KPI Definition

How to Create a KPI Definition

As Administrator User you can create a KPI in two ways.
First one: starting from the menu “KPI definition”

The KPI Definition user interface let you to add, delete and update the KPIs, so you can manage your KPIs library and link every KPI to different models and different model instances.

Second one: starting from the Model Definition
expand the panel “Kpi list” and click on to create a new KPI then drag and drop the KPI to bind it to the node of the model.

Add a KPI

Click on to insert a new KPI. The user interface for the KPI Definition is structured in four tabs: Detail, Advanced, Udp values, KPI Links.

Detail Tab

Fill the properties in the form:

• Name: name of the kpi
• Code: mnemonic code. Name and Code will be displayed where you have to select a KPI relation. Code is also displayed in the tree view of the models.
• Description: it describes the kpi
• Dataset: dataset implements the algorithm to calculate the KPI value (for example: query, script, java class,..). You can select ones from the list of existing Datasets or you have to create ones before using “KPI Definition” panel.
• Threshold Name: You can select ones from the list of existing thresholds or you have to create ones before using "thresholds Definition" panel.
• Documents: label of the analytical document. You can define an analytical document, for the KPI, that can be accessed directly from the KPI dashboard by clicking the document icon.

• The link to the analytical document is automatically produced by the current dashboard layout template that by default is composed by:
  • ParKpiResource = value of the relative resource
  • ParModelInstance = Model instance id
  • ParKpiInstance = kpi Instance id
  • OtherParam = All values of the analytical drivers filled by the user in the original Kpi Document with relative urls

• Weight: you can assign a weight to manage kpi aggregation. For example you could create a kpi for a node i as weighted media of the kpi of the nodes i-1.

Advanced Tab

The Advanced tab let you define more structured information for better describe the KPI under definition.

The Advanced tab let you define more structured information for better describe the KPI under definition.

• Interpretation: provides guidance to help the users in interpretation of KPI results and representation
• Algorithm description: it’s a free textual description. It describes the algorithm needed to the computation of the KPI. The algorithm will be implemented by the “Dataset”.
• Target Audience: specifies how relevant stakeholders may notify comments, improvements or modification of results and processes.
• Input attribute: specifies the source of the data and each attribute needed to calculate the metric (from a logical point of view)
• Model Reference: specifies process and activity linked to the measurement objective. For example, the metric could be linked to CMMI process area and specific practice or to a specific IT service task.
• KPI type: the kpi or metric could be “basic” or “derived”
• Measure type: specifies the type of the measure: COUNT TIME SIZE
• Metric scale: specifies the type of the scale: nominal, ordinal, intervals, ratio, absolute, day.

KPI Links Tab

If your KPI value depends on the result of the calculation of other KPI values, previously created, go to the Kpi Links tab and assign a KPI to each the Dataset parameters of the selected KPI. Just click on the lens icon and the list of existing KPIs will appear.