 Live from Las Vegas, Nevada, it's theCUBE. Covering IBM World of Watson 2016. Brought to you by IBM. Now, here are your hosts, John Furrier and Dave Vellante. Okay, welcome back everyone. We are here live at the Mandalay Bay Convention Center for the IBM's World of Watson. This is SiliconANGLE's CUBE. I'm John Furrier, my co-host Dave Vellante. Our next guest is Mark Osholar, who's the general manager and business analyst at IBM. Welcome back to theCUBE. Good to see you. Nice to see you guys again as well. Let's get a little tutorial. What part of the analytics group are you from before we get started? Just so people know which section you're in. This analytics is pretty big now. Yeah, so I have business analytics. Business analytics has three offerings within it. Cognos analytics, Watson analytics, and Watson analytics for social media. All right, so what's the update? Obviously the buzz is off the charts. Harriet Greene was on. Her portion of IoT went supernova, buzz-wise on the social impressions. A lot of conversations happening around Watson. What's going on in your world? Sure, so I mean, on one side of it, you have Cognos analytics, which a year ago we announced it, we released it, and it's a completely reimagined Cognos BI. So we've had users on our Cognos BI platform, obviously for many, many years. We reimagined it now for business users. So now business users can get in there, use the authoring capabilities, the reporting capabilities, the distribution capabilities, and also for the first time, self-service analytics available within Cognos analytics. So that was the big announcement last year, but the other big change with Cognos analytics is we went away from these big, monolithic annual releases, and we've moved it to rapid delivery. So every eight to 10 weeks, we do a Cognos analytics release. So you guys are agile now on this, going hardcore? It's all agile. All right, so talk about what's- It's all right, so you deliver that as a SaaS product or on-prem? It's actually, so Cognos is a really interesting piece of IP. Cognos, you don't load data into it unless you were doing self-serve. If you wanted to bring a spreadsheet and load it in, you can, but when you're going to enterprise data, you connect to the data. So in the cloud, you're not actually loading data into Cognos analytics. You're attaching to the data in place, running analytics as a service. So whether you run it on-premise or the cloud, it's one code base, same offering, but you're never loading the data in anyway. It's actually perfectly fit for the cloud. Exactly. And operationally, when you said eight to 10 weeks here- Every eight to 10 weeks, there's a new release. How are people dealing with that in terms of absorbing the new- It's a big change because the other thing we used to carry with Cognos was these kind of ugly upgrade cycles. A lot of heavy lifting to move from version to version. So getting from Cognos BI to Cognos analytics is our most straightforward upgrade to date, but now going forward, once you're on Cognos analytics, we have in-place upgrades. Literally, you can go from release to release just by doing a minor install. The benefit of that, the impact of that is that customers managing patches and upgrades saves them a boatload of time, right? That's the real value. And it allows us to keep them current. And I got one more question on Cognos. The sort of citizen analyst, if you will, that's an abstraction layer to the power below or is it a lightweight version? So it's a combination, like to me, we've defined ease of use for years now as UI, UX, right? Move buttons around, change colors, now it's easier, we instrument it, we try to see where friction points are, and I know I'm sure throughout the conference you've heard us define cognitive and everything around cognitive. For my users, I define cognitive as the next paradigm in ease of use. So I actually think the bar in ease of use is going to be defined by applications that understand your intent, can anticipate what it is you're going to do next, can make recommendations, can guide you, have natural language. So we've been putting all these cognitive attributes in both Cognos analytics and Watson analytics. It's a journey, but there's already many things available now. You've got to start somewhere, you've got to start somewhere, and the benefits of the whole bot culture of having kind of this, I call digital assistant on steroids, is about ease of use. And ease of use isn't just UI, UX anymore. It's work streams, so it's relevance too, right? Where you're ease of use and relevant. Yeah, if you have two mobile apps, and they both do the same thing, and one's easy to use and one's hard to use, and one anticipates what you want to look at and one doesn't, which one you're going to gravitate towards. It's a no brainer. So I want to get your thoughts. So we talked about Pcigano yesterday about, you know, Watson as a brand name, this is World of Watson, implies newness, right? And there's a lot of IBM stuff from 10 years, from IOD, right, first conference. So there's a spectrum in Cognos kind of fits in that, and now it's been cloudified, makes it really impressive. But you also have the social stuff. So how does the new stuff, social analytics is a new thing that's really going well. People see the conversations, they see the engagement, they see the ability to use unstructured data. And how does that come together with the pre-existing, evolving Cognos as a little bit? So yeah, we've had social media for a few years, but we really had it for data scientists. A data scientist could go in and build a social media model, analyze things. But if you're talking to a retailer and the lifespan of a new dress has eight to 10 weeks, the chance of being able to build a model that's actually going to impact your business fast enough and then make changes, tune it, you're not going to have the time now. You sold out, you missed the mark, you missed the window. Well, and the other thing about that is, you're actually doing that after the fact, but there's a lot of information out there that actually will lead you towards the appropriate trend. We were sitting down with the Twitter folks and they were sharing some really fascinating