Most Lean transformations start with tools. Huddle boards. A3 documents. Kaizen events. Visual management. The tools are visible, copyable, and produce early wins that make leaders feel like the methodology is working. The pattern is so common that it has its own diagnostic name in Lean circles -- "tool-based Lean" -- and the implication is generally that the organization is missing something deeper.
What's missing is rarely a tool. Tools are easy to acquire. What's missing is the behaviors that the tools are supposed to support, and the underlying principles that the behaviors are supposed to reinforce. Without the behaviors and the principles, the tools become artifacts. The organization can have all the boards and all the structures and still produce the same results it produced before, because the methodology that's actually changing how work gets done was never installed.
Michael Lombard, Senior Director of Operational Excellence at Cornerstone Healthcare Group, walks through what it looks like to build a Lean management system from the behavior layer up rather than from the tool layer down. The session is unusually candid about what's working and what isn't at his organization. Cornerstone is less than a year into this approach. The framing is "here's what we've learned so far," not "here's what we've figured out." That candor is what makes the session useful for other organizations considering the same approach.
Michael Lombard is Senior Director of Operational Excellence at Cornerstone Healthcare Group, a system of 19 long-term acute care hospitals based in Dallas, Texas, with locations across six states. Cornerstone takes very sick patients from intensive care units and provides specialized care for 25 to 30 days. Michael has led Lean transformation efforts in manufacturing and healthcare, focusing on behavior-driven management systems. He has been a KaiNexus customer in two different organizations.
This session was hosted by Mark Graban, then VP of Improvement and Innovation Services at KaiNexus.
The session opens with a candid framing. The word "behavior" sounds soft when you've come up through Lean as a tools-and-systems practitioner. It can suggest phobias, neuroscience, organizational psychology -- territory that feels distant from the operational work most CI practitioners do. Michael acknowledges he had the same reaction when behaviors first came up in his work. The framing shifts when you see what behaviors actually look like in practice.
He shares an example from a hospital visit in Huntington, West Virginia. A clinical team had been running daily care coordination huddles for some time. The huddles produced action items and assigned ownership. But the team wasn't being scientific about their work. They weren't learning from each step they took. They couldn't articulate why their care plans weren't advancing as fast as they wanted.
A nurse leader working on this problem sketched something on a napkin. If we do X, we predict we'll get Y. If we get Z instead, we'll learn from that. The drawing was simple, but the behaviors it implied were concrete: be rigorous about the X (the action we said we'd take), be objectively curious about the Y (the result we expected), and be enthusiastic about the Z (the unexpected result that produces learning).
Those are behaviors. They can be observed. A coach watching a huddle can see whether the team is being rigorous about whether X actually happened or being curious about why Y didn't match the prediction. The behaviors are measurable in a way that abstract principles like "scientific thinking" aren't. The napkin sketch translated a lofty methodological idea into operational practice.
The hypothesis Cornerstone tested: if we can promote the right behaviors, the right principles will be upheld, and the right long-term results will follow. Most organizations skip the behavior layer and try to install tools that they hope will produce principles automatically. The shortcut doesn't work. Tools without behaviors produce activity without learning.
The session uses Toyota's andon cord to illustrate the gap. The tool is the cord itself -- a physical mechanism for stopping the line when a worker sees a problem or makes a mistake. Many organizations have copied the tool. Few have copied the behaviors and principles that make the tool work.
The behaviors that make the andon cord effective: when the cord is pulled, someone immediately comes to help. The supervisor or team leader asks "what's the problem?" and "what can I do?" rather than "why did you stop the line?" The team works the problem before resuming production. Root cause analysis follows. The same problem rarely recurs.
The principle underneath those behaviors: workers are trusted to know when something is wrong. The system is built around their judgment, not around enforcing compliance with a predetermined schedule.
Organizations that have copied the tool without the behaviors produce a different system. The cord exists. Workers know they can pull it. But when they pull it, they get reprimanded for stopping the line, or yelled at for making a mistake. The behaviors send the opposite message from what the tool was designed to support. The workers learn quickly not to pull the cord. The tool becomes ceremonial.
