How do you check your website’s Core Web Vitals?
Use Search Console for groups of similar pages and PageSpeed Insights for a particular address. Separate real-visitor data from the lab test. They answer different questions.
What does Search Console show?
The Core Web Vitals report groups URLs by status, metric and similar user experience. The metrics are LCP, INP and CLS. Data comes from real visits in the Chrome UX Report, or CrUX, over the last 28 days.
It is a sample for assessing the site, not a complete inventory of every page. Only indexed URLs with the necessary data can appear. A group’s status follows its worst metric. Do not treat the group result as a separate measurement of every address.
With insufficient data, Search Console can create a broader origin-level group. An origin is defined by protocol, host and port, not necessarily the entire domain. If data remains insufficient, the report may be unavailable. A new Search Console property can also need time to display existing data.
Access requires the appropriate permissions for the Search Console property. That is different from access to the website’s administration panel. After a fix, you can start 28-day tracking. It monitors issue data; it does not trigger reindexing or prove an immediate fix.
What does PageSpeed Insights show for one address?
You can use PageSpeed Insights for a public address without owning the site. The tool separates CrUX field data from Lighthouse lab diagnostics, for mobile and desktop.
Field data covers a trailing 28-day period. Note whether the result concerns the particular page or the origin. If page visits are insufficient, the tool may use the broader level; if that also lacks data, no field assessment is available. This proves neither good nor poor speed.
Lighthouse loads the address in a controlled simulated environment and provides diagnostic guidance. One load does not reproduce real visitors’ entire behaviour. The lab performance score is not equivalent to a field Core Web Vitals assessment.
PageSpeed Insights field data updates daily. The CrUX BigQuery dataset is published monthly at origin level. Both use trailing 28-day periods. A changed number alone does not establish a change to the website.
Why do lab and field numbers differ?
A lab test uses a particular device, network and location. Real visits span different conditions, caching and interactions. The same address therefore does not imply the same measurement.
In Google’s documented example, LCP is 1.8 seconds at the field’s 75th percentile and 3.0 seconds in the lab. This illustrates different conditions; it is not a measurement of your site or a rule for the size of the gap.
Start from real experience when sufficient data is available. Use lab diagnostics to investigate a possible cause and fix. Record address, device, period and aggregation level before comparing results.
Questions
Do we need access to the website’s admin panel?
Not for a public PageSpeed Insights test. Search Console needs appropriate permissions for the property. That does not itself require permission to edit the website.
Why do I see “No data available”?
The page or origin may lack sufficient CrUX data, or the Search Console property may be new. An absent report is not a speed measurement and does not prove whether a URL is indexed.
Do the numbers update daily?
PageSpeed Insights field data does. The BigQuery publication is monthly. Distinguish the report date from the 28-day period it summarises.
Can I check someone else’s website?
You can run PageSpeed Insights for an accessible public address. Search Console requires authorised access to the property; a public test does not grant those permissions.
Sources
Related reading
Next step
Compare data for the agreed key pages.
The Website Review uses PageSpeed Insights for the agreed key URLs on mobile and desktop, and available Search Console group data where access is provided. Gaps are recorded. Findings feed a prioritised Fix List; the review itself does not implement fixes.
or write to contact@ergostrat.com