2013-IT02

Posted by

Capacity Cost Model: Balancing the Demand for Software Changes Against the Supply of Resources

Information Technology Track

IT-2_Presentation_CapacityCostModel_Kaldes

IT-2 – Presentation – Testing Benford’s Law with Software Code Counts

IT-2 – Handout – Testing Benford’s Law with Software Code Counts

Abstract:

A Program Executive Office (PEO) within the Department of Defense (DoD) is responsible for multiple information systems that support the Services (Army, Navy, and Air Force). This PEO is not only responsible for Operations and Maintenance (O&M) of the information systems, but also for implementing system change requests that are made by the Services’ Senior Service Representatives (SSRs).

The PEO was relatively satisfied with the status quo, but the executives within the PEO recognized that there were three problems with doing business today. (1) Currently, there was no way to scientifically indicate/prove to the SSRs what software changes they could promise. During the Configuration Control Board Meetings, the SSRs would identify their needs and then the information system program managers would sign up for changes, with little or no knowledge of how many (or how few) of the changes they could actually address. (2) Since there is a requirement for O&M, the executives and the program offices had no idea how many resources they had to dedicate to this activity before they could start making changes to the software. (3) The program managers received all three colors of money, and there was no way for the PEO to tell whether the colors of money were being spent on the services that corresponded to those colors of money.

To address these three needs, the PEO developed a Capacity Cost Model that allowed the PEO to (1) have a scientific and quantitative approach for balancing the demand for software changes against the supply of developer (and other) resources; (2) understand how many resources are being used for O&M software maintenance; and (3) compare how their colors of money were used, and for what purposes.

Equally importantly, the Capacity Cost Model allows the PEO to understand the total costs (overhead) associated with each information system. More specifically, important metrics are calculated, using actual contract data as inputs. The Capacity Cost Model also includes reports and outputs that the PEO can use to measure performance. The Capacity Cost Model is up-and-running, and it is used on a quarterly basis to guide the SSRs in their determination about which software changes to implement.

The model is complex, in that it takes a significant amount of contract data and massages it in multiple ways. From a supply perspective, the 400+ person staff at the PEO is first classified into multiple categories and sub-categories to understand their role; frequently, staff wears multiple hats, further complicating matters. Second, the staff is identified by what information systems they support. Some staff support only one system, but many staff support multiple systems (particularly overhead staff). Last, the staff is on certain contracts, which shows the color of money that corresponds to their time. As such, staff is examined from three dimensions. From a demand perspective, the forecasted level of effort associated with the O&M is first calculated and subtracted from the available level of effort. Next, the system change requests in question have developer rough order magnitudes (ROMs), but they are then burdened with total ROMs.

This is a common problem within the Military, Civilian, and Commercial environments. Frequently, software development/maintenance program offices or umbrella organizations typically have multiple customers that request updates and changes to the software. And the organization is left with little understanding of what, and how much they can promise to their customers. This is particularly true of a fee-for-service software organization. Multiple paying customers are requesting changes, and a model/approach like the Capacity Cost Model allows the organization to be more diligent and judicious about what changes they make, and even more importantly, how many changes they can promise to their customers.

 

Author:

Chris Kaldes
Deloitte Consulting, LLP
Mr. Kaldes has over 20 years of professional experience in program management, cost estimating, and economic analysis. He has developed many cost studies and investment determinations for government and private sector decision-makers in domestic and international markets including: life cycle cost (LCC) estimates, cost/benefit analyses, business case analyses, cost/performance trade-offs, cost proposal evaluations, alternatives analyses, budget analyses, Exhibit 300s, and risk quantification studies. He also has extensive experience with database design, cost model design, and risk management. Mr. Kaldes also has significant experience in system requirements analysis, DoD MAIS requirements, systems analysis, and systems design. Mr. Kaldes has been with Deloitte Consulting for the last three years and he leads the Cost Estimation Capability for Deloitte Federal.