Mark adds: this pattern shows up across Lean methodology, not just with the andon cord. Organizations copy huddles, A3s, kaizen events, gemba walks. Many of them produce the appearance of Lean without the function. The behaviors that animate the tools have to be deliberately developed, observed, and reinforced. They don't come bundled with the tool.
Cornerstone organized their behavior work using a framework from the Shingo Institute. The model groups ideal behaviors into three categories:
Align behaviors help the organization figure out what the right things to work on are. Catchball, consensus-building, connecting work to True North.
Improve behaviors help the organization do the right things better than they're being done today. Scientific thinking, structured problem-solving, experimentation, PDCA discipline.
Enable behaviors develop people and bring more of them into the improvement work. Coaching, respecting capability, engaging frontline practitioners rather than relegating them to compliance with predetermined processes.
The three categories are interconnected. Alignment without improvement produces strategic clarity without operational change. Improvement without enabling produces results that depend on a few specialists rather than building the broader capability that sustains change. Enabling without alignment produces broadly engaged improvement work that isn't connected to organizational priorities.
The Shingo Institute's resources at shingo.org are referenced throughout the session as the source for this framing. The model is one of the more practical frameworks available for distinguishing principles from behaviors from tools without collapsing them into each other.
Cornerstone's alignment challenge was familiar. The organization had more than 30 strategic objectives at the start of the year. Everyone was busy. Few of the objectives were actually getting completed.
The methodology Michael wanted to use to reduce the list was catchball -- the iterative back-and-forth between organizational levels that produces consensus on priorities. The problem: the team hadn't practiced catchball enough to be good at it. They couldn't reduce the list to a manageable size through the catchball process itself.
The workaround was a special event. The senior leadership team gathered for a structured consensus-building session. The session got the list down to about six strategic imperatives -- still more than ideal, but vastly more manageable than 30. Michael acknowledges directly that this wasn't catchball working as designed. It was a temporary countermeasure while the team built the capability to do catchball well.
The acknowledgment matters. Methodologies don't install themselves. The team had to use a less-mature workaround for a real problem while they developed the capability to use the more-mature methodology. That kind of pragmatic adaptation is what distinguishes organizations that build Lean management systems from organizations that copy them.
The improvement category surfaced a different problem. Michael shows a slide with a phrase that includes a duplicated word -- the kind of trick exercise that asks readers to count occurrences of a specific letter. Most people miss the duplication entirely on the first pass.
The point: human cognition is built around jumping to conclusions. The reflex is so deep that even practitioners trained in structured problem-solving fall back to it under cognitive load. The team at Cornerstone had this reflex, like every other organization Michael had worked with.
The Lean countermeasures to the jumping-to-conclusions reflex are familiar: go to gemba, gather data over time, run experiments, test hypotheses rather than asserting answers. These are behaviors that interrupt the default cognitive pattern. The team needed to develop them deliberately, because the default reflex would otherwise prevent the methodology from producing the results it could produce.
Mark adds a related observation: there's a difference between the mindset that says "we know the answers" and the mindset that says "we will figure things out and test them over time." The difference applies at every scale, from solving a problem about a recurring infusion pump failure to designing a multi-year Lean transformation plan. Even the transformation plan should be approached as a hypothesis to test through PDSA cycles rather than as a static document to execute.
Michael's preferred countermeasure to the jumping-to-conclusions reflex is Toyota Kata -- specifically the improvement kata. The methodology gives practitioners a structured routine for retraining their brain to be iterative, accept uncertainty, and ground improvement work in experimentation. The kata is itself a behavior, practiced repeatedly until it becomes the default response to a problem.
The third category surfaced the constraint Michael describes as the most vexing problem Cornerstone has faced.
The premise of enabling is straightforward: if the organization truly respects everyone's ability to be problem-solvers, the improvement work should bring everyone in rather than restricting it to specialists. The practical implementation is harder.
The limiting factor at Cornerstone wasn't training capacity. It was coaching capacity. The coaching chain that's supposed to develop practitioners through repeated structured practice doesn't work if there aren't enough coaches operating at enough levels of the organization. Building coaching capability is itself a slow process, because coaching is a behavior that develops through repetition, and the repetition has to happen at every level of the chain.
