Recently we had the real pleasure to sit down for an extended conversation with David Linthicum, SVP of Cloud Technology Partners and one of GatheringClouds.com’s favorite cloud thought leaders. David writes for a number of online publications, including GigaOM, TechTarget, and Infoworld, among others. Our discussion covered a range of topics including David’s approach to publishing, cloud washing, how cloud has changed the nature of IT and more.
This is the first of 2 parts. Check out Part 2 here. Let me know your thoughts on Twitter @CloudGathering.
Gathering Clouds: As a key thought leader in cloud, what governs what interests you and who are you trying to write to?
David Linthicum: What governs what interests me is people trying to solve this problem or issue of cloud internally. So, I always imagine that I’m writing to a very confused IT staffer that’s been given the edict to go figure out a way to get the company into the cloud and make things cheaper and better, per the articles the CEO has been reading on airplanes. So, it’s that person who I think is probably a good 80 % of potential cloud users out there that are probably 90 % confused.
And it’s important that you ground stuff and provide some practical advice, in terms of a how-to in words they can understand. For example, I will write about how to move existing legacy systems in this direction. Or when you actually make the move, what’s real versus what’s just fantasy out there in terms of the emerging cloud computing technology. Another important aspect is trying to put a voice to most of the people who are looking to get away from the “hype” surrounding cloud. A lot of people in this space are saying “cloud, this” or “you got to go to cloud.” For example, I just received a PR blast that said “Enterprise IT is dead.” Well, there you go. The marketing teams are taking the concept of cloud computing in new and confusing directions.
But are you going to move in the cloud direction without throwing the baby out with the bath water in terms of the value of IT, as well as not making a lot of major mistakes? Probably not, and the reality is somewhat different than perception, based on the hype out there. I’ve been in this business a long time, so I try to bring context to it, so people can kind of understand exactly what they should be working on and should be concerned with right now, versus just kind of marching where everybody else is marching; sometimes, off a cliff.
GC: So, from your perspective, what is causing the confusion? What’s causing the sort of opacity of what cloud computing brings in terms of value? It seems to me that whenever there’s some sort of hyped movement, whatever it is, especially, in business context, there’s a level of marketing buzz that goes along with it, some of it helpful, some of it not so helpful. But does that make or break how the message is disseminated?
DL: No, it doesn’t necessarily break it. And it’s an old problem. I’ve been through this sort of change many times. It was initially the PC stuff: PCs will save us and become a lot cheaper. When I first got a job, PCs weren’t allowed in the building.
And then, the market overhyped the value of PC. We came to the reality of what these things are, in terms of tools, and the practicality of the technology eventually. Of course, after that, it was distributing, computing, PC-based databases, local area networks, and client/server. Then, there was the great rise of the Internet, and with it the rise of service-oriented architecture (SOA), and now, cloud computing. With each one of these waves, there is a parallel revved up marketing machine that is always going to be there. I’m not necessarily pushing against it. I’m just trying to get people to understand the reality of the situation before the other marketing folks just kind of take the technology concepts and drive off with it.
I wrote a blog post in InfoWorld last year that is still setting all kinds of traffic and readership records. I argued that cloud computing was meaningful before Gartner defined it as 100 different things. So, if cloud computing means everything, it really means nothing.
I think that cloud computing is built around a few basic, specific concepts. It’s how we consume technology and how we self and auto provision resources, how we share infrastructure. All of these attributes together become the nut to crack for businesses to get maximum value from the cloud. So, enterprises need to keep a focus on that and start moving in that direction, versus enterprise advertising-as-a-service and all these other things “as-a-service” that Gartner has in their adoption curve.
Cloud computing is basically new models for consuming the same kinds of resources that we’ve been consuming for years and years. However, it’s going to take an architectural shift in how we deal with technology, as well as a different way of thinking around how we build those databases, deploy the applications, leverage storage, and all the things that we’re used to doing. So, the days of going off when we had a problem, buying some new hardware and software and dragging it into the enterprise and standing up yet another silo, are hopefully coming to an end. We’ll be able to share infrastructure, share information, and get a lot of problems solved that, quite frankly, we’ve been making worse over time.
