The standard question in continuous improvement is "how do we get frontline employees to participate?"
There's a second question that often comes from internal CI practitioners and process improvement teams: "how do we get executives to sponsor and support bottom-up improvement?"
Mark Graban's argument in this session is that both questions, framed that way, lead into a trap. The premise embedded in each one is that someone -- the frontline employee, the executive -- needs to be convinced or persuaded or directed to do something they aren't doing. And the natural response to that framing, for any of us trying to drive change, is to tell people more clearly, explain the rationale more rigorously, present more data, or push harder.
It doesn't work. Not for the executives we're trying to influence. Not for the frontline staff we want to engage. Not even for ourselves, when we try to change our own habits.
The session draws an unexpected connection between continuous improvement and a counseling methodology called motivational interviewing -- a technique developed in addiction counseling that has implications for how anyone trying to drive organizational change might operate. The argument that emerges is sharp: factually correct doesn't lead to easy acceptance. Telling people what to do triggers resistance even when they agree the change would benefit them. The work of building an improvement culture isn't the work of pushing change. It's the work of creating the conditions where people choose to change.
Mark Graban is Senior Advisor at KaiNexus and the author of Lean Hospitals, co-author of Healthcare Kaizen, and author of The Mistakes That Make Us.
The session opens with a story Mark uses to illustrate the limits of top-down improvement.
In a hospital where he was working on the early stages of a Lean engagement, a nurse manager was standing with him in a hallway when she got distracted. Workers were rolling out new carpet along the corridor. She hadn't been told. She certainly hadn't been consulted. The decision had been made at the executive level as a way to reduce noise and improve patient satisfaction scores.
Both of those are reasonable goals. Patient satisfaction is a metric worth caring about, and noise reduction is a real driver of it. But the executive who ordered the carpet had also separately mandated that nurses would use computers-on-wheels throughout the unit, which became harder to maneuver on carpeted floors. Two top-down mandates collided.
The patient satisfaction data showed something interesting. Scores did improve in that unit -- substantially. But they had started improving a month before the carpet was installed. The actual improvement came from many small bottom-up changes the nursing staff had been making: adjusting how they handled night-shift conversations, modifying alarm settings, changing how supplies were retrieved. Lots of small changes, none of them dramatic individually, accumulated into a meaningful shift in patient experience.
The carpet was top-down problem-solving. The patient satisfaction improvement was bottom-up. The story isn't an indictment of executives -- executives need to set direction and define strategic priorities, and that's appropriately top-down work. The story is about a specific kind of failure: the assumption that knowing the goal means knowing the solution. The people closest to the problem usually have a better view of the solution space than the people farthest from it.
The data on bottom-up improvement at scale is substantial.
Alan Robinson and Dean Schroeder, in The Idea-Driven Organization, document a pattern they've found consistent across services, manufacturing, healthcare, and government. At Coca-Cola Stockholm, the impact distribution looked like this: a small number of Black Belt projects averaged 750,000 SEK each. A larger number of Green Belt projects averaged 200,000 SEK each. And 1,720 frontline ideas averaged 4,600 SEK each. The collective impact of the frontline ideas exceeded the combined impact of the higher-profile project work.
John Toussaint, former CEO of ThedaCare and one of the most-studied healthcare Lean leaders in the United States, estimated that 80% of ThedaCare's overall performance gains came from frontline ideas. The other 20% came from major improvement projects. ThedaCare's early years were heavily focused on the projects. The realization that most of the impact was coming from elsewhere shifted the organization's emphasis.
KaiNexus customer data tracks the same pattern. About 2.5% of all submitted improvements have a financial impact greater than $10,000. About 1.4% have an impact greater than $100,000. The distribution is long-tailed -- most improvements are small, but the cumulative value is substantial, and the occasional large-impact improvement only surfaces when the volume is high enough to find it.
One concrete example from the session: an academic medical center hematology/oncology clinic noticed that patients were filling prescriptions outside the hospital pharmacy. The improvement was small -- a sentence or two in the patient encounter explaining that the in-house pharmacy was available and could save the patient a trip. No technology change. No major workflow redesign. In the first three months, 167 patients switched their prescriptions to the in-house pharmacy, generating $638,000 in revenue that had previously been walking out the door. Extrapolated to a year, that's roughly $2.5 million in revenue capture from a change that cost essentially nothing to implement.
These are the kinds of improvements that don't get found by Black Belt projects. They get found by people who are already doing the work, who see the pattern, and who have a way to act on what they see.
Here's where the session takes its sharpest turn.
