KPIs that count: How and why to measure planned coverage
Planned Coverage is simply a measure of how much of the maintenance we performed was planned maintenance. With more purpose in mind, we could say it is a measure of how much of our work had the benefit of a planner giving helpful information first, rather than the craftsperson starting the job from scratch. Obviously, we would like most of our work started with as much help as possible. Nonetheless, in practice, Planned Coverage is a bit difficult to leverage because we must first define a “planned job” and second understand that it is a lagging indicator.
A popular definition of a planned job is that a planner has identified all the steps to do the job and has kitted and staged all the spare parts needed. In practice, this definition is not very helpful because it infers perfection in steps and parts. For example, what does “all” mean? Out of a 100-step process to rebuild a pump with 15 parts, if the planner left out one step or one part, was the job not planned well or not planned at all? What does all mean?
Others add to the definition that the plan should be planned for the lowest skilled craftsperson in workforce. Does that mean we would plan every job for the greenest 10% of the workforce? In reality, we do not have the time to plan every job to such a level even if it were possible. And Dr. W. Edwards Deming reminds us that no one is perfect anyway. (Do we have to be told?) Instead, we should plan each job with as much detail as we can, subject to the constraint we must plan nearly all the work.
It is better to plan all the work and plan on collecting feedback from the craftspersons as they work the jobs to suggest how that work could be planned better in the future. We want all jobs to go through the Deming Cycle of continuous improvement. We can do that if we expect our workforce to have good skills and not be at the minimal level. Do not expect planning to replace the need for hiring, training, and retaining skilled crafts. Instead, we want to leverage the cumulative wisdom and experience of the whole workforce over the years by having the feedback go back into the plans.
Regarding kitting and staging spare parts: “it depends,” particularly on the individual plant storeroom circumstance for whether to even kit parts at all. Some factors include whether the storeroom is helpful or not, close by or not, fully stocked or understocked. Other factors include whether the part is expensive or not, the job is high priority or not, persons to do the work are plentiful or not. Not all planned jobs need kitted and staged parts.
So then, should we grade plans to see if they were “appropriately planned” for the work in question, or should we say anything a planner touches before execution is a plan? Neither.
Taking a sample of each planner’s plans and grading them would destroy our Deming Cycle of continuous improvement. Knowing that some plans will be graded, planners slow down to make each plan as perfect as possible instead of favoring getting more work through the Deming Cycle. They do not plan as much detail as they can, subject to the constraint they must plan nearly all the work.
But neither do we want to say simply that anything they touched is a plan. Instead, we should define that any time the planner makes a living job plan for a work order, the job was planned. Nearly every CMMS has a special planning module. The planner should create every plan here as a living plan and then utilize that living plan for the work order. The reasoning behind this extra effort is that the living plan that can be improved and reused; this CMMS module is where the planner runs the Deming Cycle. Easily enough we could count every work order that has a referenced living plan to be a planned job, as simple as that.
(Note that the quality of the plans cannot be managed by grading them. The “quality” of the plan is set when we hire, train and retain planners. Don’t expect planning to replace a skilled workforce and don’t expect grading job plans to replace skilled planners. Get planners with excellent communication and organizing skills. Then insist they run the Deming Cycle by using living plans and not making everything a “one-off” work plan.)
One final consideration of a planned job is that it must have estimated hours so we can schedule work. We simply say any executed job that had a referenced living plan and had any estimated labor hours was planned work.
Measuring Planned Coverage offers two options: we can use either labor hours charged or sheer numbers of work orders closed. The former would be calculated as the hours charged to work orders that had a living plan and estimated hours, divided by the total hours charged to work orders. The latter would be the number of work orders with a living plan and estimated hours closed, divided by total number of work orders closed.
Either metric is meaningful. We prefer any hour a craftsperson works to be on a planned job. We also want most work to be planned. Using sheer work order numbers might be more helpful to drive more planning for non-PM work if a lot of our work is large established PMs.
The common industry goal of 80% Planned Coverage cannot be “driven.” (Another note: This goal seems high anyway, especially if we have a lot of reactive work. And it is unclear whether industry means hours or work order numbers.) Planned Coverage is a “lagging” indicator. It cannot be managed anyway. The “leading” indicators of planners making living plans and quickly planning work are what can be managed.
Good Planned Coverage is a “result” of planners dutifully creating living job plans and demolishing the unplanned backlog. Manage planners to encourage them to get busy making living plans for all the unplanned work even if they are not perfect. Get that work planned, and Planned Coverage will naturally rise showing that we are helping more and more craftspersons get the benefit of planner input and not starting jobs from scratch.
