Rackspace Hosting, Inc. (NYSE:RAX)
Company Conference Presentation
December 10, 2013 5:15 p.m. ET
John Engates – Chief Technology Officer
-- analyst here at JPMorgan, very happy to have with us John Engates who's the Chief Technology Officer of Rackspace. I'm going to turn the microphone over to John to go through a brief presentation, and then we're going to jump in to Q&A. So with that, John?
Okay. So I have to apologize, it's my fault that we're getting a late start because I gave them the deck kind of last minute. And also had to convert it to PowerPoint. So if you see any kind of stuff like I see right now that says Boston Analyst Day, that's -- ignore that. That came through when I converted from Keynote to PowerPoint. So, something got funky there, and ignore the mistakes.
Anyway, my name is John Engates, I'm CTO at Rackspace. I've been at Rackspace for about 13 years. Spent a lot of times with our customers, spent a lot of time with the market, the analysts, the people that are the buyers, the consumers of cloud computing, and I'd spent a lot of time telling the Rackspace story. So let me start with sort of the strategy behind what we're doing at Rackspace, just to give you a flavor of how we think about cloud computing and how we're attacking that market.
So first of all, we think a lot about open technology, open source. We built our company on open source. The web in our -- throughout our history was built on open source tools like Linux and Apache and MySQL and a lot of that. So we've taken the same approach when it comes to cloud computing. We have created -- we're the founders of this project called OpenStack. We're also one of the largest contributors to that project. And that is the foundation of everything we do in the cloud. So open source is not only the technology we use but it's also the technologies that our customers use when they build their solutions, whether they be mobile apps or software-as-a-service app. We find that open source resonates with those developers and it has certainly been a good model for us to build our cloud technologies.
Another key strategy that we think differentiates us in the market is the idea of our hybrid portfolio. Rackspace's history is in hosting. We came from a model where we created dedicated servers for customers. We manage those dedicated servers. We moved through the era of virtualization. We moved into the era of cloud computing.
I think Rackspace is in a position unique in the market where we can actually blend those things together seamlessly and we can give customers a number of choices and mix them effectively so that customers can actually find the right mix or the right fit in the technology choice, and the portfolio really gives them that broad spectrum.
And then the last bit here is called Fanatical Support. And you've probably heard that or seen it at least maybe on our website, it's the brand name that we apply to the way that we deliver customer service. It's the depth of service that we offer. It's also the culture at Rackspace of how we build the foundation for great customer service. It's how we hire, it's how we train people. It's a lot more than what it sounds like.
And I think one thing I really want to make clear is that this isn't just platinum support like some of the other guys. This is actually taking service to actually another level, and I'll talk more about it in a minute, but let me go in to what I mean by fanatical support.
So throughout our history, we have obviously offered a technology product. But on top of that, we've layered in services and support and expertise that actually allows the customer to lean heavily on Rackspace. They don't have to do everything themselves. They aren't treating Rackspace just like a data center or just like a technology platform. We are an extension of their IT organization. We take on a lot of the heavy lifting of managing their applications and keeping their applications up to date and running.
A lot of our customers historically have actually not had to hire people because Rackspace does so much of the work. So it's not just, you know, a cloud server rental business, it's actually a technology operations business where we're running a lot of the stack.
In the history of Rackspace, a lot of what we did was, you know, managing the servers, managing the operating systems, doing things like patching and monitoring and keeping the servers healthy. What's changing lately is that customers are transforming their business from a traditional IT ops model to more of a dev ops. I don't know if you've heard that term much, but dev ops is the way that cloud companies are sort of looking to run operations in their environment. So that means that the developer is much more front and center than ever before. Developers are actually integrated with the operations teams now. And they're automating, they're orchestrating the automation of their environments with new tools and new paradigms.
So we're transforming fanatical support from the old IT ops world to the new dev ops world. And so that means leading with tools like the dev ops blog, actually reaching out to developers and telling them how we operate our cloud and how we operate other customers' applications. And we're sharing much more of our tools and our technologies in the open source domain, on GetHub. GetHub is a very popular place for developers to share code and share patterns for how they build and operate their environment. So Rackspace has adapted in that world and starting to lean into the idea of dev ops and transform our fanatical support to that model as well.
