Monitor usage metrics - Need help to understand the 10%, 50% and 90%

[BIForce]

Need some help to understand this statement - The performance for 10%, 50%, and 90% of the open report actions calculated for each individual day.

Does this mean ->

first 10% of users opening report performance

then 50% of users opening report performance

then total 90% of users opening report performance

?

[npergand]

The report opening time corresponds to the Xth (10/50/90) percentile of the time it takes to open the report. or, the time below which X% of the open-report actions are completed.

Think of it like sorting all "open-report" actions by time fastest to slowest. Starting at fastest, go up to the X% mark, that should give you the opening performance time.

Here is a little more detail on the metric definitions: https://docs.microsoft.com/en-us/power-bi/collaborate-share/service-modern-usage-metrics#which-metri... 

[BIForce]

Thank you so much @npergand . This is super helpful. I pulled out one example. 

Per this example: 25% of opened reports took 14 secs, 50% of reports took 19 secs and 75% of reports took 45 seconds. Am I interpreting it correctly?

Also, how come it is possible for the same report/same day to took 14 to 45 secs.

Please let me know your thoughts. Again, thank you so much for your inputs.  

[v-easonf-msft]

Hi, @BIForce 

Report performance will be affected in many ways. Poorly performing calculated columns/measures, unnecessary visuals, complex model design  and a large amount of data returned by queries will slow down your report loading and response speed.

You can refer to these official documents, which explain some factors of loading report speed in more detail:

Optimization guide for Power BI 

Troubleshoot report performance in Power BI 

Best Regards,
Community Support Team _ Eason