Managing Business Processes with Metrics

Typically, data describing project status and performance are collected and reported in a variety of ways. Sometimes project managers stop short of getting real value from the numbers they have collected. The least effective way to understand the meaning of project data is placing it in alphanumeric tables that are not only very difficult to understand but impossible to analyze for trends. Data tables should be used as a source for displaying data, not as the final presentation method. Use of metrics charts a Quantitative Management Plan (QMP) is designed to help you focus on detecting specific problems and risks to enable development of accurate, timely and cost-effiective solutions. The solutions may range in scope from making small adjustments in process implementation to doing major project re-planning.

The best method for identifying, visualizing and analyzing metrics data is to describe them in a QMP. The QMP contains information about how to collect and apply data gathered during implementation of project processes to assess the need for constructive change while a project is underway. The data featured in the QMP focuses on use of metrics charts to describe project conditions and situations. It describes why metrics are important, what they measure, how they are collected and when, how they are constructed, what the data source is, and why it is important to report them to project management on a regular basis.

Two major types of metrics are commonly described - project progress and project defects. One is a category of metrics that describe project performance, including progress in terms of tasks planned and accomplished, cost estimating, scheduling, budgeting, and stability in size and requirements. Other metrics are collected to evaluate the accuracy of planning, and to predict problems or risks if published milestones are not accomplished as planned.

The other major metrics category concerns defects. Although many defects are found during independent review and peer review of work products, too many defects are typically found during testing. Thorough and frequent review of work products during the requirements analysis, design and implementation phases vastly reduces the number of defects that escape detection and result in expensive and time consuming fixes during unit, function, and system integration testing. Defects that escape detection until after delivery to the customer or to the intended user are the most expensive to fix.

The Quantitative Management Plan focuses mostly on characterizing defects and monitoring progress in resolving them. The primary purpose of collecting and reporting defects as early as possible in the software development cycle is to ensure that managers can make corrections in project implementation and provide opportunities for re-planning as soon as possible to minimize risk to project completion. Also, data analysis provides insights into process improvement opportunities that do not otherwise exist.