 Live from San Francisco, it's theCUBE, covering Informatica World 2017, brought to you by Informatica. Okay, welcome back everyone. We're here live in San Francisco for theCUBE's exclusive coverage of Informatica World 2017. I'm John Furrier with SiliconANGLE Media. My co-host, Peter Burris, head of research at SiliconANGLE Media as well as the general manager of wikibon.com, wikibon research, check it out. Got some great research there on IoT, big data, and certainly cloud computing. Our next guest is Graham Thompson, Executive Vice President and Chief Information Officer for Informatica. Great to see you, welcome back to theCUBE. You see you John. Conference here, a lot of customers to the executive summit, dinner last night. You're kind of like the sounding board. They go to you for like the checkpoint. Hey, this is a story, Jai. What's going on in terms of, because you're living through a transformation as well at Informatica. Your customers are going through a transformation as well. We're at this tipping point. What's your take so far at the conference and is that still the case and anything you'd like to share on that would be great? Yeah, I mean, we're proud to have some of the world's best companies using our products to do meaningful and important things. And the scale that some of these companies are doing it at is just staggering. I met with someone last night at dinner and an Allegious or Talent Management organization and they process and keep up to date 55 million resumes every day and they extract the metadata from those resumes to match the right candidate to the right job. And that's interesting for them as a company, but the societal impact of that's significant. You're matching, I mean, we're all star for talent and you're matching the right talent with the right opportunity more often than not using the intelligence of the data. It's pretty interesting. And then of course, I know you had Andrew McIntyre from the Cubs on yesterday. And how can you not love that story of how an organization is great and renowned as the Cubs is using data to transform its business operations. It's really amazing. We had Bruce Chisholm, who's executive chairman of the board of Informaticals on the board at Oracle. But Peter asked me an interesting question that I'll ask you is, what's your definition of strategic data management? That's a good one. So the way I define it is if the basis of your competition is on digital assets compared to physical assets, so we're no longer dealing with plan and machinery or even capital, it's digital assets. If that is the basis of your competition, then the data that you rely on is a very foundation of that. And then it becomes strategic, just like money is strategic and the access to talent is strategic. The ability to leverage the data within your company, about your company, is strategic. And you have to be able to do it on-prem, you have to be able to do it in the cloud, and you have to be able to do it in the real world where most of us live, which is in both worlds. And that, to me, that's what makes it strategic. But let me build on that, Graham, because in many respects, the whole concept of digital transformation is, or let me step back, one of the premises of business is to try to reduce what's known in financial or economic worlds as an asset specificity. So traditionally, we've looked at assets and said, this asset's going to be applied to that use, and this asset's going to be applied to that use. And if the use isn't needed or if it's not being applied, you lose the value of the asset. One of the basic premises of digital business, and business generally, is how do we reduce asset specificity and data lessets do that? By turning an aircraft engine into a service, we have transformed the role that that asset plays in our customer's business. So it's, you're absolutely right. It's the ratio of physical to digital assets, but all businesses have to find ways to reduce their asset specificity by adding digital on top of it so they can appropriate that asset to a lot of new purposes. Do you agree with that? Absolutely. So take, so I know you talked to Sally about the data lake. So take a use case like customer support. Who in a software company knows more about the customer? What product they're running? What version of product they're running? What they're using it for because of the connectors they have? Nobody in the company knows more about that than the customer support organization. But that asset, the most profitable use of that information might be in marketing because then we can help our customers adopt something more quickly. We can help them get value from it more quickly. And it helps us because it helps us focus our R&D effort where the customers are really using the product instead of having a guess. So I think you're spot on. If you can remove the constraint on the asset to be for who paid for it, for one particular purpose and make it available to the entire enterprise and outside the enterprise, then you really start to see the value. The thing that you mentioned about digital assets, Peter and the Wikibon team talk about this all the time in their research, digital assets is the data. Whether it's content or whatever, certainly we're in the content business. Well, digital assets are data. Are data, exactly. And whether it's content or whatever aspect it is. So I got to ask you- Software. Software is a digital asset, it's data. Data is at the center of it all. So I got to ask you, there's been a lot of artificial intelligence washing going on in the industry. I call it augmented intelligence because it's truly not yet artificial by the strictest, purest definition. But machine learning is very relevant. I talked about IoT when you were at last in our studio. How is it impacting your business and customer's business? Because that's the real proof in the pudding, if you will. And customers are trying to sift through the BS that they're hearing from other folks. I'm not saying that you guys are saying BS, but what's the asset test? What's the, how do you differentiate between smoke screen and real deal? Yeah, so I think it comes down to like any other technology investment is what is the business outcome that it generated? So if you're trying to, so humans make mistakes. So if you're trying to eliminate human error from a process, a machine can execute that process more repeatedly