While the shift towards cloud computing is an undeniable part of current and future IT operations, it's unwise to plan for any cloud-related innovations without understanding where cloud computing has evolved from, what it used to be and the elements that now make it appealing.
Jeff Kaplan, Founder and Managing Director of Think Strategies, sat down with us to discuss the basics of cloud computing.
A lot of people point to Amazon, which sells virtual servers that you can pay for by the hour called EC2, as the eponymous example of cloud computing. Would you agree?
Jeff Kaplan: Absolutely. In fact, Amazon is, for the most part, responsible for having popularized the idea. Cloud computing was actually most previously promoted by the technology industry itself back in the early part of this decade, when we were referring to it as utility computing and companies like IBM, HP and Sun were doing all they could to complicate the concept with terms like 'autonomic computing.' This was a time when customers were looking for a greater simplicity and ease of use, as well as flexibility and agility from a provisioning standpoint, and the utility computing idea just didn't seem to jell with that.
Then Amazon came along, recognized that it had built some internal competencies that it had the opportunity to commercialize and basically resell in the way in which it provisions its own internal operations, much like Google has done. And with EC2, as you pointed out, Amazon gave it a try and lo and behold, there was a laden market ready to buy.
Is it the simplicity of cloud computing that attracts interest?
Kaplan: I think there are three elements. Simplicity is one. Number two is price. And related to price is on-demand provisioning capabilities, which not only allow you to buy on-demand but to also decommission those same resources almost as instantaneously. And that's really where the innovation has emerged in cloud computing, as opposed to the Software as a Service (SaaS) market, which we at Think Strategies see as a stepping stone to the success that we are now experiencing in cloud computing.
Some people I've spoken with who have worked in IT for a number of years believe cloud computing is the same as time-sharing. What are your thoughts?
Kaplan: Cloud computing is based on the same concept as time-sharing. I've heard time-sharing referred to as shared services before. It was first talked about in the 60s and 70s. The aerospace industry had excess capacity, much like Amazon does today, and tried to resell it. Those companies didn't end up commercializing it to its ultimate success; it was an innovator in the form of EDS that emerged at that point.
The same principles hold true today. If a primary vendor has figured out a best practice for managing a data center and can, in essence, resell that competence, why not allow third parties to take advantage of it?
When talking about the economics of cloud computing, what type of businesses would you say are going to see value from using these services?
Kaplan: Companies of all sizes and shapes and in various stages of their evolutions can see value. Obviously, startups are a key target and the most obvious example. Small businesses, if you look at them in a more general sense, can also take advantage because they have an opportunity to see a benefit. But the real surprises come from another category, which is the enterprise. Most people felt that SaaS and now cloud computing would not sell well into the enterprise, and the enterprise has proven them wrong.
Enterprises have found that there are tremendous opportunities to take advantage of with SaaS. Now they're increasingly looking at cloud computing as well, initially as a test-bed for new applications, a place where they can more economically pilot new ideas. They're 'kicking the tires' to determine whether or not these cloud computing vendors can really scale to meet their ongoing productivity or production needs. They're also looking to find ways to offload or what we call 'outtask' ongoing pieces of their operations to these cloud computing environments. For instance, there may be business cycle reasons where a company needs to scale up its operations periodically but doesn't want to make the permanent investment in additional hardware and software, which would lock the company in to a capital investment that it doesn't need throughout the year.
Is this the beginning of a period where computing power is becoming like a grid of resource to be used like electricity -- generalized and easy enough to consume, without having to make an investment?
Kaplan: Absolutely, but there is another aspect of this. If I am an established company that has made a tremendous investment in IT, even if that IT is working for me today (although I don't know of too many organizations that can say that they are totally satisfied with their IT operations) it doesn't give me the competitive advantage it used to give me. Today, anyone can access Amazon and get some computing power, obtain a myriad of enterprise applications on a pay-as-you-go basis called SaaS and combine it all with on-demand resources. There are a number of ways to get to market faster, easier and more economically and become a competitive threat to established firms. The ways in which the barriers to entry are being knocked down by cloud computing are enormous.