As companies adopt more cloud IT services and work with an increasing number of service providers, the tried-and-true IT towers of the past no longer serve their needs. "The old model lacks the clarity of ownership required to drive decisions on as-a-service offerings that span the traditional tower structure," says Steve Keegan, principal with outsourcing consultancy Pace Harmon. "Determining who makes the call isn't straightforward -- the server team, the app team or the database team?"
IT groups need to reorganize around processes rather than technology areas, such as infrastructure or applications. In process-driven IT organizations, "people are focused on activities that are of long-term benefit to the organization with clear accountabilities and increased efficiencies in how work gets done," Keegan says.
But such "plan-build-run" models often don't take -- for a number of reasons. "Application teams still dominate the decision-making process. Legacy applications leaders don't want to diminish their authority to drive investment decisions," Keegan says. "Leadership doesn't hold their own managers accountable for minding boundaries. The architecture function lacks the resources they need to be successful and therefore become strictly oversight with no real time to contribute and provide planning insight to project and program leaders."
Those IT groups that do successfully shift to plan-build-run have clear leadership support and make the change wholesale, instead of letting certain groups continue operating in the manner that they always have.
Keegan offers five tips on transforming IT into a process-driven organization:
1. Enlist an army of business analysts.
These IT professionals should align themselves with business relations -- not the applications -- to handle upfront planning. They will have the business process knowledge necessary to clarify changes that will be required in the applications landscape and create new and consistent business processes. "There are often far fewer business analysts than anyone believes is sufficient," says Keegan. "Business analysts provide a core competency in aligning business needs for change with the underlying systems that support that business."
2. Don't skimp on planning.
"The early project phases of scoping and planning that ultimately feed the business case are indeed worthy of sufficient investment, and are not only valuable in assuring that investments are targeted correctly, but are legitimately allocable to project investments since their purpose is scoping, defining a plan for, and justifying the expenditure of investment to achieve strategic business objectives," Keegan says.
Plan to spend about 5 to 10 percent of project time (and cost) developing a business case. "Spending more time planning creates not only more certainty that you'll be pleased with the outcome, but you will know when you break ground that you won't be making too many changes to the plan once you start," says Keegan.
3. Put architecture in charge of application planning.
"It's fine to leave application teams in charge of portfolio and roadmap planning -- so long as you don't ever want to retire an application or simplify the support and resultant cost of the application landscape," Keegan explains. "Traditional application leaders typically attempt to expand their control by increasing complexity, expanding the application suite resulting in increased licensing and maintenance cost, and expanding complexity of the resultant support -- because they are accountable for delivering individual projects, not the cost of supporting what they put in.
Architects have a mission to optimize the portfolio, reduce total cost of ownership, and leverage investment to the greatest degree possible." Incentivize efficient utilization and rationalization of the application suite, retirement and migration of less used or expensive to maintain applications and platforms, and optimal cost and quality.
4. Never start projects until sufficiently planned and staffed.
"Without an understanding of or visibility to the portfolio constraints, executives and other leaders step in to drive determination of budget and timeline that may or may not be achievable," says Keegan. Leverage the portfolio planning function to prioritize and sequence demand for as-a-service offerings.
5. Create an effective gating mechanism prior to production.
"A capability for owning service transition needs to be baked into the model to assure that what goes into production are only changes that are ready to be supported," Keegan says. Those responsible for these decisions should be independent, but aligned with both the build and run functions.
"This capability owns the build-to-run boundary, making sure all is correct (to meet business needs), captured (so it can be reproduced, and replicated), and supportable (so when phone calls reach the help desk, they know what to do). Whatever the change is, whenever it needs to be deployed, the gate into production needs to first and foremost consider the ability of the run organization to support that change when it goes in."