Performance Analytics and KPI Tracking
Introduction
Standard ServiceNow reports show you the current state of data — how many open P1 incidents exist right now. Performance Analytics (PA) shows you how that number has changed over time — whether your P1 backlog is growing or shrinking week over week, and how today compares to last quarter.
This guide covers PA fundamentals: indicators, data collectors, breakdowns, and dashboard widgets.
Performance Analytics vs. Standard Reporting
| Feature | Standard Reporting | Performance Analytics |
|---|---|---|
| Data currency | Real-time | Historical snapshots |
| Trend analysis | ❌ | ✅ |
| KPI tracking over time | ❌ | ✅ |
| Target/threshold lines | Limited | ✅ |
| Breakdown by dimension | Limited | ✅ |
| Data collection | On-demand | Scheduled |
PA doesn't replace standard reports — it adds the historical dimension they lack.
Core Concepts
Indicator
An Indicator is the KPI you want to track. It defines:
- Source table: Where to count records
- Filter: Which records to count
- Aggregate: COUNT, SUM, AVG, etc.
- Collection frequency: How often to snapshot (daily, weekly)
Examples:
- Open P1 Incidents (count of active incidents with priority=1)
- Mean Time to Resolve (average resolve time on closed incidents)
- SLA Compliance % (% of resolved incidents within SLA)
Data Collector (Snapshot)
Each scheduled collection run takes a snapshot of the indicator value and stores it with a timestamp. Snapshots accumulate over time, building the historical dataset that enables trend charts.
Breakdown
A Breakdown adds a dimension to an indicator — splitting the KPI by assignment group, priority, category, or any other field. This allows charts like "Open Incidents by Assignment Group over Time."
Widget
Widgets display PA data on dashboards:
- Score widget: Shows current value with trend arrow
- Time series chart: Line/area chart of values over time
- Breakdown chart: Bar chart comparing segments
- Heatmap: Intensity grid of two dimensions over time
Creating an Indicator
Performance Analytics > Indicators > New
Example: Daily Open Incident Count
| Field | Value |
|---|---|
| Name | Open Incidents |
| Table | Incident [incident] |
| Filter | Active = true |
| Aggregate | COUNT |
| Direction | Decrease is good |
| Frequency | Daily |
| Active | ✅ |
Adding a Breakdown
On the Indicator record → Breakdowns related list → New:
- Breakdown source: Assignment group (from incident table)
- This creates segmented snapshots by assignment group each collection run
Setting Targets
Targets allow you to draw a goal line on trend charts:
Indicator → Targets related list → New
- Period: 2024-Q4
- Target value: 50 (goal: keep open P1s below 50)
On charts, the target appears as a reference line, making it immediately visible whether performance is above or below goal.
Collecting Historical Data
PA only knows about the future by default — it starts collecting from when you enable the indicator. To backfill historical data:
Indicator → Scheduled Data Collection → Run Now
— or —
PA Admin > Data Collector Jobs → run for past dates
Note: Backfill is approximate — it uses current filter conditions applied to historical sys_created_on data, not true historical snapshots.
Building a PA Dashboard
Create a Dashboard
All > Dashboards > New
Add PA widgets from the widget picker:
Recommended layout for an Incident KPI dashboard:
Row 1: [Score: Open P1s] [Score: SLA Compliance %] [Score: MTTR]
Row 2: [Time series: Open Incidents by Priority - Last 90 days]
Row 3: [Breakdown: Open Incidents by Assignment Group]
Row 4: [Heatmap: Incidents by Category × Day of Week]
Configuring a Score Widget
- Select Indicator: Open Incidents
- Display: Current value, trend arrow, sparkline
- Comparison: vs. 30 days ago
- Threshold coloring: green <50, amber 50-100, red >100
Automated Insights
PA's Automated Insights feature (available in newer releases) applies machine learning to detect:
- Anomalies (unusual spike or drop)
- Trends (consistently increasing/decreasing)
- Forecasts (projected value in 30/60/90 days)
Enable on each indicator record → Automated Insights tab.
Best Practices
- Start with 5-7 indicators, not 50 — focus on what actually drives decisions
- Collect daily for operational metrics, weekly for strategic KPIs
- Always configure a Breakdown for actionable metrics (so teams know who owns what)
- Set realistic targets before the first collection run
- Review PA dashboards in weekly operational reviews — the data is only valuable if it's looked at
- Don't delete old indicators — historical data cannot be recovered
- Use Direction field (Increase/Decrease is good) to enable correct trend coloring
Conclusion
Performance Analytics transforms ServiceNow from a system of record into a system of insight. The combination of scheduled snapshots, breakdowns, targets, and trend widgets gives teams the visibility to understand whether their processes are improving over time — not just what's happening right now. Start with a focused set of indicators tied to your service level commitments, collect consistently, and review in your regular operational cadence.