 Okay, we're back here live at IBM's Information on Demand in Las Vegas. This is where ground zero is for big data week starting today. This is siliconangle.com, siliconangle.tv's theCUBE, our flagship program. We go out to the events and extract your signal from the noise. I'm John Furrier, the founder of SiliconANGLE and I'm joined by my co-host. I'm Dave Vellante of Wikibon.org and we're here with an IBM executive, Colin Shearer, who runs the advanced analytics solutions business for IBM. IBM's got a new announcement today, some cloud-based analytic activity that we're going to talk about. Colin, welcome to theCUBE. Thank you, very good to be here. Thanks for having me along. So we're here at IOD, John, you know, just non-stop. Great to have you. You've got some things coming out around the corner in terms of product, technology. Tell us what you're working on and we'll jump into the questions. Well, this is something we've announced today. It's a new offering called IBM Analytic Answers. And as you said, it's to do with the cloud and it's advanced analytics. And what we're doing is bringing those together to try to make the power and the value of advanced analytics really for the first time, available to the types of users who wouldn't have been able to embark on this journey in the past. We're talking, for example, smaller mid-market organizations. Like what size? Well, we've seen typically in the past, if you look at the other track record of predictive analytics, there's no question and you heard this in spades today in the keynotes. You've heard the sort of value that it generates. But typically, it's the larger organizations, the big enterprises, the household names who basically invest upfront in the software they need, getting into the infrastructure, acquiring or building up the skills that they need. That puts it beyond the scale of, say, somebody who could be an insurer with a modest number of customers. Say, maybe a million policy holders or something like that. Or a small retailer, a specialist, for example, selling in an e-commerce operation. They've got the same business problems. They're just as significant to them within their own world. But it's been much harder for them to even contemplate using this type of analysis. Because of just the sheer size and complexity or what has been the core issue for advanced analytics? There's a set of barriers, really, which is that it's the, first of all, these companies don't have deep pockets. So they're not able to invest upfront. And then even if they were able to get the hands on the software, the powerful software to do this, where do the skills come from? They've either got to train people up, get new people on board to do it, and so on. And then they've got to look at how would they plummet into their IT architectures, which are often not as advanced, for example, as some of the larger organizations. Yeah, they're likely to have emailization. Exactly. Email working. Yeah. And so these are the challenges. And then going along with those, because if these weren't enough, all of these together lead to them having a long lead time before they begin to see value. They've got to do all this stuff and then begin to get the payback that we know was there with advanced analytics. So what we're doing is we're trying to remove all of those barriers. In terms of the cost and accessibility, it's a subscription-based service. In terms of the skills gap, they don't need technical skills. So we're building solutions in the cloud that are based on IBM's expertise. We've done jobs like using advanced analytics to do insurance renewals over and over again. We're encapsulating our knowledge from that in the cloud, doing the analysis all behind the scenes in the cloud and producing only the results that feed back directly to the line of business. That's why we call it analytic answers because all they see are the analytic answers that they need to do their business with. Very little IT involvement, cloud-based, the line of business uploads the data, gets back the results. Okay, so first of all, we love big data. So there's no argument from us on any kind of new products that help get answers faster and removing those barriers, just totally true. But that assumes that they know the questions, right? So the questions is the big part, that big data talks about it and the smarter kind of answers you can get back with analytics. So how does a small, medium-sized enterprise understand the questions? Is it programmed as their processes already? Unstructured data opens up a whole new can of worms around this whole area. Just to be clear about this, we are providing the answers to a set of specific questions. Okay, so they know the questions. We're not providing, yeah, exactly. We're not providing this as some general analytical capability for them. Instead, as I said, we've taken our experience, we know the places where this can really deliver a lot of value. We've selected a set of those solutions in different industries, in different business areas. We have built those in a package form, and they are the ones that they can buy. So this talks about the payback. The customers know what they need. Yeah. It's easy to do. And they get the answers they're looking for. That's, that's, I can have a bit of myself. I should write that down for more. Okay, so it's on. We've got a date. We've got it with more. Can we go to Copyright Silk and Ingle right now? Yes, absolutely. Absolutely. So this is a forthcoming product, and when we ship it, it will have four initial answers available. Three of them are to do with specific industries. So one is insurance renewals. And that addresses the problem of when a policy holder comes up to go, as a, to renew the policy, will they stay or not? And this