Skip to main content

Process Analysis

In the Process Analysis topic, you can ask questions and start discussions about systematically examining workflows to identify inefficiencies, capture value opportunities, and get recommendations for improvements using Process Mining and AI.

  • 645 Topics
  • 1,634 Replies
643 Topics
P
Level 2
patrick.flößLevel 2
asked in Process Analysis
Error in Table while PQL Editor Preview shows results

Hello all, In the PQL Editor an output of my formula is shown in the preview. If I then go to Done and want to look at the table, I get the following error message could not find referenced aggregate COUNT(DISTINCT ). Can anyone help me? I use the PQL formula: KMEANS() in its complex form. Best Regards, Patrick 01 Preview.jpg2156449 64 KB 02 result.jpg1383276 20.7 KB

20
P
Level 2
N
Level 3
naveen.guptaLevel 3
asked in Process Analysis
Not able to edit any out of box Analysis

We are not able to edit any out of box analysis like Conformance checker, Social Analysis etc. We would like to use some of the components / formulas from these analysis into our single analysis for the users. How can we achieve the same?

20
N
Level 3
J
Level 2
john.sekutLevel 2
asked in Process Analysis
Opening a document with a button within an analysis

I would like to open a reference document contained in the File Upload Center using a button. I cannot figure out what the Document ID is though. Any variation of ID that I enter results in a 404 error. Where must/can the document be stored for accessing? (I assume its in the file upload of the Event Collection)Where do I find the Document ID?Please advise if I am way off and there is a cleaner way to do this. Thanks!

40
C
Level 1
.
Level 3
.14600836Level 3
asked in Process Analysis
Segregation of Duties

Hi all, We are using a dedicated Segregation of Duties dashboard to analyze cases where two activities are processed by the same person (e.g. Creation of PO and Releasing the PO). In this dashboard, the viewer selects two activities, which shall not be performed by the same user to see in how many cases there is a violation. See here the PQL statement: avg (case when PU_first(EKPO,"_CEL_P2P_ACTIVITIES".USER_NAME, _CEL_P2P_ACTIVITIES.ACTIVITY_EN = <%=From%> AND _CEL_P2P_ACTIVITIES.USER_TYPE <> B) = PU_first(EKPO,"_CEL_P2P_ACTIVITIES".USER_NAME, _CEL_P2P_ACTIVITIES.ACTIVITY_EN = <%=To%>) then 1.0 else 0.0 end) * 100 By checking the violations in detail, Ive recognized that there are cases where Person A creates a PO, Person B approves the PO and afterwards Person A changes some field in the PO (e.g. Net Order Price). Due to this change, the release indicator is AUTOMATICALLY changed from B (release) to R (not released). This case is shown as a viola

70
.
Level 3
.
Level 2
.14600772Level 2
asked in Process Analysis
Celonis mobile view

Hello Celonis Team, is it possible to use celonis on iphone or ipad? Could you give me some information about the mobile view from celonis. Many thanks

110
C
Level 1
L
Level 3
laura.lauraLevel 3
asked in Process Analysis
Differents currency in “units”

Hi, I would not use VBAP.NETWR_CONVERTED and have the Net Value Order in the Document Currency. When I insert as KPI VBAP. NEWTR, I would like to insert in Units the Document Currency. I tried with: VBAP.WAERS<%=VBAP.WAERS%>a variable : VBAP.WAERSa variable : {VBAP.WAERS } But it is not working. Any idea ? Thank you

20
L
Level 3
L
Level 3
laura.lauraLevel 3
asked in Process Analysis
Count Variants by company code

Hello, I would like to create a table with a count of variants by Company Code but when I use the standard KPI Number of process variants, the number not corresponding to the total I have on the Variant Explorer. Is it exist a simple formule to count the variants? Thank you

00
L
Level 3
R
Level 2
rakesh.paul12Level 2
asked in Process Analysis
Process Mining for Application monitoring tools

Hi All, Do we have any use cases for Application Monitoring tools such as Dynatrace. Can we get a process-level view with metrics of applications to see which application/interface is down because of which the process has stopped ( in real-time ). Regards, Rakesh Paul

00
V
Level 1
T
Level 3
thomas.nagel11Level 3
asked in Process Analysis
Calculation of average cases per user in a number component

Hello, I am trying to calculate the average number of posted invoices per user in a number component in the AP data model. The basic code of the KPI looks as follows: COUNT(BKPF.USNAM) / COUNT(DISTINCT(BKPF.USNAM)) The current KPI however does not yet show the real performace of the AP department because many invoices are posted via batch files or batch users and some invoices are posted by users that are doing an internship or an apprencticeship and these users do not post more than for example 100 invoices. In order to get rid of the cases which were posted by batch users, I used a component filter which works fine. But for the second issue (exclusion of cases which were posted by users that have not posted more than 100 invoices in the selected time range) I have not yet found a solution. I have tried several case when statements and also worked with ABC clusters to exclude the cases of these users, but nothing worked. Does anybody of you have an idea how this could be solved? I am

00
T
Level 3
J
Level 2
jelmer.bodeLevel 2
asked in Process Analysis
Graph - Changing activity at certain date

Hi all, We have implemented a new system, which has different activities names. I would like to have one graph, where the graph shows the data from one activity, and after a certain date (01-09-2019 (dd-mm-yyyy)) from an other activity. Using now the following formula in two different graphs, but I would like to combine them. AVG(CALC_THROUGHPUT(FIRST_OCCURRENCE[Ontvang Factuur] TO LAST_OCCURRENCE[Valideren Factuur], REMAP_TIMESTAMPS(XXCEL_AP_ACTIVITIES.EVENTTIME, HOURS))) AVG(CALC_THROUGHPUT(FIRST_OCCURRENCE[XXAIMG: Ontvangen in APRO Imaging] TO LAST_OCCURRENCE[Valideren Factuur], REMAP_TIMESTAMPS(XXCEL_AP_ACTIVITIES.EVENTTIME, HOURS))) Is this possible? Hope you can help me. Kind regards, Jelmer

00
J
Level 2
M
Level 2
maximilian.heilmLevel 2
asked in Process Analysis
Dimension combining calendar week and year

Hello all, I wonder if there is a way of combining the CALENDAR_WEEK() format with the YEAR() format similar to the way the statement ROUND_MONTH does for exmaple. The problem I have with the calendar week as the dimension is that in case of a dataset that includes data from more than one year, the KPIs do always calculate the data of all years for each calendar week while I would like them do distinguish between the weeks of the particular years. Best regards, Maximilian

10
M
Level 2
S
Level 3
sasan.babaz12Level 3
asked in Process Analysis
Match ID’s from two different tables

Hello, I got a question again I try to do an OLAP table where I need to combine data from different tables. I try to explain the case as good as I can. In my table T052U_BSEG I got the column named TEXT1 (here is a description in each column) and the column ZTERM (here i an ID which is unique for each different description, but a description can come up multiple time). On the other hand I got the table PB_CLAR where I got two columns with the IDs from ZTERM in it. I want to create a column in an OLAP table that the ID in PB_CLAR look up the ID in the T052U_BSEG and pull the right description from TEXT1 Thank you guys in advance

60
C
Level 1
.
Level 3
.14600836Level 3
asked in Process Analysis
Pull function –> how many PO items are created for one PR

Hi all, Id like to analyse the number of PO items which have been created for one and the same PR, e.g. PR 12341234 is related to 5 PO items. I tried using the pull function but always when combining PO item data with PR data from table EBAN I get following error message: Undefined exception occured. Table EBAN is not a parent of EKPO. Use non pull up function instead. I guess this should be an easy formula but unfortunately didnt figure out how to achieve the result Im looking for. Any ideas? Thanks, Oli

10
.
Level 3
T
Level 3
takayoshi.ichikLevel 3
asked in Process Analysis
Always display certain activity in Process Explorer

Hello. I want certain activities always displayed in Process Explorer. Im tackling web access log with Celonis. CVR (user ratio who apply to our service) on our website is lower than 1%, so conversion activity is very rare. When I want to see conversion activity in PE, I must adjust the slider of activity up to almost top. As you know, when I do that, the process view get complicated with other unwanted activities. I want to keep the slider of activity low with arbitrary activity always displayed. Is there any way to do that? Thanks.

20
T
Level 3
R
Level 3
robin.pianeLevel 3
asked in Process Analysis
Throughput time between activities followed directly

Dear all, How can we calculate a throuput time between 2 activities, but only when the activities are directly linked ? An example: I need to calculate the lead-time in orange. With this formula the result is not correct, because there is an other A2 activity later in the process (it could also happen that this additional A2 activity, not directly coming from A1 occurs before in the process) AVG(CASE WHEN (PROCESS EQUALS A1 TO A2) THEN CALC_THROUGHPUT(FIRST_OCCURRENCE[A1] TO LAST_OCCURRENCE[A2], REMAP_TIMESTAMPS("_CEL_ACTIVITIES_NC".EVENTTIME, HOURS)) ELSE NULL END) Thank you for your help, Best Regards Robin Celonis_Help.png87681 1.94 KB

40
P
Level 1
.
Level 3
.14600836Level 3
asked in Process Analysis
Automation rate

Hi all, were currently only considering user types to define which activities have been done automatically (batch user) and which have been processed manual (all other user types): 100.0*AVG(CASE WHEN _CEL_P2P_ACTIVITIES.USER_TYPE IN (<%=auto_user_type%>) THEN 1.0 ELSE 0.0 END) However, for our MRP runs which shall be also considered automated there is always a manual user in SAP. We can identify MRP runs by the doc type (EP) in combination with the PR creation indicator (B). Any suggestions how we can adjust above formula to also consider EP in combination with B as automated? Thanks Oli

60
.
Level 3
N
Level 3
naveen.guptaLevel 3
asked in Process Analysis
Througput time at header level

Dear All, We ahve a througput time caclulated as below CALC_THROUGHPUT(<%=from_occ%>_OCCURRENCE[<%=From%>] TO <%=to_occ%>_OCCURRENCE[<%=To%>], REMAP_TIMESTAMPS("_CEL_O2C_ACTIVITIES".EVENTTIME, <%=unit%>)) THis gives us througput time at line item level (Case key). How can I calculate the same at header level. Say at VBAK level. Thanks Naveen Gupta

20
B
Level 1
.
Level 3
.14600836Level 3
asked in Process Analysis
Amount of price change in Euro

Hi all, When analyzing price changes in Celonis I would like to show in our details tab the Euro amount of the actual price change per PO line. We are already using quite a sophisticated formula to show in a graph the price changes for different thresholds in percentage. I assume I can use this formula partially in order to show the Euro amount of the price change. Following part, I used to do that (which does not work yet): CASE WHEN ISNULL(PU_FIRST(EKPO,"_CEL_P2P_ACTIVITIES".ACTIVITY_EN, _CEL_P2P_ACTIVITIES.ACTIVITY_EN = Change Price)) = 1 THEN No Price Change Activity WHEN PU_FIRST(EKPO, _CEL_P2P_ACTIVITIES.CHANGED_FROM_FLOAT,"_CEL_P2P_ACTIVITIES".ACTIVITY_EN = Change Price) - PU_LAST(EKPO, _CEL_P2P_ACTIVITIES.CHANGED_TO_FLOAT,"_CEL_P2P_ACTIVITIES".ACTIVITY_EN = Change Price) END Any idea how to resolve? Thanks, Oli

10
.
Level 3
N
Level 3
naveen.guptaLevel 3
asked in Process Analysis
Saving my selection and coming back to it when I want to

We are looking for an option to save a selection like a variant. For e.g. Saving Sales Org, Division and Mateial groups selections as a variant. With the possibility to come back to that selection. Any suggestion, how to achieve that?

20
N
Level 3
R
Level 2
rakesh.paul11Level 2
asked in Process Analysis
Hyperlink Issue

Every time an analysis is transported from QA to PROD, old analysis is deleted and a new one is placed in the folder, as a result, the bookmarks are also deleted with transport. How we can get rid of this problem.

30
R
Level 2
.
Level 3
.14600836Level 3
asked in Process Analysis
Threshold values on PO line item amount

Hi all, were using threshold values based on PO header level in order to filter on low or high value items (e.g. below 100 or above 10k). The formula works on PO header level (EKKO table) but should be based on line item level (EKPO) - the formula looks as follows: CASE WHEN PU_SUM(EKKO,EKPO.NETWR_CONVERTED) >= 0 AND PU_SUM(EKKO,EKPO.NETWR_CONVERTED) <= 100 THEN a) 0 - 100 WHEN PU_SUM(EKKO,EKPO.NETWR_CONVERTED) > 100 AND PU_SUM(EKKO,EKPO.NETWR_CONVERTED) <= 500 THEN b) 100 - 500 WHEN PU_SUM(EKKO,EKPO.NETWR_CONVERTED) > 500 AND PU_SUM(EKKO,EKPO.NETWR_CONVERTED) <= 1000 THEN c) 500 - 1.000 WHEN PU_SUM(EKKO,EKPO.NETWR_CONVERTED) > 1000 AND PU_SUM(EKKO,EKPO.NETWR_CONVERTED) <= 10000 THEN d) 1.000 - 10.000 ELSE e) > 10.000 END I tried to exchange EKKO with EKPO to have PU_SUM(EKPO,EKPO.NETWR_CONVERTED) but that didnt work. Any suggestions how to change properly? Thanks, Oli

40
.
Level 3
M
Level 2
michael.johnsLevel 2
asked in Process Analysis
Measure Booking Process by Order Level

I need to measure the number of orders within the booking process: anything between Create Quotation and Create Sales Order Item. For background, the # of Sales Orders using COUNT_TABLE("VBAK") is 200k. Using either of the below formulas continues to give me 200k Sales Orders. I have also attempted it with/without many different component filters: COUNT(DISTINCT CASE WHEN "_CEL_O2C_ACTIVITIES"."ACTIVITY_EN" IN ('Create Quotation','Create Sales Order Item') THEN "VBAK"."MANDT" || "VBAK"."VBELN" ELSE NULL END) or COUNT (DISTINCT CASE WHEN (PU_COUNT("VBAK","_CEL_O2C_ACTIVITIES"."USER_TYPE","_CEL_O2C_ACTIVITIES"."USER_TYPE" LIKE '%' AND "_CEL_O2C_ACTIVITIES"."ACTIVITY_EN" IN ('Create Quotation','Create Sales Order Item')) = PU_COUNT("VBAK","_CEL_O2C_ACTIVITIES"."USER_TYPE"

00
D
Level 2
S
Level 3
stephanny.hartmLevel 3
asked in Process Analysis
Aligning X & Y Axis at Zero

Hi all, Im creating an Credit note analysis - taking that into consideration all the values of my documents are negative: this causes that my graphs when Im looking at values and quantities look funny since the value 0 is not the aligned as the crossing for the X axis and the primary and secondary Y axis. Is it possible to align the crossing of the X and both of the Y axis to 0? In this case, it will mean that the line will be above of the X axis, and all columns will be below. Thank you!

10
S
Level 3
L
Level 3
laura.lauraLevel 3
asked in Process Analysis
Count Table if request closed

Hi everyone, I would like to add a line with a count all requests with a closed date not null by month on my chart. image.png1215225 17.9 KB I tried with this formulation but it is not working: CASE WHEN name_table.closed_at <> NULL THEN COUNT_TABLE(name_table) ELSE 0 END

10
L
Level 3
L
Level 3
laura.lauraLevel 3
asked in Process Analysis
Count table closed request

Hi everyone, I would like to add a line with the number of the closed request opened in a month. image.png1215225 17.9 KB I tried with this CASE WHEN rm_enhancement.closed_at <> NULL THEN COUNT_TABLE(rm_enhancement) ELSE 0 END but it is not working. Someone had the same issue with a count with a case? Thanks

00
L
Level 3

Badge Winners

  • Qualified to Create and Deliver Demos
    anna.jakub12has earned the badge Qualified to Create and Deliver Demos
  • Masterclass in Implementation
    marwan.elkad11has earned the badge Masterclass in Implementation
  • Masterclass in Implementation
    lyzanne.tr24has earned the badge Masterclass in Implementation
  • Masterclass in Implementation
    elliott.thas earned the badge Masterclass in Implementation
  • Celonis Certified Data Analyst
    min.bae11has earned the badge Celonis Certified Data Analyst
VIEW ALL
Community Code of Conduct