Get A Strong Focus For Your Approach To Cloud

In discussions on cloud computing, I often talk to architects who have been told to create a "cloud strategy." This sounds appropriate enough, but there’s a devil in the details: When the task is "create a Technology X strategy," people often center strategy on the technology. With cloud, they aim to get a good definition of pure cloud and then find places where it makes sense to use it. The result is a technology strategy silo where cloud is placed at the center and usage scenarios are arranged around it. The problem with this is three-fold:

  1. Considering the full business dynamics of any given usage scenario, there is a wide continuum of often strongly competing alternatives to pure cloud (including cloud-like and traditional options).
  2. The rapid pace of market development means that business value equations along this continuum of options will keep changing.
  3. Your business needs integrated strategy for many technologies, not simply a siloed cloud strategy.

The point of a strategy is to enable good decisions over the long term. With cloud, one’s evolving decision framework should put the business decisions to be made at the center, arranging options -- pure cloud, cloud-like, virtualized, traditional -- around the business decision. This is what a good architecture strategy does, and the best way to develop a strong approach to cloud is to integrate cloud as an option into broader strategies. Consider infrastructure-as-a-service (IaaS) as an option within your overall infrastructure strategy. Platform-as-a-service (PaaS) might factor into your application platform and development strategy. Your business solution road map might have a place for software-as-a-service (SaaS). Building cloud into these other strategies: 1) establishes a good foundation for making strong business decisions; 2) offers protection against getting caught up in cloud hype; and 3) provides a stronger basis for evolving toward pure cloud through stages like server virtualization, internal cloud, and hosted cloud.

To read more, go to a longer post at cio.com or to this Forrester report.

What are you doing to protect your organization from cloud hype?

Comments

Motivation Model is Most Important Model for EA

If there's one model that all enterprise architects need to create before talking tech (be it Cloud, SOA, BPM, whatever), it's a business motivation model. As Randy says, before talking about the means (mission, strategies, tactics), we need to understand the ends (vision, goals, objectives). For cloud, the motivation might be the ability to enter emerging markets without a huge CAPEX budget, or aligning scarce IT resources with systems and services that competitively differentiate our company from the competition. Give me a business motivation model and a business capability map, and watch how quickly we can unlock the value of business technology (cloud or whatever).

Take the Fog out ?

I think the first step is a clear definition of what 'cloud' is. There really isn't one definition, even for what you refer to here as a 'Pure Cloud'. And I'm not talking about the clear differences between public and private clouds, but the whole cloud concept.
Even in you blog post, you refer to Iaas, PaaS, SaaS as components of cloud architecture, but do you need one, two, all or any of these to utilize a cloud concept for content and application management?
For me, it's sort of like the definition of a house or a home- runs the gamut from a mansion to a cardboard box under an overpass, depending on your personal infrastructure strategy =)

Definitions don't (really) clear the fog

Lawrence: You make my point precisely. You never buy a definition; you buy specific products/services, and they tend to approximate the definition. Because definitions are slippery, and furthermore because (even if the definitions were clear) many cloud-like offerings don't fit "pure cloud" definitions, the *first* step is (a) to clearly understand the stream of business decisions that need to be made and (b) to categorize those decisions for the purpose of coherent strategy and planning. For each category of business decisions, the strategy organizes and selects among the continuum of options available. Properly identified, the business decision (e.g., "select cost-effective computing resources") remains stable, and you continually adjust the strategy to account for new market offerings, fuzzy and changing definitions, and (hopefully) continually growing organizational expertise and maturity. It's the business decision focus that clears the fog, not the definition.

Buying? Selling?

If you follow the press and trade rags, G00gle, Am@zon, M$oft and 1BM are all selling "the cloud", whatever that is... no one is selling products/services and so when the C-levels go to IT or Business Services, they ask "why aren't WE in the Cloud?" It's sad but true- it's on the covers of all of the magazines and trade rags they get, both public and private sector, and it's what all the buzz is about.

The questions that SHOULD be getting asked are related to services and features, irrespective of what they're called. Where the rubber meets the road are the features, risk protection, up-time, continuity of operations in disasters, etc.

Even the concept of cost effective is questionable- it's not the same to look at lowest cost of operations as opposed to total cost of operations when you have hosted and leased services- control what you can do best, buy the rest. Manage your mission critical and sensitive data, toss your less valuable but necessary data over the fence to a lower cost alternative... don't force yourself into having to calculate the cost of risk and recovery into your TCO.

There's always risk (but I get your point)

Preach on, bro!
I would say, however, that a full business analysis does include risk calculations — as well as flexibility. The TCO method does not handle these well, so you need Forrester's Total Economic Impact method instead. Sometimes, a (well-managed) risk is worth it.

Cloud and EA

I was talking with a cloud Evangelist the other day and he made the statement that "we're using cloud, we don't need an EA".
After I re-booted my mind after a comment like that, I responded that no you need the enterprise architecture more than ever. Just because you've "outsourced" the technical architecture via the cloud. It doesn't mean that this set of services can stand alone. For most organizations it needs to be integrated into the wider enterprise in order to deliver comparable value.
As you state, you need to have a strong focus so that you can approach the service vendors from a position of strength, and not just ride along with the current.

I couldn't help myself

For further reading...

In a subsequent post, I put a bit more stucture around the distinction that I'm making here. In Your Future Needs A Type 4 Technology Strategy, I define a four part taxonomy for types of technology strategy. As an example, I explain that a dedicated cloud strategy is a Type 3 strategy, whereas the Type 4 strategy is my recommended approach of building cloud options into your existing infrastructure, application platform, and solution portfolio strategies.