Research Hypotheses

By October 2000, Escrow.com was experiencing severe and growing problems in its development efforts, including a dramatically slowed pace of delivery, increasing development costs, poor product quality, and a deteriorating state of the code base. Early process improvement efforts were largely unsuccessful at solving these core problems.

Senior management recognized that a fundamental change to the development process was needed. Spearheaded by a team of senior developers, research was conducted using XP. Afterward, we hypothesized that adopting XP would deliver the following benefits:

1. Increased rate of development, enabling the development efforts to keep pace with the increasing demand for new features

2. Reduced development costs, enabling a reduction in the size of the development team or deploying existing resources to new projects

3. Increased correlation of software to business needs, bringing the delivered releases in line with the requirements of the end users and eliminating unnecessary or low return-on-investment features

4. Reduced release cycle, enabling frequent releases of high-priority features

5. Increased product quality, reducing the quantity and severity of defects that delayed production releases

6. Increased quality of implementation, reducing the code entropy that hindered the addition of new features and increasing the maintainability of the code base

Existing project management and change control practices produced metrics that measured developer effort and defect discovery rates. Combined with code analysis, this data enabled objective measurements of the improvement from adopting XP.