Why VDI Adoption Has Stalled

By Kevin Goodman
published: Tuesday, June 16 2009

In August 2007, Gartner predicted that "By the end of 2010, all new PC deployments will be virtualized." (Brian Gammage and George Siffler III, August 8, 2007). Aside from Cloud Computing, VDI may be one of the most talked about new technologies - yet it has failed to take off to the degree that some experts predicted. Indeed, by October 2008, Gartner changed their prediction to, "...However, despite ambitious plans for many organizations, deployments of hosted virtual desktop capabilities will be adopted by fewer than 40 percent of target users by 2010 ."

Everyone knows by now that VDI is a new way of managing user environments. The idea of replacing a traditional PC with a virtualized image in the data center and having the user access the virtualized image via a remote desktop protocol is appealing to IT staff for a variety of reasons that we are all now very familiar.

So, if the move to Virtual Desktops is so compelling why has the adoption rate been so slow? One reason could be that today's VDI software isn't up to par. It is reasonable to assume that since the technology is so new that the software is immature and lacks sufficient features to make the endeavor worthwhile. Another reason could be the unexpected downturn in the economy. Despite the bearish adoption rate, however, Gartner remains upbeat. Earlier this year prediction was bullish:

"The worldwide hosted virtual desktop (HVD) market will accelerate through 2013 to reach 49 million units, up from more than 500,000 units in 2009. Worldwide HVD revenue will grow from about $1.3 billion to $1.5 billion in 2009, which is less than 1 percent of the worldwide professional PC market, to $65.7 billion in 2013, which will be equal to more than 40 percent of the worldwide professional PC market ". - Emerging Technology Analysis: Hosted Virtual Desktops February 17, 2009.

It is easy to have a bullish attitude because it is certainly reasonable to assume in the next few years that the technology will catch up and the economic downturn will reverse. However, this article examines other more fundamental reasons why the VDI adoption rate is not what was predicted. And, if not corrected, these issues could continue to hamper the VDI adoption rate regardless of the state of the software and the state of the economy.

Typically, when an organization is deciding upon a new paradigm such as VDI they choose to conduct a proof of concept or "pilot". Pilots usually take place within a subset of the user community and users are expected to perform their normal line of business applications. That is, they are expected to perform the same functions they would normally perform except they are on a non-production system. If your organization is large enough, you may choose to run multiple pilots simultaneously with competing vendors' software. If the pilot is successful, you can expect to move from pilot to production at some point in time. If the pilot fails or languishes in the pilot stage for a significant period of time then there will be no production system. It is pretty simple actually: unsuccessful pilot equals no production system which equals no VDI adoption.

So if you are contemplating a VDI pilot (or are in an existing pilot that is languishing) there are five distinct issues that could subvert your pilot:

1. Incorrect management organizational structure
2. Lack of virtualization expertise
3. Failure to properly respect the user community
4. Poor understanding of the organization's hardware and software inventory
5. No way to measure success or failure

Incorrect Management Structure

Your VDI pilot may be predestined for failure based upon your company's organizational chart. If the person responsible for the servers in your data center is not the same person responsible for managing your desktops, and these people do not have a common boss for two or three levels, then watch out - the power struggle will overwhelm any chance of the pilot succeeding. In large companies these two functional areas also can be separated by geography. For example, if your server department is in St. Louis and your desktop division is located in Des Moines, there is a good chance that nobody knows each other.

This situation can be avoided if the mandate for virtualizing desktops comes from the CIO or another person in the organization who has authority over both groups. But what frequently occurs is the VDI proof of concept is a grass roots approach initiated by a technician or well-meaning manager in either the desktop or server group. Their idea is that they will put together a quick pilot and if it succeeds they will submit a proposal to upper level management. What happens instead is that a project designed to reduce expenses and streamline support and management costs ends up in a turf battle. The battle can take place on one of two fronts: if the server team initiates the pilot it can be seen as empire building by the desktop team. If the desktop team is the initiator the server team can become territorial by refusing access to the data center and refusing administrative access to certain resources.