I’ve been speaking with some CIO’s and CFO’s at larger firms about how they can take advantage of cloud computing, especially in a recessionary economic climate. 

It’s easy for startups and small firms to take advantage of the menu of cloud computing services out there, but how does an IT manager at a mid-size or large enterprise get in the game?  There is a ton of hype out there about cloud computing – and nervous IT managers don’t want to risk their reputation or their organizations cash fiddling with systems that can be argued as "unproven." 

While cloud computing won’t solve all your woes, or cut your IT budget to zero, it does offer a nice way to trim the fat and reduce overhead, including energy footprint and physical costs (space, hardware, ping pipe and power, etc).   All this while remaining scalable and flexible; options that heretofore went out the window when the budget axe came through.

But as the saying goes, "No one’s been fired for buying IBM," so let’s look at some low-risk, high-return entry points into the cloud for today’s IT manager.  I’ll outline them below and follow up with more detailed posts later.

1. Communication and Collaboration Services:  Let’s lump in email (Google Apps), phone (outsourced Voip), and project management utilties here.  There is obvious return (no hardware or maintenance costs, increased functionality, spam protection, scalability, etc) – but how can a larger firm use this in a safe, low risk way?  Exploring a progressive migration to these types of services, targeting the temporary and flexible portion of your workforce is a good start.  More on this to come.
2. Lab infrastructure: Can you think of lab related functions in your organization that require physical infrastructure and can be moved to cloud based hosting like Amazon EC2?  This sort of infrastructure is small in comparison to your mainline production and data processing systems, but I am sure occupies at least a few percentage points of your overall IT spend.   By labs, I mean things like:
   • development servers
   • QA and test servers
   • R&D and skunkworks environments
   • Training environments
3. Time-sensitive infrastructure: How much of your datacenter spend is dedicated to overnight reconciliations, true-ups, clearing, and processing? A unique advantage of utility oriented cloud computing solutions (like Amazon EC2), is that they can be turned on and off at will.  Imagine eliminating racks of servers that performed time-based functions and only paying for the time they are in service. I’ll explore the security implications of this later because this is a no-brainer if you can work that side of it out.
4. Web infrastructure: Intranets. Extranets. Public facing websites.  Get them out of your datacenter. If they are already out, get them out of Managed Services (or at least traditional managed services).  With the advent of Microsoft Windows Server 2003 support on Amazon EC2, and SQL Server support, only the most intensive web applications will continue to absolutely need to be in a physical managed environment. 

This is an evolving list, if you have any suggestions or thoughts, please comment! I’ll continue this thread of thought as the cloud computing space quickly progresses.

Related Blogs

• Related Blogs on cloud computing
• Ten Key Reasons for Enterprise to Choose Cloud Computing
• Auto-scaling cloud computing and Denial of Service attacks « Wille …