 Live from Boston, Massachusetts, extracting the signal from the noise, it's theCUBE, covering Red Hat Summit 2015. Brought to you by Red Hat. Welcome back to Red Hat Summit everybody. This is theCUBE, SiliconANGLE Wikibon's live coverage of Red Hat Summit in Boston. Marco Bill Peter is here, he's the Vice President of Customer Experience and Engagement, gave a great keynote this morning. Marco, thanks for coming to theCUBE. Thank you. Okay, so skiing or biking, what's your favorite? Oh, actually the both, both. Like, one, do you do in winter, skiing and snowboarding and mountain biking? It's perfect, while you ride, too. Yeah, snowboard. Yeah, oh yeah. Or skiing, where do you ski mostly? Where's your favorite place in the world to ski? You know, I have to say, I grew up in Switzerland and so I always was kind of saying, oh, this US skiing is kind of like Lamo, right? Especially East Coast. New skiing. New skiing. But then they all got discovered on the river, but this year, and then it's kind of like, I'm not going to the West. Why would you do that? Yeah, I saw your sticker there, the ski to East. I saw it, yeah. But then, like this year, I have to say, I went to the best skiing place, was at Jackson Hole in Wyoming, and I just was like, I kind of like, I gave up on European skiing. Yeah, Luke over there, Luke says ski to East, but of course he lives in Colorado now. Yeah, I know that, it's kind of, yeah. So great keynote this morning. I think you actually look good as a blonde, by the way. I was blonde for a while and then, yeah. I love that story. You said, hey, it's rock solid, no bugs. You find bugs, you know, I'll shave my head. Yeah, I set the bar too low though. It was just too low. You know, it was a good competition. It's kind of like, I mean, in the end, it's for the customer experience, right? Whatever we find before it goes GA, your general availability, it's good, right? And so it was kind of a joke. I was like, well, let's do that. And then it was, I think it was easy for the team. They got 50 and they enjoyed it. They kind of hovered around 55 and then I was blonde for a while. And then just a few days before it was like, okay, this is the limit they pushed it through. So they were teasing you with the blonde? They wanted to see blonde. Which was actually hard because, you know, the blonde wasn't my thing. But then, so I wanted to shave it as quickly as possible. And they kind of enjoyed that for a while. But it was good. This is good. And you know, it helped in the end also, I think, to make the product a bit more stable. So I love your story. You know, many of us have read the Jim Collins book, Good to Great, what makes a good customer experience and what makes a great customer experience? I think great customer experience is really where you have the collaboration and the engagement, right? Good support can be many things, right? It can be the reactive support, can be like standing behind you if there's something going on. But preventing issues, really working with a client to actually build the next innovation. I mentioned the keynote that elevator company, right? I mean, you think about an elevator company, right? They are a manufacturing company. All of a sudden, they got to do this software engineer, right? You see many companies today in a very traditional business. All of a sudden, they got to be a software company, right? And they got to innovate, right? And all of a sudden, this open source thing is coming. And so I think we have made that as a red hat consumable for the enterprise, but we also got the way that we can interact with customers in our open and collaborative way. And I think that's great support, in my opinion. Yeah, so well, you sort of gave the example of, you know, good support is when, you know, there's a problem and you react very, very quickly. You're on top of it, great. But great support is you're proactive. You can predict how do you use data to be more proactive? So that's a really good point, right? And we have built internally Hadoop clusters, really like big data in a way to figure out, okay, what are the trends? What's going on? And, you know, the knowledge that we have, we also use in support, in the good support, our support is not just like you fix an issue and you move on, it's actually every issue creates knowledge. So every issue in the support engineer first searches, if they can't find it internally, they start creating basically an article. And so you create content on the fly. And this is what I mentioned in the keynote, there's millions of content that we have that describes the problems in the language of a customer, not the problem that the engineering person sees. Well, if the code goes here and here, you get this issue. It explains in the voice of the customers and like, if I do this and this, that happened. And so from that knowledge, we do a lot of, I would say, analytics as in like, what comes up and what's kind of a trend, right? And in the past, I always had that feel, okay, one customer starts calling us and then there's a few customers that call in the same issue. And actually, instead of just having the gut feel of that's happening, having actually the data model that kind of tells you, hey, this is going on. And then to create rules and to lead into the Red Hat Access Insights, actually provide these rules back to the customer as in that they can prevent it. I think that's where, in fact, to like, using data but also being really great support is if we can prevent these. And we're going to talk about your announcement, but I want to sort of unpack the philosophy a little bit more. You publish your phone number on the website. You encourage people to call you, unlike the airlines, trying to get the whole the airline last night. You can't even find the number, you have to Google six or seven. Or you're like one hour on the hold, right? It was actually two hours last night. But they had a nice