Members Log In to My ASQ Members Log In   View Shopping Cart Shopping Cart   Quality Progress Magazine Quality Progress Magazine Make Good Great
Magazines & Journals
Software Quality Professional

Printer Friendly
Issues
I Want To
Article Access Key
  • Public Article
  • Log-In to View
  • Full, Senior, or Fellow members with no subscription.
  • Full, Associate, Forum/Division, Senior or Fellow members who are also subscribers.
  • Enterprise and Site Members have access to all issues.

March 2001
Volume 3 • Number 2

Contents

PROJECT MANAGEMENT
Credible Estimation for Small Projects

by Rita Hadden, Project Performance Corporation

The author presents a repeatable process for generating defensible estimates on projects of two to six months in duration, with two-to-six full-time equivalent personnel. After initial calibrations for size and complexity, the process uses work breakdown structure (WBS) assignments to build up estimates early in the project. By making explicit the basis of the estimating (including assumptions), this method aids in initial negotiation with management and subsequent renegotiation when project circumstances change.

Key words: collaboration, commitments, key deliverables, negotiation, scheduling, work breakdown structure (WBS)

INTRODUCTION

“What’s the point in doing project estimation when the end date has already been fixed before we start?” How often does one hear this question from software project leaders and practitioners? The author can personally testify, having navigated through more than 45 organizations, small and large, that it is the rare software organization where this belief is not a cultural norm.

“Gut-feel” estimation is often practiced in software development organizations, particularly those with many small projects–small meaning projects with two-to-six full-time equivalents, lasting two to six months. The project leader for these small projects often feels he or she cannot do a credible estimate of the effort and schedule for the project prior to analysis and detailed design. Yet “up front” is when senior management wants the estimates. As a result, gut-feel effort estimates are often padded to make up for uncertainty.

Without a defensible estimation method and historical data to back up the estimates, the project leader often does not get what he or she asks for in terms of time and resources. In fact, senior management often wants to know, “Why should this project take so long to do?” A project leader who does not have an objective and repeatable way to perform estimation is typically at a disadvantage when estimates are challenged. The vicious cycle starts when senior management requests that the project leader “Find a way to get the project done in half the time with the available resources.”

How can one break this vicious cycle? Following is a simple yet credible estimation method. This method has been used by many organizations doing small software projects and is based on more than 60 small projects in the private and public sectors. It does not require counting function points or lines of code. This method can be used up front, often within three to four weeks of project start, as soon as detailed requirements, high-level design, and work breakdown structure (WBS) are available. It is designed for software organizations and individuals who want to estimate in a more repeatable manner and work to improve their project performance and operational excellence. This method enhances trust and respect between senior management, project leaders, and practitioners.

Real-world examples are used throughout this article. A real-world project is also used to illustrate how estimates using this method can be used to successfully negotiate scope, resources, and schedule with management.