This content is part of the Essential Guide: Guide to low-cost desktop virtualization
Evaluate Weigh the pros and cons of technologies, products and projects you are considering.

How to analyze the costs of a virtual desktop deployment

Part 3 of our desktop virtualization handbook can help you understand how to assess the potential costs involved with building a virtual desktop environment.

Before you jump on the desktop virtualization bandwagon, you need to figure out whether your organization can benefit from the technology. In this third segment of a four-part e-book Desktop Virtualization From A to Z, we'll help you analyze the costs of a virtual desktop deployment.

Although saving money isn't the best rationale for desktop virtualization, at some point, someone's going to ask you how much this whole thing is going to cost. So before you say that it doesn't matter, let's try to figure out what exactly this person is asking.

As you probably know, business costs are usually divided into two classifications: capital expenditures (Capex) and operating expenditures (Opex). Broadly defined, Capex includes onetime costs, and Opex covers ongoing costs. Capex includes buying software and hardware and the time and money it costs to implement and migrate from other technologies. Opex is usually just support and maintenance costs.

The first step to doing a cost analysis is to figure out whether you're going to look at Capex, Opex or both.

Entire books have been written about modeling the expenses of IT systems, and there are companies dedicated to analyzing expenses. Because there are so many different kinds of desktop virtualization, we can't do a proper expense analysis here. But we can look at some quick "back of the napkin"-style numbers to compare virtual and physical desktops.

Most people believe that the Capex of implementing virtual desktops is about the same as that for physical desktops. In other words, if you spend about $800 per desktop to buy and roll out traditional desktops, you can probably build a virtual desktop for about $800 each.

These models assume that you'll reuse existing computers, maybe as thin clients or simpler terminals for accessing your newly built virtual desktop environment. So if you have 100 users who would otherwise get new desktops at $800 each but you reuse existing machines instead, that frees up about $80,000 to buy some nice VDI servers, software and consulting.

Opex is not nearly as simple to quantify as Capex for two reasons:

  • There's no clear-cut way to figure out what's included in your Opex calculation.
  • Before a project has been implemented, actual Opex is just a best-guess estimate anyway.

For example, will Opex calculations include soft costs such as "employee productivity gained by less downtime" or "user satisfaction?" Are your operating expenses based on dollar amounts you actually pay? Have you used the power savings of thin-client devices as part of your justification, even though the thin clients plug into users' cubicles whose power is not billed back to the IT department?

Have you included everything in your cost models? For data center servers, have you based your power costs on their rated wattage, or have you included your data center's power usage effectiveness?

Of course, you can use any of these methods to nudge the cost analysis in the direction you want to prove that desktop virtualization is cheaper or that it is more expensive than traditional PCs. The mere fact that cost analysis is complex doesn't mean it should be ignored. It just means that you need to be more aware of what a cost analysis can and cannot be used for.

Unless your cost analysis can fit on the back of a napkin, it's probably crap. So go ahead and work out the basic numbers. Figure out your high-level stuff, and create a budget for the hardware and software you'll need for desktop virtualization. But beyond that, don't waste your time building a complex Excel model. There's a good chance that if you show it to someone, he won't believe it.

Just keep in mind why you're virtualizing desktops in the first place. If it's to save money, make sure you're really going to save money. And if it's for another reason, such as more redundancy or increased security, don't waste your time going nuts with return on investment (ROI) models.

Brian Madden
is an independent industry analyst and blogger, known throughout the world as an opinionated, supertechnical desktop virtualization expert. He has written several books and more than 1,000 articles about desktop and application virtualization. Madden's blog,, receives millions of visitors per year and is a leading source for conversation, debate and discourse about the application and desktop virtualization industry. He is also the creator of BriForum, the premier independent application delivery technical conference.

Dig Deeper on Virtual desktop infrastructure and architecture

Start the conversation

Send me notifications when other members comment.

Please create a username to comment.