service, they could call me back. So they called me back. They called me back at midnight. Okay, so what are the touch points? How do customers, they want to interact with, I presume, different ways. Some people want to call, some people want to chat, some people want to search. How do you handle those different channels? You got to offer all of the channels, right? And that's what we do, right? And like a few years back, it was much more heavy on phone, right? It was like, I think 60% phone volume, 40% electronic like portal or chat. These days it's 90% on the portal and 10% is maybe phone. And our actually outbound phone is much higher. So we, in many situations, if we are not clear what's really going on, we encourage the engineers to call back, you know, maybe us older guys kind of use the phone still, but it's a really powerful tool if you use it the right way. Call back a customer, hey, what is going on? Figure it out. But yeah, phone I think has an importance. It's kind of like making sure things get resolved. Electronic is powerful, because then you get also the data. You get the data, hey, what is the customer experience? It can upload the log files. We have, even if you create a support case, very simple example, but we will scan whatever customer uploads us and against known issues. And it can help against solving issues. Well, and you're getting a click stream. And now, when you say portal, does that include chat or is chat all integrated? It's all integrated in the portal. And you know, you say click stream, right? It's the analytics about, hey, who is using what, right? It's very powerful. And from a company perspective, for us it's powerful to know who is using what. And also like in some cases, we see a customer that buys from us, but then never downloads from us. It's like, well, what's going on? And so we're kind of starting to engage these customers as well as in, hey, are you struggling to download the software? Are you trying to do something? And so that's kind of more proactive engagement versus waiting till a customer struggles and to call them. And that's where analytics come in quite handy. Yeah, I mean, I'm envisioning, chat is interesting, right? Because a lot of times when you interact in chat, you can tell there's canned responses. The person's got six chat windows open. Yeah, it falls apart. Now you've described, you don't have a sort of waterfall tiering approach. You organize it by the problem or you describe that a little bit? Yeah, so we use swarms instead of the layout model, layout support model. In my opinion, it can help in certain cases. And there's some corner cases where I would say, okay, that works like language support, right? If you want to provide support in Spanish or French, right? That might actually be better for having like a language team that does it. But otherwise we actually grouped the engineers, the support engineers around small technology fields. And they're small as in virtual memory, you know, CPU loading or, you know, pick your favorite. It's not even like a product. It's really specific areas. And that allows these engineers to be much closer to the technology, but they're also aligned with engineering much better. So they know each other, right? And, you know, that's still, we're still a company that that matters that, you know, the engineer works in the same swarm than like the development engineer and they have a really good connection. If that's electronically or even by phone, that helps, it tells you this helps a lot. It's a different model. So the simple math would say, if you sell a subscription and they never call you, or they never use that subscription, you make more money. But you have a different philosophy. That's a totally, yeah, yeah, yeah, yeah, yeah. Opposite of what really happens. So, you know, I'm not happy about that because if somebody doesn't use it, I don't know if the customer's happy. I don't know if in a year the customer will renew. We don't know anything about that customer. So for us, that's why we drove this whole engagement to understand, is the customer satisfied? Will they renew, right? And, you know, the products are fairly good. So not everybody uses support. And so that other tactical support, reactive support, that's why we started providing an engagement platform and understanding who is using it. And, you know, I give you an example. Most customer or new customers, when they struggle with, you know, kind of the first two months to install the software, get it up, they will not renew because that's the hardest part, right? It's kind of well known in the industry subscription business. It's the first one month, two months is critical. And then you get them onboard and then you got to kind of look at the utilization. If you look at the utilization, it's a really good point to kind of see, okay, they are using it, they're happy, okay, renewal is, no, it's not guaranteed, but at least you know, hey, we have track record that customer uses the product. Well, the power of that subscription model from an economic standpoint is enormous. I've said, I mean, I term it as, Red Hat's renewal rate from a revenue standpoint is over 100%, right? If you look at your average prices going up, you're crossing, you're cross-selling, you renew, 90 whatever percent of the customers and they're paying more, you're at 115% renewal rate from a revenue standpoint, it's a very powerful model. It's awesome, yeah, yeah. But we also got to make sure that that happens, right? If the renewal rate is so high, you got to do something for it. There's an expectation there. Yeah, exactly. You got to know how that happens. And it was interesting to hear you describe, juxtapose your model relative to the traditional, you know, software license in the front. The maintenance model, the maintenance is like, oh, we don't have to work for this. They have to get the maintenance. Exactly, right? Yeah, and even usually that's