Part 1 of Ask Us Anything generated more audience questions than two people could answer in an hour. The questions kept arriving after the broadcast ended. Part 2 picks up where Part 1 stopped, working through the backlog of submissions plus the new questions that came in after.
What makes this session useful is the source of the agenda. The topics were not chosen by KaiNexus. They came from CI practitioners who were actively stuck on something: how to start a program without overpromising, how to read whether a culture is genuinely improving, how to run huddles when your team is spread across cities, how to convince senior leaders to take Lean seriously in five minutes, what metrics are worth tracking, and how to help a command-and-control manager become a coach.
Because the questions arrived from across the maturity spectrum, the conversation moves between very different operating realities in the same hour. The starter-program advice sits next to a discussion about whether 40 percent participation is good (it depends on what it was last year). The Lean-in-five-minutes pitch sits next to Greg's story about distilling a two-hour Kaizen lecture down to fifteen minutes after watching residents glaze over. That range is the point. Most CI programs are not entirely new or entirely mature -- they are partway through several transitions at once, and the questions practitioners ask reflect that.
This session was also a small test of change in its own right. It was the first KaiNexus broadcast on a new platform, which made the format itself an example of one of the principles that came up repeatedly: try something small, see what happens, adjust.
Mark Graban was then VP of Improvement and Innovation Services at KaiNexus. He is the author of Lean Hospitals and co-author of Healthcare Kaizen with Joseph E. Swartz, and a long-time advocate for applying Lean thinking to healthcare in ways that engage frontline staff rather than treating improvement as a specialist activity.
Dr. Greg Jacobson is CEO and co-founder of KaiNexus and a practicing emergency physician. He began teaching continuous improvement principles to ER residents in 2005, an experience that shaped both his perspective on how to introduce Lean to busy clinicians and the platform he later helped build.
The opening question was how to build momentum at the very beginning -- how to engage employees and capture their early ideas before the program has any track record to point to.
Greg's first instinct was structural. Do not try to boil the ocean. Pick a small group -- three people, ten, fifty at most -- and get them moving in the right direction. A smaller group makes it possible to respond quickly to every idea, which is the single most important variable in whether people will participate again. It also makes it possible to coach early ideas rather than reject them outright. The first wave of submissions will often look like the parking-garage idea: vague, off-scope, or wishful. The right response is not to file them in a "no" pile. It is to go back to the author, ask what underlying problem prompted the idea, and help them shape it into something testable.
Mark added that even with software in place, the work of getting started is mostly face to face. Managers need to get out of their offices, walk the gemba, sit in team huddles, and look for the moments when people are frustrated or struggling. Those moments are where ideas live. The technology is for capturing what gets surfaced -- it is not a substitute for the surfacing itself.
A subtle but important point came out of this exchange: the CI coach's primary working relationship is with the manager of that small starter group, not with the team members themselves. If the manager is not aligned, the coach will spend their time fighting an undertow. Coaching the coach, in other words, starts before there is anything to coach.
The next question shifted from starting to assessing. What signs indicate that a culture is genuinely supportive of continuous improvement, versus one that is going through the motions?
The answer split into hard and soft signals.
On the hard side: how many improvements were submitted, how many were implemented, what the impact looked like, and how those numbers were trending. KaiNexus customer data across the platform showed about 75 percent of all submitted improvements resulted in a change, with some organizations reaching the 90s and others closer to 50 or 60. The 75 percent figure does not mean three-quarters of initial ideas were brilliant. It means a working process for taking imperfect ideas, doing root cause work with the author, and shaping them into something implementable -- which is itself a sign of cultural maturity.
On the soft side: how much time leaders were actually spending on improvement work. Whatever leaders pay attention to, the people around them treat as important. A leader who runs to the gemba once a quarter is signaling the priority level of CI no matter what they say in town halls.
The most informal but most telling diagnostic was something Greg landed on by accident. Stop someone in the hallway and ask what improvement looks like in their workplace, or what they would do if they noticed something that could be better. If the answer comes back clean and confident in one sentence -- no hedging, no "I guess I'd email someone" -- that is a culture working as designed. If you get a shrug, you have your answer about that too.
Mark added a related question worth asking: what percentage of submitted ideas actually get implemented? Legacy suggestion box systems, paper or electronic, typically run at one to two percent. An effective Kaizen process moves closer to ninety. The gap between those two numbers tells you most of what you need to know about whether a program is real.
A natural follow-up: what about workplaces where morale is already low? How do you motivate people to participate in improvement when they have given up on the idea that anyone listens?
Mark pushed on the question itself. Often, the underlying cause of low morale is that nobody has ever listened to their ideas. People who want a voice and have been ignored long enough simply stop trying. The remedy is not to dump enthusiasm on them. It is to acknowledge the past honestly -- yes, the previous system did not work, and yes, we understand why you stopped bothering -- and then to demonstrate, through immediate action on early ideas, that the new pattern is different.
