This year, companies worldwide are predicted to spend almost $16.5 billion on cloud Infrastructure-as-a-Service (IaaS) — a 32.8 percent increase over 2014.
And, with 93 percent of companies already running applications or at least trying out IaaS, everyone is trying to find out whether or not all of this investment in the cloud is paying off.
Turns out, it’s quite a challenge.
Because there are so many factors to consider, cloud ROI can’t be determined with a simple formula, said Brett Gillett, public cloud practice lead for Softchoice, a provider of technology solutions and services.
“When you’re talking about straight cloud computing, pricing models can be complex,” he said. “There are a lot of moving pieces. It can get confusing, and costs can be easily overlooked.”
CMSWire asked Gillett to give us a rundown of some of those costs, as well as some tips on what companies can do to reach those elusive ROI figures.
Input/Output (I/O) Costs
One cost that customers can easily overlook, said Gillett, relates to input/output (I/O), including how much memory and disk space is required when moving to the cloud.Because users don’t have the tools to understand how their applications are using I/O, they’re not aware of how these costs impact their bottom line, he said.
“These are not hidden costs, but they’re easy to overlook. Many times customers end up with bills they don’t expect because they didn’t know about them, or they underestimated.”
To help companies get a better handle on this, Gillett recommends having monitoring tools in place to better determine CPU usage. By better understanding how their apps are working, IT teams can translate what that work would look like with a cloud provider to ensure they’re not overpaying.
“You need to right-size the environment when you move to cloud,” said Gillett. “Find out your actual use of CPU, and determine whether or not can you put it in a smaller instance with the cloud provider when you get started.”
People know they’ll save money if they move some or all of their infrastructure to the cloud, said Gillett. But how can they estimate those savings?
Infrastructure costs could include heating and cooling associated with having an onsite server, the cost of electricity, or hardware investment.
As with I/O, Gillett recommends comparing the capital expenditure of buying new equipment versus the operational expenditure of moving to the cloud, and reallocating appropriately.
For example, what would replacing a fan or hardware on-premise, versus moving to the cloud cost?
“Overall facility costs are not easy to calculate,” said Gillett. “There are good models out there for calculating electricity costs, but it’s still difficult. These are not trivial tasks.”
Finally, Gillett advised IT teams to estimate the difference in time people spend on tasks such as infrastructure management, compared with more strategic tasks, once they move to the cloud.
“When you move to the cloud, there are a lot of things you shift to the provider,” he said. “Understanding how you can reallocate human resources to other tasks is difficult because people don’t understand how much time is used to manage their IT resources.”
To help make the best use of their people, teams should look at implementing tools and best practices in problem, incident and change management.
“Having the tools is one thing,” he said. “But having people use those tools is another. People need to ask themselves: How can I use that expensive network resource to do higher value business tasks such as talking to clients to build better solutions with IT?”
One last recommendation from Gillett, whose company just happens to provide cloud consulting services:
“Find a good partner that has the kinds of skills to help you estimate cost on premise compared to costs running on cloud. There is value working with a partner if that is their specialty.”
Title image by Ales Krivec.