So continuing my discussion about my experiences talking to senior management at some of the largest financial, insurance and banking companies about cloud infrastructures, I’ve told you about a couple different kinds of customers (the “Huh?” and the “Been There, Done That”). In this installment, we’ll cover the particular kind of customer who thinks that simply wanting something strongly enough will allow them to negate the laws of business (and physics) and get what they want.

Would you like ice cream with that?

I have been around the enterprise space, all over the world but specifically in New York City the last 5-6 years, and I have run into all kinds of customers, but I think that the NYC customer is a particular kind.  Now, I say that with great respect and deference (in case any recognize themselves), as being on the wrong side of a New Yorker is not the place you really want to be.  It’s not that they are violent or mean or anything like that…it’s just that they know in the centers of their hearts that they’re right.  And you are wrong…and nothing you can say, no evidence you can present or truth you could shout from the highest rooftops will convince them of anything else.  But I digress…

I call this customer the “I want to eat dessert before dinner” type because, even though they clearly and unambiguously understand the process (and in some cases have already been through the process), they simply refuse to believe that they are constrained like every other Tom, Dick and Harry.  It is like they actually believe that because one woman can give birth to a child in nine months, they can then convince nine women to give birth to a child in only one month…I mean it’s just math, right?

Well, identifying, designing and implementing a cloud infrastructure is not at all like a pregnant woman because there is not just one very clear and understandable process for how it should be done…there are potentially hundreds.  Thousands even…and if an organization doesn’t take a deep, hard look at themselves before they start, there is the very real chance that they will pick the exact wrong way to do it and ultimately fail and possibly waste millions of dollars.

Yes, that’s millions with an M.  The fact is that for a “cloud infrastructure” to be successful, it will  have to be specific to every customer and will have to be based on exactly what they want to do.  The following examples show that, based on the desired end-state, each cloud infrastructure design will result in an entirely different outcome:

• A customer wants to increase revenues by creating a cloud infrastructure for decreasing time-to-market for their applications. Therefore they will design a self-service portal integrated to a service catalogue of available development and composite-application virtual machines that are keyed to user roles and, using automation and orchestration, are able to be quickly spun up, spun down and archived for later regression testing. This reduces time-to-development  (in some cases) by 2-3 months.
• A customer wants to save costs across the board so he designs an end-to-end cloud infrastructure where all applicable production machines are running in that context…accessing fungible resource pools of memory, CPU and storage.  This allows them to specify the lowest cost, commodity hardware (server and storage) because they have abstracted the actual machine specifications (required for an application). So if an application needs 8 cores and 256 GB of RAM, instead of specifying that as a very expensive physical machine, they simply slice and dice the cloud in order to supply that level of resource to the application.  This reduces overall cost anywhere from 5-25% (based on bulk purchasing of the physical assets).

Believe me, it’s easier picking the wrong design than the right one when the sum total of your qualifying reference is the business benefit at the end.  It’s all well and good knowing what you want (dessert!) but that does little good if you don’t know that you have to sit through a four hour, ten course meal (the process) listening to Uncle Jeb talk about his latest surgery before you can actually have that gloriously delicious smelling piece of cherry pie…with ice cream.

Getting to dessert after that meal is the same thing as getting to the ROI in a cloud infrastructure engagement after completing the investment, effort and time it takes.  Just as there are different kinds of cloud infrastructure engagements, there are different kinds of meals (Thanksgiving dinner with Uncle Jeb, a Big Mac at McDonald’s, etc.) and, as always, you get to decide what kind of meal you want to sit through and whether the dessert is worth waiting for.

Or you can know exactly what you want AND know beyond a shadow of a doubt that you’re right…and ultimately hear in excruciating detail all about how the tear in Uncle Jeb’s lower intestine was repaired…ice cream not included.

Next installment: Where do I start first?

More from Trevor: Wish you had an application store in your IT organization that worked like iTunes? Learn how to begin delivering cloud services more easily, and MUCH faster than you imagine…automatically.

In this webinar, you’ll learn the step-by-step processes necessary to successfully implement cloud solutions into your environment.