**5.2 Indicator metrics**

Mapped nodes in the graph model are either processing nodes or decision nodes (figure 3). Decision paths reflect how the users are making choices and thus directly impacts on "Edge Probabilities", based on the variations in Edge Probabilities the analysts may decide to rearrange the workflow in favor of currently used paths thus changing the branching factors as well as the underlying graph model.

Probing stations can collect node visit frequencies which will identify the popularity of workflow locations and, more specifically, certain features of the system. Feature Utilization (FU) frequency directly impacts the feature's rank. Variations in the rank are major indicators in identifying any redundant components (for elimination), under-utilized features (may need advertisements) and over-utilized features (that need to be efficient). With the changing rank in features each node's weight will vary. For certain nodes with no significant features the weight will eventually be reducded to zero and will have no contributions in the system value.

User satisfaction index (USI) reflects the outcome experience of a user or a group of users. A declining USI is a driving force to initiate major revisions into the system to meet newly emerged requirements. Overall system value is something that the stakeholders, architects, analysts and developers keep their eyes on. Based on the type of services provided they will derive a composite formula or index that best describes a system's overall performance at a particular instant of time. For example, a highly significant feature may yield a lesser system value during low usage activity for that cycle time.
