious from everything Larry's saying right now that Oracle wants a piece of that business because they see it's where it's going. It's where it's coming from. He always talks about non-differentiated, heavy lifting. And so within your customer base, you must see people realizing, coming to that realization, it's a lot of business outcomes, sort of shifting the focus from, I want to make sure I understand what you said, Donald, from pure cost to really where's the value. Right, so that's a skill set issue. That's a, what do you tell your storage admin, your network admin, your virtualization admin? I mean, wither those people, where do they go? They do something more valuable. You know, there isn't a DBA that we talk to every day that would not rather be spending time thinking about design for a new application opportunity than did last night's backup run and how do I get the snapshot up and applied to the current version of the database? Nobody wants to have to do that stuff. Is, aren't computers supposed to do this stuff for us? Why do I have to do it? So it's a focus on what is IT delivering to the organization? A machine that runs smoothly or a better return on investment, a better risk portfolio, a better customer experience, that's what I want them to do for me. I think we underestimate sometimes the understanding of this from the technical people. They're worried about their jobs, but I think there's really, I spoke to a whole group yesterday, 130 people from Z, okay? These are your real techies and have been for 40 years, okay? Most of them in the room had been there for 40 years. Cobalt guys. Okay, yeah. Well, MBS. Back in the day. Yeah. They were applauding some of the things I was talking about, how their job is going to change, taking them away maybe from some of the deep technical stuff to being able to use a Z, for instance, for big data, using a Z for analytics, for predictive analytics, for prescriptive analytics, for things like that, that people think the Z is there to do day-to-day everything and forget the new stuff's got to go in Linux. It's got to go in Unix. It's got to go in Windows. Not true. And they're starting to realize that there's value to them doing other things, not just doing APARS on a system Z. So they're embracing it. Yeah. And how about you guys, the three of us, gray hairs. How is your business being affected by these changes? How are you guys responding to that shift from sort of pure IT to much more diversity? I mean, you guys are famous for the quote, CMOs will spend more than CIOs. If you're in the IT research business, change is what we need for breakfast. Bring it on, baby, because everybody wants us to bring our perspective, our view across the market. The most frequent question we get usually is, what are people like me doing? No matter what the technology is, how is this happening out there? We're looking at this. Gartner, you're talking to everybody. Are they doing it? How soon do we need to move so we don't get left behind? Is it time? That's really what the hype cycle is about. We've talked about the hype cycle on the show before. And the market clock. And the market clock. It's all about when do you get in? When is it time to get in so that you're not too late? Most of the organizations we talk to, yes, absolutely. Most of the organizations we talk to don't want to get in too early. They understand that pioneers can usually be identified by the arrows sticking out of their backs and they don't want to go there. But they don't want to be any later than they have to be. We've seen, we've been doing value for a long time. When we get questions about no SQL today, where do I use it and why do I use it? It's not just where and it's the latest thing so it's cool and we're doing meetups and our ponytails are all interested in it and it's got the latest language go or whatever the latest one is this week. He's the oldest of the three of us to say. But they want to know what it can do. Why do they want to use it? Where's the value of doing this for our organization? And so that's what's really good about what we do. We've got all these new changes coming in. It's our job to figure out where, when and how and explain that to our customers. And then the differences in the technologies or the products or whatever else is coming along. So we've been doing value for a long time. Now that's the thing they're interested in the most. Why should I use something besides my standard? I'd have DB2 as a standard or Oracle as a standard. Why do I want to go to Cassandra data stacks or why do I want to go to Mongo or Couchbase or whatever it may be? Here's the value there. That's the answer. And by the way, sometimes that value is as simple as to your point. The guys you're hiring right now, they're not going to touch those other things unless you force them to. You brought up IMS. Can you imagine taking a kid out of college and saying we're going to teach you IMS? I don't think so. We were asking the guy from the South Africa Bank, where do you find cobalt programs? And they said, well, we have an intensive training program and actually in Africa we're lucky enough to have some underprivileged people who want to learn. You can put the terminals in the nursing homes and see if you can get some seniors. There's always that. All right gentlemen, we have to leave it there really. Always a great segment with the Gartner analysts. It's always some of the most watched. I really appreciate you guys coming on. It's great to see you. Thank you very much. All right, keep right there. We'll be back with our next guest. Check out ibmgo.com to see the digital social experience at IBM Insight right back. This is theCUBE.