When I was a kid, Miller Lite ran an endless series of commercials in which former sports greats debated the merits of the beer. One would assert that the best thing about the beer is its smooth, rich flavor. No, the other would respond, the best thing about the beer is that it's light. They'd go back and forth:
This would go on until the voiceover announcer would settle the matter by noting that what's really great about Miller Lite was that it's less filling and it tastes great. In other words, it was special because, unlike every other beer on the planet, it could square the circle and deliver two previously incompatible characteristics. What was fantastic about Miller Lite was that it combined two contradictory qualities: flavor and low calories.
I'm reminded of those commercials when I hear people talk about cloud computing benefits. One person will say that cloud computing is less expensive than traditional IT, contending that on-demand pricing, efficient provisioning and scale make it possible for cloud providers to deliver IT capability much less expensively than established practices.
After a couple of minutes, someone else will say, "Well, I don't know if it's less expensive or not, but what's really great about cloud computing is its agility." Because cloud providers offer self-service and immediate provisioning, companies can now respond to business opportunities or threats far more quickly than the old, manually-provisioned practices of IT can support.
Both proclaim their opinion as if cloud computing really provides only one of the benefits, that it's either cheaper than IT or works faster than IT. According to these positions, the whole discussion about cloud computing has to focus on their side of the discussion, with the other side not really important at all.
Moth sides of this cloud computing debate are missing the point, though. Cloud computing is special because it provides both lower costs and agility. The only reason it exists is that it provides both.
Agility, Low Cost Inherent in Cloud Computing
The key thing to understand about cloud computing is that is substitutes automation for manual effort. Instead of doling out work to a system administrator, who then manually completes the task and makes the resource available, cloud computing uses resource APIs and an orchestration engine to drive the same task. Therein lies the genius of cloud computing.
Automation also negates the "tastes great, less filling" agile vs. cheap cloud computing argument. Fact is, both agility and low cost are inherent characteristics of cloud computing—and automation is the foundation for them both. There's no way to achieve one without the other, because both require automation, and it's their combination that makes cloud computing a disruptive revolution.
It's easy to understand the intertwined nature of agility and low cost with a couple of thought experiments that consider what the IT world would look like if you could achieve one of the characteristics without the other.
What if you could get agility without an associated low cost? Actually, you could implement agility within the standard IT processes of today. It would just take a much larger IT staff so that people were always available to respond immediately to any service requests that might be submitted. You'd be able to respond quickly to any business opportunity or threat. You'd have agility, for sure—only IT would cost probably twice and perhaps four times as much as it does today, with people spending most of their time sitting around, with no work to do, waiting to respond to sudden opportunities or threats.
It's obvious that no one would ever be willing to achieve agility by doubling or quadrupling costs. So IT implements imperfect agility at a spending level that can be justified by standard business conditions. When a sudden opportunity or threat occurs, IT responds slowly, as it has to slot the new work into an organization optimized for standard conditions.
What about the opposite thought experiment? What if you could drop the cost of IT by 80 percent but it still operated at the same pace due to manual processes? It depends on whether you think there are lots of uses for IT that don't depend on agility and could be addressed if a lot more budget were available. Perhaps there are.
I would argue that an 80 percent reduction in IT cost, even with undisturbed processes, would be revolutionary. Even at the laggardly pace of most IT processes, much cheaper infrastructure would cause an explosion of demand.
Here's the rub, though: It's the very nature of today's processes—manual administration, face-to-face interactions and so on—that cause IT to be so expensive. It's impossible to imagine significantly cheaper IT without figuring out how to remove human touch.
Get over the "It's agile; no, it's cheaper" debate. It's agile and it's cheaper. It's agile because it's cheaper.
4 Ways the Affordable, Agile Cloud Impacts Enterprise IT
So what does this mean if you're an enterprise IT shop? You have a real task ahead of you. You have to figure out how to be agile and cheap, which means reworking your process to pull people out of provisioning.
Here are four things to think about:
First, think about end-to-end delivery times from resource demand to application delivery. It's not enough to put up a self-service portal if all it does is trigger a work ticket. And it's not enough to use cloud computing to streamline the developer's job—you have to streamline the whole process. In the ERP world, this is known as order to cash: How much time passes from when a customer says "I want to place an order" to the point that money hits your bank account? That kind of thinking—ERP for IT—will be critical for the future.
Second, think through what agility means. By definition, you need to be agile to respond to unexpected conditions. After all, if you could predict what you need to do, you could address it methodically. It's the unexpected that demands agility. This requires extra infrastructure resource. And, by the way, these unexpected opportunities or threats? Today they typically have much larger resource demand variation, so recognize that you'll need greater capacity available than you would forecast based on historical demand.
Here's a different way to say this. In the past, you didn't confront variability much, since the lack of agility meant these unpredictable situations never got onto your portfolio backlog. The unfortunate side effect? Most IT organizations fail to understand how the expectations of agility unleashed by the rise of cloud computing is about to crash into IT.
Third, think about the impact of established market rates for IT infrastructure that are undoubtedly much lower than your costs. You will have to match those market rates for your internal resources. Don't imagine that attempting to bundle the resources with benefits such as "enterprise-grade equipment" or "business expertise" will somehow preclude cost competition, either. It won't.
Fourth, think about where your expertise can accelerate agility. Service and contract management. Predefined, tested software stacks so developers can avoid having to locate, install and configure components. Application resource functionality delivered via APIs so new apps can easily integrate with existing apps, making it faster to get the new ones to market.
Getting caught in the agility vs. cost argument is extremely dangerous. The future of IT requires both. This carries implications regarding infrastructure automation and process streamlining. Neither cost nor agility on its own is sufficient. Delivering one or the other, but not both, won't leave the rest of IT undisturbed.
The next five years of IT is going to be exhilarating or terrifying, depending on your perspective. But one thing's for sure: It's not going to be business as usual.
Bernard Golden is part of the Cloud Computing Enterprise Solutions group at Dell. Prior to that, he was vice president of Enterprise Solutions for Enstratius Networks, a cloud management software company, which Dell acquired in May 2013. He is the author of three books on virtualization and cloud computing, including Virtualization for Dummies. Follow Bernard Golden on Twitter @bernardgolden.