The case for bottom-up improvement is rationally compelling. The data is consistent across industries. The impact distribution is well-documented. The leadership benefit -- developing problem-solvers throughout the organization rather than concentrating problem-solving in a small set of experts -- is observable in any organization that's traveled the journey.
And yet. Organizations that have access to all this evidence routinely choose not to act on it. Executives who have been shown the data continue to operate in command-and-control patterns. Frontline staff who have been invited to participate decline to engage. The factual case for bottom-up improvement doesn't translate into behavior change.
Mark's reframe: factually correct doesn't lead to easy acceptance. Personal change works the same way. A patient told by their doctor that they need to lose weight knows this is true. They've known it for years. The knowledge doesn't produce the behavior. The same dynamic operates at the organizational level.
If telling people what to do doesn't change their behavior, what does the work of change look like?
The methodology Mark introduces comes from an unlikely source: addiction counseling.
Motivational interviewing was developed by clinical psychologists working with patients struggling with addiction. The traditional counseling approach was directive -- tell the patient to stop the harmful behavior, explain the rationale, increase the volume if they don't comply. The patients didn't change. They got defensive, or they agreed in the room and reverted afterward, or they stopped showing up.
Motivational interviewing operates from a different premise: people change when they articulate their own reasons for changing, not when other people articulate reasons for them. The counselor's job isn't to provide the reasons. It's to draw out the reasons the patient already has but hasn't fully accessed.
The core concept is what Mark calls the "righting reflex." When we see someone doing something we think is wrong -- whether it's an addict struggling with their addiction, or an executive failing to embrace continuous improvement, or a frontline employee not participating in idea submission -- our natural reaction is to fix it. Tell them what they should do. Explain why. Provide the solution. When they resist, increase the volume.
The research is consistent: the righting reflex doesn't work. People meet directive change attempts with what motivational interviewing calls "sustain talk" -- the articulation of reasons not to change. The more the helper pushes, the more sustain talk the person produces. The helper is generating exactly the response they didn't want.
What works is the opposite move. The helper asks questions that draw out "change talk" -- the person's own articulation of why change might be valuable. The more change talk the person produces, the more likely they are to actually change. The job of the helper is to evoke the change talk that's already latent in the person, not to provide change reasoning from outside.
This has direct implications for anyone trying to drive organizational change.
The framework Mark walks through comes from Ron Auslan and Larry Anderson, who teach motivational interviewing concepts in workplace settings. The model works backward from observable results through the layers that produce them.
Results are what you can measure -- engagement scores, improvement volume, cost performance, quality metrics.
Actions are specific things people do that produce the results. The executive orders carpet. The manager skips the gemba walk. The frontline employee declines to submit an idea.
Behaviors are the patterns of action that recur across situations. The leader generally doesn't ask for input. The team generally responds to problems with workarounds rather than escalations.
Assumptions are what people believe about the situation, often without realizing they hold the assumption. "Staff don't have meaningful ideas." "People are too busy to participate in improvement." "Improvement is the job of leaders."
Beliefs are deeper -- what people hold true about themselves, their role, and how the world works. "I'm smart and successful and I get things done." "Leaders are supposed to have answers."
The model is useful because it makes visible the levels at which change has to happen. Most change efforts target the actions and behaviors layer. Mark argues that without working the assumptions and beliefs underneath, the actions and behaviors revert because the underlying structure pulling them is still in place.
For a leader who has spent decades being rewarded for having answers, the belief "I'm smart and I get things done" is foundational. The behavior "I tell people what to do" is downstream of that belief. Asking the leader to stop telling people what to do without working on the underlying belief produces ambivalence at best -- they may try to change, but the underlying belief keeps reasserting itself.
A useful concept from motivational interviewing: most people in the middle of considering a change are ambivalent. They say things that point both ways. "I know I need to change my leadership style, but I'm afraid staff will walk all over us if I start asking for their input." The first half is change talk. The second half is sustain talk. The same person produces both in the same sentence.
This isn't dishonesty or confusion. It's the normal psychological state of someone considering a meaningful change. The helper's job is to recognize the ambivalence and work with it rather than against it.
When a helper hears change talk, the move is to draw out more of it. Ask follow-up questions that make the person elaborate. "Why is it important to you to change your leadership style?" produces more change talk.
When a helper hears sustain talk, the move is not to argue against it. Arguing against sustain talk produces more sustain talk. The person feels the need to defend their reasons, articulates them more fully, and ends up reinforcing the position they were considering moving away from.
The directional question Mark introduces is striking. Imagine asking someone on a scale of 0 to 10 how important it is for them to embrace bottom-up improvement. They say 6. The follow-up question matters enormously.