The coaching kata is Cornerstone's primary tool for building this capability. The kata gives coaches a structured routine -- the same set of questions, the same daily cadence, the same expectation of brief sessions rather than long meetings. The repetition develops both the coach's technique and the learner's improvement capability over time. Michael notes that coaching skill develops over hundreds of cycles, not over a single training class. He's done roughly 1,000 coaching cycles over three years and finds something to improve in nearly every one.
Cornerstone identified a partial list of behaviors they wanted to develop -- catchball, gemba walks, huddles, scientific thinking, root cause analysis, coaching, andon-style problem surfacing, run chart interpretation, and several others. Each behavior had associated principles and tools.
The problem became clear quickly: the behaviors weren't becoming habits. Michael describes the situation as building on quicksand. The team would adopt a behavior, practice it briefly, and then drift back to default patterns within weeks. The behaviors weren't sticking because the underlying foundation wasn't yet in place.
The diagnostic moment came when Cornerstone realized they needed a system of ideal behaviors -- not just a list. A platform for practicing the behaviors repeatedly enough that they would actually become habitual. The system needed to support the repetition rather than depending on individual willpower or one-off training.
The framing Michael uses for what they're building: the True North Management System. The structure is intentionally underdefined. Cornerstone is less than a year in and hasn't figured out all the inputs, processes, and outputs. The system is constantly evolving as the team tests different tools and observes which ones actually support the behaviors they're trying to develop. There's no point at which the management system gets locked down. The evolution is part of the design.
The principle Michael returns to: don't focus on the management system's specific structure. Focus on the ideal behaviors that drive the business principles that produce long-term results. The system supports the behaviors. The system is not the point.
The platform's role in this approach is specific. It supports the behaviors rather than replacing them. The session walks through how Cornerstone uses specific KaiNexus features to support behaviors in each Shingo category.
Aligning behaviors. KaiNexus uses attributes that practitioners tag to their projects -- priority level, challenge area, True North metric. The act of configuring these attributes forced Cornerstone to clarify what their True North metrics actually were. Reducing the list from 30 to a half-dozen wasn't just a planning exercise. The platform's dropdown menu structure required a short list. Michael calls this "almost mistake-proofing your strategic planning."
When practitioners submit a project, they can tag it to a True North metric. Tagging is optional, but the act of providing the option nudges the organization toward alignment. Roughly 75% of projects at Cornerstone align with one of the True North priorities. That alignment didn't come from a business rule requiring it. It came from the platform creating a structural opportunity for self-alignment.
Huddle boards. The platform supports huddle boards with a standardized lane structure. Each lane includes two run charts (a primary metric and a driver metric) plus a project list. Cornerstone created leader standard work specifying how to use the structure. Empty lanes create what Michael calls a visual pull -- a space that begs to be filled. The act of filling the lanes through catchball produces the alignment conversation the methodology requires. The behavior of catchball improved noticeably once the empty lanes existed as a structured prompt.
For an organization geographically distributed across six states, virtual huddle boards solve a problem that physical boards can't. Corporate leaders who can't be on-site stay connected to frontline work. The platform extends visibility across the distance rather than replacing the daily ritual entirely.
Improvement kata templates. The platform includes a Kata project template structured around the four routines of the improvement kata: understand the direction, grasp the current condition, establish a target condition, and run experiments toward the target. Each routine has its own field. The template doesn't guarantee that practitioners use the kata correctly -- challenge statements still need coaching, current condition analysis still needs depth -- but the structure begins to hardwire the kata pattern into how improvement work gets framed.
Coaches can see when projects come in with weak challenge statements or thin current condition analysis. The visibility creates coaching opportunities. The platform supports the coaching cycle by surfacing where coaching is most needed.
Run charts. Cornerstone moved from balance-scorecard-style tables to visual run charts that show actual results against budget over time. The shift produced behavior change immediately. Tables with red/green indicators can hide unfavorable trends behind several months of green; a run chart makes the trajectory visible. Mark adds a friendly note that he'd want to apply statistical process control discipline to know whether a trend is real signal or just variation -- a topic covered in depth in his companion session on improvement metrics.
