How Financial Modeling Plays into Cloud Adoption
It’s not uncommon for enterprises to feel “stuck” in their current application ecosystem. Everything works together more or less well; the architectures used in the applications are generally well known and humming along. When promising, enticing or greatly ballyhooed new technologies come along, it can be hard to find a way to shoehorn them in.
That’s how many (even most) organizations proceed with cloud. There’s a lot of skepticism and fear still out there (yes, still), not to mention a learning curve that’s both technical and operational. So, it’s wise to start really small, do some learning, and get a small success under your belt. Yet, organizations don’t (generally) adopt new approaches on a wide scale without careful analysis and planning. And planning usually means, “show me the money.”
A first step, usually pushed for by cost-conscious executives, is to get some kind of financial modeling of the Total Cost of Ownership of current systems versus potential migration of those systems to a cloud environment. On the IT side, it’s still rather common to see people scoff at the very notion of putting anything in the cloud: why pay $1,000 a year (for example) for several years for a virtual instance running 24x7 when I can easily procure a server for that much and own it immediately? But a quick-and-dirty comparison like that will often be misleading in the extreme, in that it omits a lot of key considerations. It turns out that an appropriate cost comparison of cloud vs. on-premises is actually not intuitively obvious.
What needs to be included in order to get a more meaningful cost comparison than the gut-level, quick-and-dirty common one, then? Several major groups of cost need to be analyzed and folded into a “total cost of ownership” financial model. It turns out that running a physical server entails not just the purchase cost of the equipment, but also a myriad of direct, indirect, and overhead costs that you inevitably incur when operating and owning physical servers.
Here’s what’s usually omitted from a seat-of-the-pants “cloud is more expensive” judgment:
- your physical server’s direct costs (power, floor space, storage, plus the operations staff needed for management of those factors);
- the associated indirect costs that make your server work in your environment (network and storage infrastructure, plus IT staff time for management of those factors as well);
- the overhead aspects (non-technical, largely), consisting of procurement, vendor management, and general IT management.
Determining these non-hardware costs is anything but easy in most organizations; they are often scattered across work groups, and can be tough to tease apart. Determining the appropriate “slice” to use out of what is often a rolled-up total involves a number of judgment calls and estimated adjustment factors. But it can and should be done, in an effort to establish some level of apples-to-apples comparison in a monthly total cost of ownership sense.
Peter Kretzman is a consulting CIO and author of the blog, CTO/CIO Perspectives: Intensely practical tips on information technology management.
For more thought provoking cloud management insights visit vmware-erdos.com.