Countless words have been written about cloud computing economics. The catchphrase is summed up as "OpEx vs. CapEx," shorthand for rent vs. buy, with an ongoing and endless vociferous argument on the topic.
Most of the controversy surrounds whether it's less expensive to use pay-per-use resources offered by a cloud provider or if it's cheaper, in the long run, to use purchased assets owned by the user organization.
As you might expect, the position to which one adheres usually aligns with whether they use public cloud computing or on-premises infrastructure. There's also a third-party to this argument, which claims that it's the agility of cloud computing, not the economics, that's important.
Frankly, I've always found the last position untenable. The real issue is agility at what price; if agility's too expensive, no one's going to take advantage of it. After all, it's easy to maintain that, really, it's the comfort of a car that's important. It's only after realizing the cost of a top-range Mercedes that most of us opt for less comfort but a more affordable monthly payment.
You Need to Know What Each Use Costs
As I wrote a few months ago, affordable agility is cloud computing's real selling point. However, the OpEx vs. CapEx debate fails to capture the real revolution in IT economics that cloud computing has brought: The "pay-per-use" model forces an understanding of the real costs of offering IT resources.
That's because pay-per-use requires three things:
- Transparency.Say what you will about Amazon Web Services, but it posts its prices right on its website. This discomfits its competitors, who are more accustomed to a knowledge advantage that's leveraged to extract higher margin from smaller, less-powerful customers.
From the point of view of IT organizations, this means there's pressure to unbundle offerings and provide more granular resource offerings. In any case, what AWS has brought forward is an expectation that it should be easy to find out what something costs without needing to engage in a lengthy discussion and/or negotiation.
- Benchmarking. Once an established price regimen is available, users will compare other offerings to it. From a positive perspective, this lets other providers understand what they must meet; from a negative perspective, this creates competitive pressure for competitors to get up to scratch.
- Preparing for diffusion. Many IT organizations respond to AWS economics by creating private clouds and assigning public cloud-comparable prices to the resources provided internally. However, it doesn't take Carnac the Magnificent to realize that, inevitably, users running legacy applications in the same internal data centers will soon demand the same transparency and cost-competitiveness. With that will come the associated pressure to deliver non-cloud resources at the same costs as offered by the private cloud.
Put another way, many IT organizations are offering private cloud pricing at competitive pricing without really knowing whether, say, a virtual machine offered at $.10 per hour really costs $.10 or less; they just offer them, figuring that the overall cost of the private cloud is so tiny compared to the total budget of IT that any cost inefficiencies are immaterial. However, once all users want the same kind of transparency and prices, there's nowhere to hide on understanding real cloud computing costs and being prepared to meet available benchmarks.
IT Economics Mean Defining, Benchmarking Costs
If you're an IT organization facing this new world of transparent, benchmarked and diffused economics, what should you do to remain relevant? Here are four suggestions:
Understand your economics. Five years into the cloud computing era, most IT organizations still have very little understanding of their costs to deliver resources, especially at a fine-grained level. Many still suffer from the fact that portions of their budgets roll up to different groups, including groups outside of IT such as finance and facilities.
More troubling is that many fail to realize that aggregating the costs that make up IT spend is only the first step. After that comes the work of defining costs for fine-grained resources such as the monthly charge for a gigabyte of object storage or the hourly charge for a certain size virtual machine. IT requires the same kind of activity-based costing that the manufacturing sector adopted two decades ago when first confronted with outsourcing.
Benchmark your economics. Once you understand your costs, find peer companies and compare. Both parties can learn from one another and help accelerate progress toward meeting best-in-industry standards. Compare yourself to public cloud providers as well to gauge how you stack up.
Marry your services to business value. A lot of lip service is given to "speaking the language of business" and "getting a seat at the table." Marrying IT services to business value goes well beyond using the right terminology; it has to focus on specific business deliverables or initiatives and providing technology to enable them. The fact that many companies go outside of IT to implement mobile apps shows how much need there is for closer collaboration. To that end, the Technology Business Management Council is devoted to this issue and hosts its inaugural conference in November.
Prepare for real-time utilization analysis and demand management. If you run a capital-intensive, service-based business such as an airline — which offers a good model for what IT needs to look like, by the way — then it's critical to keep utilization high with innovative demand management measures. Amazon Web Services is the leader in this area, thanks to its clever use of reserved and spot instances to keep its assets highly utilized. Utilization analysis and demand management goes well beyond typical capacity planning, though, which is focused on ensuring that sufficient resources are available but typically assumes up-front capital recovery models.
Cloud computing is serving as a forcing function within IT, causing it to move from a semi-handcraft industry based on infrastructure investment to a fully-automated one based on offering fine-grained computing resources. The financial practices fully capable of supporting the former are no longer sufficient for an accelerated environment of transient resource use.
Bernard Golden is senior director of 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.