Notes on formal software development effort estimation and overall financial management of software projects

Abstract

Cost estimation in software engineering is one of the hardest challenges. Estimating the financial cost, total amount of man months needed in software development project or software size can be done in various methods. This research looks into some effort estimation methods and techniques. It explains the benefits and drawbacks on some of the selected techniques and methods, and gives grounds why understanding metrics is necessary before applying them into practice.

Algorithmic software effort estimation techniques are chosen from the effort estimation family and put into main focus in this report; besides being metrics by numbers, can the parameter information used in estimation process easily be combined to some data analysis techniques aiming to even more exact estimation results.

Being a competitive software house is a much larger concept than only being able to estimate the application development effort. For this reason an introduction to software development methodologies and capability maturity model is provided together with an explanation of their meaning to the software development and software effort estimation. Effort and cost monitoring are an essential part of the project management besides estimates and this study introduces one possible cost and effort monitoring model.

This research answers to the following dilemma: What does software development effort estimation mean and how does one estimate software development costs? What are formal effort and cost estimation methods? What other estimation methods exist? Why one needs to understand methods and metrics before applying it into practise; why results given by the same estimation method can vary by software houses? How to manage one’s software project performance?

This study covers only some models and can provide only limited information about the topic.