Five mistakes companies make when setting up KPIs
1. Selecting a single KPI to monitor
A single KPI will be hard pressed to monitor performance and trends effectively enough to reveal required improvements. Instead, an organization needs to formulate a family of relevant and interrelated KPIs that operate at various levels of hierarchy to meet the total needs of the business. For the maintenance industry, standard means of quantifying overall maintenance performance often include:
- Measurements of OEE (Overall Equipment Effectiveness),
Maintenance cost as a percentage of estimated replacement value (% ERV)
2. Not connecting KPIs to larger business goals
Ultimately, all KPIs in use should relate to larger corporate goals. Accordingly, the setting of organizational KPIs should be a “top-down” process. When lower level KPIs used to monitor individual departments and processes are derivatives of corporate indicators, they assume improved relevance and understanding.
Some argue that essentially all maintenance managers perform the same tasks. Typically these involve:
- Statutory checks
- Time-based maintenance
- Life cycle-based maintenance
- Condition monitoring (subjective/objective)
- Safety issues
- Environmental issues
- Inventory control
- Financial (e.g., maintenance cost per unit of production or inventory value as a percentage of estimated replacement value).
- Overall maintenance performance, assessing the quality and efficiency of services provided to the organization (e.g., unscheduled downtime as a percentage of total downtime, or maintenance re-work hours as percentage of total maintenance hours).
- Maintenance process KPIs providing timely warning that scope for improvement exists (e.g., predictive maintenance effectiveness or emergency purchase orders raised as a percentage of all purchase orders raised).
- Functional KPIs to identify specific improvements for individual activities such as preventative and predictive maintenance. These could include cost savings that result from early detection of faults (measuring PdM effectiveness) and schedule compliance.
- System-related KPIs identifying the effectiveness and utilization of maintenance management systems, such as the number of equipment items that exist within the CMMS database, expressed as a percentage of the total number of equipment items on site.
- Inventory management KPIs involved in the measurement of maintenance inventory management often parallel KPIs employed for management of other inventories (e.g., raw materials, work in progress, or finished goods).
- Personal development KPIs such as training hours per maintenance employee.
3. Not knowing what “good” looks likeBenchmarking is a continuous improvement process through which an organization attempts to identify “world class” or “best in class” practices by objectively studying other organizations and their procedures. Benchmarking projects cannot be quantified unless performance is measured and improvements are tracked. But there is more to benchmarking than just metrics. In fact, an over emphasis on metrics may cause benchmarking to lose focus.
That’s because the ultimate objective of benchmarking is to identify the practices behind the metrics. When a performance indicator can be clearly and unambiguously aligned with best practice, it becomes a benchmark that may be employed as a process improvement goal.
Benchmarking projects require performance improvements tracking through KPIs. Unfortunately, many documented metrics of public domain data sources relate to higher level KPIs. Although they may be helpful when assessing an improvement scope, they rarely provide diagnostic details to identify specific process improvements.
4. Choosing KPIs that are too hard to measure
Effective use of KPIs requires a clear and accurate measurement process. If data varies due to collection inconsistencies, the degree of change will need to be greater before it assumes significance, which also limits timeliness.
This is particularly relevant when attempting to objectively quantify subjective facets of maintenance performance – for example, attempting to rate a group’s attitude toward an aspect of maintenance by asking an assessor to provide a value for on a scale of one to ten gives variable results.