 Okay, we're back. This is Dave Vellante. I'm from wikibond.org and we're here at the IOD conference, IBM's big event. It's historically an information management conference, but it's evolved and really morphed into an analytics and big data conference, which is, from my perspective, more exciting than information management, but a lot of those information management technologies, things like classification and search are used and utilized and extensively in IBM's new offering. So we're seeing the industry's broadest portfolio in big data. There's no question about that. And IBM really forcing us to think big at this event. That's the mantra, the tagline of the event, is think big, a playoff of Thomas Wasson's think, the most famous line, the most famous mantra or tagline in IT industry history. Very good show, probably about 12 to 13,000 attendees here. A lot of practitioners, a lot of business getting done. And good show by IBM, very impressive. And we're here, this is theCUBE. This is where we bring you all the best guests that we can find, we extract the signal from the noise, and I'm here with my co-host. I'm Jeff Kelly, also from Wikibon, covering all things big data, and we're here with our next guest, another great guest, Michelle Mollat. Welcome, Vice President of Business Analytics and IBM's software group. Great to be here. Welcome to theCUBE. Thank you. Yeah, thanks for coming up. You know, being a business analytics person, I'm very comfortable with theCUBE. Yeah, that's right, indeed. Of course, I think you guys invented the... We did, the PowerCube. Looking at things in multiple dimensions. So I have to ask you, so business analytics, big data, are they one or the same? Are they different? What is that? They're the yin and the yang. I mean, really, big data is a foundation for business analytics today. A lot of the new advances that organizations are making in terms of their business outcomes are being driven from a combination of big data with advanced analytics. And you can't have one without the other, really, to truly drive business outcomes. So for us, it's about getting at new data sources into the analytic environment, like the social media, and getting things in it more volume to be able to process them in real time so that people in the operational areas can actually take advantage of it. So what do you mean by real time? How do you define that? So real time is in the moment when people are trying to make decisions, but they can't make decisions in the way that you would make a decision. So you would make a decision by taking the time to analyze whatever information as a knowledge worker that you have available to you. There's a lot of people who work in real time environments where they are being pressed to make a decision right now. They don't have the luxury of doing analysis the way knowledge workers do analysis. They have to be essentially given answers, not insight. And so that's really what we mean by real time is delivering to people the right next action in the moment. Okay, you're the cop, you're on the beat and the crimes are happening. The predictive models are working in real time. They're seeing patterns in crime. They're directing you to go to that precinct because crime is popping up and they're putting more police officers right in the moment versus three days later analyze and say, yeah, there's a lot of crime that happened in the years. That's what we mean by real time. Talk about what's changed. I mean, how did we get here? The world of decision support and data warehousing and the promise of 360 degree views and predictive analytics and it was always a struggle. It was like this patchwork of infrastructure and applications and all of a sudden the light shines and big data comes into the floor and these two worlds are mashing up. Can you talk about that a little bit from your perspective as to what's changed? What's changed is the technology can handle it now. I mean, that's fundamentally what's changed. Big data's not new. Telco's have been collecting call data record, detailed records forever. And yeah, there's more now than ever before. So what? There's still- Bigger data. Bigger data, right? But when you're getting to petabytes to zettabytes and it's all big, right? So I think what's changed is now the ability to make sense of it, to do something with it versus just taking all that big data and archiving it, which is what people were doing because they couldn't make sense of it. There was just too much there. And now they can actually make sense. The old signal through the noise analogy I think is exactly what's happening. I'm sometimes hard on the sort of traditional data warehousing BI business, even though it's created tremendous value. And I remember the days of the original days of the beer next to the diaper, right? The Walmart example. But at the same time, I said it had been hard on it because I feel like it failed to live up to the original promises. What do you think is different this time around? Well, first I don't believe it didn't live up to the original promises. So let me stress that first, because I think there was a massive lack of understanding of how to drive classic BI. So people built it, they built it from the ground up. I have this much information. I'm going to build reports that reflect that information and then I'm going to present it to my users versus starting with the business problem, thinking about what information can help me think differently about my business and what information would really impact my business and then building the BI down and predictive. Because to me, a BI is a spectrum that's more than reporting, right? It covers predictive and analysis, classic reporting and query. And now I think what Big Data does to it