Some people will respond right away. Others will keep their arms crossed. Both responses are fine. Start with the people who are willing, deliver visible results, and let others come along when they see that the new pattern holds.
Greg connected this back to research from Ethan Burris at UT Austin: when you ask employees for input and then fail to respond, morale drops below where it was before you asked. That has a direct implication for any CI program. The riskiest moment is not deciding whether to launch. It is launching without the infrastructure to respond. Asking and ignoring is worse than not asking at all.
Both came back to the same principle from different directions: this is not a project with a five-year plan. It is a lifestyle change for the organization, built through small steps that compound. The early steps matter disproportionately because they teach people what to expect from the next several years of effort.
Two related questions came in about virtual teams. How do you run daily huddles when people work from home or across multiple sites, without it becoming an excuse for everyone to disengage? And how do you get a globally dispersed leadership team to adopt Lean and model it day to day?
Greg described how KaiNexus runs internally with offices in Dallas and Austin plus people working from other locations. The mechanics are not complicated: video conferencing, shared boards, and a discipline about running the meeting against a visible artifact rather than against a slide deck or a memory. The example he gave was a customer with six different rehabilitation departments running a single 9 a.m. daily huddle on a rotating board -- every department's board comes up once every six or seven days, which keeps each one fresh without trying to cover everything every morning.
Mark added a more uncomfortable observation. Most of the difficulty people report with virtual huddles is the same difficulty that exists with in-person huddles -- it is just easier to disguise online. If people are not attending, the first question to ask is whether the huddle is providing any value to them. Sometimes the answer is a legitimate scheduling conflict, in which case fix the time. Sometimes the honest answer is that the huddle has drifted into a status meeting with no benefit to the attendees, in which case fix the agenda. Technology cannot rescue a huddle that no one would want to attend in person.
On asynchronous communication: a lot of the work happens in the platform between live touchpoints. Notes from a huddle logged in KaiNexus become reference material for people on other shifts or in other time zones. An ER night-shift worker who got off at 5 a.m. is not going to enjoy a 9 a.m. video huddle, and asking them to is not respectful. The blend of synchronous and asynchronous communication is what makes it work.
On the globally dispersed leadership question, Greg's answer was deliberately deflating. He does not see a meaningful difference between leading CI across a hallway and leading it across a hemisphere. The leadership behaviors that matter -- commitment, resources, modeling, simple and consistent discipline -- transfer cleanly. The medium of communication is secondary to whether the behaviors are real.
The next question was an elevator pitch: a Healthcare leader had five minutes to convince their senior management team to seriously investigate Lean. What are the three or four most important things to communicate?
Greg reframed the question before answering it. The framing he suggested was not "what should I tell them about Lean" but "what does this leadership team already care most about right now, and how does CI help with that?" If they are focused on length of stay, talk about length of stay. If they are preparing for a Joint Commission audit, talk about that. The opening pitch does not need to use the word Lean at all. It can use the language they already use -- continuous improvement, staff engagement, reducing variation, building capability -- and only later introduce Lean concepts once the value has shown up.
The argument he wanted to make in those five minutes was a numbers argument that does not require any Lean vocabulary. You are currently solving this length-of-stay problem with twenty people on a project team. What if you had 250 people seeing it from different angles -- the people who actually do the work and notice the small obstacles those twenty cannot see? Three to six months in, when results start showing up, the conversation naturally opens to "how did Sally's idea work" and "what was Sally actually doing differently" -- and that is when Lean concepts can be introduced in their proper context.
Mark added the point that this pitch is fundamentally about scope. Lean is not just a toolkit for frontline staff to use. The senior team also has work to do -- learning to use Lean thinking in their own management processes, strategy deployment, daily review of improvements, coaching their direct reports. Organizations like ThedaCare and Virginia Mason did not get their results by sponsoring CI; they got them because the CEO and senior team were practicing the methodology themselves.
Greg's story about teaching ER residents is one of the most useful parts of the session, because it shows what works and what does not in introducing Lean concepts to busy clinicians.
In 2005, his approach was to distill Masaaki Imai's Kaizen into a two-hour lecture, broken into two one-hour sessions. By 2008, after watching multiple cohorts of residents glaze over, he had compressed it to a fifteen-minute primer focused on a handful of principles.
The two ideas that landed most reliably with frontline staff:
The first was "low cost, low risk." Frame the kinds of ideas you want people to bring forward. Not capital projects, not parking garages -- small, testable changes that someone can try in a shift without anyone's permission.
