Tracking Chrome UX Report data over time
Trends included from Starter, $19 a month.
The Chrome UX Report, or CrUX, is Google's public dataset of real-user experience metrics collected from Chrome users who opted in. It powers the field data you see in PageSpeed Insights and Search Console. If you want to monitor Core Web Vitals over time, CrUX is where the numbers come from. The question is how you view it.
What CrUX contains
- Metrics: LCP, INP, CLS, plus First Contentful Paint and Time to First Byte
- Percentiles: 75th percentile values, plus the share of experiences rated good, needs improvement and poor
- Dimensions: form factor (phone, desktop, tablet) and, depending on the source, connection type and country
- Levels: individual URLs with enough traffic, and whole origins
- Time: a rolling 28-day window, updated daily in the API, plus monthly datasets and weekly history
Ways to view CrUX data
PageSpeed Insights
Shows the current 28-day field data for one URL or origin. Good for a quick look. No history.
Search Console
The Core Web Vitals report groups your URLs into similar pages and shows pass and fail counts over time. Useful for spotting site-wide problems, though grouping can hide which page changed.
CrUX API
Returns the latest 28-day data for a URL or origin on request. The CrUX History API returns weekly data points going back several months, which is what you need for trends.
BigQuery
The monthly CrUX dataset is public in BigQuery. It is powerful for origin-level analysis across many sites, and it needs SQL and a Google Cloud account.
Looker Studio dashboards
Google has published dashboard templates that read from CrUX data. Handy for monthly origin-level reporting.
Limits to keep in mind
Traffic threshold. URLs without enough eligible traffic have no URL-level data. You fall back to origin data.
Lag. Because data covers 28 days, a fix shows up gradually. A change you shipped today affects about one twenty-eighth of the window tomorrow.
Chrome only. CrUX reflects Chrome users who opted in to sharing usage data. Safari and Firefox users are not included.
No why. CrUX tells you what users experienced, not what caused it. Lab tests and page diagnostics fill that gap.
Building your own vs using a tool
You can build a CrUX dashboard yourself with the API and a spreadsheet or BI tool. You will need an API key, a scheduled script, a place to store results, charts, and alert logic. That is a weekend for a developer and ongoing maintenance after.
Vitals Clock does this for each URL you add: it pulls current CrUX data and the weekly history, charts LCP, INP and CLS for phone and desktop, and alerts you when a metric crosses Google's thresholds. It pairs this with daily lab tests, so you see a regression the day it ships and then watch it work through field data.
Reading trend charts
- A step change usually lines up with a release, a new third-party script or an infrastructure change.
- A gradual drift often comes from accumulated features and scripts.
- Seasonal patterns can come from different audiences, devices or networks at certain times of year.
- Recovery takes weeks in field data even after a fix, because of the 28-day window.
Questions
Does Vitals Clock use BigQuery?
No. We use the CrUX API and CrUX History API, which give per-URL and per-origin data.
Why does my origin pass while a key page fails?
Origin data averages all pages. A heavy page with less traffic can fail on its own while the origin passes.