Performance assessment, as we are using the term here, does not require collecting data. We are attempting to make forecasts or predictions: how well do we expect Web-based services to perform? Neither optimism nor pessimism is advantageous here--only realism. For each level of aspiration (i.e., modest, moderate, and elaborate), targeted measures of performance are required. These targets will most likely be "judgment calls," since no perfect prediction will be available. As part of the Internet Services Testbed, we asked agency teams to discuss and agree on these targeted measures of performance for their own service objectives. No one felt entirely comfortable with their forecasts but everyone agreed that they were using their most informed judgment at the time. For estimates such as these, informed, consensual judgment from a group of knowledgeable managers is probably the best forecasting approach that we can use in this context.

A thorough treatment of performance measurement is beyond the scope of this guide. A good discussion of how to develop meaningful measures can be found in Information Management Performance Measures: Developing Performance Measures and Management Controls for Migration Systems, Data Standards, and Process Improvement, available from the National Academy of Public Administration, 1120 G Street, NW, Suite 850, Washington, DC 20005.

A sample worksheet for identifying performance variables, measures, and targets is shown below. For examples of its use, see p. 27 and p. 32. Often, only a few performance variables are identified, and this is sufficient to drive the analysis. On the other hand, you may decide that more performance variables are required to measure fully the improvements that you anticipate from WWW use. In principle, there is no limit to the number of performance variables that could be incorporated into this assessment.

Figure 4. Blank Performance Targets