The second was a working definition of non-value-added work, taught through an example. In the ER, a patient needs a chest tube placed to re-expand a collapsed lung. The clinician goes to the cabinet where the tube should be. It is not there. Someone has to run to Central Supply down the hall, find the right tube, and bring it back. Every second of that running, searching, and returning is non-value-added work -- and possibly time during which the patient is getting worse.
Once frontline staff hear that example, the response is consistent. "My day is full of non-value-added work, and I'm happy to start identifying it." From that point, the third principle adds itself: when you feel frustrated at work, pause for fifteen seconds and ask whether there is a process problem underneath the frustration. Turn the negative feeling into an observation that can be acted on -- which becomes the seed of an improvement.
Mark made the broader point that you can get frontline staff engaged in improvement without formally teaching them much Lean at all. Start with what frustrates them. Layer in concepts -- workarounds versus root causes, busy versus value-adding -- as the work calls for them. Belt programs and tool-by-tool training are usually unnecessary at this stage and frequently counterproductive.
The question on metrics had two parts. First, beyond the number of PDSAs, the number of problems identified, the number of improvements done, the number of things not done, and the number of people coaching others -- are there other metrics worth tracking? Second, should there be targets for those metrics?
The first part was straightforward. The metrics that show up across KaiNexus customers cluster around four categories. Activity (submitted, completed, percent change). Engagement (ideas per person, percent of staff active). Impact (financial, time, satisfaction, safety -- by location and by person). Best-practice indicators (response time within 24-48 hours, distribution of ownership rather than one person carrying everything, healthy implementation rate).
The more important part of the question was about how to use those metrics. Both Mark and Greg argued for tracking trends over time rather than absolute numbers, and against turning the metrics into targets.
A 50 percent participation rate is not a number you can evaluate in isolation. If last month it was 80 percent, something just broke and you should go investigate. If last month it was 10 percent, you are moving in the right direction and the right move is to keep going. The number itself tells you nothing without the trend.
The case against targets was made specifically. Setting a quota of two ideas per employee per year predictably produces a December surge of low-quality submissions from people trying to hit their number. Setting a target that every improvement must save X dollars is one of the surest ways to kill a Kaizen program -- it filters out exactly the small, frequent, hard-to-monetize improvements that drive most of the long-term value.
Mark used the example of Joe Swartz's organization, where participation has stabilized around 40 percent and has not yet moved higher. The leadership's response is not to set a higher target or to blame the 60 percent who do not participate. It is to ask what they as leaders need to change so that the other 60 percent want to participate. That distinction -- between forcing participation and earning it -- is the difference between a program and a culture.
The final substantive question was how to help leaders make the transition from directing work to coaching it.
Mark's first observation was that someone has to coach the coaches. The Mary Greeley example came up again: Ron, their director of process improvement, has spent years coaching individual managers and helping them build good habits. There is no class or book that produces a coach. There is only practice, with someone watching closely enough to catch the small habits that get in the way -- the manager who dismisses an idea in front of the team, the manager who solves the problem themselves instead of asking a question, the manager who skips the gemba walk when calendars get tight.
Greg took the question through a different lens, drawing on Simon Sinek's Leaders Eat Last and the example of Gregg Popovich. Asking a leader to coach rather than direct is asking them to give up some control. That requires them to feel safe in their own role. Popovich has been able to take coaching risks across his tenure because his position is secure. A manager who is afraid of being fired is not going to give up command-and-control habits. Psychological safety, in other words, is not just something leaders extend to their teams. It is something they need extended to them before they can change.
Both came back to a practical norm worth stealing: praise publicly, coach privately. The worst possible coaching moment is a manager being corrected in front of their team. The best is the quiet pull-aside afterward -- "let's talk through what happened in that huddle; here's what I saw; did you see it the same way?" The football coach in the player's face on national television is exactly the wrong model.
Most of what came up in this session is about behavior, not software. The leadership commitment, the willingness to coach rather than direct, the discipline of responding to ideas quickly, the manager who gets out of the office and walks the gemba -- none of those are things a platform can do for you.
What a platform can do is remove the structural reasons those behaviors fail. The "black hole" problem Mark and Greg kept returning to -- ideas submitted to a suggestion box or an email address and never heard from again -- is almost entirely a tracking and routing problem. When every submission has an owner, a visible status, and an expected response window, the black hole closes. The cultural work still has to happen on top of that, but it has somewhere to land.
The same logic applies to the metrics conversation. The case for tracking trends over time rather than snapshot numbers is much easier to make when the data is in one place and updates automatically. The decision not to set targets becomes easier when leaders can see real participation and impact across the organization without needing to chase reports. The argument for distributing ownership of improvements across many people rather than concentrating it in one CI coach is easier to enforce when imbalances are visible immediately.
