Model Maintenance
Monitor the modelling accuracy (for example, by comparing actual to forecast resource utilization) and manage efficiency (for example, automation and data availability) of IT capacity modelling. Refine and recalibrate the model's structure, parameters, and assumptions, as required to improve performance.
Improvement Planning
Practices-Outcomes-Metrics (POM)
Representative POMs are described for Model Maintenance at each level of maturity.
- 1Initial
- Practice
- Monitor actual IT capacity headroom for selected resources.
- Outcome
- Actual IT capacity headroom of individual resources is understood and can be acted upon.
- Metric
- # of IT resources monitored for IT capacity headroom.
- 2Basic
- Practice
- Ensure basic maintenance of models.
- Outcome
- Capacity models are maintained.
- Metric
- Actual IT capacity vs. plan.
- Practice
- Generate and maintain models manually.
- Outcome
- Capacity models are reused but maintenance is manual.
- Metrics
- % level of IT capacity model reuse.
- % level of IT capacity model automation.
- Frequency of IT capacity model maintenance cycle.
- 3Intermediate
- Practice
- Review and improve capacity models regularly based on on-going comparisons between planned and actual capacity.
- Outcomes
- Capacity models are well maintained and calibrated for improved accuracy based on actual vs. plan.
- Models are reused
- Metric
- Actual IT capacity vs. plan.
- Practice
- Consider ease of maintenance as a requirement during model design and apply automation as appropriate.
- Outcomes
- Models are designed with maintenance in mind.
- The modelling process is automated in parts
- Metrics
- % level of IT capacity model reuse.
- % level of IT capacity model automation.
- Frequency of IT capacity model maintenance cycle.
- 4Advanced
- Practice
- Review and improve capacity models based on error tolerances agreed with business.
- Outcome
- Capacity models are well maintained and accuracy aligned based on business input.
- Metric
- Actual IT capacity vs. plan.
- Practice
- Structure models for ease of maintenance and apply automation to improve efficiency of model maintenance.
- Outcomes
- Models are easy to maintain.
- The modelling and model maintenance process is automated
- Metrics
- % level of IT capacity model reuse.
- % level of IT capacity model automation.
- Frequency of IT capacity model maintenance cycle.
- 5Optimized
- Practice
- Regularly check the statistical properties of modelled data for changes.
- Outcome
- Understanding of statistical changes leads to higher accuracy of model predictions.
- Metrics
- Actual capacity vs. plan.
- Frequency of analysis of statistical properties of modelled data.
- Practice
- Adapt models to reflect possible changes in statistical properties to increase accuracy of models.
- Outcome
- Understanding of statistical changes leads to higher accuracy of model predictions.
- Metrics
- Actual IT capacity vs. plan.
- Frequency of analysis of statistical properties of modelled data.
- Practice
- Review and optimize the maintenance process on a continual basis.
- Outcome
- The model maintenance process is optimized.
- Metrics
- % level of IT capacity model reuse.
- % level of IT capacity model automation.
- Frequency of IT capacity model maintenance cycle.
- Practice
- Review and action opportunities for further or improved automation on a regular basis.
- Outcome
- The level of automation is adapted to special needs of the organisation.
- Metrics
- % level of IT capacity model reuse.
- % level of IT capacity model automation.
- Frequency of IT capacity model maintenance cycle.