The Humphrey Award nominee's productivity improvement must, to an exceptional degree, be significant, measured, sustained, and shared.
- To be significant, the work must have a demonstrated impact on the target organizations' software engineering results (e.g., productivity increases, defect density for software reaching integration and test decreases, or rework percentage decreases, and by how much). It should also consider, as pertinent, the impact on the target organizations' management, workforce, and system development and maintenance processes (e.g., cycle time to address emergency tickets decreases and percentage of tickets correctly fixed on first try increases, and by how much).
- To be measured, the work must include the collection of data used to guide the work and clearly demonstrate its impact. Concrete software engineering process improvement goals, factors, and metrics must be defined. The work must involve assessing the cause-and-effect relationship between changes (e.g., to the organization’s processes, training, and tools) and their impact (to reduce defect density, reduce cycle time, or increase productivity).
- To be sustained, the work must lead to a continuing impact on the target organizations' software engineering activities. In particular, the work should result in well-documented support (e.g., in the form of coaching, training, tools, measures, and obtaining feedback on changes) for effective process performance and continuous software engineering process improvement.
- To be shared, the insights, experiences, and proven practices stemming from the work must have been made available beyond the target organizations. This could, for example, be to other parts of larger organizations within which the target organizations are subunits. It could, additionally or alternatively, be throughout software engineering process improvement communities in which the nominee and target organizations participate.