For distributed teams, the asynchronous documentation that Greg described is what makes virtual huddles workable in the first place. A note logged during the morning huddle in one site becomes context for the team at another site that does not run on the same schedule. The blend of synchronous and asynchronous work that he and Mark kept coming back to depends on having a shared place where the asynchronous part actually lives.
None of this replaces the leadership work. It just removes the excuses for not doing it.
How small should "starting small" be when launching a continuous improvement program? Small enough that one CI coach can respond to every idea within a day or two and have a real coaching conversation with the author. In practice, this usually means three to fifty people in one or two departments, with the manager of that group genuinely on board. Trying to launch across an entire health system or plant network from day one almost always crashes because the response infrastructure cannot keep up with the volume.
How quickly should leaders respond to a submitted improvement idea? Within 24 to 48 hours, ideally faster. The first few times an employee submits an idea, they are testing whether the system actually works. Silence or a delayed response teaches them not to bother. The response does not need to be a yes -- it can be a question, a clarification request, or a coaching conversation about scope. It just needs to be timely and human.
Is a 75 percent implementation rate good or bad? It depends on what you started from and what your peers look like. Across KaiNexus customers, the platform-wide average is about 75 percent, with some organizations in the 90s and others in the 50s and 60s. Legacy suggestion box systems typically run at 1-2 percent, so any move into double digits represents real progress. What matters more than the absolute number is whether it is trending up over time and whether the implementation rate includes the coaching work that turns rough ideas into testable ones.
Should we set targets for the number of improvements employees submit? Generally no. Targets like "every employee must submit two ideas per year" tend to produce a year-end surge of low-quality submissions and signal that the program is about hitting a number rather than improving the work. Targets that require every improvement to save a specific dollar amount filter out exactly the small, frequent improvements that compound. Track participation rates carefully, but treat low participation as a question to investigate, not a number to force.
How do you run an effective daily huddle when team members work remotely or across multiple sites? The technology choices are less important than the underlying structure. Run the huddle against a visible board that everyone can see. Keep it short and focused on exceptions, not status. If people are not attending, ask honestly whether the huddle is providing value to them -- the same question you would ask about an in-person huddle that nobody wants to come to. For 24-hour operations or teams across time zones, supplement the live huddle with asynchronous notes logged in a shared system so off-shift staff are not excluded.
What if a manager is not bought in to coaching improvement work? Coaching the manager comes before coaching their team. A CI coach trying to drive improvement in a department whose manager is disengaged or skeptical will spend their time fighting an undertow. The better approach is to invest in the manager first -- helping them see the value, practice the behaviors, and feel safe enough to give up some command-and-control habits. Psychological safety for managers is a prerequisite for them to extend it to their teams.
How do you introduce Lean to senior leaders who only have a few minutes? Speak the language they already use. If they care about length of stay, talk about length of stay. If they care about Joint Commission readiness, talk about that. Make the argument that the difference between solving the problem with a twenty-person project team versus engaging two hundred and fifty people who see it from different angles is a numbers argument, not a methodology argument. Lean concepts can be introduced in their proper context months later, once results have surfaced.
What's the simplest way to introduce continuous improvement to frontline staff? Two ideas tend to land. First, frame the work as low cost, low risk -- small changes someone can try in a shift, not capital projects. Second, define non-value-added work with one concrete example from their environment. Greg's example with ER residents was the chest tube that should have been in the cabinet but was not, sending someone running to Central Supply during a trauma resuscitation. Once people have that frame, they recognize non-value-added work everywhere in their day.
How do you handle improvement ideas that are clearly not feasible? Do not reject them outright. Go back to the author and ask what underlying problem prompted the idea. The first version of an idea is usually a solution looking for a problem. The second or third version, once you have done a few rounds of root cause questions with the person who submitted it, is often a real opportunity. Across KaiNexus customers, most ideas that eventually got implemented did not start in their final form -- they were shaped through that catchball process.
What are the soft signs that a continuous improvement culture is actually taking hold? Stop someone in a hallway and ask what improvement looks like in their workplace, or what they would do if they noticed something that could be better. If you get a clean, confident, one-sentence answer, the program is real. If you get a shrug or "I'd email my manager," you have your diagnosis. The other soft signal is how much time leaders are spending on improvement work, because whatever leaders pay attention to is what the rest of the organization treats as important.
How do you re-engage employees who have been burned by previous suggestion box programs? Acknowledge the past honestly and demonstrate, through immediate action on early ideas, that the new pattern is different. Some people will respond right away. Others will keep their arms crossed and watch. Both responses are fine. Start with the willing, deliver visible results quickly, and let the holdouts come along once they see that the new pattern holds. Forcing participation in this situation usually makes things worse, because it confirms their suspicion that this is another corporate initiative being done to them.
.svg)
Copyright © 2026
Privacy Policy