Going from strategic to implementation

Part 1 and Part 2 described a strategic framework to think about your desktop strategy. This final part completes that framework and extends the discussion from strategic framework to implementation.

As you build a strategic approach to application delivery, adaptive service-oriented connections between applications and people are created. These connections make IT a core enabler of business evolution. You can then enable your users to choose devices and securely access applications.

Each organization is different, so how you choose to mix and match the various options available to you will vary. The flexibility to choose the best fit for your use case enables you to benefit from the best of all worlds.

If you’ve read all the other parts in this series, you’ll hopefully have a much better understanding of how the concept of connecting applications to people has evolved and how powerful it can be. You’ll also be better informed to judge what the cost is to you of staying with the status quo. Like I said in part 1, the net present value of doing nothing can’t be assumed to be a zero cost option.

It should also be clear now, that there are many choices that you could make. A one size fits all approach is going to be very limiting. This is why the products have evolved around use cases in today’s Flexcast delivery story.

So how do you get there?

At Citrix our experience has shown that this should be broken into three steps.

1) Identify your business priorities.
2) Time to value.
3) Transformation roadmap.

Business priority seems like an obvious thing to consider, however it’s one of the most common reasons I see customers stuck in endless discussions on value. They treat desktop virtualization as a PC refresh project as opposed to understanding what business, technology and user opportunities exist in their respective organizations. When people talk to me about cost, my answer is often a question along the lines of what value are you are trying to create? That helps create a much more pragmatic conversation around the value of any particular Flexcast model.

Time to value. Endless pilots, analysis paralysis, engineering religion and limited progress is another thing I see all the time. Often this is just a symptom of not understanding what the business, technology or user drivers are. Once this is understood it becomes much clearer as to when a particular project needs to be completed by. In some cases this is very ambitious and in others very conservative. There is no wrong answer. It is key to understand what is needed in the context of ones own capability to deliver. It seems so obvious to point out in a blog, but I see this as a sticking point with customers week in week out. So watch out!

The transformation roadmap.

It’s important to understand that today most people are coming from a device centric mindset. I can usually tell when people say, “I want to migrate my machines to Windows 7.” It’s subtle, but the real challenge in my mind is not about migrating machines when in most cases they will be replaced irrespective of virtual desktops or physical desktops. What’s really happening is users are being migrated from a physical distributed world to a more people centric centralized managed world. I sometimes think of it as moving from a model of pushing configuration to distributed endpoints to a model of pulling configuration from distributed endpoints.

Sometimes that’s too big a leap of faith. So the question becomes does one have to transform everything at once? Well no.

You can start by simply implementing centralized desktops. Often people refer to this as 1-1 desktops if they are using the VDI model. What they really mean is that most of the management practices they had in the distributed world are fully intact in the data center. There is a cost to do this, but new capabilities are added. Even when starting this way, there is still a lot of organizational process change that takes place. That in my personal experience requires attention and focus to get right. However the organizational changes that will be required long term to transform to a service delivery organization are put in motion.

Other’s start with the hosted shared model or implement a pooled VDI model. This takes those customers further down the path of optimizing their desktop plant. Technology innovation industry wide in this area continues to enable broader adoption of an optimized model. This is optimization on both the CapEx and OpEx front. Some refer to this as the layer cake. I’ve talked about it in the past as the stateless desktop.

Once a few business use cases are successfully satisfied, many are ready for broader deployments using several Flexcast models in conjunction with their streamlined organization. At this stage of maturity one is pretty far along to transforming from a device centric distributed computing organization to an IT service organization that is nimble and flexible and ready to adopt emerging client and cloud computing models.

Start where it makes the most sense

What I have just described is the path that Citrix currently expects most customers to head down. That however does not mean you have to follow these steps sequentially. These are just guidelines, and some customers will want to run several projects at different transformation state stages. There is nothing wrong with that, pick what makes the most sense.

In conclusion, what ever path you decide to take I think it’s important to ask yourself, what is a desktop? What does it mean for you today and what should it mean tomorrow? As you think about this, understand that there is no one size fits all approach. Question the wisdom of those who preach otherwise. Realize that consumerization is already here, and most likely you will need to find ways to embrace it with the governance requirements of your organization. Think like a service provider and shed the shackles of distributed legacy IT thinking that is inflexible. And finally remember that doing nothing is not free.