information with us. They were talking about women's hair colors and nail polishes, et cetera, and the various color tones that come out. And they were talking to us about the names that come out. So I think the next, if I got it right, I think the next wave is going to be a set of animal names that come out around describing these colors. So they said, so now you kind of wonder where do these things come from? And they said, South Korea, two years before it hits North America. So sometimes when you're doing this demand planning, this demand forecasting, instead of kind of doing a little bit of guessing and let's launch something and let's see if people like it or they don't like it, you can actually look at the history of the trend, the profile, food's another great example. Food trends originate in San Francisco. So when you look at quinoa, kale, you name it, beets, right? All these hot soups. So you see origination points with the data. So the point is you can get the data if you know what you're looking for. Beets to make and beets not to make. We did a cool one at the holidays around top gift ideas. And you had like hoverboards, drones, Star Wars, you had all these top gift ideas. And hoverboards was a really fascinating one because if you went into like New York, you saw the electronic shops, they were stocked with these things. Everyone was talking about them. Now you look around and you don't see the same stockpiles. You also don't see a whole bunch of people on it. Why? When you look at what people were actually talking about, just as many people were saying things about I like it, that's something about danger. Right? So when you start looking at things beyond just the impressions and the noise and you say what are people actually talking about? What's their behavior? What's their tone? Do they have an intent to purchase? It's much more meaningful. Much more insightful. Well you can give the collective intelligence of the crowd, if you will, to synthesize demand data. Yes. We had a startup on, yeah, doing the financial stuff. Alpha Motors. Alpha Motors is really phenomenal. So these are new business models. How's the customer attraction on that? Because now some customers will be like, well I'm not that geeky. I'm running retail or I'm a small, medium-sized enterprise. I just want it to work. Yeah. That's kind of where the Watson, I think, resonates in the messaging I see now. How does that put into practice? What's some fun? So this is a lead offering for us. Why? Because a lot of companies are debating which workloads do they move to the cloud. Watson Analytics for social media brings all the data. You don't load any data into it. It hooks into the boards, the reviews, the blogs, the news feeds, Twitter, Facebook public, all the information that's out there, and it pulls it in. Then it allows the business user to say, all right, what is it I want to analyze? My topic might be Hilton, Hilton Hotels. We'll show the business user a stream of what everyone's talking about. Then they can start to say, oh, I don't mean Paris Hilton. I don't mean Perez Hilton. I don't mean this convention of the Hilton Hotel. You can start to include and exclude terms. Then you can put in, okay, what themes do I want to analyze? And almost every user out there wants to know what are people saying about my brand, right? So that's one of those spots where a lot of people start is simply brand perception. And not only what are they saying about my brand, what are they saying about someone else's brand, right? And are things that are going on with someone else's brand about to affect my brand. So when I license Watson Analytics for social media, you're saying I get those data feeds. You get those data feeds. You also get Watson Analytics with it. So we will allow you to create those models. We give you a whole bunch of kind of quote unquote, canned reports, canned analytics, and then we create the data set. And we allow you to use Watson Analytics with the data set to do any additional analytics you want to do. And visualization's a part of that, right? Visualizations, natural language, natural language exploration, all of that. You made some enhancements to the visualization. I don't know, I can't remember when, but I remember I saw a demo at one of these conferences. Yes, and we always are. Which is, okay. So yeah, one of the other. That's a key part of it. I mean, the Viz is increasingly important for, especially for the citizen analyst, right? And we have our own visualization engine. So you've probably heard a lot about D3 as one of these open source offerings. So we have our own called Rave. And we've just released Rave 2. And what's interesting about Rave 2 is Rave 2 embraces D3. So if you have, if there's a great cool visualization out there and you want to use it, you can go and grab that visualization and use it within Rave 2. And then Rave 2 enhances D3. It does the accessibility. It does the legend wrapping. Like some of the things that D3 doesn't do, it makes it enterprise ready effectively. But it also gives you the capability to then author your own visualizations. So whether the user wants to use Rave 2 or D3, we're now agnostic in terms of how they want to get there. Okay, so you're saying it's easier to use, getting to simple. I remember when we first started playing around with it a couple of years ago, it was like, ooh, it's pretty complicated. Yes, it's very easy to use. And I would say the other thing we've done is we've adopted this paradigm between Watson Analytics and Cognos Analytics of kind of looking at the commodity items and doing build once, use twice. So Watson Analytics and Cognos Analytics actually now share the same visualization library. They both use Rave 2. If the Watson Analytics team goes and builds a cool network diagram, they check it in, it's available to the Cognos team. The Cognos team, our big announcement here, is geospatial mapping that's coming in Q4. The Cognos team is adding that to the visualization library. We partnered with Mapbox and Pitney Bose for that. The Watson Analytics will pull it out of the visualization library. So now instead of two teams building everything in duplicate, they build it once, they check it