CAS-PA — Frequently Asked Questions

Formula indicators and external indicators should be excluded from Historical Collection jobs. Historical collection is designed to backfill data for automated (real-time) indicators by running them over past dates. Formula indicators are calculated on-the-fly from other indicators, so backfilling them is unnecessary and can cause incorrect results. External indicators pull data via JDBC from third-party systems and should not be included in PA historical collection jobs either.

Indicator frequency determines how often data is collected (daily, weekly, monthly, quarterly, yearly). Breakdowns allow you to segment indicator data by specific fields (e.g., category, assignment group, priority). When you define a breakdown on an indicator, PA collects data for each unique value of that field at each collection interval. The frequency of the collection job must match or be a divisor of the indicator frequency. For example, a daily collection job can feed a monthly indicator, but the system aggregates the daily snapshots into monthly values.

An anti-signal in Performance Analytics indicates a negative or undesirable trend in a KPI. When KPI Signals are enabled, an anti-signal is triggered when the current KPI score moves in a direction that is considered bad for the business (e.g., increasing incident count, decreasing resolution rate). Anti-signals are automatically detected based on the KPI's baseline and trend analysis, and responsible users receive notifications. You can configure the baseline calculation period, number of scores used, and trend method to fine-tune signal detection.

A formula indicator calculates its value using a formula that references other indicators (automated, manual, or other formula indicators). It supports mathematical operators (+, -, /, *, %) and comparison operators (>, <). The formula uses double bracket notation [[Indicator Name]] to reference other indicators. Formula indicators are recalculated in real-time based on the underlying indicator values and do not require separate data collection jobs. They cannot be directly assigned to a breakdown — only the indicators they reference can have breakdowns.

The valid PA widget types are: Time Series (shows changes over time, visualized as line/spline/step/column/area charts), Breakdown (groups scores by breakdown elements, visualized as scorecard/pie/funnel/pareto/treemap), Score (shows aggregate scores against targets, visualized as latest score/speedometer/real-time score), List (lists metrics for multiple indicators), Pivot (groups one indicator by two breakdowns, visualized as heatmap), Text Analytics (visualizes word frequencies as word clouds), and Workbench (shows multiple indicators and their relations to monitor workflows).

The pa_viewer role is required to create personal targets in the Analytics Hub. Users with the pa_viewer role can view the Analytics Hub, create personal thresholds and targets for indicators, and manage the thresholds and targets they created. The pa_target_admin role allows users to create and manage ALL targets (including those owned by others), but for personal targets specifically, pa_viewer is sufficient.

To activate KPI Signals for an indicator, navigate to the indicator record, go to the KPI Signals related list, and click 'Activate KPI Signals'. You must be designated as a responsible user for that indicator. After activation, configure signal detection settings including the baseline calculation start date, number of scores used for baseline, and trend method. You can also assign responsibility for specific breakdown elements. Automated detection jobs run periodically to identify signals and notify responsible users. Formula indicator signals require a separate scheduled job.

To enable real-time updates for a score visualization, you need to use a Score widget type with the 'Real-time score' visualization option selected. Additionally, the underlying indicator must have 'Real-time' collection enabled (set the collection frequency to 'Real-time' or ensure the data collector job runs frequently enough). The widget will then display the current KPI value along with the previous value and the change percentage, updating automatically as new data is collected.

Platform Analytics Workspace is the unified entry point that provides a personalized dashboard experience with quick access to your most-used PA content, including scores, widgets, and reports. Analytics Center is the central repository and governance hub where administrators manage all PA content — including indicators, breakdowns, widgets, dashboards, color schemes, and targets. Analytics Center provides a comprehensive overview of all Performance Analytics content and its usage, while Workspace is focused on individual user consumption and exploration of analytics data.

The 'State' breakdown should be excluded for 'Number of Open and Overdue Incidents' historical data collection. This is because the state field changes frequently as incidents move through their lifecycle. Collecting historical snapshots with state-based breakdowns can produce misleading trends since an incident's state at snapshot time may not reflect its actual behavior during that period. Instead, use stable breakdowns like category, priority, or assignment group that provide more consistent historical trends.

