Why 94% of Manufacturers Aren't World-Class (And What to Do About It)

A few months ago, I was walking a floor with a plant manager who'd just finished presenting his continuous improvement results to corporate. He was proud—and he should have been. They'd hit 72% OEE, up from 64% two years earlier.

"That's pretty good, right?" he asked.

I didn't have a simple answer. 72% is above average. It's respectable. It's also leaving about 13 points of capacity on the table compared to the facilities that have figured this out. On his throughput, that gap represented roughly $8 million a year in potential output.

That conversation made me want to dig into OEE data more seriously. Not the vendor pitch deck numbers, but what's actually happening out there. What separates the plants that hit 85%+ from the ones stuck in the 60s and 70s?

Where Most Facilities Actually Land

The honest global picture is sobering. According to Evocon's OEE research and industry benchmark data I've analyzed, the average discrete manufacturing facility runs somewhere between 60% and 75% OEE. The number has crept up over the past five years—from about 58% in 2020 to around 68% now—but most facilities remain firmly in the middle of the pack.

That last number is the one that should get your attention. Roughly 6% of manufacturing operations achieve world-class OEE. The other 94% either haven't figured out how, haven't prioritized it, or have accepted "pretty good" as good enough.

Context Changes Everything

Before anyone feels bad about their numbers, context matters enormously. I've seen food and beverage plants agonize over 70% OEE when they should be celebrating—mandatory sanitary cleaning eats 15-20% of their available time, so their world-class threshold is really around 75%.

Meanwhile, electronics manufacturers routinely hit 88%+. Their processes are highly automated, their products are consistent, and they've had decades to optimize. Comparing yourself to them isn't fair.

Make-to-order shops—custom fabrication, specialty vehicles, that kind of work—struggle the most. High variety, constant changeovers, every job different from the last. Fewer than 5% of make-to-order operations achieve world-class numbers. If you're running a job shop at 70%, you might actually be exceptional.

What I've Noticed About the Top Performers

After spending time in facilities across the OEE spectrum, I've started to see patterns. The world-class operations aren't doing anything mysterious—they're just doing basic things with unusual consistency.

They actually know their numbers. This sounds obvious, but the gap is real. World-class facilities track OEE in real-time, by machine, by shift, by product. They can tell you what's happening right now, not what happened last month.

Average facilities? Many calculate OEE monthly, if at all. Some still rely on operator logs that nobody trusts. By the time they see a problem in the data, it's been bleeding capacity for weeks.

They fix the right things. OEE breaks down into three buckets: availability (is the machine running?), performance (is it running at full speed?), and quality (is it making good parts?). Most facilities have one bucket that's particularly leaky.

The world-class operations figure out which bucket is their problem and attack it relentlessly before moving on. Average facilities try to fix everything at once, spread resources thin, and improve nothing much.

They prevent more than they repair. This one is almost cliche at this point, but the data backs it up. According to the Siemens True Cost of Downtime 2024 report, facilities using serious predictive maintenance—not just sensors on a dashboard, but actually acting on the signals—see 30-45% less downtime than reactive operations.

I know a tire plant that went from 76% to 84% OEE over 18 months mostly by getting ahead of breakdowns. The investment paid for itself more than three times over in the first year.

They train like it matters. Japanese manufacturers average around 120 hours of training per employee per year. Most Western plants are under 40. World-class OEE requires operators who can spot problems early, do basic maintenance themselves, and make good decisions without waiting for supervision. That takes investment.

The Mistakes I See Repeated

There are a few failure patterns that show up constantly:

Optimizing the wrong machine. Your constraint is the only equipment where OEE improvement translates directly to throughput. Improving OEE on a non-bottleneck just builds inventory faster. I've seen teams spend months improving a machine that wasn't limiting anything.

Gaming the numbers. OEE is depressingly easy to manipulate. Exclude certain downtime categories. Redefine what counts as a defect. Run only your easiest products when you know someone's watching. The result looks great on a PowerPoint and means nothing operationally.

Buying technology before fixing the process. I can't tell you how many facilities I've seen install fancy monitoring systems on processes that were fundamentally broken. Now they have real-time visibility into their chaos. The sensors didn't help.

Ignoring the people. Technology doesn't improve OEE. People do. If operators don't believe in the program, don't trust the data, or don't feel ownership over the outcomes, they'll find ways to work around whatever systems you implement. I've watched million-dollar initiatives die because nobody bothered to get the floor bought in.

A Realistic Path Forward

If you're sitting at 70% and want to reach world-class, here's what the journey actually looks like. Be prepared for this to take years, not months.

First, get visibility. You cannot improve what you cannot see. Real-time OEE tracking on your constraint equipment is step one. Not monthly reports—shift-by-shift dashboards that people actually look at. Most facilities gain 3-5 points just from this. Awareness changes behavior.

Then, pick off the easy wins. Pareto your downtime reasons. Attack the top few. Implement quick changeover techniques. Fix the chronic minor stops that "everyone knows about" but nobody's addressed. Another 5-8 points is typical here.

Build real capability. This is where it gets harder. Predictive maintenance. Standardized procedures. Cross-training. Root cause analysis that actually gets to roots. Expect 4-6 more points over a year or two.

Then sustain. The last few points to world-class are about culture more than systems. Zero-loss mindset. Continuous improvement as a daily practice, not a project. This is where the 6% separate from everyone else.

My Honest Take

World-class OEE is achievable. About 6% of manufacturers prove that every day. But it's not about buying the right software or finding the right consultant. It's about doing fundamental things—measurement, prioritization, prevention, training, discipline—with a consistency that most organizations struggle to sustain.

The gap between average (68%) and world-class (85%) looks like 17 points on a chart. In practice, it represents a completely different approach to how you run operations. Different priorities. Different habits. Different expectations.

That plant manager I mentioned at the start? He decided 72% wasn't good enough. Last I heard, they were at 78% and still climbing. Not by doing anything revolutionary—just by getting serious about the basics and refusing to accept "pretty good" as the finish line.

That's the real question: Is your current OEE good enough? If not, are you willing to do the sustained work to change it?