I mentioned hybrid at the outset. Hybrid is another, like I said before, key aspect of how we differentiate in the market. A lot of our competitors, you know, Amazon is obviously the biggest, Google, Microsoft, others in the market, are really only public cloud. They're only offering virtualized multi-tenant cloud services. I think Rackspace, again because of our legacy of -- or a long history I should say of offering dedicated servers, we can actually tie that together with a unique technology that we call RackConnect. We can blend dedicated bare metal servers, private clouds based on open stack and our data centers, public and our data centers, and even private clouds hosted in customer premise data centers. All of that could be pulled together with the RackConnect networking technology and it gives customers actually a better technology platform than any one of those alone. We'll talk more about that later as we go into customer case studies.
Another thing that I'm particularly excited about that we've recently done is started to tackle the big data market. Big data is something that a lot of our customers are experimenting with, using in their own environments, building into their applications. We launched recently a Hadoop service that is based on Rackspace plus Hortonworks technology where customers can consume the Hadoop platform as a service. We also have a MongoDB service that customers can use. Those are basically going to make up the components that are the foundations for our customers' applications.
I'll leave it at that and we'll maybe go in to an example of what big data really means at Rackspace. Again I apologize for how the slides went or if they don't work perfectly, that's the connection or the translation here.
So this is supposed to be a mobile application called MapMyFitness. Underneath that little crazy logo is a mobile application called MapMyFitness. They're a customer of Rackspace that creates a tremendous amount of data that goes into the cloud. They track people's runs or their rides or their fitness. They have 17 million users in their network of MapMyFitness. Each one of those users is creating, you know, potentially four or five routes that they run. Those have to be stored. New users are coming on regularly. Now they've got up to 80 million routes that are going into this network. That translates into a dramatic number of activities per day and basically a year, annually, they're doing 1.4 billion activities per year that are going into their platform. That translates into large amounts of data that are going into the cloud.
So what I'm getting out here is that cloud is a foundation for a lot of new kinds of applications, mobile apps, software as a service applications. And when I say big data, it's not just big data analytics in the sense of a Hadoop cluster that analyzes your structured or unstructured data. It's also just the massive amounts of storage and, you know, data that needs to be stored to make these applications real.
Another great customer or example of big data is the large Hadron Collider at CERN. We're actually helping CERN run a private cloud in Switzerland to power the large Hadron Collider. They basically create 25 petabytes of data per year. That needs to be analyzed. This is actually 25,000 gigabytes, what 25 petabytes equates. And, you know, just the idea that more and more of these modern applications are going to create massive amounts of data, that's what gets me excited about what we're doing in the private cloud and the public cloud and how this hybrid stuff comes together.
So we didn't spend much time on public cloud, but Rackspace has a comprehensive public cloud offering in the market today based on OpenStack. We have cloud computing, cloud storage, a cloud network that basically creates a software-defined network, cloud storage in the form of object storage. We have a cloud databases product which is MySQL as a service, plus the Mongo that I mentioned, and also the Hadoop service.
And then private cloud. Private cloud is a very interesting offering because it's actually allowing us to talk to markets that we've never had access to before, people that want to do things in their data centers. Rackspace has never had a real relevance within customer premise data centers until we started offering private cloud, and now customers actually hire us to help them build private clouds in their data centers. We've done that from Fidelity, we've recently announced that Workday is now a customer helping -- having Rackspace help them operate their private cloud in their data center.
Let me -- we'll go back to private cloud for just a second. Let's just spend one more minute on public cloud and then we'll get to that.
So another thing that we did this year that I'm particularly excited about is we launched a new capability called Performance Cloud Servers where we took all the spinning discs out of the servers. We introduced a 100% solid-state disc into the servers now. We raised tremendously the headroom in terms of RAM, the processing power in terms of CPU, and the network bandwidth. We have 40 gigabits per server now to every single host.
This has unlocked a tremendous number of possibilities for our customers in terms of what they can do with the public cloud. And then again in concert with the other services, the hybrid cloud, I think it opens up a lot of new capabilities for us. We have now twice as much CPU, up to 2.1 times of CPU, 59 times more I/O to the servers, 3-1/2 times more network bandwidth, and overall 2.8x overall performance. Now these numbers don't mean anything until you put them into context of an application, and I have a case study to share in a minute here.
