R Studio
Incident Ticket Analysis:
-
Root Cause Analysis: Utilize R Studio to perform in-depth analysis of incident tickets, identifying underlying root causes and patterns.
-
Data-Driven Insights: Leverage advanced statistical techniques and data visualization tools to uncover actionable insights, enhancing incident management processes.
-
Proactive Solutions: Develop and recommend data-informed strategies to mitigate recurring issues and improve overall system reliability.
​
In ServiceNow create a Report (List) of all incidents created and configure it fitting for your purpose. Then export the Excel (.xlsx) file. You either then first start cleaning, transforming, and so on in Excel first before you import it to R Studio. Depends on what works the best.
​​
Before starting any Exploratory Data Analysis (EDA):
- As for me it where several sheets I had to read each sheet into a separate data frame, check the structure, rename it, list it and than combine them.
- Than convert to datetime so it all would be the same format, calculate resolution time in days, hours, minutes and seconds, calculate the average resolution time, convert the format, and add new columns.
Analysis of Configuration Items with Highest Volume -
Bar Chart of Top 20 Configuration Items
​
-
Data Preparation: Gather and clean data to ensure accuracy and consistency, laying a solid foundation for analysis.
-
Exploratory Data Analysis (EDA): Utilize advanced techniques in R Studio to explore and understand the data, identifying key patterns and trends.
-
Visualization: Create insightful bar charts to visually represent the volume of configuration items, facilitating clear and effective communication of findings.
​
Based on this visualization, you can conclude what Configuration Items are handling the most volume and might need more resources or attention.
-
Identify Key Configuration Items: Focus on the configuration items that have the highest number of incidents. These are the areas where improvements can have the most significant impact.
-
Training and Knowledge Articles: For the top configuration items, consider developing targeted training programs and comprehensive knowledge articles. This can help reduce the number of incidents by empowering users with the information they need to resolve common issues themselves.
-
Automation Opportunities: Look for repetitive tasks or common issues associated with the top configuration items that could be automated. Automation can reduce the workload on support teams and improve response times.
-
Resource Allocation: Allocate more resources to the configuration items that contribute to the majority of incidents. This could mean dedicating more support staff, investing in better tools, or enhancing monitoring and maintenance practices.
-
Trend Analysis: Use the data to identify trends over time. Are certain configuration items consistently problematic? Are there seasonal spikes in incidents? Understanding these patterns can help in proactive planning and mitigation.
Cause and Effect Diagram - "Fishbone Diagram
​
-
Identify the Effect: Start at the head of the fish, which represents the problem or effect you are analyzing. In this case, it’s “Low FCR rating from L1 agents.”
-
Examine the Main Categories: The main branches of the fishbone represent broad categories of potential and input them into R.
-
Develop Action Plans: For the prioritized causes, develop specific action plans to address them. For example:
- Man: Implement comprehensive training programs and set realistic expectations.
- Machine: Improve system maintenance schedules and reduce downtime.
- Method: Standardize processes and streamline procedures.
- Material: Ensure knowledge articles are complete and records are consistent.
- Measurement: Improve data accuracy and calibration processes.
- Environment: Stabilize organizational changes and enhance network connectivity.
- One of the plots shows a significant outlier that affects the regression line. In the ticket data, outliers (e.g., tickets that are reopened or reassigned multiple times) could skew the analysis. Identifying and understanding these outliers can help in addressing specific issues that cause frequent reopens or reassignments.
​
- Another plot shows a curved relationship, indicating that a simple linear model might not be appropriate. Similarly, the ticket data might have non-linear relationships that needs to be explored using more complex models.
​
- One plot shows a consistent linear relationship, suggesting that if the data follows a similar pattern, there might be a straightforward correlation between certain factors and ticket reopens or reassignments.
​
Conclusion
- Further Analysis: Need to conduct a deeper analysis to understand the specific reasons behind ticket reopens and reassignments. This can involve looking at the types of issues, the teams involved, and the time taken to resolve tickets.
- Training and Knowledge Articles: When certain patterns or outliers are identified, targeted training and detailed knowledge articles can help reduce the number of reopens and reassignments.
This Scatter Plot shows a positive correlation between the number of times tickets are reopened and the number of times they are reassigned.
Conclusions
- Positive Correlation: The trend line indicates that tickets that are reopened more frequently also tend to be reassigned more often. This suggests a relationship where unresolved issues might lead to both reopens and reassignments.
- High Reopen and Reassignment Counts: Tickets with high reopen counts often have high reassignment counts. This could indicate complex issues that require multiple interventions or expertise from different teams.
- Outliers: There may be specific tickets that are reopened or reassigned significantly more than others. These outliers can skew the overall analysis and might need special attention.
Recommendations
- Targeted Training: Focus on training for the teams handling the tickets with the highest reopen and reassignment counts. This can help reduce the need for reopens and reassignments by improving initial resolution quality.
- Knowledge Articles: Develop detailed knowledge articles for common issues that lead to reopens and reassignments. This can empower support staff and users to resolve issues more effectively.
Automation: Identify repetitive tasks or common issues that can be automated. Automation can help reduce the workload on support teams and improve response times.
- Root Cause Analysis: Conduct a root cause analysis for tickets with high reopen and reassignment counts to identify underlying issues. Addressing these root causes can help prevent future occurrences.
​
This Box Plot titled “Resolution Time Distribution per Month” provides valuable insights into the average resolution time of tickets across different months.
​
Conclusion
-
Median Resolution Time: The median resolution time varies slightly across the months, indicating some consistency but also room for improvement.
-
Interquartile Range (IQR): The IQR, which represents the middle 50% of data points, varies between months. A wider IQR suggests more variability in resolution times, while a narrower IQR indicates more consistency.
-
Outliers: Some months show potential outliers (points beyond the whiskers), indicating tickets that took significantly longer to resolve. These outliers can skew the overall analysis and might need special attention.
​
Recommendations
-
Targeted Training: Focus on months with higher median resolution times or greater variability. Providing targeted training for support teams during these periods can help improve resolution efficiency.
-
Knowledge Articles: Develop and update knowledge articles to address common issues that lead to longer resolution times. This can empower support staff to resolve tickets more quickly and consistently.
-
Automation: Identify repetitive tasks or common issues that can be automated. Automation can help reduce the workload on support teams and improve response times.
-
Root Cause Analysis: A root cause analysis needs to be conducted for months with significant outliers to understand why certain tickets took longer to resolve. Addressing these root causes can help prevent future occurrences.
-
Continuous Monitoring: Regularly monitor resolution times and update training and knowledge articles as needed. Continuous improvement can help maintain and enhance resolution efficiency over time.
​