Podcast: Why everyone is faking their KPIs (but shouldn't be)

Joe Kuhn, CMRP, former plant manager, engineer, and global reliability consultant, is now president of Lean Driven Reliability LLC. He is the author of the book “Zero to Hero: How to Jumpstart Your Reliability Journey Given Today’s Business Challenges” and the creator of the Joe Kuhn YouTube Channel, which offers content on starting your reliability journey and achieving financial independence. In our monthly podcast miniseries, Ask a Plant Manager, Joe considers a commonplace scenario facing the industry and offers his advice, as well as actions that you can take to get on track tomorrow. This episode offers insight into why fake numbers fake real progress.

Below is the transcript of the podcast:

PS: For all of our listeners, this is really your chance to ask those burning questions that you want to ask maybe your plant manager or leadership, but you're unsure of how they might react. Ask us those tough questions. Joe has 30 plus years of experience working in plants as a manager and in maintenance, but he's retired from that, so he can share all the secrets with us. And he can answer those hard questions or touch on those more controversial topics that might be difficult to bring up in your everyday operation. So Joe, you're definitely not afraid to answer those tough questions, right?

JK: Exactly right. And working through these is a big part of the culture change that people need to go through. But it's a big part of my story, and that's why I enjoy sharing.

PS: Alright, so today, we're going to dive into one of those, maybe more of a problematic topic for people. And so last month we talked about celebrating a planned maintenance culture rather than championing the emergency maintenance saves that you have to make as a result of poor planned maintenance. So as part of that discussion, you laid out an interesting scenario related to planned maintenance and KPIs or the important key performance indicators for maintenance. So in this example, we're looking at the plants assigned planned work for the next day. This specific job that we're talking about in reality is two hours of work for two people. But the job is planned as two hours for eight people, because they know they're going be pulling people off to do emergency work instead. So sometimes PMs don't get done entirely, and yet, they still get marked as complete, or emergency work gets marked as planned work, so the PM KPIs aren't truly reflecting the actual work environment.

Joe, how common is it for plants to fudge their KPIs like this? And why might they be doing that? Aside from this example, what are some other maintenance and reliability KPIs that plants might doctor in their favor? And ultimately, why is faking KPIs not a good idea? What are you actually losing by pretending to have better numbers than you do?

JK: Well, there's a lot in there. One correction I will make, and I know you just slipped up on this, but instead of planning a job for eight people, two hours, they plan the job for two people, eight hours. Okay, so they just add a lot of time to that. Unfortunately, this is very common, and one of the things I used to call it, I used to call it managing by Facebook. It's like people's Facebook life is not exactly the same as their real life. You don't see everybody struggles. You just see all the nice things. And the problem with KPIs, I'm telling you, when I go into a plant, the plants I worked at, I believe less than 50% of what the KPIs told me. 50%, okay. And for some of them, that's a little generous.

Now, you may say, why? Why is this going on? It's a great place to start. Top management, I blame them at least 50% here. Top management, they may not know anything about maintenance. They go to a seminar, or somebody gets in front of them says, ‘Hey, you’ve got to get your PM compliance up, your preventive maintenance compliance up, you’ve got to get it to at least 90%.’ Then, they go back to their plant, and they see that it's at 50%. So they say, ‘It needs to be 90%. You got 30 days.’ Well, you've got 50 years of culture there. You've got all kinds of issues going on. You may have 80% unplanned work, could be 20. I don't know what the planned work is. But getting PM compliance to that level that quick, with the same number of people on the same budget, the same culture is a difficult task.

What you do is you start redefining. Put your thumb on the scale, let's just call it that. Put your thumb on the scale and say, ‘Well, we're doing this PM. It's only two hours long, but they always find something so we want them to correct it. We don't know what it's going to be, so let's just add six more hours to the job and then we'll be sure to get it done.’ And that's all called PM hours. And it seems innocent. It seems innocent, and it placates top management.