So again back to the hybrid cloud, I think what really is exciting about when we put together a private cloud, public cloud and our traditional dedicated server offering is that customers can actually make a better overall total cost of ownership than they can even at the best case on some of the public clouds, our public cloud included. If a customer tries to run all of their application in a public cloud, there are some trade-offs they have to make. Sometimes they have to over-provision, sometimes they have to make performance trade-offs. But when you can blend dedicated bare metal servers and the control that you get of a private cloud with the scalability and the elasticity of a public cloud we believe, and we've seen many cases, that you actually get a better overall total cost and much better reliability in performance.
This is sort of what's so great about private cloud. This is why people choose private cloud. Sometimes it's the security reasons, sometimes it's the cost reasons, sometimes it's for location reasons. It's all of the above. And again that's why we've seen such interest in the Rackspace private cloud product.
We've taken what we believe is the best of our public cloud capabilities, our expertise, our scripts, our tools. And we've distilled them down into a product that customers can now take and consume on premise in their data centers. And that's very attractive to a lot of customers.
This is a case study of a company called Digital Film Tree. They were actually featured at the most recent OpenStack conference in Hong Kong. And they talked about what they're doing on the Rackspace hybrid cloud.
They run an application that film producers and television producers use to collaborate amongst their directors and their film teams and all of that. Think about it sort of like a way to share the shots and the flow of the show or the movie with all of the team even if they're in remote locations. So think about what that means in terms of a cloud application. All of the video has to be uploaded to the cloud and also be stored in the cloud, has to be transcoded into usable formats for tablets and for laptops and has to be sent around the world. So this is basically being used by a large number of people that are very dispersed. Perfect application for the cloud.
So Digital Film Tree took the Rackspace public cloud and also our private cloud. They blended those together. They're a video software production company that's turning their business into a SaaS business where they can deliver this as a service to their customers. They find value in the public cloud, the private cloud, the scalability. And they actually said 10x performance with our new public -- or performance cloud servers. They found 10x the performance over our previous platform. So those numbers that I threw up there, they don't really mean much, again, until you put them into this context. But a customer that's able to actually get that kind of performance, do it all on the hybrid cloud and do something that's, you know, being used by a large number of people around the world, I think that's really what our hybrid cloud means to customers.
So that's all the presentation. We can go into Q&A. Maybe I can sit down and answer some questions for you guys.
Sit down, relax. Let's try to get through a bunch of these, because I've been excited about this presentation all day.
Let's start with -- especially for those that maybe are just new to the whole cloud experience, so they get thrown, there's Google, there's Amazon Web Services, there's Microsoft Azure. What the heck is the difference? What is it that someone as a customer gets out of Rackspace that they don't get out of Amazon, or vice versa?
Yes. Well, so if my microphone here -- there we go. I can hear myself now.
So -- okay. So fundamentally, what Amazon has done is built an infrastructure-as-a-service offering that is the most -- I guess the leader in the market, certainly. What they haven't done is really taken that to the next level and offered customers a service wrapped around that. You're kind of on your own when you're with Amazon. You can rent as many servers as you want, you can use as much storage as you want, but you have to have the operations team, the engineering team, the dev ops folks to run that on a day-to-day basis.
I think the real difference between Rackspace and all of those other guys, Google, Amazon and even Microsoft, is that Rackspace takes an active role day to day in the operations of our customers' applications, in their infrastructure, the way they use the infrastructure, the way they optimize that infrastructure.
The other guys, they don't monitor your sites, they don't know what applications you're running. They don't really care much about what applications you're running. They're just renting capacity.
Rackspace is doing that, plus we are actually operating these environments with our customers hand in hand.
Would it be fair to then characterize Rackspace or let's say Amazon is more of a do-it-yourself and Rackspace is more of a do-it-for-me or do-it-along-with me --
Well, certainly you do it along with me, yes. I mean we don't build applications for customers. We don't write custom smartphone apps or web applications, but we will run them for you and with you. So sometimes I explained it to customers as if we're -- you're the pilot, we're the co-pilot. We're in the seat next to you, so that when you want to go home at night and when you want to have a nice weekend and when you want somebody to be there with you on Black Friday and Cyber Monday on the phone with your ops and your dev teams, that's Rackspace. We're actually able to pick up, you know, a lot of the work on a customer's behalf so that they don't have to have as big of an operations team, they don't have to have as many experts.
A lot of times when you want to take advantage of something like Hadoop or Cassandra or MongoDB or one of these new tools, having one guy that knows it a little bit isn't oftentimes enough. It's -- that's enough to get you in trouble, that's enough to get the application built. But it's not enough to run that application in production at a high scale when you're betting your company on it. So oftentimes they don't want to run with no safety net and with no co-pilot. And that's where Rackspace really does pick up the slack.