Asking "why did you say 6 instead of 10?" produces sustain talk -- the person tells you all the reasons they're not at 10.
Asking "why did you say 6 instead of 0?" produces change talk -- the person tells you all the reasons they're not at 0.
Same number. Same person. Same conversation. Different question, different direction, different outcome. The work of evoking change is largely about which questions get asked.
The session is most pointed when Mark applies this lens to the people who do continuous improvement work -- including himself.
The internal CI practitioner trying to influence executives is in essentially the same position as the addiction counselor trying to influence a patient. The directive approach -- explain why improvement matters, present the data, increase the volume when the executive doesn't respond -- triggers the same righting-reflex dynamic. The executive produces sustain talk. The CI practitioner produces more directive material. The relationship degrades.
The reframe Mark offers: the question shouldn't be "why aren't they doing what I'm telling them to do?" It should be "what would create the conditions where they choose to change?"
Several patterns follow from this.
Lead with humility. John Toussaint, in Management on the Mend, names this directly. CEOs are rewarded for having answers, and it feels good to provide them. The shift to recognizing that you don't have all the answers is hard, but humility produces freedom -- the freedom from having to pretend. The same principle applies to CI practitioners trying to influence change. The willingness to not have the answer creates space for the leader to develop their own.
Draw out, don't push in. Telling leaders why bottom-up improvement matters produces sustain talk. Asking leaders why they think improvement might matter to their organization produces change talk. The substance of what gets said is similar. The direction of who's saying it is opposite. And the second pattern produces actual behavior change in a way the first one doesn't.
Accept the autonomy of the person you're working with. This is the hardest move. Motivational interviewing as a practice requires the counselor to accept that the patient might choose not to change -- and to mean it. The forced version doesn't work. If the helper is internally desperate for the patient to change, the desperation leaks through and produces sustain talk. Genuine acceptance of the other person's autonomy is what makes evocation possible.
In the workplace, this translates to something Mark articulates carefully: "I accept your choice to not participate in Kaizen." Not as a passive-aggressive surrender, but as a genuine recognition that participation is voluntary and the helper's role is to work with what the other person brings, not to force what the helper wants.
The methodology has a phrase that Mark notes maps closely to language Lean practitioners use: "the spirit of MI."
The four components of the spirit, as motivational interviewing defines them:
Partnership -- the practitioner and the person they're working with are partners, not an expert and a recipient.
Acceptance -- the practitioner accepts the autonomy and inherent worth of the person, even when the person isn't changing.
Compassion -- the practitioner is genuinely oriented toward the well-being of the person rather than toward being right.
Evocation -- the practitioner draws out the person's own reasons for change rather than providing reasons from outside.
Mark notes the parallel to the "spirit of Kaizen" or the "spirit of Lean" -- the recognition that using Lean tools without the underlying mindset produces mechanical compliance rather than cultural transformation. The same is true here. Asking motivational interviewing questions without the underlying spirit produces stilted, formulaic conversations that the other person sees through immediately.
A reframe worth pulling out separately. The Peter Senge line Mark cites: "People don't resist change. They resist being changed."
The distinction is important. When a CI practitioner encounters resistance from an executive or a frontline staff member, the natural read is "this person is resistant to change." The motivational interviewing read is different: "this person is responding to being changed."
The implication is that the resistance is information about the helper, not just about the person being helped. If we're encountering more resistance than we expected, it's worth asking whether we're pushing change rather than evoking it. The Stephen Parry line Mark cites makes this sharper: "Resistance to change is directly proportional to your lack of leadership."
The line is provocative on purpose. It points the question inward. The CI practitioner who is frustrated that executives won't embrace continuous improvement might be encountering the natural consequence of trying to push change top-down on people who haven't yet articulated their own reasons for wanting it.
The substantive content of this session is about the human and cultural side of building an improvement program. The platform plays a supporting role rather than being the focus, but several connections are worth naming.
The ROI data from KaiNexus customers grounds the case for bottom-up improvement empirically. The 2.5% of improvements producing over $10,000 in impact, the 1.4% producing over $100,000, the long-tailed distribution that produces most of the cumulative value through high volume rather than through occasional big projects -- this data is what allows CI practitioners to make the factual case for engaging the frontline. Whether the factual case alone produces change is a separate question, but the data has to be there.
The platform's idea capture, workflow, and tracking infrastructure enable the practical work of running an improvement program at scale. A team that decides to engage 1,000 employees in continuous improvement needs infrastructure to handle the resulting volume. Without it, the program stalls at whatever the central team can personally process.