GC: So, do you see the value of the cloud being the same as it was when you published that piece a year ago? Have you seen that value shift? How do you define that value overall?
DL: I think the value’s the same. What I did in the cloud space since ’99 wasn’t called “cloud computing.” We were calling it just “internet delivery of enterprise IT resources.” I just put a paper up online that I wrote back in 1998, which was the basic blueprint of an infrastructure-as-a-service (IaaS) cloud. I wasn’t really attempting to make the point that I was thinking about this space before anyone else, but rather to highlight that it’s been an evolving science. The ability to get to a state where very expensive resources can be shared amongst millions of consumers is amazing. And that’s what cloud computing promises: the ability to build out very successful and very efficient compute resources on other people’s technology that you can access as-a-service. Therefore, if you’re focused on the economies of scale, you end up with something that’s going to be much more advantageous, flexible and agile to serve the business. That’s really the core objective of what we’re looking for. As long as we kind of keep our collective sights on that, I’m okay with everything and anything coming into the space.
I think where people have a tendency to get derailed is when cloud computing suddenly became popular. Everybody wanted to get into the space. All the major software and hardware vendors just, in essence, just renamed everything as a “cloud.” And that’s where the confusion came in.
There was a recent survey that found that most people who are consuming cloud-competing resources are finding that people are BS’ing them in terms of what the technology being bought really does as far as cloud capability. They’re not really providing a cloud-based resource. Auto provisioning, self-provisioning, it’s just a piece of software where some companies happen to scrape off the enterprise names, stick in cloud and now, they’re trying to sell it back to the enterprise IT guys who often have to, quite frankly, play private detective to figure out what’s real and what’s not real.
If you look at what I write about, it’s really getting into the functional business level conversations, calling out those sorts of falsehoods and then making sure people understand the details behind it. In other words, what they need to do to be successful in the cloud. It’s not really saying that this is a bunch of crap. Instead, I write about ten steps you need to go through to understand the true value of the cloud resources out there, or here are five things you need to do to evaluate whether or not these technologies or services are usable or not. And I think that’s what’s missing.
There are very few people out there that are doing it this way. Everybody else is kind of grabbing pom-poms. And I don’t think it’s a good time to grab pom-poms. Or perhaps, everybody shouldn’t grab pom-poms. Someone should be in there trying to make sure we’re making the tough decisions and the tough calls. Eventually, it’s going to come around to that anyway. People are starting to deploy these technologies right now. And they are making mistakes because the reality of what they bought and what they’re trying to do are very different than the perception that they had when they got into it.
GC: On the topic of cloud washing, as you are discussing it, how prevalent do you see the practice being? There are plenty of actual cloud providers and cloud technology providers. But what governs why a company might go to a provider that is actually cloud washing over an actual cloud provider? There must be some sort of due diligence on the part of a client who wants to consumerize this product. Or are they really just going along with what’s being sold to them?
DL: No, they’re actually being very smart and doing their due diligence. I mean I do that for a living. So, my whole mission in life, as I was writing the blogs, is working with the Global 2000 and the government to help better screen technology providers to figure out what they’re trying to do and trying to sell. That’s a good time to gather my data points as well.
But it is a problem. And I think it’s going to continue to be a problem, simply because the large technology providers can’t turn on a dime. And they are, in essence, trying to survive in this new IT paradigm. Quite frankly, it’s going to take years before the technology is ready for it to be a true multi-tenant, publicly-provided solution. So they’re in there trying to razzle-dazzle potential clients right now. And to be honest with them, I’ve been on the razzle-dazzle side of things when I was a CTO, and I understand how that works. For example, you’re a tech provider who is under funded, and you don’t have enough resources yet to make your technology cool. Therefore, you have to “fake it until you make it.” And that’s what they’re doing.