I know you're the CTO, not the CFO, so I want to be conscientious of it, but just to ask the question that I know is on investors' minds. So in that context, a lot of us have trouble comparing you. We see all these price cuts right now [ph]. Amazon cut here, Google just has the new Google Compute Engine with lower prices. How do you guys make sense of what's going on out there from the price versus the service that -- or solution that's being offered?
Well, so I use a lot of analogies when I explain things, or try to use things that people can grasp. And I like to use, in this case, you know, why does everyone not shop at Wal-Mart if Wal-Mart has the cheapest prices? Why do you guys sometimes go to Target, sometimes go to Nordstrom, sometimes go to Best Buy, sometimes go to, you know, if I'm buying a new TV, I don't always just go to Amazon, I go to the guy that's the specialist in HiFi and stereos and TVs.
So there are different buyers in the market. There are buyers who value the cheapest possible cost. Sometimes if you're a startup, and Amazon has thrown a huge startup program kind of offer at you and said, free cloud for the first year. It's attractive, you know, and developers sometimes need that. But other times when they grow up a little bit, as the companies evolve and have expanded, they find that sometimes Amazon's incremental posted price on the website doesn't necessarily translate into the overall cheapest price or the overall best value.
So we try to differentiate on things that the customer values, not just the lowest possible cost. And again there are things that we do that Amazon and Google and Microsoft just don't do at all.
Give one or two examples of that in particular. I know you have --
Well, so on -- I mean just a really good example is going deep on a database platform. If you don't have a database administrator on your staff and you have a corrupt cable that's causing your application to fail miserably on your biggest day of the year, there's no one in Amazon that you can call to fix that corrupt cable. There is no one that is going to rescue you from yourself. If you delete something by mistake and you don't know how to get it back, there's no one there that is going to -- if you've got the sophistication to set up the Amazon cloud to protect yourself from yourself, then maybe it's okay. But a lot of companies don't have that sophistication.
The biggest poster child for Amazon is Netflix. Netflix has spent a tremendous amount of engineering effort to engineer around the inconsistencies of the public cloud, the ups and downs, the failures that just are a matter of course. But they're unique because they invested in that engineering team. They spent a tremendous amount of overhead to do that. Most companies don't have that level of sophistication and they need to lean on somebody else to do that work for them or with them. And those are the kinds of things that Rackspace does every day.
Well, let's talk about OpenStack. Beyond the, okay, it's a cloud platform it's a management platform. Specifically for these investors to understand what exactly does OpenStack do, what is it needed for from the customer, and what are the other options out there in the marketplace? What does an Amazon, a Google or Microsoft Azure, what do they do if they're not using OpenStack?
Okay. Okay, so, you know, if you think about building a cloud, there is a data center, there's a bunch of servers, a bunch of storage, and then there's some software. The software is the magic of making the cloud into a cloud. The software pulls those resources together. It makes the -- computing a pool of computing, it makes the storage a pool of storage that can be even carved up and served to a customer via an API. An API is a programming interface that the customer uses to control the cloud or write or read from the cloud. Without that software, a cloud is just a bunch of servers. And that's what we used to do. We used to dedicate individual servers to a customer. They will put one operating system on that server and they would build a website.
Now what we do is we rent capacity by the hour and OpenStack in our case is the software that does that orchestration of all those computing resources. It's what tracks which customer is using which virtual machine. It's what gives us the ability to automate the provisioning of servers rather than having individual technicians go out and do it by hand. It's what allows us to scale horizontally on our -- on a lot of the platforms and automate that scaling, rather than having individuals go copy data from server to server by hand.
Now Amazon and Google and Microsoft have all built their own proprietary cloud software technology. They -- Google probably had it ever since their days as a search engine company from the early days. Amazon built it for their own purposes and then turned it into a service offering. Microsoft has -- is the latest one probably to the game and they've built up some of that recently to get into that as their business.
They did have to sort of pivot. Azure 1.0 is very different than Azure sort of today in terms of what cloud computing means. And so they've had to do some rework.
