Open topic with navigation

Change Failure Prediction

Change Failure Prediction iCube enables you to create ad-hoc reports that help you to track all historical changes and the probability of a current or future change being successful. You can generate reports to view the probability of success or failure for a change and the risk associated with each change request.

The Change Failure Prediction iCube contains attributes and metrics for variables commonly associated with unsuccessful Changes. This can be customized by including additional predictor variables that are found to be significant while building a model.

Attributes

ClosedExpand this link to view the list of available attributes and their descriptions:
Attributes Description
Assigned To Person primarily responsible for working on the change request
Assignment Group Group or team of the person primarily responsible for working on the change request
Caused Incident Flag Flag indicating if a closed change has caused one of more incidents. This is not applicable for open and canceled changes.
Caused Problem Flag Flag indicating if a closed change has caused one or more problems. This is not applicable for open and canceled changes.
Change Change Request number as mentioned in ServiceNow
Change Failure Risk

The risk of a change request failing based on the failure probability. The risk buckets are configured as follows but can be customized:

  • LOW - Failure Probability <=5%
  • MED - Failure Probability >5% to <=25%
  • MED-HIGH - Failure Probability >25% to <=50%
  • HIGH - Failure Probability >50%
Change Number Same as the Change attribute. However, this attribute also provides a form to display the ServiceNow source URL.
CI Class Class of the Configuration Item (CI) associated with the change request
Close Code Closure code associated with the Change Request. This is only applicable for closed changes and not for open and canceled changes.
Failure Flag

This flag indicates the definition of Change Failure.

First level check is based on the Closure Code of a Change Request:

  • If the Closure Code is “Successful”, the Flag value should be “N”
  • If the Closure Code is “Unsuccessful” or “Successful with Issues”, the Flag value should be “Y”

Second level check is performed on Change Requests with Closure Code as “Successful”:

  • If a Change Request has caused one or more Incidents, the Flag value should be “Y”
  • If a Change Request has caused one or more Problems, the Flag value should be “Y”
  • If a Change Request has caused one or more Outages, the Flag value should be “Y”
Failure Probability

Captures the failure probability for an open Change Request as predicted by the Change Failure Prediction model

This is applicable only for “Open” Changes (Changes where standardized DWH Code = “Open”)

For changes in Closed, Canceled, Post-Implementation Review states, Failure Probability will be populated as “Not Applicable”

Short Description Short description of the Change Request from the source
State State of the Change Request from the source

Metrics

ClosedExpand this link to view the list of available metrics and their descriptions:
Metric Name Description Formula Expected Value
AG Prior Changes {Change Number}

Count of all changes closed by an assignment group before the start date of the current Change

 Sum(AG Prior Changes) >=0
AG Prior Failure Rate {Change Number}

Percentage of failed changes that were previously closed by an assignment group before the start date of the current Change

Note: If the Planned Start Date, Actual Start Date, or Opened On date is null, then the value is displayed as Data Inconsistent.

Sum(AG Prior Failure Rate) >=0
Assigned To Prior Changes {Change Number} Count of all changes closed by a user before the start date of the current Change Sum(Assigned To Prior Changes) >=0
Assigned To Prior Failure Rate {Change Number} Percentage of failed changes that were previously closed by a user before the start date of the current Change Sum(Assigned To Prior Failure Rate) >=0
Change Backlog Count of all Changes that are currently in the backlog queue Count (Opened Changes) where the Backlog flag is set to Y >=0
CI Class Prior Changes {Change Number} Count of all changes closed before the start date of the current change and that are associated with the CI Class of the current Change [CI Class Prior Changes] >=0
CI Class Prior Failure Rate {Change Number} Percentage of failed changes that were closed before the start date of the current Change and that are associated with the CI Class of the current Change [CI Class Prior Failure Rate] >=0
CI Prior Changes {Change Number} Count of all closed changes associated with a Configuration Item before the start date of the current Change Sum(CI Prior Changes) >=0
CI Prior Failure Rate {Change Number} Percentage of failed changes out of closed changes that were previously associated with a Configuration Item before the start date of the current Change Sum(CI Prior Failure Rate) >=0
Closed Changes Count of all closed changes Count (Opened Changes) WHERE Change State is 'CLOSED' >=0
Failure Rate Percentage of unsuccessful changes [Unsuccessful Changes] / [Closed Changes] >=0
Incidents Caused by Change Count of Incidents caused by a closed Change Sum(Incidents Caused by Change) >=0
Open Changes at Risk Count of currently open changes with a medium or high risk of failure

[Sum (Open Changes) where Change Failure Risk = MED-HIGH] + [Sum (Open Changes) where Change Failure Risk = HIGH]

 >=0
Opened Changes Count of all opened changes Count (Opened Changes) >=0
Problems Caused by Change Count of problems caused by a closed change Sum([Problems Caused by Change]) >=0
Unsuccessful Changes Count of all closed changes marked as unsuccessful or failed Count([Closed Changes]) where the Failure Flag is set to Y >=0

Note: If the Planned Start date, Actual Start date, and Opened On date for the Change Request is null, then:

© 2022 Digital.ai Inc. All rights reserved.