PageSpeed Insights Update Thread

[Paul Irish]

We'll update this thread with any user-visible changes to PageSpeed Insights, including the API, score calculation, etc.

Please start a new thread for any comments or questions.

[Paul Irish]

From Feb 1 to Feb 4 2019, PSI had a bug that led to lower performance scores. This bug is now resolved.

The headless Chrome infrastructure used by PageSpeed Insights had a bug that reported uncompressed (post-gzip) sizes as if they were the compressed transfer sizes. This led to incorrect calculations of the performance metrics and ultimately a lower score. The mailing list thread titled [BUG] [compression] Avoid enormous network payloads / Defer unused CSS doesn't consider compression covered this issue in greater detail. Thanks for Raul and David for their help.

As of Monday Feb 4, 5pm PST, the bug is completely addressed via a new production rollout.

[Paul Irish]

Today, we're upgrading the Lighthouse version used by PageSpeed Insights v5 API (and the PSI site) from v5.0.0 to v5.2.0.  You can view the release notes for v5.2.0 and v5.1.0 on github for more details on what is included in these versions.

These upgrades should have little to no effect on performance scoring.

[Jasmine Yan]

As of December 3, 2020, PageSpeed Insights (PSI) is using the Chrome User Experience Report (CrUX) API as its source of field data. Previously, PSI was using its own custom data pipeline similar to CrUX, but this change brings the two in closer alignment. In the API, this affects the loadingExperience and originLoadingExperience objects. Specifically, the values of percentiles and proportions of field data metrics should now be identical across APIs.

Note that PSI users who have been monitoring their metrics over time may have noticed a small jump when the data source of the API was switched over. Most users should not notice any significant changes in their results, but please start a new thread to let us know if you see anything unusual or have any questions.