and more accurately than a human. It's not about reducing costs, that's only semi-interesting. It's about enabling outcomes that weren't possible before. So you think about healthcare industry. You know, we, everyone talks about self-driving cars and how safer it will be if the cars aren't dependent on a human. But one thing I read recently is we kill more people in the U.S. by prescribing the wrong drug or the wrong dosage than we do on the roads. So humans are, they work hard, but they make mistakes. If we can have the machine do that job because a human can tell it how to do the job and it can learn over time, then you eliminate that error and we're able to do things that we can only imagine. Yeah, machines rarely get tired. They rarely lose attention, blah, blah, blah, blah. And it's all those things we would like to, we'd like to, and that's where the augmentation is, is because, and there will be other forms of artificial intelligence. The algorithms have been around for a long time. The hardware now can support it and the data's being generated to apply it. Yeah, the data's available and the cost of compute is approaching zero. So we're able to do things that the government could only do before. Graham, I want to get your thoughts on data integration. Certainly we saw yesterday the news with Google Spanner. You guys were one of three companies that was early on before they announced their general release of Spanner worldwide, the distributed database, horizontally-scaled database, big deal. But you guys were also on the front end of that, as it says in their blog posts. And you guys are really strong at data integration. What are some of the challenges that the customers face with integration? What are the key things? Because that seems to be, whether you're going multi-cloud or hyper-cloud today, which is a gateway to multi-cloud, which is happening pretty fast. Data integration is pretty important. Yeah, so as a CIO, this is something that is a very hot topic for me. And it's not a new hot topic. It was a hot topic 15 years ago when we went nuts and deployed all these client-server applications because they were cheap and easy. And then you had to think about, oh, these different disconnected applications don't serve an end-to-end process anymore. Now I have to stitch them all together. That was hard, but it was all on-prem and you had access to it all. It was all programmed. Right, whereas now, like you said, you've got Salesforce, you've got Workday, you've got great people, you've got your on-prem stuff, you've got applications that you're hosting on, someone's PAS Cloud and the IAS Cloud and the SAS Cloud. But to execute an end-to-end business process to generate an outcome, you have to tie it all together. So instead of thinking about- And it's not under pressure, you can't touch it and you have to, it's not on, you'll have it. Right, so you can't hand code that. You could, but I would argue that would be an unintelligent way to do it, which is where microservices APIs come in. So you can leverage the R&D efforts that the great software vendors like Salesforce create for us and then you use microservices to plug into that instead of having an army of people hand-coding interfaces, which is what we used to do at 15 years old. Actually, a human error point. I mean, it could be SpaghettiCo, all kinds of errors could happen. Yeah, but also the maintenance of that is just virtually impossible given the speed and the fact that human beings are now thinking about new ways of doing things. You just can't keep up with that. Exactly. I mean, the coding thing's a big deal. We used to call it back on the day, SpaghettiCode, because it's like all this integrated, purpose-built coding for one purpose to glue it together. Right, and then you change one data element and you have to rewrite and retest the whole thing. The guy leaves or the girl leaves. That's a nightmare, right? With APIs and microservices, you're decoupling that. That's kind of what I think you're getting at, right? Exactly, and that's what the whole iPad space is about. You can decouple the user experience from the data and just have what does the user have to do, and then the microservices and APIs will take care of the work behind the scenes between the applications. And that really lets, there's this concept of a citizen integrator. So 15 years ago, it was kind of a modern thought to have business people write reports. I think it won't be long before we'll be able to give the business teams the ability to do integration between applications without depending on me. I was sure with the young developer the other day, I'm like, hey, you know, your coding is like me doing PowerPoints. I'm like, what do you mean? It's so easy. I'm like, oh, it's not that easy. Well, we've been building macros, good or bad, inside, for example, things like Excel for a long time. And one of the primary drivers, in fact, of a lot of the BI stuff was citizen coders building macros and said, I need the data to make my little macro run. Now, I don't want to say that that is, that's not what we're talking about. We're talking about something that's considerably more robust where we can be very, very creative in thinking about how we might use the data and then being able to discover it and find it and very quickly and with a low code orientation, being able to make the actual application happen that has a consequential impact in the marketplace. So, Graham, you're in a company that is trying to help customers move through some of these transitions. You're in a crucial role because we know where the data is, we know how to integrate it. We're discovering where the data is. We have tools that's going to help us. We're learning how to integrate it. But one of the big challenges is to get the business to adopt new orientations to the role that data's going to play. That, to me, is one of the key roles of the CIO. Having worked with a lot of CIOs over the years for a very, very simple example. Agile development does not line up with annual budget and finance. How are you, with Informatica, helping to accultrate executive teams to think through new processes, new approaches to doing these things so that the business is better able to use the data so that consequential action happens as a concept of these great insights that we're generating? Yeah, so the whole change management effort is a huge and complex