CrUX data is updated everyday (barring any issues!) and represents an aggregation of the previous 28-days worth of data.
So if you look at the top of
PageSpeed Insights for example and check back tomorrow, you may well see different data.
It's true the BigQuery dataset is only published once a month, but that's intended for more long-term analysis.
The 28-day period was chosen for a number of reasons, including:
- Smoothing out the rough with the smooth. RUM data is notoriously volatile and differs as the cohort of users varies, the day of the week, particular pages trending one day but not the next...etc. Pulling back a bit gives a more smooth number to help site owners understand their actual site performance, rather than it jumping up and down. It is also why we look at the 75th percentile.
- Privacy reasons and ensuring we have enough users to be able to publish the data. We don't publish the limits here, but you can imagine if we have 1 user on day one, a different user on days 2 and 3, another user on day 4...etc. then no one day would have enough users, but across 28 days this is more likely.
This can be furstrating, when you;ve implemented a site change and are keen to see the impact. You can look at the histograms to see if the stats is trending in the right directions
as I wrote about here (long before I joined Google but the advice still stand).
But the best way to get more detailed information (including the timeframe, and segmenting the data in other ways), is to collect Core Web Vitals data yourself through a RUM solution.