is it also now, it's brought the conversation to the boardroom, it's brought the conversations to the line of business, and it's making people think that way. And that is a huge change. Because before, to get a marketing person to engage in a conversation about data was virtually impossible. It was too hard though, right? But now you've got Harvard Business Review writing about Big Data. You've got all the big pubs writing about Big Data. You guys talking about Big Data. We started. Right, so I think that is putting people into a mind that says, Big Data should be my companion in my journey and analytics is the way to make sense of that data. So there are two P's and a pod, if you will. Right, okay, so fair enough. Of course, from a reporting standpoint, huge success, right? It got us out of the Enron disaster. And then, but the predictive analytics side is getting a new face now, isn't it? Sure, because the more data that you can bring into a predictive model, the more accurate that prediction becomes. Jeff Jonas's puzzle pieces. Exactly, right? So you bring more types of data. So it's not just more data, because that was probably not the best way to say it. It's more and different types. Then you become much more informed. The customer example's the best, right? Because just understanding transactional records if there's volume, okay, it tells you a little bit. Add on all their comments. Add on the social sentiment. Add on all the other things that are going on in the blogs. All of a sudden, now you have a much more complete view of that customer, and that's really changing the business. What are some of the more exciting things that your customers are doing with the data and with the analytics? Yeah, I think the most exciting examples are the center stone example that we saw this morning in the main stage, where they're using it to improve mental health. Because they're using predictive to be able to optimize their operational environment, predicting when patient loads would be greater, optimizing every single part of their business through analytics, and being able to accept more patients. So they're actually fundamentally making a societal improvement. So I always like the ones where analytics is helping people to do better for people. But, and then we have fraud reduction. It's one of my favorite examples because people see the results instantly. You know, when you can reduce fraud in your business, you see money piling up, and then it gets people in the line of business supporting it, and then they start thinking, well, if they've saved $200 million, like Santam debt did, well maybe I could too, in this area, or maybe I could use it to drive new revenue opportunity, it gets people thinking analytically. So fraud's an exciting, hot area. So I was talking about fraud. So what's the breakthrough there? Is it that I don't have to sample anymore? I can actually work in the whole, the entire data set? Absolutely, that's exactly it. Because sampling, I mean sampling was good, but sampling was a sample, right? And, and... Oh, I missed that one. Right, I missed, yeah, I missed that one. That was, happened to be the big one. But it's the anomalies that you want to find. Maybe we'll get it next time. Exactly, I mean that, being able to parse all the data is, that's the technology advancement that has taken us from, where we've always been with fraud, which was the sampling world to now the, there's no sample in real time, records are being processed for, for fraud all, you know, everywhere. Many different industries are using it, right? So how do you approach the, I guess the behavioral challenge of changing the behavior of your customers who maybe, you know, have built up their own processes over the years, a lot of times using IBM technology to start thinking analytically, as you put it. The idea of getting a recommendation, saying this is the next step you could take, you know, rub some people the wrong way, like a machine telling me what I should do next, it's, some people, you know, push against that. So how do you go about, not just from a technology perspective, but from a cultural behavioral perspective, get people to start understanding the power of predictive analytics and how it can help your business and it's not, it's not something to be afraid of. Well we've, we spend more time talking to line of business than we do IT today and I'm in the marketing side of the world, so this is a fundamental shift from five years ago where we would have spent, probably 30% of our time talking to line of business and the rest to IT or Lobbit, would I like to say, but now we're talking to line of business and when we talk to them, we talk to them about what their peers are doing in their industry and then you don't have to, you don't have to ask for their attention, they are all over it because everybody is suffering competitively, everybody, right? Everybody's being asked to do more with less, everybody's at the breaking point and they need innovation. They can't, they can't cut more and do more and so they're desperate in a lot of environments to find the secret sauce, if you will, to help them innovate to grow again and then when we show them that analytics is doing it for other companies, you know, then all of a sudden, and we have, you've probably seen the MIT Sloan Study that showed the difference between the analytic haves and have-nots, we use that really effectively as a door opener to get people to say, look, where are you in this spectrum? Because the chasm's getting wider, which side do you want to be on? So give us an example, pick any industry you want and talk about the conversations that you're having with the line of business versus the spinning disk conversation with IT. Well, so there's, we have