The impact tracking the platform supports addresses a specific motivational issue. Most frontline workers are intrinsically motivated to improve their own work, but the visibility of impact -- both for the individual employee and for the program as a whole -- reinforces the motivation. Submitting an idea that disappears teaches the employee not to submit again. Submitting an idea that gets implemented, tracked, and acknowledged teaches the opposite.
The spread mechanism that makes improvements visible across departments and sites supports what Mark calls "spreading continuous improvement." A team that implements an improvement worth $85,000 at one site can have that improvement adopted (or adapted) at other sites in the organization. The cumulative impact depends on the visibility, not just on the original implementation.
None of this substitutes for the leadership and coaching work Mark describes. The platform amplifies the practice; it doesn't replace it. But for organizations trying to do this work at scale, the infrastructure matters.
Mark Graban is Senior Advisor at KaiNexus and a recognized voice in Lean management and continuous improvement. He is the author of Lean Hospitals, co-author of Healthcare Kaizen, and author of The Mistakes That Make Us. Mark has worked with organizations across healthcare, manufacturing, and service industries to build cultures of continuous improvement grounded in respect for people. His ongoing exploration of how leaders and CI practitioners can become more effective at evoking change rather than pushing it has shaped much of his recent work, including his interest in motivational interviewing as a discipline for anyone trying to drive organizational change.
Why doesn't telling people about the benefits of bottom-up improvement produce behavior change?
Because being factually correct doesn't lead to easy acceptance. The same dynamic operates in personal change -- a patient who has been told by their doctor that they need to lose weight already knows this is true, and the knowledge alone doesn't produce the behavior. At the organizational level, executives who have seen the data on bottom-up improvement continue to operate in command-and-control patterns. Frontline staff who have been invited to participate decline to engage. The factual case is necessary but not sufficient. What produces change is the person's own articulation of why change might be valuable to them -- not someone else's articulation of why it should be.
What is the "righting reflex" and why does it work against change?
The righting reflex is the natural urge to fix what seems wrong with someone -- to confront them with reality, provide the solution, and tell them what to do. When the person resists, the reflex is to increase the volume. The research from motivational interviewing is consistent: this approach triggers what's called "sustain talk" -- the articulation of reasons not to change. The harder the helper pushes, the more sustain talk the person produces. The helper ends up generating exactly the response they didn't want. The reflex feels right because it's well-intentioned, but it works against the outcome the helper is trying to produce.
What's the difference between change talk and sustain talk?
Change talk is the person's own articulation of reasons to change. "I want my organization to perform better." "Bottom-up improvement would let us catch problems we're missing." "I think my leadership style could be more effective." Sustain talk is the person's own articulation of reasons not to change. "I'm not sure now is the right time." "I'm worried staff would take advantage." "We've tried this before." The research shows that the volume of change talk is the best predictor of actual behavior change. The helper's job is to evoke change talk and avoid generating sustain talk by pushing too directly.
Why does the difference between "why did you say 6 instead of 10?" and "why did you say 6 instead of 0?" matter so much?
Because the two questions point in opposite directions and produce opposite types of talk. "Why did you say 6 instead of 10?" asks the person to articulate why they're not higher -- which produces sustain talk, the reasons not to change further. "Why did you say 6 instead of 0?" asks the person to articulate why they're not lower -- which produces change talk, the reasons to change at all. The person and the situation are identical. The direction of the question determines what the conversation generates, and what the conversation generates determines whether actual change happens.
How is the spirit of motivational interviewing similar to the spirit of Lean?
Both depend on the practitioner operating from a particular mindset rather than just executing a technique. Lean tools applied without the underlying mindset produce mechanical compliance rather than cultural change. Motivational interviewing questions asked without the underlying spirit produce stilted conversations that the other person sees through. The four components of the spirit of MI -- partnership, acceptance, compassion, evocation -- map closely to Lean principles of respect for people, collaborative problem-solving, and developing capability rather than directing it. The shared core is that effective change work requires being genuinely oriented toward the other person rather than toward being right.
What does this mean practically for a CI practitioner trying to influence executive sponsors?
A few specific moves follow from the framework. Stop trying to convince executives that bottom-up improvement is important -- they may already know it, and the convincing triggers the righting reflex. Instead, ask questions that draw out their own articulation of why improvement might matter to their organization. Accept that the executive has the autonomy to choose not to engage -- and mean it -- because forced acceptance produces sustain talk. Look for ambivalence and work with it rather than against it. When the executive expresses both change talk and sustain talk in the same conversation, follow up on the change talk rather than arguing against the sustain talk. Recognize that the work is slower this way, but it produces durable behavior change in a way that directive pushing doesn't.
.svg)
Copyright © 2026
Privacy Policy