Try it free

Alert from business events or flow

  • Latest Dynatrace
  • Tutorial
  • 6-min read
  • Published Sep 08, 2026

This use case shows how to configure alerting on business events using Anomaly Detection - new Anomaly Detection and problem notifications to automatically notify you when a business metric deviates from its expected behavior.

Target audience

This article is intended for business analysts and process managers who monitor how their businesses are performing in real time and want to be alerted when key business metrics such as revenue, order volume, conversion rate, or business exceptions move outside of the expected ranges. For this use case, we assume that you have a basic knowledge of how business events are captured and, for the second approach, some familiarity with Dynatrace Query Language (DQL).

Before you begin

This guide covers two alerting approaches for business events. Use the section that matches your situation:

  • Alerting from a business flow—the fastest option if you already have a business flow configured. Business Flow Business Flow pre-fills the DQL query and opens the anomaly detection configuration for you.
  • Alerting using Anomaly Detection—a custom alert that you build from scratch, where you provide the DQL query and the alert parameters yourself.

Additionally, you should determine which:

  • Business KPI or metric you want to alert on.
  • Analyzer type fits the metric's behavior (static threshold, auto-adaptive threshold, or seasonal baseline).
  • Threshold value and the direction of the alert (above or below).

Alerting from a business flow

If you already have a business flow configured, you can create alerts directly from the built-in KPI metrics of Business Flow Business Flow. The app pre-fills the DQL query and opens the anomaly detection configuration for you.

1. Verify the business flow

Open Business Flow Business Flow and confirm that your flow is configured and processing events correctly. Make sure that the steps show data and that there are no gaps.

2. Select a KPI to monitor

At the top of the business flow page, locate the four KPI metrics and decide which one you want to alert on:

  • Errors / Business exceptions—alert when the error rate or the count of business exceptions changes.
  • Avg. duration—alert when the average flow duration deviates from expectations.
  • Revenue (business performance indicator)—alert on revenue changes in the flow.
  • Conversions / Fulfillments—alert when the conversion or fulfillment rate drops.

3. Select New alert

Select the KPI you want to monitor, then select New alert.

This opens the anomaly detection configuration wizard with the DQL query already filled in. Business Flow Business Flow creates a query automatically based on the selected KPI and business flow.

Known limitations

Some KPIs have limited alerting support for certain scenarios. Depending on your flow, you might encounter the following errors:

  • Errors—The query for anomaly detection could not be generated. This can happen when the underlying data or event configuration doesn't support generating a valid query.
  • Avg. duration, Revenue, and Conversions—Anomaly detection for this KPI is only available for business flows with an average duration of 30 minutes or less.

Alerting on key metrics is only possible when the average duration of the monitored business process is above 30 minutes, because Anomaly Detection - new Anomaly Detection analyzes a 60-minute rolling window on defined metrics. See Sliding window for details.

If your flow has an average duration exceeding 30 minutes, only the Errors KPI will likely be available for alerting. If none of the KPIs work, use the Alerting using Anomaly Detection approach, where you build the DQL query manually.

4. Configure the alert

In the anomaly detection configuration wizard, which opens with the pre-filled query:

  1. In Define alert condition, choose the analyzer type: Static threshold, Auto-adaptive threshold, or Seasonal baseline.

  2. Set the parameter for the analyzer you chose:

    • Threshold for a static threshold
    • Number of signal fluctuations for an auto-adaptive threshold
    • Tolerance for a seasonal baseline

  3. Set the Alert condition direction (Alert if metric is above, Alert if metric is below, or Alert if metric is outside).

  4. Optional Select Preview to verify that the alert behaves as expected.

  5. In Add details, set the Title and the Event name.

  6. Select Create.

Check the Status of the new configuration shortly after creation to ensure there are no errors in the execution.

Alerting using Anomaly Detection

Use Settings Settings to create a custom alert based on business event data. You need to prepare a DQL query and the alert parameters before you start.