conversations mostly around four areas, customer analytics and that varies from industry to industry, so customer analytics and banking, it's some call center conversations, it's about marketing outreach, so optimizing your campaign so that you're not sending you and you and me the same proposal when we really have nothing in common in terms of our financial world, necessarily, right? So segmentation, targeted marketing from a customer analytics perspective there, telco, so much a telco right now is call center, you know, that proactive approach to people to say, now you should be doing this because you understand this customer so well, so treating the customer as an individual. And then when we take it to insurance, it's really more in the fraud area that we talk to insurance right now because they spend so much time putting investigators on claims that they really shouldn't be investigating. And conversely, they miss ones, they should, right? So being able to target which claims should be fast tracked and getting those through right away and paying people right away, their customer side goes through the roof. And then being able to reduce the amount of fraudulent claims that go through the system also saves them money. So it's a double benefit for them. Proof customer satisfaction, decreased fraud. Huge business impact. The other, there's two other areas though that we're really big into right now. One is finance. And we talked to finance about basically becoming not just the financial watchdog, if you will, but actually becoming what we call a strategic co-pilot for performance. So today, finance is mired down in report. They have to generate all this information for everybody. Instead of thinking about how do I help this organization drive the business. And now, by helping them to automate some of the processes that they spent a lot of time on in spreadsheets today, they can save that time, start thinking more about the business strategically, looking at predictive analytics, do things like predictive forecasting. So now they can say, based on where our metrics are today, guys, in a year from now, we'll be here. It's probably not a good place or this is a great place, whatever the case may be. But that helps them to be partners in the business versus just the guys always laying out the finances. And then the other is risk. And risk analytics is not just for banking and insurance anymore. There's a lot of conversation on the IT risk side, people preventing a loss of identifiable personal information, a big one. So risk has become a more generic conversation or not generic, but more generalized conversation by industry as well. So those are the four things we talked a line of business about. Can you talk about the data sources and how that's changing? Yeah. Well, the data source, we have a lot more people looking at unstructured content than ever before. And that's a given. In terms of social, most of our customers want to do something with social. Very few are actually actively engaged in real meaningful business outcomes, business outcome-driven social projects. So we're at the early stage in that area in terms of really driving substantial outcomes, but everybody wants to do something. Yeah, because everybody's out there, you talk about hearing about sentiment, everybody's listening, it's like, okay, green. But so what? Yeah, so what? What do I do with that? But so, the customers I've talked to want to get down to, okay, well, who's ready to buy? Right, right. And how do I engage with those folks? I remember the stories of when TV came out, all the radio executives said who want to watch a bunch of guys talking on the radio on TV? Doesn't make any sense. And I feel like there's the same thing here with social. It's like, well, how do I email blast? You know, well, you don't. You got to have new processes. And it seems like customers really haven't figured that out yet they will, but they're geared toward the past and really not geared toward this new social realm. You see that? Well, I think there's two things. One is that they have to start their initiatives with a view to the actions they want to take. So just saying I want to listen is meaningless. If you listen for what purpose? I'm listening to improve my product management. I'm listening to improve my marketing. I'm listening for some other purpose, but you have to listen for a purpose. And that's where we see the initiatives that succeed versus the ones that fail. When they tied into, for example, their operational system. So we've got customers that are looking at social sentiment. They're also looking at inventory levels, right? So they've got the two connected. So then it becomes more of an operational activity, not just a nice, oh, that's really interesting. Customers are not saying positive things. But now they can say, well, look at this response over here. We need to be able to change our forecast because these products are hot, right? And being able to track it back closely. And then on the social sentiment, I think the other thing that's really required is that customers really need to focus on social sentiment. Sorry, I just lost my train of thought completely. Well, it happens to me all the time. Well, there's something you mentioned. So we kind of heard from, we had Paul, as a copalist on earlier today, he mentioned something similar. You have to go into these engagements with a business problem. Right, a business problem. He said the science experiment approach, not the way to go. But what's interesting about that is what we're hearing in the Hadoop world, big data world is some of the data scientist types are saying, that's exactly what I want to do. I want to experiment. I want to just play in this sandbox. So how do you balance those