I may be over simplifying by not understanding the question but is not count of cases at Node A (100) your denominator? Sorry If I have not understood your question

I'd need to test on my environment, but I guess I'd start with ACTIVITY_LAG to count cases to the next and then using GLOBAL() in the denominator.

Yes, it sounds very simple and I need really 100 as denominator, but the problem is how could it be expressed by PQL for Process Explorer.

Thanks a lot for your idea!

It works with GLOBAL(COUNT(SOURCE(...))) but only for the discribed situation with two levels in process (A-> B/C). If there are futher nodes outgoing from B or C the denominator is still the same (100). So it shows the "general" rate. I would define this rate for every "parent" node.

Hmmmm... not sure what you say... you can calculate the % of cases that goes from 1 activity directly to others. That you can show on Process Explorer as a Process Explorer KPI.

I seriously doubt you can calculate something like "% of cases that, from A, goes to B and then C" at least as a Process Explorer KPI.

You can, of course, create an OLAP or Pivot where you can show the % of cases that goes from one activity to each of the others, either directly or indirectly. But not in Process explorer.

But maybe I am not understanding you