By default, a newly created indicator is visible only to the creator ('Visible to Just Me'), and 'Visible by All Roles' is set to False. This means the indicator is private and not accessible to other users until the creator explicitly configures broader visibility options. This default ensures security and control over potentially sensitive or unfinished indicator definitions.

The 'Indicator Usage' feature (accessible from the widget record) shows all indicators used by a specific widget and its associated tabs. This helps administrators understand dependencies between widgets and indicators, making it easier to manage changes, troubleshoot issues, and perform impact analysis when modifying or deleting indicators. You can access this from the widget form under the Related Links section.

The sn_pa.incident facts table should be used for Incident Close Code breakdown in ServiceNow Performance Analytics. This facts table contains incident data aggregated for PA queries and includes the close code field, allowing you to create breakdowns by incident close code. The standard PA facts tables (sn_pa.incident, sn_pa.problem, sn_pa.change_request, etc.) are optimized for PA queries and should be preferred over querying the base tables directly.

When a KPI target is missed on the review date, the system marks the target status as 'Missed' and this status is reflected in the scorecard visualization. The KPI Signals feature can also detect this miss and generate an anti-signal notification to responsible users. The target status options typically include 'On Track', 'At Risk', 'Missed', and 'Achieved', depending on how the target thresholds are configured. Review dates can be configured to trigger status evaluations at specific intervals.

A syntactically incorrect PA formula would be one that uses improper bracket notation or invalid operators. For example, '{indicator:open_incidents} / {indicator:total_incidents} 100' is incorrect because the correct syntax uses double square brackets: [[open_incidents]] / [[total_incidents]] 100. Other common syntax errors include: using single brackets [indicator_name], using curly braces {indicator_name}, omitting the double brackets, or using invalid operator combinations. Valid formulas use [[Indicator Name]] syntax with operators +, -, /, *, %, >, <.

User Experience (UX) filters in Performance Analytics are created through the dashboard widget configuration. When editing a PA widget, you can add filters that allow end users to dynamically filter the data displayed in the widget without modifying the underlying indicator definitions. UX filters are based on breakdown fields and are presented as dropdowns, date pickers, or multi-select widgets on the dashboard. They enable interactive exploration of KPI data while preserving the integrity of the indicator definitions.

The Field Attribute in the Breakdown Source form defines how a specific field's values are mapped and categorized within a breakdown. It allows you to create custom groupings, labels, and sorting orders for breakdown elements. For example, you can map multiple numeric values to a single category label (e.g., mapping severity values 1-3 to 'High Priority'). Field Attributes also enable you to exclude certain values from the breakdown and define the display order of breakdown elements in visualizations.

The globalTarget() function in ServiceNow Performance Analytics returns the global target value for an indicator. This function is used within formula indicators to compare current scores against organizational-wide targets. It can also be referenced in widget configurations to display target lines on charts. Other related functions include target() which returns the target for a specific breakdown element, and baseline() which returns the calculated baseline value.

When 'Collect breakdown matrix' is unchecked on an indicator, PA collects breakdown data independently for each breakdown level but does NOT collect the combined cross-tabulation (matrix) of multiple breakdowns. This means you can view each breakdown separately (e.g., by Assignment Group OR by Category), but you cannot drill down to see one breakdown nested within another (e.g., Assignment Group within Category). Enabling 'Collect breakdown matrix' is required for multi-level hierarchical breakdown analysis, but it consumes more storage and collection time.

A Bucket Group in ServiceNow Performance Analytics is used to categorize and group indicator scores into predefined ranges or buckets for visual display. It allows you to define color-coded thresholds (e.g., Red for 0-30, Yellow for 31-70, Green for 71-100) that are applied consistently across multiple widgets and dashboards. Bucket Groups provide a standardized way to represent performance levels visually, making it easier for stakeholders to quickly assess KPI status at a glance without interpreting raw numbers.