in. The connection to the data is a really big deal. That turned out to, that's a great architectural feature. Yes. It's got flexibility and you don't have to move there. I have economies of scale. I've got an army on this stuff. Yes. How about the design side of it? What are you guys doing when you're investing? Because one of the things we're hearing, and again, more and more, every year, it's the same drumbeat, but louder and louder. Total user experience into the design process of the product. Digital, everything, yes. So give us an update on that. What's it like inside the factory there on this? Yeah, Watson Analytics and Watson Analytics for social media have gone 100% digital so people can get on there. They can do trials, gives you, in the case of Watson Analytics, 30 days to try our best edition of it. Even at the end of that edition, we'll downgrade you into a free MIM edition. You can continue to use it, decide later on that you want some of the professional features to come back in. But what we've also done specifically around design is not only have we pushed everything digital, but we've instrumented everything so heavily behind the scenes. So we actually see how users use it. You start to get into multi-variant testing. So you're getting some product feedback on the engagement of the app. Data-driven product feedback. Customer can't tell me. I can just see it. I see there's a friction point. I see there's a hot spot. Is it just close to the user? Are you cool with that? Do they understand that you're monitoring their clickstream? They do, yep, exactly. And they know that they're getting improvements based on what we're monitoring. We're not looking at the data. Do they opt in for that? Or do they have to check a box? No, no, it's, if you want to use the offering, this is what we do. All right, so is it available in the marketplace at IBM.com? It is. That's how it is. Okay, so IBMgo.com is where the replays are for this. Sign in with your IBM ID. You can do the offers in the marketplace. And is there a free trial in the marketplace? There is a free trial for both of those two offerings. There's also a trial for Cognos Analytics up there as well. So people can go and try a trial instance of that, see what they think. But Cognos is a little bit more on the traditional side. It's usually looking for groups of users, not necessarily individuals, where Watson Analytics, Watson Analytics, for social media, often we start with one single user who decides to go and adopt it and then it spreads. And it's a monthly subscription? Monthly subscription, start it, stop it, whatever you want to do. So if you just have a workload in a project, you want to come in for one month and use it? You can. And it's sort of on or off, or is it sort of small, medium, large, feature-wise? Yeah, it's basically there's a free edition, there's a plus edition, there's a professional edition, defined by data volumes, defined by the data connectors that we have, but not defined by the analytical capabilities. But all the same analytical capabilities across all three. So it's not like, oh, to get the profile and patterning I have to upgrade. No, we give you all the capability. A lot of data that you're analyzing. Yeah, yeah. All right, so what's next? What's on the roadmap? What new features are coming out that you guys are excited about? Can you share in the purview, making it easier and relevant? What's coming out? Sure, so like I said earlier, I mean the big one and it got massive applause here is the Geospatial. So we'll start with that. Weather data? Weather data is coming as well. So weather data clients are using already. They have it now. One of the things we did with the weather company when we first acquired them is instead of just giving customers longitude and latitude data, we created these smart data packages that can be easily joined in. So right now there's a process where if people are on Cognos or Watson Analytics and they want to grab weather data, yeah, they can go and grab it. But we're going to make it an in offering, offering as well, right? So we're actually going to have it available through the offering that if you want to play with this data, see if it's actually insightful to what you're analyzing. That's how we're going to bring it more closely, more closely coupled to the offer. Mark, thanks for sharing your insight here on theCUBE. Putting all the data out there, breaking down the analytics. I'll give you the last word, share for the folks watching. What's the one thing that they should know about that they might not know about around the business analytics business that you're managing? Yeah, I would say it's all about smarter insights. And for me, when we're talking about smarter insights, it's about removing the bias. So when you look at the traditional products that are out there, what do you have to do? You have to bring a question and you have to bring an assumption to test on what you think the answer is and you see if the data backs it up. But really you're stuck in correlations, right? And correlations aren't always causal. When you use something like a Watson analytics and what we're doing with Cognos analytics, you get this unbiased environment where it's statistically patterning the data, we're guiding you, we're showing you interesting things that you wouldn't pick up on your own. So it's services to you first. It's services, insights first, you react to that versus probing a question and leading it to its own. But it's a new mental model on how you trust the data. Yeah, like in Watson Analytics, we'll watch the users too. In Watson Analytics, we just say, just ask your question, don't bring any biases, just ask your question and see what we find. Or if you don't even know what question you want to answer, ask because it's a new data set, we'll actually recommend what the most interesting features are of that data set and you just start exploring through. Mark Ultscholler here inside theCUBE, general manager of the Business Analytics Group at IBM. Thanks for joining us. Thank you guys, thanks for having me. There's live here, the Mandalay Bay for theCUBE. I'm John Furrier with Dave Vellante. We'll have more live coverage after the short break.