Posted by Software Estimating in an Agile Environment
From the Journal of Cost Analysis and Parametrics: Volume 10 | Issue 2 | April 2022
Downloadable File: JCAPv10i2-SoftwareEstimatingAgileEnvironment-Goljan
Abstract: Defense organizations are moving towards agile methodologies as a preferred approach to software development. The desire to implement agile methods is discussed in the 2019 Defense Innovation Board (DIB) report, which identifies speed and cycle time as the most important metrics for software development (McQuade et al., 2019). This movement toward agile methodologies provides a conundrum for defense cost analysts. These cost analysts are proficient in developing software estimates based on commonly accepted defense sizing metrics such as Source Lines of Code (SLOC). But the agile environment is unique. The agile mentality relies on flexibility and working in small iterations. Utilizing metrics like SLOC are often discouraged as it constrains the team to a pre-conceived work estimate and because it can incentivize the contractor to develop inefficient code (Bhatt, Tarey, & Patel, 2012). As a result, agile programs require cost analysts to potentially adopt new methods for proper cost estimation. For example, agile programs may use techniques such as level of-effort estimates which incorporate the number of team members and the expected duration of time to work on a new requirement (Rosa, Madachy, Clark, & Boehm, 2020). Due to the DoD’s lack of experience and familiarity with agile, the objective of this article is to investigate the current state of agile software cost estimation and provide recommendations for cost analysts.
The DoD has only recently implemented agile software development, but the agile concept itself dates back to 2001 with the publication of the Agile Manifesto (Regan, Lapham, Wrubel, Beck, & Bandor, 2014). Since its inception, agile practices have become widely adopted throughout private industry (Randall, 2014). The private sector’s 20 years of experience provides an opportunity to uncover best practices for cost analysts in an agile environment. To study this, we first conduct an extensive literature review regarding the recommended agile software cost estimating models and techniques. The question then becomes, “how do the recommended techniques
align with the methods defense cost analysts are currently using?” To answer this, we collect data from 11 agile Air Force software factories to determine what practitioners actually do. Comparison of the two results will provide defense analysts with insight on differences between current DoD practices and those advocated by the published literature.
Authors:
Captain James Goljan, is a cost analyst at the Space Systems Command, Los Angeles AFB, CA. He holds a BS in Operations Research from the United States Air Force Academy and a MS in Cost Analysis from the Air Force Institute of Technology (AFIT). His primary research interests include agile software development, optimization models, and cost analysis.
Dr. Jonathan D. Ritschel is an Associate Professor of Cost Analysis in the Department of Systems Engineering and Management at AFIT. He received his BBA in Accountancy from the University of Notre Dame, his MS in Cost Analysis from AFIT, and his Ph.D. in Economics from George Mason University. Dr. Ritschel’s research interests include public choice, cost analysis, and economic institutional analysis.
Scott Drylie, Ph.D., is an Assistant Professor in the Department of Systems Engineering and Management at AFIT. He holds a BS in Economics from Montana State University, a M.Ed in Education from University of Nevada, a MS in Cost Analysis from AFIT, and a Ph.D. in Economics from George Mason University. Lt Col Drylie’s research interests include Smithian political economy, organizational behavior, public choice, and cost analysis.
Dr. Edward D. White is a Professor of Statistics in the Department of Mathematics and Statistics at AFIT. He received his BS in Mathematics from the University of Tampa, MAS from The Ohio State University, and Ph.D. in Statistics from Texas A&M University. His primary research interests include statistical modeling, simulation, and data analytics.