Why Your Talent Problem Isn't a Talent Problem

W. Edwards Deming said 94% of problems in organizations are system driven. Only 6% are people problems. We all nod when we hear that. We love the quote. We put it in our slide decks. And then we go right back to building performance improvement plans.

The Work Institute found that 75% of voluntary turnover is preventable. Three out of four departures did not have to happen. And yet we're spending our energy on the 6% while ignoring the 94%.

Imagine you're a surgeon and your patients keep dying on the table. You blame the patient. You get a new patient and they die too. At some point, do you start to wonder if maybe the problem isn't the patient, maybe it's the operating room?

What You'll Learn

The faulty assumptions that get leaders stuck:

• "We need to hold people accountable for results" assumes performers control the variables that determine success. Research shows 70% of the variance in team engagement relates to managers.
• "A good performer can succeed anywhere" assumes talent is portable. A role that burns through three leaders in 18 months is a role design problem, not a talent problem.
• "PIPs help underperformers improve" assumes they're developmental tools. In reality, 67% of employees say performance evaluations are based on subjective observations, not clear metrics.
• "High turnover means we need to hire better" ignores that 71% of voluntary turnover stems from poor management, not bad hires.
• "Fixing individuals is faster than fixing systems" ignores the math. A PIP plus recruiting plus onboarding takes 15 months. Redesigning a broken system takes 6 weeks.

The four questions to ask before any PIP or exit:

1. The pattern question: Is this the first time the role has failed to deliver?
2. The load question: Was this role designed for a human or for a superhero?
3. The attribution question: Are we measuring people or the systems they're in?
4. The capital question: Where is your time and money actually going?

The plays for next week:

• Run a failure audit on your last three exits
• Build a system load assessment for critical roles
• Change performance conversations to start with what the person was asked to do and what they had to do it with
• Run stay interviews before exit interviews

Key Quotes

"A bad system will beat a good person every time. If your system makes failure likely, you will keep finding people to blame until you run out of people."

"We're not saying individual accountability doesn't matter. We're saying most of what we call individual failure is actually a system failure wearing a name tag."

"Fixing people is a low altitude mandate. Fixing systems is a high altitude mandate."

"If replacing your top performers with average performers would break the system, you

Transcript

W. Edwards Deming, the guy who rebuilt Japanese manufacturing, says something that should haunt every CHRO. Ninety-four percent of problems in organizations are system driven. Only six percent are actually people problems.

The Work Institute’s 2024 retention report found that 75 percent of voluntary turnover is preventable. Three out of four departures did not have to happen. And yet, organizations are spending their energy on the six percent while ignoring the ninety-four.

Hey there, senior leader, and welcome to the Talent Sherpa podcast. This is where senior leaders come to rethink how human capital really works. I’m your host, Jackson Lynch, and today I’m joined by my co-host, my friend, a man who enjoys leg day and the founder of Propulsion AI, Scott Morris.

I want you to imagine for a moment that you’re a surgeon and your patients keep dying on the table. You blame the patient. You get a new patient and they die too, and you blame that patient. Then the fourth one dies.

At some point, do you start to wonder if maybe the problem isn’t the patient, maybe it’s the operating room?

That’s what most CHROs are doing with talent. They spend six months performance managing somebody out. They recruit a replacement. They watch them hit the exact same wall.

I’ve done this more than once. And it took me an embarrassingly long time to ask if the problem was ever the person.

Before we get going, if you like this podcast, please subscribe, leave a comment, or drop us a review on your preferred platform. It’s how we grow the pod and keep the content sharp for other senior leaders.

Every CHRO has heard this before. When someone underperforms, you coach them. You document it. You put them on a plan. And if that plan fails, you exit them. That’s the playbook. It’s in every HR textbook ever written.

And it sounds reasonable. You’re holding people accountable. You’re being fair. You’re giving them a chance to improve.

Except here’s what’s actually happening in the data. Gallup found that 42 percent of employees who voluntarily left their organization said their manager or the company could have done something to prevent them from leaving.

Nearly half.

And 45 percent said nobody talked to them about their job satisfaction in the three months before they quit.

Think about that. We have all this infrastructure around performance management, and nearly half the people who leave say we could have kept them if we had bothered to ask the right questions.

W. Edwards Deming said ninety-four percent of problems in organizations are system driven. Only six percent are people problems.

We all nod when we hear that. We love the quote. We put it in our slide decks. And then we go right back to building performance improvement plans.

The Work Institute found that 75 percent of voluntary turnover is preventable. Three out of four departures did not have to happen. And yet we’re spending our energy on the six percent while ignoring the ninety-four.

We’re not saying individual accountability doesn’t matter. We’re saying most of what we call individual failure is actually a system failure wearing a name tag.