Rackspace took a little different perspective on things and took a different path. We being open source guys from the early days, we believed in open source from our early routes, we decided that we would build our cloud platform in an open source way. So we donated all of our original cloud code that we built to a project called OpenStack. We founded an OpenStack project, then we handed over the code, and then we created a community around OpenStack to allow us to have -- help in building a great cloud platform, and also allow our customers to take that software, and not only run it in our cloud, but run it in their own private clouds as well, in their own data centers.
So today when you look at OpenStack, it's not just Rackspace. It's a little bit -- I mean I'll use another analogy. It's a little bit like, you know, the other guys are running a proprietary platform like Apple does with iOS, we're the Android to their iOS. We're the Android that everyone can use and everyone can build a cloud on and everyone can contribute to if they want to. And they can make it into whatever they need. You know, Amazon themselves uses Android to make their Kindle Fires and Samsung uses Android to power your phones. And those are platforms that now create ecosystems that developers want to build for. And we think what we've done with OpenStack is built an ecosystem of a lot of different providers, a lot of different service and solution providers, but also technology companies -- HP, Dell, RedHat, lots of them, Cisco. A lot of companies are involved now in OpenStack because they see it as an alternative to a proprietary cloud platform.
What really hit home for me is the most recent OpenStack summit was held in Hong Kong, a few weeks ago I was in Hong Kong, and company after company after company from Mainland China came to Hong Kong to present on what they were doing with OpenStack. China doesn't wait for Amazon or Google or Microsoft to come and give them a cloud. They are going to make it one way or the other. And they're doing it with OpenStack. They're doing it with an open source platform.
So even if Rackspace doesn't win all of those deals or those opportunities, we're in the right ecosystem. We're in the ecosystem that is going to be one of the winners. And we are the biggest cloud player in that ecosystem today. So we think we're in a good spot inside of an ecosystem that developers are going to build for, that is going to be a winning platform. And that's what gets me excited about OpenStack.
Wait a minute. So a Chinese company goes and builds a cloud with OpenStack, why do they need Rackspace? How do you --
They may not. They may not. They may, but they may not. So if you're a sophisticated Chinese company and you have the capabilities that Amazon or Google has in-house, you probably don't need us. But again there's a lot of companies that are reaching out to Rackspace and saying, help us with our cloud. Help us build a cloud. Give us your expertise. Lend us your expertise. That's what Workday did with us. They said, we want to have a cloud platform in our data center similar to what you do in yours. We want it to be as automated, we want it to scale like yours does. Help us with that. And that's the kind of service that we offer.
I want to pull in two companies that you mentioned. So you talked about, wait a minute, so this takes a bunch of servers and pulls it together into, you know, so you're giving it up. Wait a minute, I thought that's VMware, I thought that's a virtual machine, what --
It is a virtual machine.
-- what's the difference? And does VMware use OpenStack?
VMware is a contributor to OpenStack, off the back of their acquisition of a company called Nicira. Nicira was a really big contributor of OpenStack in terms of networking and that software-defined networking layer that's inside of OpenStack. Now I didn't mention that up until now, but OpenStack does need a very flexible networking technology to allow customers to stitch together all of their different components into a common network. So obviously you're going to need to be able to create networks for those applications to sit on top of.
And so software-defined networking is an aspect of this, hence, why those guys got involved. Now VMware, how do they play into this?
Well, technically they see OpenStack as a bit of competitor in a lot of regards. I mean it does compete with their sort of core virtualization offering. But a lot of people would wonder, why didn't we just use VMware to build our cloud? Well, they didn't really -- they didn't really understand the cloud market; they were focused on the enterprise. The enterprise had very different needs than a large-scale public cloud. In fact, the developer is who really drove the demand for cloud computing, whereas VMware themselves never really focused on developers. They focused on the CIO and the ops team that ran servers and infrastructure, but they never really thought about it from a developer standpoint.
The developer values the automation and the API that he can write scripts and codes against to automate his world. And VMware just didn't get that for the first few years of cloud computing and they sort of missed the -- they missed that sort of pivot of what virtualization was going to enable. They kept going after that enterprise, enterprise, enterprise, but the cloud went this direction and the developer just never really saw as much value in what VMware was doing.
So that's why we didn't build our cloud on VMware way back five, seven years ago. We just started building our own technology. And today, you know, VMware is still retooling and they've done vCloud Director which was their attempt to go build cloud on top of VMware, but it still didn't quite resonate as much as platforms like Amazon or OpenStack have, and therefore I think a lot of large companies have just sort of taken a step over what VMware has done in terms of building clouds and have moved on.
