At its most basic level, grants making is a series of business processes made up of highly interdependent work flows. Many of these processes are performed in parallel, others play out in specific sequences. They range from very flexible to highly rigid. Some processes are carried out within single departments; others span multiple organizations across the enterprise. Understanding how work is done within and among the organizations in the enterprise is therefore a critical building block of the research agenda.

The essence of workflow is role-based routing that reflects complex organizational policies. The proposal preparation, submission, review, and management stages of the grants-making process (illustrated in highly simplified form in figure 2) are governed jointly by the policies and practices of many organizations. As the enterprise evolves from paper-based and manual workflow systems to sophisticated electronic work flow systems, we need increasingly detailed and comprehensive models of how work is done.

Work-flow technologies deal with work processes as a collection of activities that support a specific business process. This overall business process is treated as consisting of two work-flow components, both of which are critical to smooth and effective operation. These are control flow and data flow. The mechanisms and sequences of events that determine the path and pace of the work are known as the control flow. Some activities in a process can be performed in parallel, while others must be done in sequence, thus there are multiple paths or flows possible for a business process. Data flow refers to the movements of data and documents that accompany or make up each activity in a business process.

For work-flow technologies to be effective in grants-making they must meet several criteria. They must enable researchers to focus on the content of a proposal, reduce overall human effort, minimize the number of steps in a process, and minimize the time to process a proposal. Although many current workflow systems meet some of the above criteria, none appears to address fully the complex interplay of organizational policies and interorganizational relationships involved in the grants-making process. Accordingly, future research should focus on the development of next generation workflow systems that include features such as a meta model or language for describing complex group activities and tasks. Further, they must capture complex organizational policies and handle the cross-organizational implications of policy or process changes. Research is also required to develop a framework for dynamic workflow schema changes that allow for changes to be made "on the fly" rather than necessitating wholesale upgrades or re-engineering of systems.

By its nature, the grants-making process is collaborative and hence needs to support collaboration tasks within and across organizations. Most current workflow technologies lack this support and therefore present another area for future research: the development of robust architecture and applications that facilitate seamless integration of interorganizational workflow and collaboration support systems.