Manage custom KPIs

Custom KPIs let you define and monitor business-specific metrics at the workload level, connecting technical performance to business outcomes. Available exclusively for Intelligent workload, KPIs provide visibility into metrics that matter most to your business.

Default KPIs

When you create an Intelligent workload, New Relic automatically provides the following two default KPIs focused on your focal transaction. These KPIs are uneditable but can be hidden if not needed:

Workload P99 Latency

The maximum time it took for the fastest 99% of executions of the focal transaction to complete. Only 1% of requests were slower than this value. This KPI monitors tail latency and helps prevent user frustration, ensuring that even during peak loads or background processing spikes, the workload does not suffer from significant bottlenecks.

Workload Success Rate

The percentage of successful executions of the focal transaction, calculated as the number of successful requests divided by the total number of requests.

Add a KPI

You can add KPIs to your Intelligent workload in two ways: create custom KPIs or add existing KPIs from other workloads.

Create a custom KPI

Create new KPIs to track the metrics most important to your business:

1. Open the summary page of the Intelligent workload.
2. Click Add new KPI You can either create a new KPI by writing a NRQL query to define a custom metric or using an existing KPI from other workloads.
3. To create a custom KPI:
   • Name: Give it a descriptive, business-friendly name
   • NRQL query: Write a query that calculates the metric
   • Group (optional): Organize KPIs by functional area (e.g., "Performance", "Business Metrics")
   • Category (optional): Further classify KPIs (e.g., "Latency", "Throughput", "Revenue")
4. Click Preview to verify the query returns expected results
5. Click Save

View and monitor KPIs

After adding KPIs, up to 3 can be displayed on your workload summary page as tiles. Each KPI tile shows:

• Current value: Real-time KPI measurement
• Trend indicator: Visual showing if the metric is improving or degrading
• Time series chart: Historical performance over the selected time range
• Anomaly detection: Highlights when KPIs deviate from normal patterns

The 3-KPI limit ensures your summary page remains focused and easy to scan. Use the Manage KPIs page to view the complete catalog of all KPIs available for your workload.

Create KPI alerts

Alert on KPIs to notify teams when business-critical metrics deviate from goals:

1. Click the KPI you want to alert on
2. Select Create alert condition
3. Configure the alert:
   • Threshold type: Above or below target value
   • Threshold value: Numeric value that triggers the alert
   • Duration: How long the threshold must be breached before alerting
   • Notification channels: Where to send alerts (Slack, PagerDuty, email, etc.)
4. Give the alert a descriptive name
5. Save the alert condition

Edit or delete KPIs

The Manage KPIs page provides a central location to create, edit, delete, and view all KPIs for your Intelligent workload:

Edit a KPI

Modify KPI details as your business goals evolve:

1. Go to Manage KPIs in your Intelligent workload
2. Find the KPI you want to modify
3. Click next to the KPI name
4. Select Edit
5. Update any of the following:
   • NRQL query: Modify the query logic
   • Name: Change the KPI display name
   • Group: Update the functional grouping
   • Category: Adjust the classification
6. Click Preview to validate changes
7. Click Save

Delete a KPI

Permanently remove KPIs that are no longer relevant:

1. Go to Manage KPIs
2. Find the KPI to remove
3. Click next to the KPI name
4. Select Delete
5. Confirm deletion

Hide default KPIs

If the default KPIs don't align with your monitoring needs:

1. Go to your workload summary page
2. Find the default KPI tile
3. Click
4. Select Hide from summary

Hidden KPIs remain in the catalog and can be shown again later.

Next steps

Enhance your Intelligent workload monitoring:

• Use AI insights: Get AI-generated summaries referencing your KPIs
• Configure workload alerts: Set up notifications for workload status changes
• Learn NRQL query syntax: Master query writing for complex KPIs
• Create custom dashboards: Visualize KPIs alongside other metrics