just cost, right? They're optimized for cost. I've been in that business as well, but it's not a good, you know, if you're an executive of support, you know, I don't want to be in janitorial services, right? It's like, you want to be in a company that puts that in the business, right? That's in the core business and makes it actually something, hey, this is meaningful, and so that's why I'm so happy to hear, but I think it's also the right DNA to actually look at customer success. Now, the disruptor here, oftentimes, you know, somebody said yesterday they, I think it was Cormier, they sort of ignore you and then they laugh at you and then they fight you and then you're in. The Gandhi quote, yeah. The Gandhi quote, right? So, they ignore you, but now that you're a larger company, you're having great success, you're disrupting people, they start to pay attention. Yeah, they don't ignore us. What others, right? And then they start to copy you. You left that part out, Gandhi did. So in our business, they copy you and they try to, you know, mark it around you. So do you see that model, that old sort of license and maintenance model? It's changing, cloud is changing it. People, I think the large companies are realizing, wow, the subscription stuff, if we can manage through that transition, not so bad. We're actually make more money. So they're starting to realize it. So do you see that changing? Do you see more competitive? Yeah, I mean, I think you see it as well, right? I mean, I think there was last week in an earnings announcement from a company in California that tried to describe the whole after the earnings and kind of like the subscription model, right? And I think for a lot of traditional companies, that's really hard. It's not explaining the subscription model, but kind of even mentally move from a license model, high margin maintenance model to like a subscription model. And especially in our case, that there is around open source, that is pretty cool. And so what you see is a lot of proprietary companies are trying to, you know, they don't ignore us anymore, right? 10 years ago, 15 years ago, if Linux was like, kind of like, yeah, just we'll let them do, the Raleigh company will let them do something. Yeah, cute, sounds great, the Penguin, right? Sounds great, not anymore, right? If OpenStack is in. But so then you have this proprietary company, they're mixing the models, right? They're kind of like, they still have the license model in mind, but then they throw some open source somewhere. And I think that's, I think I mentioned that, you know, it doesn't lead to the pure DNA that you actually embrace it. And I think for OpenStack and for the industry, it's really tricky. If everybody does their own OpenStack distro, I think Paul described it quite well. We'll end up in the same thing we had in Unix. And I was in the 90s doing Unix as well, but every Unix was differently. And then having a strategy around containers that you can move a container from one OpenStack implementation, they won't work, right? And so I think the next few years will be really interesting to see what shakes out. Yeah, I mean, the whole container conversation makes it interesting from a support standpoint. If I got, you know, the whole cloud makes it pretty interesting. You know, if I have hundreds of thousands of containers, you gave the example of, you know, the virus, the swarm, I mean, the heart bleed virus. Wow, now, you know, hundreds of thousands of containers, you guys got to be on top of your game. And you got to know what's in that container. Yeah, you better know what's in that container. You better be able to crack it open. I think that will be a big stretch, right? I mean, I think this leads to what we saw in Linux as well, right? You have these repositories, right? These RPM finds where you can download, open source software. If you as a corporation do that, you better have a strategy around, hey, how do you manage this open source coming from everywhere? And this I think is with containers as well, right? If you're a bank, you want to know what's in that container, what application, what kind of around the container, you know, if you don't know that, it's like not even a support problem. It's like a whole compliance issue. How do you manage as a company? Well, how do you manage those containers? That's a big, that's kind of the next wave. Yeah, and how do you avoid to make containers heavy by putting too much stuff into the container, but go to a modeler, you can certify them and inspect them as in like knowing what's in there. Yeah, the advantage of containers is they're lightweight, but then you start putting all those capabilities in, you're going to make it bloated, right? And then you add complexity, which makes your job harder. Well, your job's going to be, complexity is going up. I mean, even though the end user you want some simplicity, but yeah. Well, I think some stuff will be easier because with containers, you can probably split the application down to higher resiliency by actually having like multiple functions kind of going on at the same time. So maybe from a reactive perspective, maybe the impact is not as high if you use containers correctly, but I would say the first few years this will be an interesting journey. And like a lot of customers, they are not bleeding edge, right? There's like, we have a lot of customers experimenting with containers, OpenStack, but you also have a lot of customers. They got to run the old IT, right? Or Gartner calls it the bimodal IT, kind of the old stuff. Most customers, by definition. Yeah. Yeah, we're talking about bimodal IT for a second. You know, bimodal IT, good concept, probably Gartner does a good job of sort of creating these good concepts, but the problem I have with bimodal IT is like, create more stovepipes. Yeah. We've got the new stovepipe, and we've got the new stovepipe. Which stovepipe do you want to be in? Nobody's