Catch the slowdown an uptime check misses
A new homepage image, a third-party script or a release can make a page noticeably slower while every request still returns a successful status. Your uptime check is doing its job, but it is answering a different question.
PageLantern’s page-speed monitoring adds scheduled measurements of a public page in a fresh desktop browser. The staged paid beta gives you a way to track loading performance and investigate changes without writing a browser script.
See measurements and the evidence behind them
Each run records available loading measurements, including Largest Contentful Paint, First Contentful Paint and navigation time to first byte. It also records layout movement during the bounded load window. Missing measurements remain Not measured rather than becoming zero.
Open a run to inspect available screenshots, capture notes and a request timeline. Filter requests by resource type, URL path or failure to find evidence worth investigating. A slow request is a useful lead; it does not by itself prove the cause of a slowdown.
Set a budget that avoids repeated warnings
Performance budgets use consecutive valid scheduled measurements to confirm a slowdown and recovery. The default opens an alert after two measurements above four seconds LCP, then recovers after two at or below 3.6 seconds. You can adjust the loading and recovery thresholds.
The separate recovery threshold helps prevent repeated alerts around one cutoff. Performance state remains separate from availability: a slow measurement does not lower your uptime percentage or mark a public status page down.
Choose a schedule and inspect your first run
When the beta is enabled for your workspace, open Add monitor and select Page speed. Enter the public page URL, choose an hourly, six-hourly or daily schedule, and review the budget. Creation queues the first scheduled measurement.
Run check queues an additional diagnostic measurement. Manual runs appear in history but do not confirm or recover scheduled alerts. Use trend CSV exports for the selected measurement series, or export the currently loaded run rows when investigating individual tests.
Know the beta’s boundaries
Access requires an eligible paid workspace and explicit beta enablement. Tests use one desktop profile from the home region; authenticated pages, mobile presets and additional regions are outside this beta. These are synthetic measurements, not real-user Core Web Vitals or a Lighthouse score.
Runs share the browser-minute allowance with Browser checks. A shorter schedule consumes more allowance. If a run cannot be measured, the gap remains visible; skipped or invalid runs are never presented as a healthy result.