Remember those early days of the cloud when the IT press ran article after article debating if the cloud was really cheaper than on-premises IT computing?
Part of this was pushback on all the relentless marketing telling the world the cloud saves gobs of money. Those that took the bait often found the opposite—hidden costs snuck up and those cheap cloud apps turned into money pits. Some even reverted back to on-premises computing— not always an easy task.
In 2014, Gartner analyst David Cappuccio found that close to a third of SaaS customers went back to on-premises computing. The CompTIA IT Industry Outlook 2014 report similarly found that 24% of companies that moved apps to the cloud moved them back to on-prem.
That backwards migration still happens, according to a piece in Solutions Review, but nowadays apps that move off the public cloud simply revert to a private cloud.
Cloud vendors today still extoll the economic virtues of public clouds. What they don’t brag about are hidden or surprise charges, and costs that rise with usage and data transport. Many of these hidden charges are not the cloud vendors’ fault, it is that IT doesn’t manage cloud workloads and pay for capacity they simply would not need if they kept their eye on the cloud ball. The lesson here is that a well-managed cloud is almost always cheaper than on-premises.
That is actually saying a lot. For a well-tended cloud to be cheaper, it has to be a real bargain. In the on-premises world, the price of the computing (like servers) and storage hardware that keep on-prem software running is getting cheaper all the time. This is all courtesy of Moore's Law, which has been inviable since postulated way back in 1965. Brainchild of Intel genius Gordon Moore, the law maintains that processors, through ever-shrinking circuitry, will double in power every 18 months. That leads to two outcomes: either the price of computing hardware keeps shrinking, or what you get for the same price doubles every year and a half.
This tension between the unpredictable costs of poorly managed clouds and dramatic boosts and equally dramatic cuts in the cost of on-premises has kept the great cloud debate very much open.
Indeed, if the cloud is not done properly, it is not necessarily cheaper—and in fact can be more expensive. These days, it is an increasingly rare exception to the positive cloud economics rule.
We are here to settle that debate: The cloud is almost always cheaper—if done right!
And smart IT pros are learning to do the cloud right. They are learning through experience (and from their mistakes) how to migrate, what to migrate, how to smartly and economically adapt existing workloads to the cloud and keep a keen eye on managing expenses and crafting advantageous contracts.
These IT pros size their cloud properly from the get-go—but more importantly, tightly control workload expansion without giving up a lick of function. That is what we call a properly governed cloud—offering all the benefits of worry-free computing and less cost to boot.
While microprocessors continue their march to greater performance glory, cloud economics are likewise advancing. Processors gain through increasingly smaller circuitry, while the cloud gains by going in the opposite direction—getting bigger and thus benefiting from incredible ever-improving economies of scale for SaaS vendors and cloud providers alike.
Both now enjoy massively efficient server farms that push virtualization to the limit, taking maximum advantage of every computing dollar spent.
With these economies, cloud services can be increasingly affordable, and at the same time offering easy, near limitless scalability that simply can’t be matched by on-premises infrastructure.
All the economies of scale in the world won’t lower prices unless there is competition for your cloud dollars. Thankfully, the cloud market is hotly competitive with the big cloud infrastructure and services guns—Microsoft, Amazon, and Google battling for every penny.
SaaS providers are fighting the same war, going tooth and nail to land your business. All cloud services enjoy the same aggressive competitive dynamics—from bare-metal Infrastructure-as-a-Service (IaaS), richer Platform-as-a-Service (PaaS) and a layer above that, Software-as-a-Service (SaaS).
One of the biggest cloud selling points is moving from the locked-in, money upfront capital expense (CapEx) model to the operating expense (OpEx) approach which offers greater economic flexibility, and when the cloud is done right, predictable and attractive monthly pricing.
Not only is OpEx generally cheaper, the cloud is far nimbler. Cloud services advance at warp speed and IT enjoys innovative new features without new infrastructure investment. Contrast this with the on-premises approach where IT locked into (“stuck with” is sometimes a better way to phrase it) hardware infrastructure for three to five years, or until fully amortized. All the while, computer technology is racing ahead, making all this hardware yesterday’s news.
Trying to keep pace with processing needs is like Sisyphus: Once your infrastructure boulder makes it to the top, it comes rolling back down as application needs once again burst. With the cloud, IT no longer struggles to scale IT operations.
With on-prem, more capacity only comes from more server and storage hardware. This gear has to be bought, installed, configured, patched, maintained, and fixed when broken. Not just that, but your gear and the software that it runs on must be defended from cyber criminals.
And you have to run all this stuff, with larger server rooms costing a king’s ransom in electric and cooling—never mind the cost of real estate.
Swapping all this for the cloud means you can scale at will. If computing needs increase all of a sudden, you can handle that easily with cloud scaling, bursting, or other ways of acquiring additional cloud processing.
And what about security? Can you trust the cloud? Cloud providers, with all that scale, can afford scads of top tier security gurus and splurge on layers and layers of the best security software money can buy—and skillfully apply their expertise and security hardware and software across their infrastructure.
Not all IT work is created equal, and let’s face it, some tasks are more fun and satisfying than others. The cloud, simply put, does away with a lot of thankless grunt work, replacing it with strategic, value-driving tasks.
This changes the entire way your IT staff is structured and deployed. Instead of paying for staffers to patch and update servers and fix them when they inevitably break, your people can focus on strategic, business-building work. Instead of buying new gear, how about grabbing up a cool new app that gives your company an edge? Where do all these hot new apps run? IN the cloud, of course!
Amazon Web Services (AWS) was the solution many turned to in the early days of cloud migration, and at first these pioneers patted themselves on the back. Before too long, the bills got bigger, and soon instead of saving money the cloud became a money pit. Some cancelled their AWS contracts and moved back to in-house computing—not realizing there was a better economic solution.
The trouble was AWS workloads increased, which is in part due to the natural increase in processing as applications grow. The real problem, though, was a failure to properly size workloads and smartly govern the AWS environment.
Providers such as AWS took notice, modified workload costs and pricing models, and helped customers size their workloads more properly.
With more flexible pricing and deeper IT cloud savvy, the cloud is realizing its economic potential, and the question of whether the cloud is cheaper is increasingly answered yes. The lesson is to apply proper governance and implement tight controls, and that way avoid the cloud cost pitfalls that beset early adopters and still haunt those with a laissez-faire cloud attitude.
What to ensure better management control over core business processes? Try a free trial of MOVEit Cloud today.
Doug Barney was the founding editor of Redmond Magazine, Redmond Channel Partner, Redmond Developer News and Virtualization Review. Doug also served as Executive Editor of Network World, Editor in Chief of AmigaWorld, and Editor in Chief of Network Computing.
Let our experts teach you how to use Sitefinity's best-in-class features to deliver compelling digital experiences.
Learn MoreSubscribe to get all the news, info and tutorials you need to build better business apps and sites