UPSIDE DOWN EXCELLENCE

Quality that's simple

We close our Paul O’Neill series this week. This was a five-part look at how one CEO transformed a company by doing the fundamentals exceptionally well. Our final theme: persistence.

Let’s recap the framework we've built:

1. Start with common ground. O’Neill chose safety because everyone could agree on it.
2. Measure it. Data exposes progress and problems.
3. Make it visible. Metrics seen become metrics understood.
4. Align incentives. Celebrate the behaviors you want more of.

Now the fifth insight: stick with it.

Quality systems aren’t built in a day. Neither is trust. Neither is culture. Improvement takes time and steady reinforcement.

A story of slow, steady change.

One client of mine kept running into first-time production mistakes. Not because people were careless, but because their process was siloed. Each department worked independently and tossed work to the next group with no verification step.

The fix was clear: they needed an APQP process.

But the lead engineer resisted. He believed his system was flawless and the problems were someone else’s fault. We've all met this engineer before.

Here’s how we approached it:

• Common ground: Everyone agreed we didn’t want to send bad parts to customers.
• Measure it: Weekly and monthly counts of part launches and issues.
• Make it visible: Every management review included the metric.
• Align incentives: Good numbers earned praise. The team wanted to succeed and to be recognized.

And then we added the final ingredient: persistence.

We met every week for two months. We solved small problems. We clarified missing information. We tracked where tools or data weren’t available. Slowly, resistance softened but not immediately.

The breakthrough came when we won a job with complex requirements the engineer didn’t fully grasp. During the feasibility review, the team used the APQP process to answer his questions. For the first time, he saw its value.

Not because the process was brilliant.

Because we stuck with it long enough for the value to become undeniable.

Framework application:​
When facing resistance, ask three questions:

1. Have we agreed on common ground?
2. Are we reviewing data consistently?
3. Are we reinforcing the right behaviors?

Then keep going. Most transformations fail not because the idea was bad, but because people quit too soon.

Stick with it.

You’ve got this.