CIS-DF Hub Part 7 Performance Analytics
⬡ Part 7 · Topic 4

Performance Analytics

Standard reports show what's happening now. Performance Analytics captures metric snapshots over time so you can track trends, compare periods, and forecast. It requires deliberate setup — Indicators and Data Collectors must be configured before historical data accumulates.

📋 5 sections ~16 min read 🎯 ~7% exam weight 🏷 Performance Analytics · Indicators

Why Performance Analytics Exists

Standard ServiceNow reports are point-in-time. If you run "How many open P1 incidents are there?" today, you get today's count. You can't run it for last Tuesday — that data is gone unless you took a snapshot. Standard reports show current state, not history.

Performance Analytics (PA) solves this by running scheduled jobs (Data Collectors) that count records matching a condition at a specific time and save those counts as Scores. Over time, these scores build a time-series dataset that enables trend analysis, period comparison, and KPI tracking.

🧩
Analogy — Fitness Tracker Standard reports are like stepping on a scale once. Performance Analytics is like a fitness tracker that records your weight every morning. Both tell you your weight right now, but only the tracker can tell you: "Did you gain or lose weight this month? What's your trend?" PA is the daily weigh-in for your CMDB metrics.

Indicators and Scores

An Indicator is the definition of what to measure. It specifies:

  • Table — which table to query (e.g., incident)
  • Filter — encoded query to count (e.g., state=In Progress)
  • Aggregate — COUNT (number of matching records) or SUM (sum of a numeric field)
  • Direction — is higher better (e.g., "resolved incidents") or lower better (e.g., "open critical incidents")?
  • Target — optional threshold for green/amber/red scoring

Each time the Data Collector runs for an Indicator, it records a Score — the value of the indicator at that moment, timestamped. Scores are stored in pa_scores.

Exam Trap — PA Requires Historical Data Collection to Visualize Trends You cannot retroactively fill in Performance Analytics history. If you create an Indicator today, you only have trend data from today forward. "We want to see the last 6 months of CMDB health trends" requires that the Indicator was set up and running 6 months ago. This is a common gap discovered when organizations start using PA for the first time.

Data Collectors

A Data Collector is a scheduled job that reads Indicator definitions and calculates scores. It runs on a configured frequency (hourly, daily, etc.) and stores a Score for each active Indicator.

Data Collectors are separate from Indicators — one Data Collector can process many Indicators. If a Data Collector job fails, no scores are saved for that period (a gap in the trend line).

Breakdowns

An Indicator can be further split into Breakdowns. A Breakdown is a dimension for segmenting the Indicator score. For example:

  • Indicator: "Count of open incidents"
  • Breakdown: "by Priority"
  • Result: separate score tracked over time for P1, P2, P3, P4

This enables charts like "Show me the trend of P1 incidents over the last quarter" without creating a separate Indicator for each priority.

PA vs. Standard Reports

FeatureStandard ReportsPerformance Analytics
Data freshnessLive — current state at query timeHistorical snapshots at collection intervals
Trend chartsNot supported (no history)Core feature — trend over time
Period comparisonNot supportedCompare this month vs. last month
Target/threshold trackingNot supportedGreen/amber/red targets built in
Setup requiredJust create a reportDefine Indicators + Data Collectors first
Historical backfillN/ANot possible — only forward-looking

CMDB Health and Performance Analytics

CMDB Health scores (from CMDB Workspace) are separate from Performance Analytics indicators, but they can be connected. You can create a PA Indicator that queries the CMDB Health score values over time, enabling trend charts for health metrics like:

  • "How has the completeness score of cmdb_ci_server changed over the past 3 months?"
  • "Is the overall CMDB health score improving after our data quality initiative?"

Without PA, CMDB Workspace only shows the current health score — you'd have no way to know if the 78% score today is better or worse than 3 months ago.

💡
Tip — PA Premium Includes Pre-Built CMDB Indicators ServiceNow includes pre-built Performance Analytics content packs for CMDB Health (available with PA Premium license). These include ready-made Indicators, Data Collectors, and Dashboards for CMDB completeness, correctness, and compliance trends. Starting from pre-built content is faster than building from scratch.

Common Exam Traps — Performance Analytics

  • Standard reports = current state; PA = historical time-series (trend charts)
  • PA cannot backfill history — data collection must run for the period you want to trend
  • Indicator = definition of what to measure; Score = saved value at a point in time
  • Data Collector = scheduled job that calculates scores for Indicators
  • Breakdown = segment an Indicator by a dimension (e.g., by Priority, by Team)
  • Scores stored in pa_scores table
  • PA Trend report type requires PA data — not available for ad-hoc table queries

Practice Questions

4 questions · Select an answer to see the explanation immediately.

Performance Analytics Quiz 1 / 4

A CMDB Manager wants to see a trend chart of CMDB Health scores over the past 12 months. The standard Report Builder only shows today's scores. What is needed?

💡 Explanation

B is correct. Performance Analytics is designed for historical time-series tracking. Configure a PA Indicator that captures CMDB Health score, and a Data Collector that runs periodically (daily, weekly) to save the score as a PA Score in the pa_scores table. After collecting data over time, PA can visualize it as a trend chart. Standard Report Builder (C) only shows current table state — it can't show how a value changed over 12 months because the historical values aren't stored in the source table. PA cannot backfill history (D) — it can only collect going forward from when it's configured.

Performance Analytics is configured for CMDB Health score tracking, but the Data Collector was only activated 2 months ago. The team wants to show a 12-month trend. What constraint exists?

💡 Explanation

B is correct. Performance Analytics can only track data going forward from the moment a Data Collector is activated. There is no mechanism to retroactively compute or import historical PA scores. If the Data Collector started 2 months ago, only 2 months of pa_scores data exists. The trend chart will have data for those 2 months only. This is a critical operational implication: you should configure PA indicators and start collection as early as possible — even before you have a use case — so historical data is available when you need it. Increasing collection frequency (D) just creates more data points in the 2-month window, not historical data from before activation.

A PA Indicator tracks "Number of Open P1 Incidents." A Breakdown is added by "Assignment Group." What does this enable?

💡 Explanation

B is correct. A PA Breakdown segments an Indicator's score by a dimension. Adding "Assignment Group" as a Breakdown creates child score series for each group — you see the overall total AND the per-group breakdown, all with historical trend data. This enables analysis like: "Network team's P1 incident count has been increasing quarter-over-quarter while Database team's is stable." Without a Breakdown, you'd only see the total count. Breakdowns are PA's aggregation/segmentation feature — they're not filters (A), they're additional scoring dimensions on top of the Indicator.

What is the key difference between a standard ServiceNow Report and a Performance Analytics Trend chart?

💡 Explanation

B is correct. The fundamental distinction: Standard reports query live table data and show its current state — they reflect what's in the database right now. PA Trend charts visualize how a metric changed over time by reading from the pa_scores table, which contains point-in-time snapshots collected by Data Collectors on a schedule. If you deleted 100 incidents today, a standard report would show 100 fewer incidents; a PA Trend chart would still show the historical peak because those scores were saved in pa_scores. PA is the only mechanism in standard ServiceNow for "what was the value of this metric last month/quarter/year?"

CMDB Workspace & NLQ →