going to want to be in the old stovepipe, so I'm not sure that that's the right organizational model. Plus, plus, it leads to like, I mean, the data. How do you connect the data from the old model to the new model, right? I mean, that's, I don't think the model, I mean, it's an interesting model. It kind of describes what you see, but it's not as simple as... And you got to have a bridge. I mean, Red Hat is a bridge to that new world. I mean, you don't want to rotate to that new world. Exactly, right? And I think a lot of customers want to have, I mean, I think that's what the keynote from Paul talked about, you know, like some applications only run on bare metal, right? I was talking to a customer this week. Oh, it doesn't run on virtualized. But then you have some, you got to run it bare metal, virtualized, public cloud, private cloud, right? And having that capability to move. I think that's more describing than the old IT versus the new IT. And I think we'll see. Some companies move to the public cloud. Some actually private cloud is probably more appealing for data security or customer data security privacy. It will be interesting to... Well, we love to talk about disruption in our business, but the customers of yours, that's where the really interesting disruption is going on. I want to talk about access insights. Announcement that you made today, what's it all about? Take us through that. So access insight is basically very simple. It's taking, you know, I gave the example in the keynote about navigation system, right? It tells you, hey, there's traffic jam in Boston, happens every day. This is the route you can avoid it. It's very specific, it's specific to your car. It's not specific around the whole Boston area. And basically that's what we want to try to basically get our knowledge from our interaction down to certain rules that if you're driving from there to there and you're hitting this traffic jam, this is the detour that you got to take. And that's what we want to provide in a way for our, you know, Red Hat Enterprise Linux first, six and seven, and we'll deliver that, the console in satellite, the portal cloud forms. But basically what it does is a customer connects to us and provides minimal data around packaging, like an inventory of, hey, this is the packaging, this is some configuration data. In a secure way, you know, many customers already are connected to us with satellite. So this is kind of an extension from that. And then we match that footprint, that inventory against known issue that we just found. And basically give exactly a dashboard back. This is a security issue you're hitting, availability issues, reliability issues. And right now it can be around Linux, but it can also be a combination of our Red Hat Enterprise Linux and the firmware that you have and the hardware and kind of describe, if you do this, we had a firmware issue a few months ago with one of the hardware vendors that now we can be proactive and say, this box is exposed, that box is not exposed. And that's the specific recommendation, that's the power of it. It's not just a broadcasting as in like, you got a new, oh, we, we found, we Red Hat have found the new security issues. Go figure out yourself if you're exposed or not. That's not good enough. We want to be specific on that. Either figure out if you're exposed or just spread it everywhere and who knows what else is going to mess up. And those aren't good choices. I mean, how many notifications do you get and I don't know what your desktop is, right? But you get like hundreds, you got to update this and this. It's like, why, right? And if he can be more specific, you got to change this configuration file because you will hit this problem. Jim Whitehurst uses that example from the airline industry from, I don't know, which seven, eight, seven or one of these airplanes that has, they got to reboot the systems every 208 days. And they know it, obviously, they got to reboot. And there's many kind of frightening, right? If you know that they got to be, you got to check out the seven. Yeah, which airline is that? Yeah. And how long is that flight going on? And so that's the same. I think there's a lot of issues that we know it's time-based or the configuration I mentioned today. This is stuff a customer could know, but they don't do anything and they run into it. And we're giving them the specific recommendation. Hey, this is something you got to do. And Access Insight is available today? Today, yeah. They can download it, they can play with it. And in the next RHEL 6.7 version and 7.2 version will be implemented and they can use it in production. But we have customers in beta for the last seven months. And yeah, I'm excited. I want to see next year how this really turns out. It sounds like the ways of, you know. The ways? The ways of IT. You had to be using ways because your normal GPS wouldn't work that. No, actually, I had, I have a good car. So it was the regular one. But yeah, ways would be the one that actually works. Tesla. It was electric. It wasn't Tesla. All right, good. So now, I have to ask you, so how was that, how's the pricing model work? It's just part of the subscription or is it an upgrade to the subscription? We're still figuring that out. We will, I'm sure we will have a certain element that's free or included in subscriptions. And there's probably an opportunity that we'll do actually, hey, some advanced services around it. But for us, it's more important. Right now it's important. Let's drive this as a subscription value and then we'll see where we take it. All right, Mark, we have to leave it there. It's really a pleasure hearing you in the keynote this morning. Thanks so much for coming on theCUBE and sharing your insights. It's been great. Great, thank you very much. Thank you. Right there, everybody will be back with our next guest. This is theCUBE, we're live from Red Hat Summit. We'll be right back.