Measuring predictive maintenance program success

John Schultz

One of the key decisions in predictive maintenance is selecting the right metrics. Why? First off, what gets measured gets improved. Or conversely, what doesn’t get measured never will be improved. Tracking and reporting on key metrics lets you focus squarely on the behavior changes you want. Plus, you can shine a bright light into dark corners for everyone to see. Second, if you don’t measure something, you can’t prove it ever happened. As you may have heard, “In God we trust. Everybody else brings data.” If you want to prove the success of your program to upper management, you need to have clear, compelling data to back it up.

Before you begin planning which metrics to use, consider a few points:

The less you know, the more you must measure. If you know little or nothing about a process, everything appears to be random. So, you have to measure it until you understand it.

Keep it simple. Metrics should be simple, straightforward and meaningful.

Align activities with goals. Choose measures that translate your long-term reliability strategy into tangible goals and short-term actions. Ask yourself: To achieve the goals, at what must we excel?

Get buy-in. Don’t just come down the mountain with your new decree on a pair of stone tablets. Get feedback from key people in operations, finance, safety, quality and maintenance, and ask them, “Here’s what I’m considering. What do you think?”

Create a common language. Choose metrics that create a common language among diverse team members.

Avoid metric overload. It’s important to measure the right factors, but don’t get too metrics-happy. Too much information-gathering and reporting dilutes the purpose. You’ll spend more time measuring and reporting than making changes.

Phase them in. If you pick 12 metrics and try to focus on all of them at once, that’s 12 rounds of discussion and debate. Instead, take two or three at a time, test them and make sure they work. You can get quick wins for the organization by creating smaller deliverables along the way.

With that said, the No. 1 metric to consider is: What percent of your total maintenance man-hours is driven by predictive maintenance? As I wrote last time, best practice reliability programs generate an average of 50 percent of their work from predictive maintenance inspections and corrective work. We know predictive maintenance is more cost-effective than preventive or emergency maintenance. So, shouldn’t predictive maintenance account for at least half of your work?

This figure shows the cycle to
improved equipment reliability.
(Click here to enlarge)


Here’s a list of the top 10 metrics to track. Your usage will depend on the maturity of your program.

  1. Predictive maintenance effectiveness, or the number in hours of corrective work identified by predictive maintenance divided by hours spent on PdM inspections. You should be able to identify 1.5 to 2.5 hours of PM to every one hour of predictive maintenance applied.

  2. Mean time to implement predictive maintenance recommendations and the percent of predictive maintenance recommendations implemented, say on a 12-month rolling average. It is a better method. Your target should be less than 45 days showing a 12-month rolling average greater than 80 percent.

  3. Percent of work flow generated by predictive maintenance. You can be effective with results of 50 percent or greater.

  4. Total planned work in the planner’s backlog that was generated by PdM work. Your target should be four to six weeks of backlog, 50 percent of which should be predictive maintenance.

  5. Predictive maintenance schedule adherence. You should approximate 90 percent or more.

  6. Productivity per predictive maintenance analyst per technology. How many machines or inspection points do your analysts handle each month? Your targets will vary based on operating context and access.

  7. Asset heath. You want more than 80 percent of your assets to be defect-free. It’s critical to have the right technologies and monitoring standards in place to identify all possible defects.

  8. Maintenance costs as a percent of the asset base. You should aim for less than 2.5 percent.

  9. Percent of maintenance labor dedicated to performing PdM inspections. Your target is between 10 and 15 percent. The graphic on the preceding page will help to explain this.

  10. Overall equipment effectiveness and asset utilization. The pace-setter targets are typically above 92 percent in most industry verticals, but take into account your operating context when defining your targets.


Some people spend plenty of time tracking cost avoidance. In other words, “If we hadn’t caught this problem, it would have cost X dollars in emergency repairs.” But here’s the problem: You’ll never find cost avoidance on a company’s financial statements, so it has little real value to management. What’s important to measure are the factors that lead to:

  • Lower labor costs
  • Lower spare parts inventory
  • More throughput capacity
  • Less energy consumption
  • Better safety
  • Improved product quality

All of these directly impact financial performance. Focus on them and you can’t go wrong.

John Schultz is a partner with Allied Reliability and a Certified Maintenance and Reliability Professional through SMRP. As the largest consulting, engineering and service firm focused on predictive and preventive maintenance, Allied Reliability serves more than 200 plants and facilities in the U.S., Canada, Europe and Latin America. For more of John’s insights, get a free copy of “PdM Secrets Revealed! How to Improve Your PdM Program or Start One from Scratch” at www.alliedreliability.com. For answers to your specific questions, call John at 812-841-9252 or e-mail him at schultzj@alliedreliability.com.

Subscribe to Machinery Lubrication

About the Author