W. Edwards Deming spent decades studying why organizations produce bad results. His conclusion, which has been validated over and over across industries and decades, was this: roughly 94% of problems are caused by the system, not the individual.
Ninety-four percent.
That's a number worth sitting with for a moment if you've ever looked at an employee's work and thought "what is wrong with this person?"
The default assumption is backwards
When something goes wrong in a business, the natural instinct is to look at who. Who dropped the ball? Who missed the deadline? Who gave the client wrong information?
It feels logical. Someone messed up, so the problem must be the someone.
But ask the next question. Why did they mess up?
Usually the answer isn't "because they're incompetent" or "because they don't care." Usually it's something like: they didn't have the information they needed. Or the process was confusing. Or they were handling three other things at the same time because nobody defined priorities. Or they were trained by watching someone else do it once, six months ago, and they've been winging it ever since.
Those are system problems wearing people-problem costumes.
What system problems look like from the outside
A team member keeps making errors on client deliverables. From the outside, this looks like carelessness or lack of skill. From the inside, it might be that the quality check step was removed "temporarily" during a busy period and never came back. Or that the template they're working from is outdated. Or that they're producing work based on information from a intake form that doesn't ask the right questions.
An employee seems disengaged. Shows up, does the minimum, goes home. Looks like an attitude problem. Could be that they've been given responsibility with zero authority. They can't make decisions. They can't improve their own workflow. They've learned that initiative gets overridden by the boss anyway, so why bother?
Someone's always behind on their work. Seems slow or overwhelmed. Might be that they're the convergence point for three different workflows and nobody realizes how much is landing on them because the workload was never mapped. They're not slow. They're buried.
How to tell the difference
There's a pretty reliable test for whether you're dealing with a people problem or a system problem.
Ask yourself: if I put a different person in this exact same role, with the same tools, same information, same process, same support, would they produce a better result?
If the honest answer is "probably not," it's a system problem.
If you put a sharp, motivated person into a broken system, the broken system wins. Every time. They might fight it for a while. They might find workarounds. But eventually, the system shapes the behavior.
This is true in reverse too. Put an average performer into a well-designed system with clear expectations, good information flow, and appropriate support, and they'll often surprise you.
The cost of treating system problems as people problems
When you assume performance issues are personal, you go down a predictable path. Coaching conversations. Performance improvement plans. Eventually, maybe, letting someone go.
Then you hire a replacement. The new person starts strong because everything is new and they're motivated. After a few months, the same patterns emerge. Because the system hasn't changed.
I've watched businesses cycle through three or four people in the same role before someone finally asks whether the role itself is the problem. That's years of hiring, training, frustration, and turnover. All because the default assumption was "wrong person" instead of "wrong system."
And the human cost is real too. People internalize it. They start believing they're the problem. Good people leave your business thinking they weren't good enough, when really they just couldn't succeed in a system that was set up poorly.
What to do about it
Start by looking at the systems around your team, not at the team itself.
When something goes wrong, resist the urge to figure out who's at fault. Instead, trace the problem back through the process. Where did the breakdown start? What information was missing? What step was unclear? Where did the handoff fail?
Build this into how you debrief. Not "who messed up?" but "what happened, and what about our system allowed it to happen?" The language matters because it determines whether people get defensive or get curious.
Then fix the system. Make the information available. Clarify the process. Design the handoff. Create the checklist. Whatever the system gap is, close it.
You'll find that many of your "people problems" quietly resolve themselves. Not because the people changed, but because the system stopped setting them up to fail.
This isn't about letting people off the hook
I want to be clear: some performance problems genuinely are about the person. Sometimes someone isn't right for the role. Sometimes there's a skills gap that training can't close. Sometimes the fit just isn't there.
But you can't accurately identify those cases until you've fixed the system. When the system is broken, everything looks like a people problem. It's only when the system is working well that you can fairly evaluate individual performance.
Think of it this way: if your team is playing a game where the rules keep changing and nobody has the full rulebook, you can't judge anyone's performance in that game. Fix the rules first. Then see who can play.
If you're seeing patterns of underperformance across your team, or if you keep losing good people, it might be worth looking at the structural side of things. The post about the bottleneck problem covers one of the most common system failures, and our Culture Optimization service is designed to help businesses build systems that bring out the best in their people.
If any of this resonates and you want to talk through what you're seeing, reach out for a conversation. No diagnosis without context.