Time between two activities based on activity attribute


I need to calculate the time between two activities that are not directly followed by each other taking an event attribute into account.

Calculate time between A and B where:

  • Both activities A and B happen more than once in the process
  • To get the correct throughput time, we need to consider an activity attribute (e.g., campaign_id). In our case, if activity A happens three times, it will have three distinct campaign ids, so we need to match those ids with the ones in activity B.

How can I implement that in Celonis?

I tried using SOURCE() and TARGET() from the example in this post , but it seems it only works for “directly followed by” activities, so there was no data for my chart, as A is never directly followed by B in this case.

Thanks a lot in advance!


I was able to reproduce the solution with the OLAP table as described on this post
However, I need a column chart where the dimension is based on a timestamp (e.g., round day, or week, or month). The problem now is that whenever I change the dimension other than the source activity, the KPI returns wrong values.

Anyone had a similar case or know how to help?

Thanks and best,

Update: issue solved.