The enterprises are actually better than I thought at finding these things out. I think they’re good at understanding what’s going on via the survey, like I just mentioned. And it sounds like people are confused, in many instances, but they’re making the right calls and they’re taking very slow steps to make sure the technology is the right technology to get. But that doesn’t mean someone’s not going to buy something that they consider “cloud.” Virtualization is the biggest example of that.
I have this one conversation a couple of times per week, where a client will say that they don’t know if their business should go to cloud. They ask what cloud can do for their bottom line. And I always hate those questions. But I’ll go through it in terms of the holistic reasoning down to the strategic vision. And often these companies will say they have virtualization, therefore, they already have cloud. Well, it’s completely different, in terms of concepts. And that goes to the confusion out there, in terms of what cloud computing is and what its value is, and ultimately how it’s different from the traditional ways of doing computing.
So, it’s a bit of me being the designated buzz kill, and sitting up there telling people what I think they need to hear, in terms of how to make your way in the cloud to understanding exactly what’s real and what goes into allowing the technology to become more efficient and agile, while becoming less expensive compared to all of the things you’re trying to do versus just buying another version of the same old enterprise technology we’ve been using for years.
The other important thing is to understand how this stuff exists in the context of strategic planning and architecture. That’s the biggest piece missing in companies’ thinking. It’s one thing to tactically just go out, sign up for cloud storage with a service provider and then link them into your existing infrastructure. It’s another thing entirely to try to figure out how that use of cloud computing goes into a longer term vision, how much benefit you’re going to bring to the business with that vision, how you’re going to execute towards that vision, how to better your infrastructure a bit at a time over the next few years. And that’s where people typically stub their toes.
We don’t like strategic planning and architecture in this country. It seems to be a European and almost an Asian concept. But here in the US, we’re so used to going and buying something to solve the problem that we typically don’t think about how that solution is going to exist in the larger strategic vision and planning of the business or the business of IT. And I try to focus on that as well. My last three books are on that exact topic.
GC: How do you see cloud having changed the way businesses leverage IT, as well as where IT sits in that given business? It seems to me it places much more of an onus on innovation and a proximity to revenue in the hands of IT than, perhaps, it did maybe ten years ago.
DL: Yeah, quite frankly, IT’s getting their feet held to the fire. Ten years ago, it was a freaking party. You could spend as much money as you wanted, as long as the business was doing well. The internet was driving a lot of that. And so we had to change our architecture to take advantage of this new technology. And suddenly, the bubble burst and budgets were slashed. IT, in many instances, are considered profit centers, as well as cost centers. And so when I talk to people today, it’s not a conversation around experimentation. Businesses are very adamant about making the right decisions and moving IT to a position within the organization where it’s going to directly benefit the business. And companies can both create a business case before they leverage the technology, but also, report a business case after they leverage the technology. And both paths are going to be successful. Oh, and by the way, management is auditing them. If they’re not performing and providing a benefit to the business, then, they’re fired. This happens all of the time.
So, it’s a very different world than it was 10 or 15 years ago when we were, in essence, leveraging technology to strategically direct the company. Now, people are very adamant about that technology providing value. And that has paralleled the rise of cloud computing. I think part of the reason that there’s an interest in cloud computing is because the pressure is being put on the IT organization to do a lot more with less. And there are only a couple of ways to do that.
Number one, you start reducing the services you’re providing your business but if you do that, you’re fired. Number two, you figure out cheaper ways to bring those same services to your business, which if you’re slick and innovative, there’s ways to do that. Cloud is the newest, shiniest tool in the shed that can make that happen.
