benchmark

Detect web performance regressions by capturing baseline navigation timing data and Core Web Vitals like TTFB, FCP, and LCP, then diffing current PR runs against those baselines. This flags regressions based on predefined thresholds automatically.

Track Core Web Vitals and bundle sizes for single page applications by collecting resource loads and bundle information per commit. This establishes baselines and analyzes performance drift over time across multi-page flows.

A page load performance audit includes structured metrics for TTFB, FCP, LCP, DOM metrics, total network requests, transfer size, and bundle sizes. These metrics anchor performance expectations against baseline runs.

Performance benchmarking works with staging URLs and multi-page flows by using a browse daemon to collect navigation timing data and resource loads. You can quantify improvements against existing baselines directly on staging deployments.

Set thresholds to flag performance regressions by establishing a baseline of resource usage and page load times, then comparing current runs to those baselines. Predefined thresholds automatically surface regressions and performance drift.

You need a performance baseline to anchor expectations for page load times and Core Web Vitals. Comparing current runs to this baseline on every PR helps detect subtle performance regressions before they reach production.