The intermediate target line Michael describes -- a red line set as a stepping stone between current performance and ultimate target -- is a behavior support, not a metric. Big challenges feel insurmountable. Breaking them into bite-sized intermediate targets gives the team courage to try things. The behavior the run chart supports is the iterative experimentation that produces breakthrough improvement over time rather than the demoralized resignation that comes from chasing distant targets.
Coaching chains through teams and follower functions. The platform supports formal coach-learner designations through sponsor and facilitator roles. The sponsor practices the coaching kata. The facilitator practices the improvement kata. The formality matters because coaching only develops through repeated structured practice, and the practice requires explicit role assignment rather than informal mentoring.
The follower function lets subject matter experts see what's happening across multiple facilities. A corporate-level expert can follow projects in all 19 Cornerstone hospitals and help cross-pollinate ideas. Second coaches -- people coaching the coaches -- can follow how well coaching cycles are operating across the organization. The visibility supports the entire coaching chain rather than depending on relationships that happen to exist informally.
Michael closes with one of the most honest statements in any of the recent webinars in this catalog. Practice does not make perfect.
He has done approximately 1,000 coaching cycles over three years. In roughly 999 of them, he can identify something he did wrong as a coach. Every cycle produces another entry in his "what to improve next time" column. The mastery he's working toward isn't perfection. It's habit. The behavior of coaching has become permanent enough that it happens automatically. The execution still has gaps. The gaps are the next round of learning.
The reframe matters for practitioners considering this approach. Building a behavior-driven Lean management system isn't a project with a completion date. It's a discipline that runs continuously. The right measure isn't whether the system is perfect. It's whether the behaviors are happening reliably enough to produce learning, and whether the learning is informing the next round of practice. Permanence over perfection.
The companion principle: Cornerstone continues to learn. Michael's slide title isn't "Cornerstone has figured this out." It's "Cornerstone continues to learn." The framing is itself a behavior worth modeling. Organizations that present their Lean journey as a completed transformation rather than an ongoing discipline produce conference presentations that don't match what's actually happening on the ground. Organizations willing to be honest about what they're still figuring out produce learning conversations that compound across the broader CI community.
Why focus on behaviors rather than tools?
Tools produce activity. Behaviors produce results. Most organizations have copied Lean tools without developing the underlying behaviors and principles that make the tools effective. The pattern is so common that it has its own name -- "tool-based Lean" -- and the implication is that the organization is missing something deeper. What's missing is rarely a tool. It's the behaviors the tool was designed to support. Building those behaviors requires deliberate practice over time, supported by a management system that creates the conditions for the practice to happen repeatedly.
What's an ideal behavior?
An observable, measurable practice that supports a Lean principle and produces the results the organization is trying to achieve. The session shares an example from a clinical huddle: be rigorous about the X (the action the team committed to), be curious about the Y (the result expected), and be enthusiastic about the Z (the unexpected result that produces learning). Those are behaviors. A coach watching the huddle can see whether the team is practicing them. The behaviors translate a lofty principle ("scientific thinking") into something operational.
What's the relationship between behaviors, principles, and tools?
Principles are universal patterns that have been proven across many contexts -- like scientific thinking, respect for people, or building quality at the source. Behaviors are observable practices that support those principles in a specific organization -- like measuring twice before cutting, running experiments before asserting solutions, or going to gemba before drawing conclusions. Tools are the artifacts that make the behaviors easier to perform -- like A3 templates, run charts, kata coaching cards, or huddle boards. The right starting point is results: what results do we need? What principles produce those results? What behaviors uphold those principles? What tools support those behaviors? Skipping from results directly to tools produces fool's gold.
What is the Shingo model and where can I learn more?
A framework developed by the Shingo Institute that organizes Lean methodology into principles, behaviors, and results. The model groups behaviors into three categories: align, improve, and enable. The Institute offers training, certification, and the Shingo Prize, which recognizes organizations that have demonstrated genuine cultural change rather than tool-based superficial improvement. More information at shingo.org. The session references Jake Reimer specifically as a Shingo instructor whose teaching has been particularly useful for Cornerstone.
What is Toyota Kata and how does it fit into a behavior-based system?