So let's talk a little bit about RedHat. So you need an operating system, right, to run applications.
Isn't RedHat kind of one of the de facto standards that's out there? How does that -- how does RedHat play in? It seems like Amazon Web Services uses RedHat at certain versions, or you as well.
And is there a difference for customers?
Yes. So RedHat certainly is a popular commercial flavor of Linux. They are by no means the largest Linux distribution in the cloud. Most of the three Linux distributions eclipse RedHat in the public cloud -- ours, Amazon's, others. They are, you know, RedHat's role in Linux has been to make it useful to the enterprise by qualifying hardware and making sure that nothing broke when they upgraded from version. They back/forth bug fixes and patches and those kinds of things. There's a lot of value to an enterprise in what RedHat does on Linux.
There's less value when you talk about a large-scale company like Facebook for example or a company like Google. Google and Facebook to my knowledge don't use RedHat Linux. They use Linux, but they don't use RedHat Linux.
Rackspace's public cloud, most of the customers don't use a commercial Linux. They use free Linux. And it's because in the public cloud there's less need for that enterprise qualification of this particular piece of hardware with this particular piece of storage gear with this networking technology and the drivers that make all that work, because the public cloud is a fairly consistent platform, and all of our servers look alike. All of our storage looks alike. All of our networks behave the same way. Customers can't go in and fiddle with one particular component of our public cloud. Therefore if the Linux works, the Linux works. There's no reason necessarily to worry about somebody kind of giving me that safety net because once the Linux is up and running, it's not going to break in our environment.
So I think RedHat is still valuable to the enterprise market, but there are so many companies that are very different than that enterprise market. And I, you know, I'll just keep thinking back to all of these companies that are building, web software companies, SaaS companies, people that are not the traditional IT department but are building businesses that happen to use a lot of IT. Most of the companies that I would use as case studies, the MapMyFitness guys or the Digital Film Tree, they don't need a commercial Linux, they just need Linux.
Let me bring questions in from the audience as well. Question over here.
So if the key differentiator between AWS and the Rackspace offering is the service component, what's to prevent new entrants to provide that service? I mean Amazon's proven that they don't really care about profits.
Yes. Look, the question is what's to prevent other people from getting into the market.
And let's layer in that, HP, a great example, supporter of OpenStack, HP, big company, lots of resources. Why can't they come in and do everything you do, do it cheaper and squeeze it?
Yes. Well, look, I think that there will be competitors that will go into the service. There will be people that do that as well. I think Rackspace, as a company, that's all we do. This is all we do as a company, service and support and expertise on top of hosting and cloud computing. We're not focused on selling other products or services. We don't have a lot of other lines of business that distract us from our mission and our purpose of fanatical support. Our CEO wakes up every day coming to work thinking about the customers that rely on Rackspace to run their businesses within that context.
I think we're always -- it's a little bit about, you know, you asked the question of why can't they do it? Why did Amazon buy Zappos? Because Zappos had something special, they had a great culture, they had something that Amazon couldn't replicate, so they bought it. They were scared of it. They were scared of that passion and that culture that was behind Zappos.
I think Rackspace and Zappos were cut from the same cloth. If you visited either of our offices, you'd see the crazy people, you'd see the crazy -- and we had it before Zappos existed. We were doing that, you know, we just aren't a household brand so we don't have as much opportunity to showcase that, but that is how it is behind the scenes.
And that translates into a service experience that's hard to replicate. You have to have people that come to work every day and are fired up to deliver great service. You can't have the smartest people but that have a bad attitude come to work every day. They don't deliver a great outcome or a great service experience. We have competed for many years very successfully against much larger companies that theoretically should have been able to replicate what we have. There were the telcos, there were other hosting companies. There were companies that were far larger than we were. But we had built such a strong foundation for service and support and that culture behind the scenes that makes that real that we think kind of, bring it on, you know, bring on your best attempt to do that. And we still think we're going to beat them at that.
Now we may -- they may be good enough. They may have more customers than us. They maybe have things that, you know, maybe a larger reach. But the customer that choose Rackspace choose Rackspace because of a great service experience, a great outcome, services that the other guys don't have.