gives them back the answer for each of them. The probability of will they stay loyal or not? And if they're at risk of being disloyal, if we've got the information to support it, what's the right sort of incentive? They would persuade them to stay loyal. For retailers, we have what we call purchase analysis and offer targeting. The purchase analysis part, all the need is something as simple as point of sale data. It finds the combinations of things. The market basket combinations tend to be brought together. It gives them information on those which could form new sorts of offers, new promotions. But it can go further than that because if they can link those purchases to a customer, say a registered webshopper or to a loyalty card holder, for example, then it will help to match individual offers to specific customers. So they're able to know exactly what to offer to whom and get much, much higher conversion rates at individual level than more general marketing campaigns. The third one takes us out of the commercial sector and is actually about student retention. And here it allows educational institutions at different levels to analyze data on their students and see which ones are performing below their predicted potential. When that is the case, understand that these may be on the route to dropping out, for example, and work out what the best intervention is in order to be able to keep them, get them back on track and keep them in institution. And the fourth one actually goes across industries. It's what we call prioritized debt collection. And any type of organization could have their accounts receivable department challenged with in today's economy, rising levels of debt and challenges in collecting it because they only have so many resources often being cut all the time. How do they know who to go after to maximize their returns? In the case of this solution, what we're providing them with are three pieces of information. For each debt or debtor, is the probability they'll pay, how much they're likely to pay and what is going to be the best treatment that we persuade them to pay. And that means that they know which ones are debts that should be written off but also which ones they can pursue and where they should take the scarce resources to ensure their strategy. You're automating the whole thing. You're automating the whole thing for business. Absolutely. I said they just get back to the answers. At the individual case level, it's actionable for this debt collected by just simply sending an SMS reminder. Okay, sounds easy. So just take me through how they would do that. So they engage with the services team. What would someone do to get started? What they do, getting started, as you said, is about an engagement initially but a very, very lightweight engagement because as I was telling you, what we're starting off with are types of solution we've done over and over again. We have, IBM has done something like 20,000 analytics engagements. From that, we have distilled for these what you might think of as assets, templates, blueprints or whatever. So we're already sitting there with the skeleton of how to do, say, an insurance renewals application. When a client signs up for analytic answers for insurance renewals, what we then do is engage with them. We understand what's different from their data, from the more general model we might have looked at, what they have, what they don't have, what additional data they have that might be valuable to it. We then get that to define a historic data sample. We take that historic data and we build the predictive models in the cloud and get it up and running very quickly. And from the point when that is up and running, they have a very simple interface where they simply are on an ongoing basis in the case of an insurer, maybe every month those policies are coming towards renewal, upload the set of data for those policies, download it, enhance it with the results. Simple as that. So let's start with that engagement. So you say you engage with the client. So what does that mean? You have some kind of kickoff meeting where you get everybody in the room that's necessary. There's a lot to be done remotely. We will very rarely, I think, be out on something. We want to keep this a light way. So I don't have to see you in my shop. No. Okay, if I don't want to see you, no offense, but I'm busy, right? Okay, small, mid-sized businesses. Okay, so you do that remotely. So how do you initiate that initial engagement? You say these are the... No, we will be doing it. We have a certain expectation of, as I said, we know how to tackle this and we know what are typically the data items that are going to be predictive for this. So we start off, we set the conversation by saying, okay, here's the sort of thing we're looking for. Do you have these? Do you have things that match that? Do you have things that are close to it? We might have to manipulate a bit to get that. But also leaving it a little bit open to say, what other data do you have? Because it's very easy for us with our technology to add in incremental data that might be valuable. And it's really just, from that discussion, comes the definition of two things. One is to say, okay, now tell us, what historical data can you give us? And we work with them to say what's an appropriate, is it last year's data of renewals or whatever it is? And to also define on an ongoing basis, what shape of data will you set up as your questions? What records will you upload to us? And that's really all we need to establish. Okay, so they send you that data, however they get it to you. And then you ingest it. We ingest it, this is the historic data, we take it away. And basically we have the, using the templates we've got, we tailor them specifically to