Since Anomaly Detection - new Anomaly Detection is deprecated, we recommend configuring custom alerts in Settings Settings instead, according to the instructions below. For details, see Anomaly Detection app.

1. Create a new custom alert

  1. Go to Settings Settings > Analyze and alert > Alerts > All alerts.

  2. Select Custom alerts.

  3. Select New alert > Create your own custom alert.

2. Create a query that returns a timeseries

  1. In Query, enter the DQL query you want to use. The query must return a timeseries. For example:

    • To count business events per minute:

      fetch bizevents
      | filter event.type == "payment-processed"
      | makeTimeseries count(), interval: 1m

  • To get the average order value per minute:

    fetch bizevents
    | filter event.type == "order-completed"
    | makeTimeseries avg(orderValue), interval: 1m

  • Validate that the query returns a chart before you proceed.

Choose the right interval and aggregation

The interval parameter controls data granularity. 1m is the default value. You might want to adjust it based on event volume. For example, if you have:

  • Low-volume events—use a longer interval (for example, 5m or 15m) to avoid noisy, sparse data points.
  • High-volume events—1m works well for detecting rapid changes.

Prefer aggregated data over raw data: the more you aggregate, the less your alerts are impacted by noise. For example, alert on the overall order volume rather than on per-city breakdowns to avoid alert spam.

3. Define the alert condition and analyzer

In Define alert condition > Select use case, choose the analyzer type:

  • Static threshold—a fixed value. The alert fires whenever the metric crosses the configured threshold. You should use it for well-known, stable KPIs with predictable ranges. Enter the exact threshold value, or select Suggest values to let Dynatrace Intelligence recommend one based on historical data.
  • Auto-adaptive threshold—Dynatrace adjusts the threshold automatically, based on the recent behavior of the data. You should use it when the expected range changes gradually over time.
  • Seasonal baseline—Dynatrace learns recurring patterns (for example, daily or weekly cycles) and detects deviations from the expected seasonal pattern. You should use it for metrics with predictable periodic behavior, such as order volumes or user activity.

For details on each analyzer, see Analyzer type and parameters.

4. Set the condition

In Set a condition:

  1. Set the analyzer parameter. The field that appears depends on the analyzer you selected in the previous step:

    • Static threshold—set the Threshold value and, if applicable, its Unit. Either enter the value you want to use, or select Suggest values for Dynatrace Intelligence to recommend one based on historical data.
    • Auto-adaptive threshold—set the Number of signal fluctuations. Dynatrace calculates the threshold for you, so there's no Threshold field.
    • Seasonal baseline—set the Tolerance of the seasonal model. The higher the tolerance, the broader the confidence band, and the fewer events are triggered.

    For details on each parameter, see Analyzer type and parameters.

  2. Select the Alert condition direction:

    • Alert if metric is above—triggers when the value exceeds the threshold or the confidence band (for example, an error rate spike).
    • Alert if metric is below—triggers when the value drops below the threshold or the confidence band (for example, an order volume drop).
    • Alert if metric is outside—triggers on a deviation in either direction. Available for the seasonal baseline analyzer.

  3. Optional Select Preview to check if the alert behaves as expected based on recent data, and confirm that the configuration looks correct.

5. Add the details

  1. In Add details, configure the following:

    • Title—a descriptive name for the custom alert configuration (for example, Low order volume alert).
    • Event name—the name that appears in Problems app - new Problems when this alert triggers. You can use placeholders for dynamic naming.

    You can enter { to have Dynatrace Intelligence suggest placeholder names with the desired values (for example, {alert_condition}). For more information, see Event template.

  2. Select Create to save the alert configuration.

Check the Status of the new configuration shortly after creation to ensure there are no errors in the execution.

Related topics

  • Anomaly detection configuration
  • Business Flow
Related tags
Business ObservabilityBusiness FlowBusiness FlowAnomaly Detection - newAnomaly Detection