For AI agents: A markdown version of this page is available at https://docs.datadoghq.com/synthetics/browser_tests/test_results.md.
A documentation index is available at /llms.txt.
The test details page opens after a Synthetic browser test executes and is organized into four tabs: Activity, Test Runs, Performance, and Properties. Use these tabs to monitor uptime, inspect individual runs, review aggregate performance metrics, and manage test configuration. When a run fails, see Failed results for troubleshooting tools such as AI failure summaries and screenshot comparison.
Test activity
On the Activity tab, you can see:
The Global Uptime graph, which displays the total uptime of all test locations in a given time interval. The global uptime visualization displays red only if the alert conditions configured for a test are triggered in the given time interval. Since location uptime is computed based on the final test result after retries complete, fast retry intervals directly impact what appears in your total uptime graph. For more information about uptime monitoring, see the Website Uptime Monitoring with SLOs guide.
A Timeline of alert triggers, recoveries, and test modifications.
A Summary panel for the selected timeline event, showing what happened, the failing result, and suggested next steps for investigation.
Test runs
On the Test Runs tab, you can see all individual runs of your test. Filter by status (passed or failed), run type, location, or device, and click any row to inspect that run in detail.
Every executed test step contains a screenshot of the step action, a link to the session in Session Replay, the step description, starting URL for a given step, step ID, step duration, and page performance information.
Errors and warnings
Click the Errors pill to access the Errors & Warnings tab and examine a list of errors separated by error type (js or network) and status (the network status code).
The Errors & Warnings tab displays a list of errors separated by error type (js or network) and status (the network status code).
The error type is logged when the browser test interacts with the page. It corresponds to the errors collected between the time the page is opened and the time the page can be interacted with. The maximum number of errors that can be displayed is 8, for example: 2 network + 6 js errors.
Resources
Click the Resources pill to access the Resources tab and examine the combination of requests and assets, including the total step duration time under Fully Loaded and the CDN provider serving the resources.
You can filter resources by type and search by name in the search bar. The maximum number of resources that can be displayed is 100. Resources are ordered by the time when they start and display the first 100 in Datadog.
Resources tab columns
The following describes the column headers on the Resources tab:
Relative Time
The point in time when the resource began to load during the test step.
CDN
The CDN provider that served the resource. Hover over a CDN provider’s icon to see the raw cache status. Datadog detects Akamai, Cloudflare, Fastly, Amazon Cloudfront, Netlify, Google Cloud CDN, Imperva, and Sucuri.
Resource
The URL of the resource.
Type
The type of resource (HTML, Download, CSS, Fetch, Image, JavaScript, XHR, or Other).
Method
The method of the request.
Protocol
The protocol of the request.
Status
The HTTP response status code.
Duration
The time needed to perform the request.
Size
The size of the request response.
For Fetch and XHR resources, click on a resource row to view its request and response headers and body. Payload details are only available when Capture network payloads is enabled in the test’s advanced options.
Backend traces
Click the Traces pill to access the Traces tab and explore APM traces associated with the browser test. While the UI is similar to the Trace View in the Trace Explorer, one browser test step can make multiple requests to different URLs or endpoints. This results in several associated traces, depending on your tracing setup and on the URLs you allowed in for browser tests in the Synthetic Monitoring Settings page.
If your test reaches the maximum execution time, the timeout message indicates that the total duration includes both test steps and system overhead. As a result, the reported test duration may differ from the sum of individual step durations.
Test performance
On the Performance tab, you can see aggregate performance metrics across all runs of your test:
Browser success rate cards for each browser type (Chrome, Firefox, Edge), displaying the percentage of passing runs in the selected time interval.
Average Test duration by browser type and Average Test duration by location & device graphs, which display the time each browser, location, and device takes to complete the test in a given time interval.
p75 Largest Contentful Paint and p75 Cumulative Layout Shift graphs, which display the 75th percentile of these Core Web Vital metrics aggregated across runs.
The Properties tab contains the configuration details, ownership information, and integrations associated with your test. Use the left navigation to switch between sections.
Properties tab sections
The following describes each section available on the Properties tab:
Ownership
Displays the test owner, editor, creation date, last modified date, environments, teams, and tags. Tests also link to an out-of-the-box Synthetic browser test dashboard.
Execution
Shows the test frequency, alert conditions, and retry behavior.
Monitor
Contains the Synthetic test monitor name, priority, configured recipients, and notification message.
Lists scheduled downtimes that pause execution of this test, for example during planned maintenance windows.
Configuration as Code
Exports the test configuration in formats such as Terraform for managing tests as code.
Failed results
A test result is considered FAILED if it does not satisfy its assertions or if a step failed for another reason. You can troubleshoot failed runs by looking at their screenshots, checking for potential errors at the step level, and looking into resources and backend traces generated by their steps.
AI failure summaries
When a browser test run fails, Datadog generates an AI failure summary to help you identify the cause and next steps for investigation. Each summary includes:
A short explanation of what failed, grounded in run data such as network errors, assertions, and screenshots.
A classification of the failure as either a true failure (a real problem with your application) or a test misconfiguration (an issue with the test setup).
Suggested next steps for troubleshooting.
AI failure summaries appear on the test run details page for any failing browser test run. Treat them as a starting point for investigation, not as authoritative root cause analysis, because LLM-generated content can contain inaccuracies. Use the 👍 and 👎 buttons on the summary to share feedback and help improve future results.
Compare screenshots
To help during the investigation, click Compare Screenshots to receive side-by-side screenshots of the failed result and the last successful execution. The comparison helps you to spot any differences that could have caused the test to fail.
Note: Comparison is performed between two test runs with the same version, start URL, device, browser, and run type (scheduled, manual trigger, CI/CD). If there is no successful prior run with the same parameters, no comparison is offered.
Common browser test errors
Element located but it's invisible
The element is on the page but cannot be clicked on—for instance, if another element is overlaid on top of it.
Cannot locate element
The element cannot be found in the HTML.
Select did not have option
The specified option is missing from the dropdown menu.
Forbidden URL
The test likely encountered a protocol that is not supported. Contact Support for more details.
Alerts from your Synthetic test monitors appear on the timeline in the Activity tab, where you can review alert triggers, recoveries, and test modifications alongside the global uptime graph. To search for alerts from Synthetic tests in the Events Explorer, navigate to Events > Explorer and enter @evt.type:synthetics_alert in the search query. For more information, see Using Synthetic Test Monitors.
Further Reading
Additional helpful documentation, links, and articles:
