What measures are calculated?
PPM uses actual data from your source systems to calculate a large number of measures that you can apply to assess process performance.
Which measures can be calculated depends on the data supplied by the source system. When configuring your PPM system, the system administrator specifies the measures available to you.
In PPM, you can create additional user-defined measures based on the existing available measures.
The measures that you can use for your analyses are provided in the Measures bar. Measures that you are using in your analysis are listed in bold.
The following measures are calculated.
Process-independent measure
Process measure
User-defined process measure
Function measure
User-defined function measure
Relation measure
User-defined relation measure
Organizational unit measure
Ranking measure
The following function and process measures can be calculated in PPM:
Function measures
Number of processors
Processing frequency
Processing span
Processing time
Cycle span
Cycle time
Number of functions
Wait time
Process cost rate
Number of processors (function)
The Number of processors measure corresponds to the average number of different processors who have executed a function within the set of process instances being analyzed.
Your system administrator defines how to proceed with functions that have no processors assigned. The following settings are possible:
The measure value 0 is assigned to functions without assigned processor. It is assumed that these functions do not have any processors.
The measure value 1 is assigned to functions that have a processor assigned. It is assumed that each of these functions is executed by a different processor.
You are analyzing two process instances in which the Create customer order function occurs.
In the first process instance, the Create customer order function exists twice. The first function was created by the Order processing organizational unit and modified twice by another organizational unit. The second function was created by the Order processing organizational unit and not modified again.
In the second process instance, the Create customer order function exists once. It was created by the Order processing organizational unit and modified once by another organizational unit.
The value of the Number of processors measure for the Create customer order function is 1 (=(2 +1) / 3).
Processing frequency (function)
The Processing frequency (function) measure corresponds to the average number of processing operations of the function within the set of process instances being analyzed.
You are analyzing two process instances that contain the Create customer order function.
In the first process instance, the Create customer order function exists twice. For the first function, the number 3 is specified as the attribute value of Number of executions. The original document has therefore been modified twice. The document for the second function was not modified, the attribute value of Number of executions is 1.
In the second process instance, the Create customer order function exists once. For this function, the number 2 is specified as the attribute value of Number of executions.
The value of the Processing frequency measure for the Create customer order function is 2 (=(3 + 1 + 2) / 3)
Processing span
The Processing span measure corresponds to the average time span between the earliest start time and the latest end time of all functions of a type in the process instances. If no start or end time is known for any of the functions, the measure is not calculated.
You are analyzing three process instances of the Order processing process in which the Create customer order function exists. These three process instances were executed in October 2004. In one process instance, the function was executed twice.
The following start times were determined in the individual process instances for the functions.
Process instance | Start time | End time |
|---|---|---|
Example 46. 1 | Example 47. 5.10.04; 12 PM 01 min. 29 sec. | Example 48. 5.10.04; 12 PM 03 min. 30 sec. |
Example 49. 1 | Example 50. 5.10.04; 12 PM 51 min. 16 sec. | Example 51. 5.10.04; 12 PM 59 min. 47 sec. |
Example 52. 2 | Example 53. 15.10.04; 13 PM 38 min. 06 sec. | Example 54. 15.10.04; 13 PM 40 min. 33 sec. |
Example 55. 3 | Example 56. 30.10.04; 14 PM 15 min. 25 sec. | Example 57. 30.10.04; 14 PM 20 min. 05 sec. |
In the first process instance, the processing span of the Create customer order function is 10 minutes 32 seconds.
In the second and third process instances, the processing span is 2 minutes and 27 seconds, and 4 minutes and 40 seconds, respectively.
Therefore, the following value applies in October 2004 for the processing span of the Create customer order function in the Order processing process: 5 minutes and 53 seconds (=(10:32 + 2:27 + 4:40) /3).
Processing time
The Processing time measure corresponds to the average time span between start and end of processing of the functions of a type. This measure can be calculated on the basis of various indicators.
First, the Processing time attribute of the functions of a type is evaluated. If this attribute is not specified for a function, the processing time is calculated as the difference between the end time and start time of the function. If only one of the said times is available, or if the calculated difference is negative, the function is not considered in measure calculation.
Example
You are analyzing two instances of the Returns process, in each of which the Create returns delivery function exists once. Both process instances were executed in September 2004.
For the first function, the Processing time attribute has the value 15 minutes. For the second function, the value is 27 minutes.
In September 2004, the Processing time measure for the Create returns delivery function produces a value of 21 minutes (=(15 + 27)/2).
Cycle span
The Cycle span measure corresponds to the average time span between the latest end time of all function instances of this type and the earliest end time of all preceding functions within a period under analysis.
The Create returns delivery function of the Return process was executed three times in April 2004.
Function | End time of preceding function | End time of Create returns delivery function |
Example 60. 1 | Example 61. 06.04.04; 10 AM 40 min. 00 sec. | Example 62. 06.04.04; 11 AM 35 min. 15 sec. |
Example 63. 2 | Example 64. 15.04.04; 09 AM 15 min 30 sec. | Example 65. 15.04.04; 10 AM 20 min. 45 sec. |
Example 66. 3 | Example 67. 29.04.04; 16 PM 30 min. 00 sec. | Example 68. 30.04.04: 09 AM 30 min. 00 sec. |
The cycle span of the Create returns delivery function in the first process instance is 55 minutes and 15 seconds. For the second and third process instances, 65 minutes and 15 seconds and 17 hours are calculated.
In April 2004, this produces a cycle span of 6 hours, 20 minutes and 10 seconds (=(55:15 + 65:15 + 1020:00)/3) for the Create returns delivery function of the Return process.
Cycle time
The Cycle time measure corresponds to the average duration of the execution of a function. It comprises the average processing time and the average wait time of a function. The cycle time of a single function is calculated as follows:
Cycle time = End time of function - Latest end time of preceding functions.
If these values cannot be determined for a function, or if the calculated cycle time is negative, the function is not considered in the measure calculation.
You are analyzing four process instances of the Returns process, in each of which the Create returns delivery function exists once. One of the said process instances was executed in September, the other three were executed in October. The sole preceding function is Create return.
The following end times were calculated:
Process instance | Create return | Create returns delivery |
Example 71. 1 | Example 72. 21.09.04 10 AM 04 min. 00 sec. | Example 73. 21.09.04; 10 AM 15 min. 00 sec. |
Example 74. 2 | Example 75. 05.10.04; 11 AM 14 min. 00 sec. | Example 76. 05.10.04; 11 AM 25 min. 00 sec. |
Example 77. 3 | Example 78. 15.10.04 09 AM 00 min. 00 sec. | Example 79. 05.10.04 09 AM 20 min. 00 sec. |
Example 80. 4 | Example 81. 27.10.04; 11 AM 24 min. 00 sec. | Example 82. 27.10.04; 11 AM 37 min. 00 sec. |
September: 11 (=11/1) minutes
October: 14 minutes and 40 seconds (=(11 + 20 + 13 )/ 3).
Number of functions
The Number of functions measure corresponds to the sum total of the occurrences of a function in the process instances under analysis.
Based on the current filter settings, in September 2004, four process instances are considered in measure calculation for the Return process.
The following table shows how often the Create returns delivery function was executed in the process instances:
Process instance | Occurrences |
Example 85. 1 | Example 86. 0 |
Example 87. 2 | Example 88. 1 |
Example 89. 3 | Example 90. 0 |
Example 91. 4 | Example 92. 1 |
The number of functions of the Create returns delivery function in September 2004 is 2.
Wait time
The Wait time measure of the functions of a type corresponds to the average time span between the end of processing of the preceding functions (latest end time) and the start of processing of the functions under analysis. The wait time of the individual function corresponds to the difference between the start time of the function under analysis and the latest end time of the preceding functions.
In May 2004, the Post goods issue function is executed once in each of three process instances. For the first function, the following points in time are available:
Example 95. Start time of the Post goods issue function | Example 96. 03.05.04; 10 AM 40 min. 00 sec. |
Example 97. End time of the preceding function | Example 98. 03.05.04 09 AM 15 min. 00 sec. |
This results in a wait time for the function of 85 minutes.
For the other two functions, wait times of 110 and 180 minutes, respectively are calculated.
In May 2004, a value of 125 minutes (=(85 + 110 + 180)/3) is produced for the wait time of the Post goods issue function.
Process cost rate
Product of the execution time of a function and the average cost rate of all organizational units assigned to the function.
The execution time of a function describes the time required for its one-time execution.
The average cost rate is calculated from the sum of the cost rates of all organizational units assigned to the function, divided by the total number of processing operations. Calculation of the sum total incorporates each individual execution of the function.
A function has been executed multiple times if
multiple organizational units are assigned to the function.
the same organizational unit has executed the function multiple times. Multiple execution by the same organizational unit is indicated by the Processing frequency attribute of the connection between the function and the organizational unit.
Process measures
Number of processors
Number of processing steps
Function frequency
Number of processes
Process frequency
Process costs
Process cycle time
Number of processors (process)
The Number of processors (process) measure corresponds to the average number of different processors in all process instances under analysis. Processors belonging to an organizational unit excluded from measure calculation are not included in the calculation.
Your system administrator defines how to proceed with functions that have no processors assigned. The following settings are possible:
Functions that have no assigned processor will not be considered. It is assumed that these functions do not have any processors.
Functions that have no assigned processor are treated like an additional processor. It is assumed that each of these functions is executed by a different processor.
All functions that have no assigned processor will be considered only once in the entire process. It is assumed that these functions are executed by the same processor.
Based on the current filter settings, in October 2004, four process instances are considered in the measure calculation for the Order processing process.
The following table shows how many processors executed functions in a process instance:
Process instance | Number of processors |
Example 101. 1 | Example 102. 4 |
Example 103. 2 | Example 104. 5 |
Example 105. 3 | Example 106. 4 |
Example 107. 4 | Example 108. 7 |
For the Number of processors measure of the Order processing process in October, a value of 5 is produced (=(4 + 5 + 4 + 7) / 4).
Number of processing steps (process)
The Number of processing steps measure for a process instance corresponds to the sum of the processing frequencies of all functions of that process instance.
Based on your filter settings, in September 2004, two process instances are considered for the Return process.
The table below shows the processing frequency of the functions in the individual process instances:
Function | Process instance 1 | Process instance 2 |
Example 111. Create return | Example 112. 1 | Example 113. 1 |
Example 114. Create credit memo | Example 115. 1 | Example 116. 1 |
Example 117. Create returns delivery | Example 118. 2 | Example 119. 0 |
Example 120. Post goods issue | Example 121. 1 | Example 122. 0 |
Example 123. Sum | Example 124. 5 | Example 125. 2 |
For the Number of processing steps measure of the Returns process in September 2004, this produces the value 3.5 =((1 + 1 + 2 + 1) + (1 + 1) )/ 2
Function frequency
The Function frequency measure corresponds to the average number of function instances of a type executed per unit of time within the set of process instances under analysis per day.
In October 2004, the Order processing process was executed three times. In the three process instances, the Create customer order function was executed three times.
The function frequency of the Create customer order function with the By month scaling is assigned the value 0.0967 (=3/31) executions per day for October 2004.
Number of processes
The Number of processes measure corresponds to the number of process instances.
Based on the current filter settings, in September 2004, four process instances are considered in measure calculation for the Return process.
For the number of processes of this process in September 2004, a value of 4 is produced.
Process frequency
The Process frequency measure of a process type corresponds to the average number of process instances of a type within the set of process instances under analysis per unit of time.
Based on the current filter settings, in October 2004, four process instances are considered in the measure calculation for the Order processing process.
The Process frequency measure of the Order processing process in October 2004 is thus assigned the value 0.129 (= 4 / 31) processes per day.
Process costs
Sum total of the process cost rates of all functions of a process instance.
Process cycle time
The Process cycle time measure corresponds to the time span between the start time of the first function of a process instance and the end time of the last function.
For example, if only end times are specified for functions, the measure corresponds to the average time span between the end time of the first function and the end time of the last function.
Example
Based on the current filter settings, in October 2004, three process instances of the Order processing process are included in measure calculation.
The following table provides an overview of the start and end times of the relevant functions in the individual process instances.
Process instance | Start time of first function | End time of last function |
Example 131. 1 | Example 132. 03.10.04; 14 PM 11 min. 42 sec. | Example 133. 03.10.04; 14 PM 30 min. 12 sec. |
Example 134. 2 | Example 135. 15.10.04; 14 PM 40 min. 17 sec. | Example 136. 15.10.04; 15 PM 02 min. 53 sec. |
Example 137. 3 | Example 138. 30.10.04; 07 AM 10 min. 39 sec. | Example 139. 30.10.04; 07 AM 18 min. 49 sec. |
This produces the following processing times for the process instances:
Example 141. Process instance | Example 142. Processing time |
Example 143. 1 | Example 144. 18 min. 30 sec. |
Example 145. 2 | Example 146. 22 min. 36 sec. |
Example 147. 3 | Example 148. 08 min. 10 sec. |
For the process cycle time of the set of process instances of the Order processing process under analysis in the month of October, a value of 16 min. 42 sec. (=(18 min. 30 sec. + 22 min. 36 sec. + 08 min. 10 sec.)/ 3) is produced.