A structured practice routine developed by researcher Mike Rother from his study of Toyota's management practices. The improvement kata is the four-step pattern practitioners follow: understand the challenge, grasp the current condition, establish a target condition, and experiment toward the target. The coaching kata is the structured set of questions a coach asks daily at the learner's storyboard. Both are behaviors -- not techniques to apply occasionally, but patterns practiced repeatedly until they become habitual. The kata methodology gives Cornerstone its primary mechanism for developing scientific thinking as a default response to problems rather than as a special discipline.
What did Cornerstone learn about catchball?
That it requires practice to do well, and that organizations rarely have the catchball capability when they need it most. Cornerstone tried to use catchball to reduce 30 strategic objectives to a manageable list. The team wasn't yet skilled enough at the methodology to produce the consensus. The workaround was a special event with structured consensus-building. The event got the list down to a manageable size, but it wasn't catchball working as designed. It was a temporary countermeasure while the team built the underlying capability. That kind of pragmatic adaptation -- using a less-mature workaround for a real problem while developing the more-mature methodology -- distinguishes organizations that actually build Lean management systems from organizations that copy them.
Why is coaching such a difficult bottleneck?
Coaching capability develops through hundreds of cycles of structured practice. The repetition has to happen at every level of the organization, because the coaching chain depends on coaches being available at every layer. Building coaching capacity at scale takes years, even in organizations committed to the methodology. Michael describes coaching as the most vexing constraint Cornerstone has faced. The platform supports the coaching chain through formal sponsor and facilitator role designations, follower functions that let second coaches monitor cycles across the organization, and visibility into coaching cycle frequency. The platform doesn't substitute for the coaching itself, but it makes the coaching infrastructure visible and supportable.
How does an organization start if they're not yet practicing any of this?
Start small with one behavior. Michael emphasizes scientific thinking as a "keystone behavior" -- the one that, once established, makes most of the other behaviors easier to develop. The structured practice routines that develop scientific thinking are the improvement kata and the coaching kata. Practitioners pair up, with one in the learner role practicing the improvement kata and one in the coach role practicing the coaching kata. They meet daily for brief cycles. Over hundreds of cycles, the scientific thinking pattern becomes default rather than special. From there, other behaviors -- catchball, gemba walks, root cause analysis, andon-style problem surfacing -- can develop on top of the scientific thinking foundation. Trying to develop all the behaviors simultaneously usually fails because the underlying foundation isn't yet in place.
Should leaders coach each other before rolling out kata to the rest of the organization?
Generally yes, though Michael notes there are multiple workable approaches. Senior leaders can practice kata themselves as both learners and coaches, which builds their personal capability while signaling the organizational commitment. An alternative is to create an advanced team that includes a mix of senior, mid-level, and frontline leaders, run a focused project for six months, and let kata practice cross departmental boundaries. Each approach has trade-offs. The deeper principle: don't focus on how good the leadership team's kata practice is initially. Focus on the number of repetitions they can get. Quality develops through volume of practice rather than through training intensity.
Why is making decisions for the long term so hard?
Because the short-term financial signals are so much clearer than the long-term ones. A leader can see exactly how much money the organization would save this afternoon by sending nurses home early. The same leader can't see exactly how much money the organization gives up over years by demoralizing the staff and undercutting their capacity to participate in improvement work. The Toyota Way's first principle is making decisions based on the long-term perspective even at the expense of the short term. Most organizations adopting Lean don't focus on this principle as deliberately as they focus on the tools. Building the discipline of long-term decision-making is one of the deeper struggles in Lean management. The coaching conversation Michael describes -- helping leaders see how short-term decisions cost the organization over longer horizons -- is one of the places where Lean coaching can do the most useful work.
What does "long-term results, not results this month" mean operationally?
Michael describes this as the hardest translation challenge for Lean coaches. Short-term metrics are easy to track and obvious in their implications. Long-term metrics are diffuse and contestable. The coaching work is helping leaders see how today's behavior produces tomorrow's results, even when the connection isn't immediately visible. The example Mark shares: sending nurses home early to save labor costs today produces lower idea submission tomorrow, less quality improvement work next week, and lower retention over the year. The leader can see the labor savings on the daily report. The downstream costs don't show up there. The Lean management system tries to make those downstream effects visible enough that leaders can choose differently.
.svg)
Copyright © 2026
Privacy Policy