I mean likely, you know, Amazon, if they got into the business, they would set up some offshore call center somewhere, they would, you know, do it in India potentially. They would make it cheap to lower the cost. But most of my experiences with big call centers offshore that didn't care about me are pretty bad experiences. And our Rackspace customers actually give us really high marks in terms of their experience every day. They have the cell phone number of people at Rackspace, they know people at Rackspace. They create sort of the different kind of level of relationship that a company like Amazon will try to automate or try to just ignore or not do if they can help it.
I think Apple's tagline was fanatical footwear.
I'd like to think that, yes.
One of the things you mentioned though, that a newer element for people to understand is bare metal. What is a bare metal server? Why is it attractive to a customer? And will everybody have a bare metal offer?
Okay. Well, so, yes, that is a new term but it's not a new concept. Bare metal is what we had forever until virtualization came along. Bare metal was one server with one operating system on it. Virtualization took that to one server with dozens of operating systems running on it.
Now imagine if you play out those dozens of operating systems in a public cloud where Customer A could be next to Customer B, Application A could be next to Application B. You end up with what they call a noisy neighbor problem, meaning that this customer's running an application that's actually affecting your application negatively. The other problem is what they call a nosy neighbor problem which is potentially your security risk is such that you don't want other people sharing your server with you.
So performance is one reason, security is another reason, that people sometimes gravitate back to having a dedicated server or a bare metal server where there is no virtualization layer. Bare metal just means one server, one operating system.
So the customers that find that are gravitating to bare metal, they want predictable performance where they're not competing for resources with somebody else. They want reliable performance so that they can say, I want redundancy, I want multiple network connections. I don't want that server to fail. The way that we dealt with failures in the bygone era is that we actually built more redundancy into the server. Sometimes that's valuable to a customer. Instead of building your redundancy across servers which is complex and takes a lot of engineering, sometimes it's actually easier and less costly to just build a little bit of your redundancy into the server itself and not have to build these complex applications that automatically fail over from one server to the other. That is a very complicated thing.
Amazon will not write you an SLA on uptime unless you've built your application across multiple regions of their cloud. They don’t have an SLA that applies until you've got multiple regions. Sometimes customers need an SLA at a much lower and less sophisticated level than that and they're just not willing to step up and write the checks that it takes to build that level of engineering into their application.
Some applications, legacy applications especially, were not built for the cloud era. They don't across multiple data centers. I mean you pick most enterprise applications that are available today running in a data center, they can't run across multiple data centers effectively because they weren't built for that.
Other questions from the audience?
Two areas I want to jump in to really quickly. One, you touched upon kind of the Zappos [ph], Terremark [ph], the guys that are part of the telcos.
How have you seen their solutions evolve or not evolve in the competitive landscape?
Yes. Well, look, honestly when one of our competitors, one of our really good competitors gets bought by a telco, we're kind of -- we're happy about it, because it usually waters down the best of what that company was. A lot of the good people kind of leave the company after they get acquired. A lot of times the strong talent that built that platform technology-wise or built that talent from a service or support perspective, it gets lost in the big giant telco.
So honestly it's always been a good thing, even though they have more dollars and more reach, they oftentimes find that it just becomes sort of an offering that is lost in the massive portfolio that is a modern telco.
And the last area is security. You touched upon a couple of times. For managed hosting and cloud, what do you do for security? Do you have a bunch of firewalls where you protect like the perimeter? Do you have your own intrusion prevention? What is it that you do to provide security? Or is it up to the customer to provide their own security within your cloud environment?
Well, it's all of the above. Everything you mentioned is part of the equation here. I mean security changes somewhat in the cloud, but it, you know, there's a lot of the things that are -- that have to be done no matter what and no matter where the application sits. You have to have good application security, you have to have good code that actually doesn't leak your passwords out on the net. There's nothing that the cloud can do if your code is terrible and it's easy to bypass your security.
So let's assume that the code is good, now what else do you do? Well, you have to -- you like use firewalls. You have to put a perimeter firewall around the environment. So we can do that in the cloud in a number of ways. We can do it with bare metal servers, with a real physical firewall, like a Cisco firewall, and we can do it in a public cloud with a virtual firewall. You have to have intrusion detection systems, you have to have log management. You have to have a lot of the same kinds of things. Sometimes they're just done using software versus an appliance, but it's still very viable -- I mean it's still very required, very much required in that environment. And we attack it with a lot of different tools and sometimes it actually includes Rackspace delivering those capabilities as a service to the customers, like intrusion detection and log management or both services that we offer to customers. But it is all of the above.
All right, great --