OPEX versus CAPEX: How the Cloud is Changing the Way We Think about IT Spending

Many years ago, my first boss had a habit of walking around the office shouting, “GET YOUR HEAD OUT OF THE CLOUDS.”  I haven’t talked to him in a while but I’ve got a feeling that he may be rethinking that business philosophy when it comes to IT expenditures.

Technology was, for the most part, a return on investment play: we invest in this infrastructure capital (CAPEX), it costs a certain amount to maintain and run (OPEX), and we can produce some income with it. As we enter the Cloud Era, we’re already consuming Software as a Service (SaaS) and Infrastructure as a Service (IaaS), which means more companies are shifting some of their biggest expenses into the OPEX bucket.

OPEX Makes Budgeting Simpler

Every business wants to know if it’s making money, but calculating return on investment with complex capital and operating variables is a huge challenge; even if the initial calculations are correct, it’s a constant struggle to validate whether those returns were realized.  With the OPEX model, monthly recurring expenses are right in front of you, easily calculated, predictive, and provide more reliable inputs to profitability models.

“We’re already in the new era of transition to OPEX, but it will be a long transition. There are a number of challenges to overcome on the path: applications, process, security, emotion,” says Jonathan Butz, a VertitechIT Executive Project Officer and Healthcare Strategist. “Many businesses are making aggressive moves with transitions measured in quarters; others in years. Healthcare is probably further out, but Cerner and Epic are already delivering hosted EMR as a service.”

The transition is most visible in the startup community: Butz says that among sixty companies attending a recent event Expo in the startup community, not a single one owned any infrastructure. “They have cell phones, laptops, and services,” says Butz. “That’s it. If they’re building apps, none of them own the server on which their code runs. None of them owns email or file servers. Everything they need is IaaS and SaaS.”

Having most of your expenses in the OPEX bucket can be appealing for venture capitalists as well. VCs don’t want to invest $10 million in a startup to buy infrastructure just to prove the viability of an idea. Instead, they can use Amazon Web Services (AWS) or Microsoft Azure and pay for their use, able to pull the plug at any time without unused capital that cannot be recovered.

Butz adds, “it’s really enabling the startup community to consume resources as they’re needed, which is a huge boost to new ideas, allowing for an unprecedented era of innovation and experimentation. There are thousands of little tech services out there functioning and delivering value that would never have existed without this possibility.”

Decisions, Decisions…

Take email: it’s almost impossible to justify hosting email on-premises anymore. Companies like Microsoft, make it easier (per user/per month fee) and eliminate the need for servers and the personnel to keep them up and running. For some regulated or particularly security-sensitive entities like law firms, cloud-based email may be challenging but even these issues can be addressed through supporting technologies.

AWS is a powerhouse in the field of big data and analytics. “If you want to do big data analysis or genomics research in-house, you can do it if you’re huge and well-funded,” Butz says. “Those companies can get a return on that kind of capital, or at least fund it with grants. Smaller or regional health facilities often cannot afford the capital needed to do these kinds of projects – you can only get a return if you use them all the time throughout their useful life. So cloud is the only option.”

OPEX vs. CAPEX:  Do the Math

For every use case, there’s a rational economic assessment any organization can perform to determine whether consuming in a CAPEX or OPEX model makes sense. You can run most technology either on- or off-premises, and there’s a point of break-even between the two. The OPEX model costs more per individual transaction; it’s likely to be a little more expensive per terabyte stored, data transferred, per unit of compute – therein lies the margin of the company delivering the service to you. But there’s a point of utilization for capital that could make the on-premises option more attractive. You have to do the math.

In addition to the economics, Butz says there are some key questions to ask in determining whether the OPEX or CAPEX model is the most beneficial option:

  • Is there anything about the nature of this infrastructure or operational competence relevant to your business – that provides a competitive advantage?
  • How sensitive is the information? Is there a relative benefit to having it on-premises vs. moving it off?

However, he points out that we’re rapidly approaching a point at which the security concerns surrounding the cloud are less valid, or perhaps not a concern at all. Microsoft, Google, and AWS are all building infrastructure to allow companies to comply with as many standards as possible, from HITECH to HIPAA and more. Whether your applications do comply with those standards is still driven by your use of the service, the architecture and governance you develop and maintain.

Not all apps lend themselves well to the application models available. Many industries have legacy apps that can’t simply be moved to the cloud with the expectation of successful operations – but, there are a lot of hosted offerings in the marketplace capable of delivering whatever kind of architecture people need to suit the demands of their applications, which they can pay for month-to-month and consume as a service. Determining the right choice is a matter of examining the needs of your applications, your users, and the economics.

“We’re in another great era of transformation, an era of the most substantial change in how we use and consume technology in the history of IT,” says Butz. “The shift from the mainframe to client server and desktop computing was also a significant transition, but we were still buying and operating assets on-premises. Mainframe time-sharing was a thing once, but that faded with the rise of client-server. This is the first time we’re able to go completely back to an OPEX model, marking a major transition in the way business is done across many industries.”

So if you’re an executive and happen to notice that guy or gal in the corner looking out the window with their head in the clouds, offer encouragement.  They may have an important contribution to make to your bottom line.