Interlocking KPIs can help to ensure that system users get a well-rounded view of the organization’s performance. The big picture helps the team to focus the scorekeeping system on process improvement, rather than suboptimization or management appeasement. This is especially important in organizations that have frequent flavor-of-the-month improvement programs. In this environment, middle management often develops amazing expertise in manipulating scoring systems to create the appearance of great progress. This kind of showmanship is very tempting if another initiative will replace the current one in a month or two. Deep, lasting progress typically takes longer than a few months. So does building trust and understanding in a KPI system.
The maintenance KPIs discussed so far in this series have been:
- percent of emergency work
- percent of on-time completion
- percent of PM on-time completion
- percent of overtime
- percent of work scheduled
- maintenance backlog.
The interplay among these values is a good place to start the discussion of interlocking KPIs. Our first KPI, the percent of emergency work, identifies the amount of work that must be performed without proper planning. Please remember that unplanned work consumes three to five times the resources that the same work would have burned if it had been planned with proper lead time. Thus if emergency work is 15% of the department’s orders, it is sure to be consuming half of the maintenance resources. If unplanned work is consuming half the department’s resources, the percent of on-time completion will take a beating, or overtime might go completely out of control.
One way to hide the impact of unscheduled work would be to keep resources unscheduled so that they will be available for emergency work. However, since we are tracking percent of work scheduled, if half of the resources are being consumed by unplanned work, we will see it in percent of work scheduled. To look at it another way, if all the maintenance resources are scheduled and 15% unplanned work comes in, half the planned work is not going to be completed on time without a tremendous amount of overtime. More likely, department management will allow PM work to slip, since nobody is asking for it. Then we catch the problem with the PM completion KPI.
Whether repairs or PM work are allowed to slide, the backlog will begin to grow fast. We will see it in the weeks of backlog KPI.
|J. Stanton McGroarty, CMfgE, CMRP, is senior technical editor of Plant Services. He was formerly consulting manager for Strategic Asset Management International (SAMI), where he focused on project management and training for manufacturing, maintenance and reliability engineering. He has more than 30 years of manufacturing and maintenance experience in the automotive, defense, consumer products and process manufacturing industries. He holds a bachelor of science degree in mechanical engineering from the Detroit Institute of Technology and a master’s degree in management from Central Michigan University. He can be reached at firstname.lastname@example.org or check out his Google+ profile.|
This is how interlocking KPIs create a system that models the overall performance of the maintenance operation. While these values are built on historical performance, they can also be viewed as forward-looking indicators that forecast the direction of backlog growth and the cost-effectiveness of the maintenance operation. Also, whether it is seen initially as overtime growth, poor maintenance completion, or poor PM completion, the degradation of machine performance will be predictable. Customer service, OEE, and availability will respond next to a maintenance slide. If we are tracking a proper set of interlocked KPIs, we can prevent this chain of events from being a surprise. Hopefully, it will be prevented altogether if we have seen it in our forward-looking metrics.
Note that if any of the KPI set had been eliminated, it would have been easy to get good KPI grades without controlling maintenance. For instance, if we can’t keep up with the work, but we’re not tracking overtime, we can schedule overtime freely. This is particularly easy in budgeting systems where the OT column only carries the overtime premium, instead of total payroll cost. Overtime work costs three or four times the premium, but this is not visible in the typical accounting system.
Similarly, if we’re not tracking the percentage of work scheduled, then we can get great completion numbers by just scheduling half the workforce and half the work. Eventually we’ll see it in the backlog, but then we’re as likely to hire more maintenance staff as to schedule the ones we have.
An interlocking set of KPIs is not a substitute for managing our areas, but they make a great dashboard to tell us about the near-term results and long-term prospects of our current management decisions. KPIs should make us harder to surprise and equip us to support good financial decision making by top management. Just think, if you are scheduling all of your people and getting good completion of the work, a growing backlog or continued emergency work suggest that you either need more resources or you need to choose the work better. This could be the start of staffing and training improvements that have been overdue in past budgets. A careful examination of the nature of the failures and a detailed review of the growing backlog may also provide useful clues to what next steps the company ought to be funding. Sounds downright strategic, doesn’t it?