I recently wrote an article on how cloud computing can help healthcare. Healthcare is now facing huge regulation challenges. No matter where you sit on Obamacare, it’s going to change the processes in how organizations deliver healthcare. And they’re typically the least funded IT organizations out there. So, what are they going to do? Well, they’re going to have to either “get with it” and figure out newer, cheaper and more effective ways to do it. Or they are, in essence, going to fail. So this is a place where cloud computing “saves the day” because suddenly, businesses are able to buy infrastructure that they’re already paying for today, the data center space, and all of the traditional ways of doing it. I’m not going to make the argument through economies of scale and cloud computing. But from a utility point of view, it’s going to be a lot less expensive and a lot more reliable even than the existing technology they’re leveraging.
If other businesses are thinking about how healthcare is able to reengineer itself, that’s a huge use case for cloud computing. The problem is, that enterprise IT is a very conservative group. They don’t like their data on a public cloud. But eventually, they’re going to have to get over that. They’re going to have to figure out that the security mechanisms and ways to deal with privacy concerns, regulatory pressures and compliance issues, do sit in the clouds. Because I just don’t think they’re going to have any choice.
GC: So, for the way that cloud is changing businesses, it’s similarly affecting the C-level technology offices, whether the CTO or CIO. Is cloud a blessing to them? Or does it potentially create a situation where they’re managing a lot of desperate infrastructure usage because different groups within an organization can have access to different sorts of cloud usage. Does it threaten them? Does it mean that they have to evolve? Or is just sort of a necessary evolution of how the technology impacts a role over time?
DL: That’s a great question for a couple of reasons. Their perception is that cloud is a trap. They won’t say that in meetings or in front of their boss, but if you get them out for beers, they’re going to say, “What’s my job going to look like in five years, once we implement this technology?” The reality is that it’s going to have very low impact on IT leadership and even IT staff positions. We’re not going to make as many data centers that are owned by companies. We’re going to have larger data centers that are shared through cloud-based infrastructure. We’re going to have different roles within these organizations, probably more people dealing with security and compliance issues as well as cloud managers.
But take a step back: we’ve been changing IT roles for the last 30 years. We’re going to change roles around the use of cloud computing, much like we have during other waves of emerging technology. For example, the rise of the PC, complex distributing computing systems, and certainly, the internet was a change of roles and responsibility. We had to create new roles for new technology, and change many of the existing roles to adapt to the evolution of the technology. Cloud computing is just another evolutionary step. I don’t think it’s necessarily a revolutionary thing, which is going to change things over time. So, I don’t think a ton will change from a macro/leadership level. The tactical roles will change, as they always do.
There will probably be as many IT people needed within organizations, perhaps more. But, a lot of those jobs and positions will move out to providers or they’ll be recast or they’ll move up the stack. I don’t think it’s going to be this huge deficit that everybody’s planning for. And plus, it’s going to take a long time. Cloud computing first became relevant in 2007/2008. Here it is 2013 and there are a few projects out there. People are leveraging it. But Amazon Web Services (AWS) is still just a billion-dollar company. They’re not a $10 billion, $15 billion company. So, we’re nowhere near the level that International Data Corporation (IDC) was predicting two or three years ago. And it’s going to take a very long time to evolve in that direction. It’s growing, but at a reasonable rate.
So, those fears are largely unfounded. I don’t think the day, when people come into work and realize that everything’s been outsourced to AWS or Rackspace, is going to happen. Perhaps we will get, if we’re lucky in the next five to ten years, 20% of our processing outside the firewall. And then, some of it even moved to private cloud or hybrid cloud infrastructures internally. It’s not going to be this massive push that everybody’s thinking. You just can’t move that fast. It’s too expensive and risky.
GC: Is the enterprise the ultimate goal for outsourced cloud providers? And if that is or isn’t the case, what is it going to take to break cloud at that scale?
DL: Yeah, that is where the value is to be brought. Enterprise systems are largely inefficient, largely siloed, and very difficult to deal with. When considering moving them out into the cloud, it’s important to remember that they’re very static, not agile. And that’s where the cost comes in. Cost really doesn’t come in to maintaining the system. Cost comes in when the CEO goes to the CIO and says, “I want to buy another company. How long will it take to integrate the two systems?” And the answer is five years. The CEO will say “I can’t buy the company if it’s going to take you five years to integrate the systems.”
