IT-as-a-Service: Models for Consumption, Operations, Technology

In a week where products surely dominate the news (it's VMworld, in case you live under a rock) I want to share some non-product insights I've made while working with EMC's IT group regarding running IT as a Service.

First, running IT as a Service doesn't mean outsourcing IT. It's about transforming corporate IT into thinking of itself as an internal service provider. Catering to internal Line-of-Business customer needs. Providing pricing and rate sheets. Offering financial transparency. Even marketing against and competing against (external) services.

IT as a Service isn't just "more virtualization"

The market is equating virtualization with cloud, and worse, equating cloud with IT Transformation. And by IT Transformation, I mean the re-casting of the IT department to operate like an internal service provider, and to act like a business. In my opinion, Cloud (use whatever definition you prefer) is really just a critical enabler for IT transformation.

In both EMC's and VMware's similar 3-step "Journey to Your Cloud" models, at first glance you'd think that the end-goal is more virtualization and more automation. But what we at EMC (and a number of customers) have found is that the end-state is where the technology revolution winds down, and the organizational evolution begins to spin-up. Just having the technology in place is necessary, but not sufficient, for a transformation of how IT operates. So, as I've said, Cloud is the means, but not the end.

Our IT department has over 79% of services virtualized and running on standard platforms. And arguably the virtualization initiative is beginning to be completed. (Aside: The new VMware 5.0 product line supports even bigger VMs and leverages CPUs with more cores - allowing for production-scale databases to now be virtualized). Manpower is now being diverted into the NEW areas needing resourcing: Business Transformation of IT, and the services packaging, marketing, and management that comes along with it. Another way of putting it:  Now that the technology is posing less of a problem to design and manage, more time is being spent working with EMC's Lines-of-Business to cater to their needs and enhance their business agility.

3 Tenets of IT Transformation

After conversations with our CIO Sanjay Mirchandani and our VP of Infrastructure, Jon Peirce, it's clear that this final phase of the "journey" is only partly comprised of technology. At the core, there are 3 broad areas of focus when transforming IT: Consumption models, Operations models, and Technology models.

• IT Consumption Models
  Essentially this describes how supply is separated from demand, and that service capacity is created just-in-time. Services are generated from an approved inventory stored in a service catalog (self-service, if you're in IT), with each service having clear pricing, SLAs etc. The pricing could be variable and  "consumption based" that is, it's not only metered, but based on both true cost as well as opportunity cost for access to the infrastructure. But expect new challenges for IT finance.  Lastly, the consumption model can also include services brokering - that is, it can include services generated from outside IT as well as those generated within/by IT. Both sources are equally valid, so long as IT still provides common governance, access, pricing and secure delivery to internal LoB customers.
  
  
• IT Operations Models
  Another non-technological shift is how IT operations changes, morphs and grows. One would expect that the operations mechanisms become more automated with less human intervention. But the real shift in operations is the *mindset* of IT, shifting from a "technology builder" to a "service delivery" organization. This fundamental shift focuses on delivering services (internal, external etc.) to the internal LoB customer to meet their needs/requests. A business mindset might also mean that IT will have to grow roles that sound like "service manager" "service product manager" and "client marketing manager" skill that IT will need eventually to acquire/develop. Which ultimately implies a good dose of Change Management will be necessary - that the IT organization, skills, roles, goals, etc. will shift over time. (a very excellent blog by Chuck Hollis dives into this topic)
  
  
• Technology Models
  Last but not least is technology - but this part of the story isn't so much  about new technology so much as it is about how technology is used... and by whom.  Necessary, but not sufficient, is the use of a virtualized, automated, and converged infrastructure (cloud componentry).  Sans jargon, I mean an infrastructure that is virtual and pooled, so that it can be composed on-the-fly as business conditions require. Since this departs from the traditional vertical stack model where IT personnel *skills* are specialized, new skills will be in demand. We'll need more IT generalists, IT staff with horizontal *services delivery* skills, not point-product skills, and with comfort around automation. Oh - and the technology *roles* and *organizations* will change too to be consistent with the new model.
  

And finally, the need for Change Management

No, I don't mean Change Management in the context of ITIL/ITSM. Rather, it's in the context of organizational design and development (think: Management Consulting jargon). With all these new models, Change Management is critical to orchestrating the *non-technical* shifts in IT.  Our own EMC IT department employs a number of these types (ex-BCG, McKinsey, etc. folks) who help our staff develop new skill-sets, morph org structures, and create new engagement models for the business owners.

IT transformation does not simply happen as a result of new technology. Rather, all of the products you buy are simply enablers to help you get to the really Hard Work. But the payoff is enormous.