thing to overcome. But I have a personal passion about making sure that you always remind people why they're doing it. Too often, as product people or technologists, we get into the how and the what, and we forget the why. And as soon as it gets difficult, people abandon because it starts to get too hard, it starts to get painful. And if they've lost sight of the big why, they're not going to roll their sleeves up and gut it out and get through the process. So that's the first thing you have to do is remind them that the prize at the end is worth the pain. And it will be painful because no longer are you optimising just your function. You have to think about what happens upstream from you, what happens downstream from you, and try to optimise things at the enterprise level. And that's not the way most people were brought up, it's not how they're measured, it's not how they are compensated. But that's what's really required if you're going to make that transformation and think end to end. But it's also, and even our language, we talk about innovation in this industry as though it was synonymous with just creating something new. Certainly our research very strongly shows that there's a difference between inventing something which is an engineering act and innovating around something, which is a social act. Exactly what you just said. How do we get people to adopt things and change behaviours and fully utilise something and embed it within their practices so that we get derivative innovation and all the other stuff that we're looking for? Yeah, there's no easy recipe, people are different. So people require a different story in order to have them buy in. Some people are lost frame people. You've got to explain, here's what's going to be bad if you don't do this. Other people are gain frame people where you can say, if we can accomplish this, we'll be able to do these great things. And it would be great if everyone was the same and one story worked for everyone, but it doesn't. So it's almost a feet on the street. Go talk to people and just keep reminding everyone why you're doing this and why it's going to be worth it. A little bit of behavioural economics there. Yeah, I want to ask you one final question. I mean, you mentioned client server and how it was easy on prem in the old days to get your arms around things, which is the IT practice. That was the way it was done. Enter cloud a little bit more complex. But to take that a little step further, I want to get your thoughts on something. You live through the world of server sprawl. More servers, more glue. Yeah, you get your arms around it, but then it got bloated, IT got bloated. And that's one of the catalysts for going to the cloud is efficiencies, bottom line costs, but now top line revenue now is a mandate. So now we have SaaS sprawl. So with APIs, a little more security concern, but your thoughts on the, now we have a SaaS-ification happening or API economy. So you have a lot more APIs. There's microservices coming on the scene. It's emerging very quickly, still emergent. Embryonic, if you will, some will say not so, but I think it's embryonic still. But okay, server sprawl, client server, VM sprawl. Now you got SaaS sprawl, your thoughts on this dynamic and how a CIO tackles that. Yeah, so it's the modern equivalent of your legacy technical debt. So it's a modern mess instead of an old mess. But it's the same problem. You know, you have to stitch these applications together and it's made worse by the ease of consuming these SaaS applications. So one business function can go off and buy an application that's just for them. And then the adjacent business function goes off and buys another application that's just for them. And before you know where you are, your single sign-on page has three pages because you've got so many applications that you're using to run your business. So I think we have to be more thoughtful and think, you know, not make the same mistake that we made after 2000 when we went nuts on all these client server applications and make sure that we're thinking about the end-to-end business outcomes. So the unification layer is what? Identity, is it the data? I mean, how do you think about that and just conceptually? Well, I think you still need a sensible portfolio of applications. I don't advocate that you just go buy every great application that's out there. If your business doesn't compete based on the capability that that application provides, you're going to know business innovating, just be as good as the next guy. But if you compete based on something, go pick the very best application you can, but deploy it thoughtfully, make sure it's integrated and make sure it serves the end-to-end. If you're at a point. I'm also fascinated by the role that Claire might play here at going and looking at the metadata associated with some of these SaaS applications to help us identify patterns and utilization. I think Claire and the thing that was announced here actually could have an impact on thinking about some of these. Claire Boyant App is a great one. Claire, I mean, she, he, I mean, it's vendor neutral. That's a whole different story. Only kidding. Final thought, Graham, on this show. Just color perspective, what's your thought so far on the show vibe of the folks who aren't here? What's it like? So when you and I met a couple of weeks ago, we talked about the fact that I just joined the company just after last year's show. So I have nothing to compare it to, but the energy level is phenomenal. The feedback from the customers I've talked to just reinforces that we have really, really important customers and we are really important to them. The customers are the ones driving this digital transformation and we're proud to be helping them. And every conversation I've had with customers has really reinforced that and it's great. Can't wait to get back to the office and get to work. As we say, the KPI, the metric of the transformation world is not quadrants or category winners. It's customer wins. Absolutely. And I think that's a great point. Graham Thompson, executive vice president and chief information officer, Informatica, sharing his insight. He is integral part of their transformation as well as customers Informatica. World coverage with theCUBE continues. I'm John Furrier with Peter Burris with Wikibon.com. We'll be back with more. Stay with us after the short break.