Which data tells you whether a release slowed a page?
Use a scheduled test with the same device and connection settings to compare changes you can reproduce. Calibre can also show Chrome UX Report data where eligible. To see how visitors to your site experience a particular page, configure real-user monitoring and add the browser snippet, a small piece of code that sends measurements from the page.
What can a test or real-user report capture?
Use only a page you’re authorized to test and that the test browser can reach. Before enabling measurements on a page you control, review the actual allowed domains, sampling and data collection. Visitor measurements can include page paths and labels for elements. Mask sensitive path values such as account IDs and verify the resulting fields with fictional data in an isolated copy before collecting real visits. Use a paper test if that setup is uncertain.
What does the plan count and what can you export?
Starter lists 5,000 real-user sessions and 5,000 synthetic tests per month. Each scheduled URL and test-profile combination uses a test per run. The Starter visitor collection pauses at its session limit, and synthetic tests stop when their separate allocation runs out. The pricing table lists 90 days of Starter real-user retention. For synthetic tests, Calibre documents time-series metrics and individual snapshot exports. The command-line tool can also save summaries and trends from real-user monitoring as JSON. Check whether those summaries include the detail you need before relying on them.
How many checks would a simple monitoring plan use?
On paper, choose two fictional pages and test profiles for phone and desktop. A daily schedule uses four synthetic tests a day, or 120 in a 30-day month. Write down one metric you would watch and the number that would prompt an investigation. Count real-user sessions separately and check whether the visits you measure represent the audience you care about.