Unify Automations tracks two categories of run metrics that feed the Insights dashboard: run counts (how many runs happened and how they ended) and run duration (how long runs take). Together these tell you the volume, reliability, and speed of your automations. Insights comes in two views: a per-automation dashboard for drilling into one automation's history, and a module-level view that aggregates metrics across every automation in your workspace.
Overview
The Insights dashboard is a read-only view of automation performance over time. Rather than listing individual executions the way the Runs screen does, Insights aggregates across many runs into metrics and trends — totals, success/failure rates, and durations over a period — so you can judge overall health at a glance.
The Runs screen and Insights serve different purposes: use the Runs screen to investigate "what happened in this run?"; use Insights to answer "is this automation healthy lately?" Because the metrics come from recorded runs, they only reflect executions that were logged — if you turn run logging off, new runs will not contribute to Insights.
Note: Both the per-automation and module-level Insights views are configured dashboards, so the exact charts and visualizations depend on how your workspace has set them up. However, all charts — on either view — are built from the same two underlying metric categories: run counts and run duration.
Run Count Metrics
Run count metrics tell you how many times an automation executed and how those executions ended. Each run is recorded with a status, and Insights rolls these up into totals and breakdowns over the selected time period.
Run Statuses
Every run is assigned one of the following statuses, and run count metrics break down totals by these categories:
Status | Meaning |
|---|---|
Successful | The run finished cleanly with no errors. |
Failed | The run encountered one or more errors and did not complete successfully. May read "Failed with N errors." |
Completed | The run finished, but one or more steps failed along the way (shown in a warning color). |
Cancelled | The run was stopped before finishing. |
In Progress | The run is still executing. |
Waiting | The run is queued, waiting on a delay or an external response. |
By reviewing how runs distribute across these statuses, you can spot trends such as a rising failure rate or a growing proportion of cancelled runs, and investigate before the problem escalates.
What Counts Tell You
Run count metrics answer questions like:
How many times did this automation run over a given period?
What proportion of runs succeeded versus failed?
Is the failure rate trending upward or downward over time?
Are there spikes in cancelled or waiting runs that suggest a bottleneck?
Run Duration Metrics
Run duration metrics capture how long each execution takes. Every run records its total execution time, and Insights surfaces this as trends and aggregates so you can monitor whether your automation is getting faster or slower.
How Duration Is Recorded
Duration is tracked at two levels:
Level | What Is Captured |
|---|---|
Run level | The total execution time for the entire run, displayed as the Duration column in Run History and as a duration pill in run details. |
Step level | Each step records its own timing: "This step took {time} to execute" with the exact From (start time) and To (end time). Steps may also show "This step was cached & took {time} to execute" or "This step was not executed." |
Insights aggregates run-level durations across many executions, letting you see whether run times are stable, improving, or degrading over time.
What Duration Tells You
Duration metrics answer questions like:
How long does a typical run take?
Are run times getting longer, which might indicate a performance problem or growing data volumes?
Are there outlier runs that took much longer than usual?
How Metrics Are Collected
All metrics displayed in Insights come from recorded runs. This means three conditions must be met before metrics appear:
The automation must be deployed.
The automation must have actually run at least a few times.
Run logging must be enabled — runs are not counted while logging is off.
When run logging is enabled, each execution records the data that feeds these metrics: the run's start/end time and duration, its execution status (successful, failed, waiting, cancelled), and step-by-step details. If run logging is disabled, new runs will not be recorded and will not appear in Insights — on either the per-automation or the module-level view.
Per-Automation Insights
Each automation also has its own Insights tab, which applies the run count and duration metrics described above to that single automation's history.
Accessing the Insights Dashboard
Open the automation you want to inspect.
Select the Insights tab within the automation detail view.
The dashboard loads with charts and totals drawn from that automation's recorded run history.
Run Volume Chart
The run volume section of the dashboard shows how many times the automation ran over the selected period, with runs broken down by their outcome status — for example, how many succeeded versus how many failed. This lets you see total activity at a glance and identify whether the proportion of failed runs is growing or stable.
Duration Trends
The duration section shows how long runs are taking over time. Use it to detect regressions — if the automation was consistently completing in a few seconds and is now taking much longer, the trend will show the shift even before individual failures appear. Duration data reflects wall-clock time from trigger to completion for each logged run.
Error Analysis
Failed runs contribute to the count breakdowns visible in the run volume section of the dashboard. By comparing the proportion of failed runs to total runs over a period you can identify failure hotspots — time windows where failures spiked — and correlate them with deployments or upstream changes.
To investigate the detail of a specific failure, switch to the Runs tab and open the individual run record. Insights shows the aggregate picture; Runs gives you the step-level trace and error message for a specific execution.
Module-Level Insights (Across All Automations)
The automations area also provides a module-level Insights view that aggregates run data across every automation in your workspace, giving you a portfolio-level picture of automation health without having to visit each automation individually.
Note: Accessing the module-level Insights view requires the view-insights permission for the automations area. If you cannot see this view, contact your workspace administrator.
Use the module-level view for a portfolio perspective: which automations are most active, and which ones have the highest failure rates. Once you spot a pattern or a problem area, navigate to the individual automation's Insights tab or Runs screen to investigate further.
Accessing Module-Level Insights
Go to the Automations area of your workspace.
Select Insights from the top-level navigation within the automations area.
The dashboard loads with aggregated metrics drawn from run data across all automations.
This is a read-only view. You cannot edit automations or run configurations from this screen.
Available Metrics
The module-level view aggregates the same run counts and run duration metrics used in per-automation Insights, but rolled up across every automation in the workspace:
Metric | Description |
|---|---|
Run counts (by automation) | Total executions per automation over the selected period, broken down by run status. Lets you see which automations are most active and which have the highest proportion of failures. |
Failure concentration | Where failed runs concentrate across your automation portfolio — which automations account for the most failures — so you can prioritize investigation. |
Run duration | Execution duration aggregated across runs, surfacing automations whose run times are outliers relative to the rest of the workspace. |
Filtering & Comparing
The module-level Insights view is built as a configured dashboard, so the available filter and comparison controls depend on how your workspace has set up the view. In general, you can narrow the data by time range to focus on a specific period — for example, the past week or month — which is useful when investigating an incident or reviewing performance after a release.
To compare individual automations in depth, use the module-level view to identify which automations deserve attention, then navigate to each automation's own Insights tab for a focused breakdown of that automation's run counts and duration trends.
Notes
Insights is read-only across both views — you cannot edit automation data, run configurations, or automations from either dashboard. It is for monitoring, not changing the automation.
Give a newly deployed automation time and enough runs before relying on its Insights data — trends need several executions to become meaningful, and a dashboard will appear empty until the automation is deployed, has run, and has run logging enabled.
Use the module-level view as your first stop during an incident — it quickly shows which automations have elevated failure rates — then drill into that automation's own Insights tab or Runs screen to investigate further.
Make a habit of reviewing Insights periodically, both per automation and at the module level, to catch slowly degrading automations before they become critical — a gradual rise in failures is often easier to spot in the aggregated view.
Ensure run logging stays enabled for every automation you want to monitor — any automation with logging off is invisible to both the per-automation and module-level views, creating blind spots in your metrics.
Combine the module-level view with alerts (configured at Automations › Settings › Alerts) so that significant failure spikes trigger a notification automatically, rather than relying on you to check the dashboard manually.