Aggregation
Google is sending a lot of Web Vitals data to their CrUX database. The data for these vitals is aggregated into the 75th percentile, which means it's compiled from multiple sources to form a comprehensive view.
Advantages and disadvantages of aggregating data
This comes with advantages and disadvantages. It gives low-latency access to real-user experience data at page and origin granularity:
- You won't be able to filter beyond these dimensions;
- But fetching data via CrUX API's will be faster, and presenting the data in PageSpeed Insights and Google Search Console will be more straight forward;
- Privacy is taken care of, as aggregated data won't allow you to filter as much and prevents you from tracing an experience back to an individual.
You will need to resort to Real User Monitoring, for example via RUMvision, to get more insights.
Aggregation in RUMvision
It's mostly the data in the competitor module that is showing aggregated numbers as that data is coming from the CrUX API. Your own monitoring data within RUMvision is not aggregated and allows for extensive deep diving.
On web.dev, Google wrote a section about data aggregation in regards of pagespeed as well.