How we measure. Every query, every limit.

How a change is judged

Before OptiMo changes a page, Google PageSpeed measures it on mobile and on desktop. After the change, Google measures it again, and a confirmation run checks the result. If Google confirms the page got slower, the change is rolled back automatically and no credit is charged. Otherwise the change stays, with Google's before and after numbers in the page report, and one click undoes it.

The verdict is a lab measurement: Google loading the page under fixed conditions. Where Google has real-visitor data for a page (the Chrome UX Report), the page report shows it next to the verdict.

What a lab measurement cannot tell you

A lab run is one simulated visit. Google's Core Web Vitals assessment uses real visitors over 28 days, judged at the 75th percentile, so a change that wins in the lab shows up in field data only as new visits replace old ones. Lab results also vary from run to run, which is why one result is confirmed before it counts.

Queries behind published numbers

None yet. When an article or report publishes a number from our data, its query appears here, with the date it ran and what it cannot tell you.