their data. We build the predictive models and we build the associated analytical processes to apply those models, for example. And that's it, that is the kernel of what then sits as their analytic answer solution in the cloud. And then you deploy that as a cloud-based offering and how do you charge for that? Purely subscription-based. Okay, and you? Okay, so that is, we are not intending to charge separate setup costs or anything like that. Basically it's going to be subscription-only service. There's no setup cost. That's our intention. Okay. Just a question for you, a little bit changing. I think you guys all have the product value proposition is just global marketplace. You guys have done 20,000 Brazilian analytics, you're automating it. Great, good A-plus for IBM. And good for customers. Talk about the global marketplace. What's different around the US and different countries? Obviously there's privacy concerns everyone always wants to know about. You guys have also run into all that. So take us through the follow the sun through the North America, Europe, and Malaysia. I think you see variations. I mean the point you've made about privacy, you see variations about how far you can move data, for example. So in some levels you're at the border, the country border, some cases the large states. So for example the EU has laws about data itself. But going with that, and again, looking at and relating this back to what we're doing with analytic answers, one of the things that we're trying to do is to really step as far as we can away from concerns about confidentiality and privacy of data. So we need to do this sort of stuff. The one piece of information we don't need to know about you is who you are. I need to know lots of things about you, like what your insurance policy covers and how long you've had it and what claims you've had and so on. I don't need to know anything but identify you as an individual. So we're actually in one of the positions we're taking with us is we will not accept personally identifiable information. So we're immediately removing the risk of the organizations of moving data. So right up front, firewall that, you don't even want it. We don't want it, we don't want it. The other thing that I would say is a characteristic of the market or maturity, let's say, around the world takes us into the difference between the U.S. and Europe and the growth markets. Now in the U.S. and Europe, you will certainly find a maturity in analytics. You'll have large enterprises are certainly well up to speed on it and there are pools of talent around them. In the case of the growth markets, you often find them in a state of accelerated development and they are trying almost to leapfrog in many cases about developed markets. And which ones would that be? If you look at ASEAN, for example, look at the tiger economies and such like, you see ones that have a vision of moving, not just following the same generations as some of the other, as some of the more developed countries have, but actually trying to move very quickly on them. So they're looking for an accelerated path to value and the analytic answers approach, the way it applies to that is begin to get results very quickly. You don't have to reach out, try to find the trained resources, try to develop it. So they can begin to turn this on and can begin to get value much more quickly than you could do by traditional methods. So from that perspective, we think it's a very good fit for the developing markets. And so back to the sort of offering, if I may, your fees are a function of what? The volume of data that they're doing, the size of company, how do you... It's for each of these offerings, what they get for their subscription at the base level is they can process a certain number of records per month. Okay, so it's a monthly subscription and if they want to do more than that, there are different levels they can scale up to. But it is still aimed, as I mentioned, at typically smaller enterprises or departmental use. So typically, you're not going to find enormous supermarket chains using this. You will find smaller convenience chains or small retailers using this. And we set the actual numbers of what those records are based on what constitutes a modest size retailer, a modest size insurer and so on. But it's classic cloud. They can dial up or dial down their volume with you as a function of their ROI. Absolutely, but you can do it. But if you were, say, a Walmart or somebody like that, you know, it probably wouldn't make sense to dial up the complete analytic answers approach. There are other ways you could do that. You could afford to bring it in-house and take a more sophisticated approach. Because you mustn't forget, this is not just about scale. This is a first step on the journey. This will give an organization that has never done analytics before, and the immediate advances are beginning to deploy, the results of predictive analytics. And we know the sort of thing that can do for them. But it's still going to be a first step. It's always going to be possible to go further. Now, for some organizations, they want. They will stay at that level and the one that bought the analytic answer for insurance renewals may next buy the one for policy cross-sell and for customer acquisition and for claims risk scoring or things like that, for example, they stay at the same level. Other ones, seeing the power, might want to migrate into something that they can take more control of themselves. So they've now justified the investment to tool up, as it were, with the software themselves and to skill up with the people to use it. So we see different paths and trajectories, but it's an important first step to begin to gain value for them and get them on the analytics