In a cloud enabled company, if you get into a service-based infrastructure and start moving some processes and key services into the platforms that are outside the enterprise, they would be able to communicate effectively. With cloud, it’s really easy to change that infrastructure, to change that architecture to adapt to any kind of core changes, business processes, including an acquisition. And that’s where the money’s made. It’s the ability to kind of get into new businesses and opportunities, versus saving a couple bucks here and there.
A lot of the enterprise architectures that I see are severely broken, typically, through lack of planning. They’re very siloed and you can’t get the information out of them that you need to drive the business or change them around core business changes. Cloud makes infrastructure much more changeable, more cost effective. And it’s much easier to deal with, in terms of culling information from it. You need to drive your core business processes for the simple reason that they’re typically newer systems, they’re service oriented in nature, they’re able to scale out too much larger data capacities. And for the type of storage and processing capacity a business needs, cloud computing just going to provide more benefits. You’re able to scale up when we need it. It’s a fraction of the cost of traditional IT tools. It’s also more reliable than most internal systems. Forget all the outages that make the newspapers. How many times are internal systems down? And provider’s technical teams seem to be better at running this infrastructure than most organizations are.
The other evolution is that the newer enterprises are out there. We have a tendency not to think about those because they’re not in the Global 2000 yet. But I’m seeing some enterprises that have come along in the last ten years that have 80% of their stuff out in the cloud. And as those guys get bigger, they’re going to have the core strategic and cost advantage over people who are not in the cloud. And that’s going to be a core selling point of cloud computing, just kind of watching those guys take off.
I used to be CTO of Grand Central Communications, which is an early cloud company. We had five or six clients that did everything on the cloud. I mean, they’re dinky companies. But it was amazing to me that they had absolutely no hardware, other than some PCs in their offices. And everything else was outsourced. And they were on Salesforce.com and Grand Central, of course, at the time. And a few other storage-as-a-service providers that were in the early days at the time. It was back in 2003-2004. Old school cloud for sure.
GC: Do you believe a larger company can be 100% cloud? Is it feasible to completely rent? Or is there a point at which you’re meant to own?
DL: Yeah, I think that right now, there’s no reason you can’t leverage public cloud services for the majority of your infrastructure. I mean, do I think we’ll hit 100% of where they’re not going to want to bring some stuff in house? No, I think there’s always reason to keep stuff on premise or to own something. A technical reason, business reason, compliance reason, security reason, privacy reason, and so forth. But there’s no reason that we can’t have the majority of a lot of the compute that occurs within enterprises today out on the existing set of public clouds. Going 100% cloud may be a bit of a stretch, but I find it’s never going to be correct if we say something’s going to be absolute.
But I truly think in five years we are going to see some medium size or even emerging businesses that will have the majority 80, 90 % of their processing occurs on a public cloud environment. And that’s because they kept their teeth on it. It’s in their culture to go off and do that. And they’re going to leverage other clouds whenever they can. There are going to be some instances where there has to be on-premise systems for all the reasons that I just mentioned.
And I think we’re also going to see that those businesses will have a key strategic advantage. So, the companies who can turn on a dime, they’re the ones who can tell you why people are buying their products and for what reasons, and ultimately how it relates to the terabytes of stuff that are appearing on social networking trends. These companies will be able to leverage their information in much more innovative ways. And by the way, the bill for IT is going to be one half of what their closest publicly-traded competitor is paying with a traditional IT shop.
You know the traditional IT organizations are going to be asked to start moving increasingly in that new direction. And that’s going to be very much like it was distributed computing. Or else, we’d still have these massive, expensive, ugly mainframes today that I had to maintain when I was in college. I hated those things. Compare those mainframes to these very complex, powerful distributed computing systems, internet-based computing systems which drive IT departments today. It’s an evolving thing.
By Jake Gardner