two things? How do you promote kind of that kind of exploration? But at the same time, if that's all you're doing, you could quickly lead to the backing of executive leaving and the project going down in drains quickly. So how do you balance kind of the exploration, trying to do new things with, hey, we have a business problem, let's solve that quick win and then move forward? Well, I think every organization should take a multifaceted approach to that. Do pure data research just for the sake of seeing what's out there. And I had a conversation with a client where she was really struggling with naming her project and I said, look, if you don't have a defined business problem right now, then just run the data through some predictive models. Let's just look, let's do some data mining. Let's just look at the data to see what patterns emerge because from those patterns, you might actually define a business needed or a business opportunity, right? And that, so you want to do that, but then you want to limit how much you spend on that, but you need to do that too. Because I think people are going to find gold in them that are heels, right? And then the ones that are defined make them really defined, not just listening, make them have business outcomes, set metrics, set goals for them, just like any other project, right? And I think too many people are not doing that, they're treating it as if it's some kind of, you know, unaccountable thing. They're just listening and they put wordles up and you know, and it's all very interesting, but if it's not driving the business forward in some way, then why are we investing in it? Michelle, what's your bumper sticker on YIBM? What do you tell clients? There's just nobody that has a foundation for big data as well as an analytic platform, as well as the know-how to really guide clients on changing culture because changing culture is as important as having the foundation and having the analytics. So I think it's the fact that we have a holistic approach that you can bring us your biggest problem, your biggest challenge we can address it, you can bring us something that you've already defined and we can give you the piece parts to solve it, right, we can do the full spectrum. Can you do that affordably for mid-sized businesses and smaller organizations as well as the Fortune 500? Absolutely. On the analytics side, we have solutions today that are priced for the mid-market, they're optimized for the mid-market, meaning you don't have all the customization abilities because they don't need all the customization ability. They're going to be configured for a certain amount of users, for a certain scalability and they can take those, put them in for 25 grand, get going and as they scale and grow up as an organization and they need additional capability they can add it on. Excellent. So in terms of products, obviously there's a lot of, everyone knows IBM's acquired probably dozens at this point of analytic companies. A few. I think the number was 15 billion or something over the last. A dozen? Yeah, a few. A few. So I'm curious how you go about rationalizing all those products and packaging them in such a way that they complement one another and actually are focused on a business problem or an industry. And I think, because I think employee hearing from some customers or from some members of our community is why, IBM's got this wide breadth and depth of tools and technologies but it's almost overwhelming. So I'm curious from your perspective, with all these, you must look at this portfolio and say, okay, what can we do with this? So what's your strategy? Right, so the strategy is too prong. So on the technology side, the capability side, the strategy is to acquire and integrate and from an integration point, even in the due diligence part, we look for companies that have common technology platforms. So the integration is not apples and oranges. Immediately following integration or acquisition, we attack what is our integration roadmap? So for example, IBM acquired Cognos and then acquired SPSS right afterwards. SPSS was integrated within the Cognos platform from a metadata perspective within a year of acquisition. So models could be shared within the BI environment where metadata could be brought into the predictive analytic environment. So the strategy is to integrate on the capability side and it doesn't happen overnight for sure and I know a lot of our clients would like it to happen overnight but the roadmap is to make the tools work from a UI, from a capability, from a metadata perspective seamlessly over a period of time as fast as we can basically. But the other area we're looking to integrate is from a functional area because that's where clients really don't even necessarily care about the capability. I want to solve a problem in finance. I want to solve a problem in my customer analytics area. So we've been looking at those as more holistic solutions. So we've been integrating ourselves to deliver things like next best action for Telco and that is a complete solution top to bottom, hardware software services and we will over time integrate further along those horizontals as well as the capabilities. Excellent. All right, Michelle, well listen, thanks very much for coming on theCUBE. We're out of time, we should continue but this has been a great event and really thank you for sharing your perspectives and you got a great story and check this out so this will be on demand up on YouTube for those who don't know siliconangle.com slash YouTube, check that out. Check out wikibond.org for all the research. Go to siliconangle.com for the news of the day. Keep it right here but we're right back after this short break. This is live from IBM IOD in Las Vegas. This is theCUBE. Great.