Top management, even now, and this was started about the time I retired, people have apps on their phone, and their phone has these 27 KPIs for the plant, say five of them are maintenance. ‘I want them all great. Why is your PM compliance red? Fix it. Okay, fix it.’ Put your thumb on the scale, it's pretty easy to fix. Sometimes it's just to survive. Sometimes it does show improvement. I really think it's trying to preserve the current culture. I think a lot of people, unfortunately, in maintenance, I hate to say the word leadership, but it's more they're in the management of their current state. They know this system. They've been doing it for a while. They refuse to lead their way out of it. Okay, so what do you get if you get the expectations from top management to change this KPI? Well, we're doing all good stuff down here. I'll just modify the definition, that KPI. Put my thumb on the scale a little bit and can modify it.

Hey, I'm sorry, this is blunt, but this is what I found. I've been in over 40 locations, and this is common. This is common. If you don't think it's happening at your plant, there’s a good chance you're wrong. A good chance you're wrong. You got to go and see. And when you go and see, and you audit PMs, you'll see ‘Hey, the PM is supposed to be two hours. Corrective work is not part of the PM.’ Unless their PM is to change the brake pads out, but a lot of PMs are to inspect this, look at this, adjust that, not, ‘Oh my gosh, we found this oil leak. Oh my gosh, we found this other problem.’ And you're supposed to write a different work order for that, and that is unplanned work. The vast majority of people don't do that, and quite often when you're doing that PM you schedule it two people, two hours, and it's only two hours long. So what's the other eight? Or six? Sorry. What's the other six? Well, could be emergency work, could be some honey-do jobs out there. It's not tracked and you lose it. And you think on paper, when you're doing your planning and staffing, you think that the PM takes that full amount of time, and it doesn't. So that's the why it's a dance to appease top managements, kind of a dance to maintain the culture that you’ve got.

Now, second part of your question was, what are some other KPIs that I have horror stories on? Another one that was big at my plant was we tracked R&M dollars (repair and maintenance dollars) per metric ton we create. Every plant’s got something like this R&M dollars per, a million dollars in sales, per linear foot, whatever your product is there you create. What part of that cost of goods sold is R&M. Well, that's easy to manipulate very easy. You may decide to start calling some expenses that you have today, tomorrow you can call them production expenses. So for example, say you want to change out some rolls on a conveyor. Well, you're not changing those rolls out because of a maintenance issue, you're changing them out because of a quality issue. That's a production issue. They're the ones that need the quality. These rollers aren't quite making the turn. They're still turning. There's nothing wrong with them, but we’ve got an increased quality standard for this one customer. So we're changing out the rolls for production reasons. So that's a production cost. That was very common in an aluminum rolling mill where you had to change the rolls, because you would change with the product. So if you're running a narrow product and you'd go to wide product, you'd have to change the rolls. And maybe people did that yesterday, that was a maintenance cost. Today, it's a production cost, extremely easy. Just to take money out of one pocket, put in another pocket and make your R&M dollars change as you're relating it to production. Extremely easy to do. Schedule compliance. This is related to PM compliance. ‘We created a schedule last Thursday. We executed it this week, and we add hours into the PM. We add hours into the outage time and boy, we just did exactly what we plan to do.’ That's because you plan to do unplanned work. That makes sense.

Another one is this. I’ll tell a really bad story here. I won't say who but a really bad story. One of the things I was trying to drive really hard at my company was what percent of work was from predictive maintenance. So we were trying to get away from time-based maintenance PMs. Do this every month. Do this every quarter and doing more work from infrared, oil sampling, things like that. So predictive technology.

One of the plants I went to counted grass cutting as predictive maintenance because they only cut it when it is so tall. So they had a measure. It wasn't based on time. ‘We don't cut the grass every week. We cut it when it's tall. That's condition monitoring. That's predictive maintenance. Literally, this was in there.

Another way that you can inflate predictive maintenance is say you have two full-time predictive maintenance technicians. They're going out and doing the IR routes, the vib routes, the oil analysis. Well, you put them on a standing work order, but the standing work order says ‘Hey, Joe and Anna, you're the predictive technicians 40 hours a week. All your time is charged to that.’ But you only let Joe and Anna do that job two days a week. And three days a week, you're pulling them off. At the end of the week, we're counted as 100% PdM that inflates the PdM metric. Super easy to do. Super easy to do.

