Posted by Enterprise Resource Planning Systems: Sizing Metrics and CER Development
Software Track
SW07_Presentation_ERPSystems_Brown
Abstract:
Over the last 10 years, Enterprise Resource Planning (ERP) systems have seen increasing use within government agencies. One of the key components to any software cost estimate is a sizing metric. Unlike estimates for a traditional AIS, ERP estimates cannot use a traditional metric such as source lines of code (SLOC) because (1) ERPs are almost always based on commercial off the shelf (COTS) software; and (2) ERP implementation requires a significant amount of effort outside the categories of software development and software procurement. Reports, interfaces, conversions, and extensions (RICE) objects are often suggested as the sizing metric of choice for ERP. However, this approach also has significant cost estimating limitations. First, it doesn’t allow the analyst to treat each component of RICE separately. For example, what if the count of interfaces is thought to be more significant as a cost driver than the count of reports? What if data are available for only 3 out of the 4 RICE components? Secondly, a simple cost per RICE object metric doesn’t provide the benefits of a parametric approach, such as a measures of goodness of fit, significance testing, and quantified uncertainty analysis.
This paper presents research investigating three alternative ERP sizing metrics: number of users, number of interfaces, and number of legacy system migrations. Each metric is defined within the context of ERP implementation. Each of these metrics are generally available during the early phases of ERP development, which make them ideal for cost estimates when no specific technical solution or ERP vendor has been selected.
Using statistical analysis, each measure is evaluated using data from DoD ERP implementations. Combinations of metrics are also analyzed using multivariate analysis to develop proposed cost estimating relationships (CERs) for ERP investment and ERP sustainment. Each CER is evaluated for goodness of fit, statistical significance, and uncertainty.
Author:
David H. Brown
Technomics
David Brown is a Senior Cost Analyst at Technomics. His primary areas of expertise are IT life cycle cost analysis, database development and management, and acquisition program management. Since joining Technomics in May, 2010, Mr. Brown has developed Enterprise Resource Planning (ERP) cost estimates for clients within the Government Accountability Office and Department of Homeland Security.
Prior to 2010, Mr. Brown served as a Research Fellow cost analyst for LMI. At LMI Mr. Brown developed IT cost estimates for a variety of government agencies including the Defense Commissary Agency, Coast Guard, NASA, and GSA. Mr. Brown presented a paper to the 2010 SCEA conference on the use of life cycle cost estimates in OMB-300 reporting, and to the 1996 SCEA conference on work in the area of activity based costing.
Mr. Brown is a SCEA-Certified Cost Estimator/Analyst (C/CEA). He holds a bachelor’s degree in Engineering and Policy with concentrations in Economics and Computer Science from Washington University in St. Louis.