Risk based analogy for e-business estimation

  • Elias Pimenidis

Student thesis: Doctoral Thesis


Over the past fifteen years, new technologies have enabled the evolution of e-business and as new trends on web based applications develop at a very fast pace, projects to implement new and adapt existing e-business systems are becoming more complex in their structure. At the same time, organizations involved as stakeholders are becoming more demanding in the delivery of fast, accurate and timely results. Where interconnected e-business systems, such as Value-Added-Communities, are concerned the key criteria of success, in addition to the quality of the technical work, are the timely completion of the project and an accurate estimate of the launch time. Given the nature of such e-business development projects, estimating their duration cannot be based on traditional estimation methods that target the estimation of human effort originally and then attempt to convert this to cost and duration. This research explores a new approach aiming to improve the accuracy of estimates for ebusiness development projects, obtained at the early stages of the project. The key objective is to modify the estimation by analogy approach by using risk as the key element that along with project size and complexity are used to identify analogues and determine analogies amongst projects.
The author established the behaviour of risk with the variation of certain project attributes that define a project’s size and complexity at the very early stages of the project, before detailed requirements are identified. The variation of risk was subsequently employed, with a modification to the estimation by analogy method, to establish analogies on suitably identified analogues to the project to be estimated. The ISBSG project data repository has been used as source for potential analogue projects in testing the proposed method. Four project cases were considered. New estimates have been obtained using the modified estimation approach method and the results were compared to the original estimates for the projects, obtained using traditional methods.
Date of AwardJul 2011
Original languageEnglish

Cite this