PM compliance. This is lesson one that I want to break apart a little bit more than the first example we gave. If you're 90% PM compliance, is that good or bad? Well, some people will say good. I say, well depends on the criticality the assets that you're applying the PMs to. Say the most critical asset bottleneck of your plant is the 10% of PMs you didn't do. You did all the PMs on the ladders and the catwalks and the drinking fountains, but on lubrication, ‘Oh, we didn't do any PMs on lubrication this month. Is that a big deal?’ So doing PMs by criticality is so important. Just a raw number, unless it's 100% doesn't mean a lot to me. Which PMs are you getting done? So PM compliance based on criticality to me that's very, very important.

So another number that can be managed, the last one I'll mention, which is a big metric for maintenance organizations, is percent unplanned downtime. Okay, how can you manage that? ‘Oh, well, this bearing is starting to make noise. Yeah, we can hear it with our ear. We need to take it down in the next few days. Today is Friday; we'll run across the weekend. We'll call that a planned job that we're going to do on Monday. We're going to change out that bearing, planned on Monday.’ Is that really a planned job? Or did you have a failure that you now decided to call planned?

Any KPI can be manipulated. I can give a thousand more stories. But in the end, you're just fooling yourself. The first step to creating a culture of reliability is being honest with yourself. And this is a big one. This is the one that people talk about behind closed doors. When you go out audit, it's very easy to see and uncomfortable for people to talk about, but it's a great place to start seeing reality, what's really happening on a PM. Are you really conducting it with precision? Does it take the amount of time? Does it take the amount of people? Is it the right frequency? I can't say that I've ever seen that a plant gets an A+ on a first audit, never get an A plus. Most of them are lucky to get a C on that.

PS: I love, we love your horror stories. I'm sure you have more, but everybody loves those and everybody has got to stop planning their unplanned work. I thought that was kind of a good phrase that you throw out there. I think in some cases this is a check your ego at the door issue, and when it's about ‘I’ and not the team or the culture as a whole, you're there to just please somebody upstairs with a false sense of security. That's not going to help your bottom line. It's not going to help the culture. It’s not going to help the maintenance practices in the long run. You have to know the reality. You can't doctor it in your favor in any way to solve problems.

JK: Managing your KPIs: a couple summary points here. It just gives everybody a false sense of security. You think you're doing the PMs, and you're not. You think you're doing predictive maintenance at a certain level, and you're not. What that results in is bad decisions. For example, say that you've got a conveyor, and the unplanned downtime is just killing the plant. But you're doing all the PMs. You think you're doing all these PMs. You're doing all this predictive maintenance on it, then you say, ‘Hey, we need to spend $500,000 on a new conveyor, and you really weren't doing the maintenance on it. You weren't doing the problem solving or you weren't doing the predictive routes. You weren't lubricating it, so it forces you to make bad decisions. It masks the waste in the system. And these KPIs, this is a really sore spot for me, because I just don't believe them, and when I said 50%, earlier, I'm being kind. Bottom line. This is common sense practice in plants, and it's to appease top management. And it's to keep the current culture. And if that's what you want in your organization, just keep on believing. It's a great place to start, validate your KPIs. Validate them. I will be shocked, if you say they're 90% accurate. I'll be shocked with 90%.

PS: That's a little disconcerting maybe. Here are all the different ways that you can manipulate those numbers. I mean, there's as many ways to make them false as it is, or probably easier to make them false, than it is to find the actual numbers that are really representative of your business.

About the Podcast
Great Question: A Manufacturing Podcast offers news and information for the people who make, store and move things and those who manage and maintain the facilities where that work gets done. Manufacturers from chemical producers to automakers to machine shops can listen for critical insights into the technologies, economic conditions and best practices that can influence how to best run facilities to reach operational excellence.