A question I have is how can I focus my effort in a process mining project on the right things.
Via the root-cause-miner we can show dependencies/relations with certain dimensional values and hopefully at some point also measures (see ARISPMIYU-I-47).
I would like to take it one step further. Say we determined that a relation exist between 2 measures. Before i actually start improving something i want to make sure that (1) there is no adverse effect on other measures in the process, (2) how much should i actually improve
An impact analysis might help to answer these questions.
The idea is as follows:
User selects the Measure to be analysed
Measure is shown as a slide-bar where a user can increase or decrease the value of the measure.
Below that a list of other selected that where we want to determine impact.
Initially both the selected measure as the measures we want to determine the impact on all show the average value.
However the moment the user starts playing with the slide bar (increase/decrease) we see also the results of the measures in the list change based on the impact we determined.
This should give users a very clear picture on the potential impact of their improvements. For example; If i focus my effort on decreasing this particular cycle time i will see an improvement in measures x,y,z however it might lead into problems with measures a,b,c where the results actually get worse...
In a sense this is similar to simulation however this is purely measure focused and does not go into any actual simulation of process structural changes