
Hosted by Mark Graban · EN

What does it really mean when a problem “comes back” after a team thought it had been solved?In this episode, Mark Graban reads his article, “We Thought We Solved That: The Art and Judgment Calls in the PDSA Cycle.” He examines how a frustrated executive’s question—“Why did the problem come back?”—can either support learning or trigger blame, depending on how it is asked.Mark explores five possible explanations for a recurring problem: the wrong countermeasure, poor implementation, failure to sustain the change, changed conditions, or the possibility that the problem never actually went away. He also explains why Toyota’s language of “countermeasures” is more useful than declaring a problem permanently “solved,” and why process behavior charts can help distinguish real change from routine variation.The larger lesson: reaching the Study step of PDSA requires more than checking whether an action item was completed. Leaders must examine what actually happened, what the data show, and what the organization needs to adjust next.Read and share the original article at leanblog.org/audio471.

A construction worker named Sean Dow was working on the 22nd floor of Pfizer’s former Manhattan headquarters, a massive office building now being converted into residential apartments with additional floors being added.Dow noticed cracks in a concrete slab. When he investigated further, he found a structural column visibly bending. But when he raised the concern, the initial response from the general contractor was reportedly, “Don’t worry, you’ll work on the floor below.”Dow kept pushing. Workers were evacuated, and investigators later found floors sagging by as much as four inches.The story illustrates an important distinction in psychological safety. People remain silent not only because they fear punishment, but also because they believe speaking up will accomplish nothing.A genuine speak-up culture requires concerns to feel both safe to raise and worthwhile to raise. The real test is not whether leaders say they welcome concerns. It is what they do in the first minute after someone brings one forward.Do they dismiss the concern and offer reassurance without investigating? Or do they respond with curiosity and say, “Show me”?An organization that depends on unusually persistent workers to overcome a dismissive first response is not operating a strong safety culture. It is operating on luck.

A title change is not a culture change. In this episode, Mark Graban draws on his early experience at GM in the mid-1990s — where "foreman" became "team coordinator" overnight without anything else changing — to explore why renaming supervisors with Lean-sounding titles so often fails to deliver Lean results.Read the blog postThe discussion centers on Ford's Process Coach role: what it's supposed to be, what it often is in practice, and why the gap between those two things is a leadership system problem, not a training problem. Mark also looks at why Toyota's Group Leader model works where Ford's equivalent often doesn't — and why a senior UAW worker has rational, concrete reasons to turn down a promotion to Process Coach even if they're the most qualified person on the floor.If your organization has rebranded its supervisors without redesigning the conditions those supervisors work in, this episode is worth your time.

Blog post with linksThree years ago, The Mistakes That Make Us came out. Around the same time, Elisabeth Swan published Picture Yourself a Leader. Both books' third birthdays felt like a decent reason to get together and talk.On Thursday, May 7, at 1 PM ET, Elisabeth and I are co-hosting a live event on LinkedIn called “Still Learning: Mistakes and Leadership Lessons.” We will talk about what readers have shared with us, what has held up, what we might write differently now, and the leadership questions that keep coming back. Attendees will have a chance to win books and a few other things.

Most AI tools answer your question with a 500-word essay full of numbered steps. You nod, close the tab, and carry on doing what you were already doing. The Lean Hospitals Coach is built around the opposite instinct -- asking questions before giving answers, the way good coaching actually works.Check out the blog postIn this episode, Mark walks through how the tool works, why it runs on Claude instead of ChatGPT, and what makes coaching mode fundamentally different from the "here are 7 steps" approach that every other AI defaults to. He also covers the two knowledge sources (Book Search and Book Plus), the two response styles (Tell Me and Coach Me), and how the combinations create different experiences depending on what you need.Mark is opening 50 founding memberships at $49/year -- price locked for life -- and hosting a LinkedIn Live demo on Tuesday, March 10 at 11 AM ET where he'll take audience questions and run them through the coach on screen, unscripted. You can also try the full product free for 48 hours at leanhospitalsbook.com/start.

Read the blog postTL;DR: A sound check, live song requests, and a naming regret — what watching Brandi Carlile perform taught me about specific problem-solving, vulnerability, and continuous improvement.My wife and I got to see the amazing Brandi Carlile perform near Chicago on Friday night.She is a multi-Grammy award-winning singer, musician, and songwriter — though calling her a solo artist would be a mistake...

The blog postTL;DR: Deming and Toyota's Fujio Cho asked the same uncomfortable question: why do management systems destroy motivation in people who started out wanting to do good work? The answer points to practices leaders can actually change.Check part 1 of this series in episode 464,

Read the blog postTL;DR: In a 1993 speech, Toyota leader Fujio Cho said organizations can create their own Lean systems, but success depends on three principles: leaders going to the gemba, asking “why” to learn from problems, and respecting and motivating people — not copying Lean tools.

The blog postWhat if a book could become an interactive coach instead of a static reference?In this episode, Mark Graban shares a behind-the-scenes look at his experiment turning the award-winning book Lean Hospitals into an AI-powered chat assistant embedded directly on his website. What started as a Friday afternoon curiosity quickly evolved into a working WordPress plugin, a subscription model, and a new way to deliver improvement knowledge on demand.Mark walks through how non-developers can use AI tools to write functional software, what he learned comparing different AI coding assistants, and why the real breakthrough isn’t the technology — it’s the ability to access proven Lean thinking at the moment of need.He also explores the broader implications for leaders and organizations: Could AI assistants trained on your own standards and practices reinforce daily management, support problem solving at the gemba, and scale coaching without more training sessions?This episode is both a practical case study in rapid experimentation and a thoughtful discussion about the future of learning, leadership, and continuous improvement in the age of AI.Key themes include:Turning expertise into on-demand guidanceUsing AI to prototype software without coding experienceSubscription models for knowledge deliveryPoint-of-use support for leaders and frontline teamsWhy technology alone won’t create a Lean culture — but can reinforce the right behaviorsIf you care about scaling improvement capability, preserving organizational knowledge, or simply experimenting with new ways to learn, this episode offers a candid look at what works, what broke, and what might come next.

The blog postIn this episode, I explore the 1987 NUMMI Management Practices Executive Summary — a confidential General Motors report that documented why the joint venture between GM and Toyota was succeeding so dramatically.What’s striking is how clearly GM’s own study team understood the real drivers of NUMMI’s performance. It wasn’t tools. It wasn’t discipline. It wasn’t copying Toyota’s production techniques.It was leadership.The report describes a management system built on mutual trust and respect, problem-solving at the source, quality built into the process, and supervisors acting as coaches rather than enforcers. Nearly 40 years ago, GM documented that NUMMI’s success came from management philosophy — not Lean tools.And yet, insight proved easier than action.In this episode, I walk through the document’s key sections, including NUMMI’s basic principles and five major management strategies, and reflect on why translating those lessons into broader cultural change proved so difficult.If you’re interested in Lean leadership, psychological safety, or the origins of what we now call continuous improvement, this historical document offers powerful — and still relevant — lessons.