journey. Well, because I see it as a way for the smaller businesses to compete with some of the larger ones. I mean, you're talking about the retail purchase analysis. I mean, Amazon, you know, product matching, every time I come in there, they're making recommendations, whereas the smaller websites are just pleased by my stuff. So this gives them a capability that they wouldn't be able to have otherwise. Absolutely, with the limitations of scale. So for example, one of the things we're not doing with analytic counsel through our overall predictive analytics technology can certainly do this, as you've heard in today's keynote show, we can deliver in real time. We can deliver real time on massively streaming data, for example. We're not doing that with analytic answers. It's a batch process. You upload a set of data, you get back a set of results. So the contrast would be that Amazon, while you're on the website, is firing offers at you, right, left, and center. In the case of analytic answers, it could be a small e-commerce operation that every visitor who's checked out in that last week, they upload the records and it generates the recommendations for what to offer them, which goes in a follow-up email, for example. You're essentially instrumenting the business processes for them, with this. It's not so much figuring out a Watson solution. You guys are coming in saying, hey, small, medium-sized enterprises have known business challenges. You've seen 20,000 zillion cases. So, okay, let's automate it, instrument it, and turn key. And I think a very important, I think that's a very good way of putting it. I'll add something important to it. We talk about this being, solve your business problems one answer at a time. And that's not just meant to be a sort of slick and trite way of phrasing it. The beauty of these solutions. Even though it is, by the way. Thank you. The beauty of this is, it comes down to the individual case level. This isn't giving you a general idea of, we should be giving 10% off insurance renewals for all the customer, or whatever it is. It is actually telling you, in this case, here's what's predicted to happen, and here's what we prescribe you should do about it. Now, what that means is, every time you're pressing one of these cases, you're unable to make a better decision. When you make a better decision and take the right action, the odds are you will drive a better outcome. And the ROI meter just ticks up and up and up. It's incremental, it's very highly measurable. We're here at Collins Shearer learning a lot about some of the new stuff IBM has for small means. It's not so much fast moving, it's prefabricated, but it's designed to instrument the businesses with the answers, this answers product. I guess my next final question, if Dave has a final question, we'll get to it as well, but my final question is, what's next, okay? They get addicted to the system, providing business value, they're going to want to do more. So talk about what's next in the next couple years. How they build on how about from an IBM perspective, and then how you see the market trends that are orbiting around that. Well, I'm very glad you brought up a future and then talked about the IBM perspective and beyond, as it were, because we, as I said, this is a forthcoming offering. When it comes out, there will be a small kernel, a small set of answers. And we've handpicked those as ones we know are good wins for a range of different types of company. The vision going forward is like an app store. So think of this arranged by many different industries. Like expense reports, but for like business processes. And many different functions within those, so within customer analytics. Things to do to grow customer value, to retain customers and so on. Operational analytics, finance and risk, all of those things. And we want to make a lot of these available for customers just to subscribe to the ones that are going to be relevant for them. And like I said earlier, you might start off with renewals, go to cross-salon ups, or go to acquisition, go to claims processing and so on. Now, where this plays broader than IBM is we want to use this model going forward. Our intention is to make it as rich at a set of choices and options for our customers as possible. Partners are going to be fundamental to that. So we anticipate, and we haven't yet defined and published the partner program, so we anticipate that we will make it possible for partners to be able to build their own answers apps. And we will let them know. That's a good approach, because what you're doing is, you're not fighting off more than you could chew, as they say, you're getting a beach head around known solutions, and then you kind of grow from there. And leverage the ecosystem too. And we know there are partners who've got specialized knowledge and experience in places, even we haven't necessarily come here. Yeah, let them do that. Yeah, and they can bring those in. So we'll see them bringing in certain niche answers, but also just a range of choices, different approaches. Try them out. Colin Shearer. My last question is, when will this be available? Well, as always with future looking statements, we're not committing to specific dates, but our intention at the moment is to bring this out within the next quarter. Okay, great. Okay, Colin Shearer, global executive for advanced analytics solutions for IBM, business analytics, thanks for coming inside theCUBE. Great session, great new product. Again, automating and instrumenting business and hopefully customers is the future of big data. We love it. It's very complex, you're making it easy. Appreciate it. This is SiliconANGLE TV, siliconANGLE.com and wikibond.org coverage of theCUBE from IOD. We'll be right back after this short break.