I once had a CEO ask me why we kept churning through sales leaders. Third one in eighteen months. My first instinct was better interview questions or better job design.

He looked at me and said, or maybe we need to look at what we’re asking them to do.

He was right. And I felt like a real idiot.

That shift in thinking is what this episode is about. Not whether people matter, but where we aim our energy when things go wrong.

Let’s call out a few faulty assumptions that get leaders stuck.

The first one is: we need to hold people accountable for results. The assumption underneath that is that individual performers control the variables that determine their success.

The idea is good. The assumption is bad.

Research shows about 70 percent of the variance in team engagement relates to managers themselves. When you trace performance problems back to their source, you find structures, resource allocation issues, unclear expectations, and incentive misalignment.

The person in the role is often operating rationally inside a system that makes failure likely. Holding them accountable for system design choices they didn’t make is theater. It’s not management.

The second assumption is that a good performer can succeed anywhere. The assumption underneath that is that talent is portable and independent of context.

I believed this for way too many years. But it’s wrong.

A role that burns through people, three leaders in eighteen months, is not a talent problem. It’s a role design problem. It’s a clarity problem. It’s a system problem.

SHRM has said replacing a leader costs about 200 percent of their salary. Technical roles can be around 80 percent. If you keep paying that cost for the same role, you’ve built a job that only works if you hire superheroes.

And superheroes don’t scale. They don’t transfer. They don’t survive turnover.

This becomes a planning error, not a talent strategy.

The third assumption is that PIPs help underperformers improve. The assumption is that performance improvement plans are developmental tools.

In reality, most organizations use PIPs as documentation for termination. Employees know it.

Lattice research shows 67 percent of employees say performance evaluations are based on subjective observations, not clear metrics. When the plan is vague and the outcome is predetermined, you spend six months on paperwork while the system problem persists.

The replacement will hit the same wall.

Another assumption is that high turnover means we need to hire better. The assumption is that recruiting is the lever for retention.

Seventy-one percent of voluntary turnover stems from poor management. It’s not bad hires. It’s bad systems producing bad outcomes.

You can upgrade your ATS. You can write better job descriptions. None of that fixes how the work actually gets done.

The last assumption is that fixing individuals is faster than fixing systems.

It feels attractive. It feels actionable.

But the math doesn’t work.

A PIP takes three to six months. Recruiting a replacement takes another ninety days. Onboarding to full productivity takes another six months. That’s roughly fifteen months of cost for one role.

Redesigning a broken system can take six weeks and a fraction of the money.

We default to fixing people because it feels like action. But activity is not progress if you’re solving the wrong problem.

A bad system will beat a good person every time. If your system makes failure likely, you will keep finding people to blame until you run out of people.

Before you put anyone on a plan or start recruiting, ask four questions.

First, the pattern question. Is this the first time the role has failed to deliver?

If multiple people fail in the same role, you’ve found a system bottleneck. Keep a failure log, not just a people log. Document role context, resources, dependencies, and reporting relationships.

Second, the load question. Was this role designed for a human or for a superhero?

Most frameworks measure output, not system demand. How many direct reports? How many projects? How many approval gates? How many cross-functional dependencies?

If replacing your top performers with average performers would break the system, you’re relying on heroics.

Third, the attribution question. Are we measuring people or the systems they’re in?

When someone misses quota, how much of that was actually in their control? Territory quality, product fit, handoffs, and external events often matter more than individual effort.

Fourth, the capital question. Where is your time and money actually going?

If you believe the ninety-four six ratio, your investment should reflect it. If you’re spending ninety-four percent of your energy on people problems, you’ve inverted the ratio.

Fixing people is a low-altitude mandate. Fixing systems is a high-altitude mandate.

If you want to act next week, start here.

Run a failure audit on your last three exits. Look for patterns.

Build a system load assessment for your most critical roles.

Change performance conversations. Start with what the person is being asked to do and what they have to do it with.

Track every PIP, termination, and long time-to-fill role. Turn it into a capital allocation conversation.

Pick one broken system and fix it using an agile approach this quarter.

And run stay interviews before exit interviews, especially with people you think are underperforming.

You can hear the problems while people are still on payroll, or you can hear them on the way out. That’s a choice.

Here’s the takeaway.

When people fail repeatedly in the same role, you don’t have a talent problem. You have a system problem.

Individual performance is mostly an output of system design.

Fixing systems is faster and cheaper than cycling through people.

And the CHRO who fixes systems earns a seat at the capital allocation table.

The ninety-four six rule isn’t manufacturing theory. It’s a talent strategy.

Stop fixing the six percent while ignoring the ninety-four.

And keep climbing.