In this new series, John Dues and Andrew Stotz discuss John's model for improvement. This episode includes an overview of the model and how John uses it for goal-setting and planning in his school.

0:00:02.4 Andrew Stotz: My name is Andrew Stotz and I'll be your host as we continue our journey into the teachings of Dr. W. Edwards Deming. Today, I'm continuing my discussion with John Dues, who is part of the new generation of educators striving to apply Dr. Deming's principles to unleash student joy in learning. The topic for today is building an improvement model. John, take it away.

0:00:24.8 John Dues: It's good to be back, Andrew. Yeah, so we sort of wrapped up this last series. We had a six-part series on organizational goal setting. And we, if you remember, we talked through those four conditions that are important for organizational goal setting, especially healthy goal setting, where before we set a goal, we understand sort of how capable our system is. We understand how our data is varying within our system. We are looking at our system and seeing if it's stable or unstable. And then, of course, we want to have a method for how we go about improving. And so you kind of have to have an understanding of those four conditions before you set a goal.

0:01:03.6 JD: And I thought sort of as an extension of that, or possibly a new series, we could kind of take a look at an improvement model that would help us sort of better set ambitious goals. Because when we did those four conditions, it kind of leaves you wondering, well, how ambitious should my goals be? Should I still do stretch goals, those types of things? And I think this improvement model that we're building here at United Schools sort of addresses that. And it's something we're building.

0:01:34.4 JD: And so I think the listeners kind of get like a little bit of behind the scenes on what it looks like now. I think we'll see a version of it. And perhaps through this dialogue, through the series, we'll even think about ways to improve it.

0:01:48.4 AS: Can I ask you a question about that?

0:01:49.6 JD: Sure.

0:01:50.0 AS: One of the things, I do a lot of lectures on corporate strategy and workshops, and the lingo gets so confusing, vision, mission, values, and all kinds of different ways that people refer to things. But when I talk to my clients and my students, I oftentimes just tell them a vision is a long-term goal. And it could be a five-year or a 10-year goal. And because it's long-term, it's a little bit more of a vision as opposed to, you can see it very clearly. Like my goal is to get an A in this particular class, this particular semester. Whereas what I try to say is, a vision is: I want to be in the top of that mountain. And I want us all to be at the top of that mountain in five years. And I kind of interchangeably call that a long-term goal and a vision. And I'm just curious what your thoughts are on long-term versus short and medium as we go into this discussion.

0:02:53.8 JD: Yeah. I think as we get into the model, we'll actually see both of those things, sort of a long-term sort of goal, sort of a more intermediate thing, and then how you work back and forth between those two things. So I think that's a good segue.

0:03:08.4 AS: Let's get in it.

0:03:08.4 JD: Yeah. And so just maybe just a few other things about the model before we get right into it. So one thing to know I've come to appreciate is when when I say a model, I just mean something visually representative that helps us understand and communicate how we think things should be functioning in reality. So when I say improvement model, I'm actually like talking about a diagram on a piece of paper that you can put in front of everybody on your team. So everybody has an understanding for how you're approaching goal setting in this case.

0:03:38.1 AS: Would you call it an improvement visualization? Or what's the difference between what you mean by model and like something that I would call, let's say, a visualization?

0:03:49.5 JD: Yeah, I'd say it's a type of visualization when I say model.

0:03:52.8 AS: Okay. Excellent.

0:03:53.8 JD: Yeah, that's what I'm talking about. And I think you'll see it when we get into the model that definitely there's credit due to Mike Rother and his concept of Improvement Kata because this model heavily borrows from the work that he's done, if you're familiar with that four-step Improvement Kata process.

0:04:15.1 AS: Yeah. Very.

0:04:19.7 JD: But anytime, whatever the thing is that you call like key performance metrics, key metrics, whatever you call that thing that we all set in our organizations, there's always this gap between what we want and what we're currently getting. And this model gives us the scientific way of thinking and working to close basically that gap. In this world, the gap between the voice of the process and the voice of the customer, how do we close that gap? So that's sort of what the model is addressing. So I'll share my screen so you can see that and anybody that's watching can see what the model looks like. And I'll just kind of leave that up as I'm talking about it, put it in slideshow.

0:05:08.7 AS: Great. We can see that now.

0:05:14.6 JD: Great. So we can just start by just kind of giving an overview, especially for those people who are listening, but you can kind of picture like a path going up a mountain and that path has twists and turns. It has obstacles. In this particular diagram or model, there's rocks in the way of the path. There's a water hazard, there's trees in the way, there's a roadblock. And as you go, it's kind of strange because you're working your way up. And I'll explain this all as we go through it kind of one step at a time. But as you're working from left to right in the model, this four-step improvement model, you have a team over on the left. This team's working on a goal that you're setting. And then over on the left, you actually have step two, which is grasp the current condition. And then you have this big crack in the path that's called the threshold of knowledge. And I'll talk about what that is.

0:06:11.1 JD: And sort of the next step is actually step four, experiment to overcome obstacles as you're working left to right. You go further up this path, up this mountain. And number three, the step three is establish your next target condition. And then when you get all the way up the mountain and you have this challenge or direction. So that's what you were just talking about. So what's that long-term thing that you're trying to accomplish? We call that a challenge or direction. So the steps that you're taking actually chronologically are you're going to do number one first.

0:06:43.2 JD: You're going to set that challenge or direction, but it actually is the thing that you're working toward. That's the sort of beginning with the end in mind. So that's why it's way up on the mountain, but you're going to do that first. And the next thing you're going to do is go all the way back down to the start of the path and grasp whatever that current condition is in your organization. And then you're going to run experiments on the way to trying to get to the next sort of intermediate step, that next target condition. So four steps, and then you have this team working on it.

0:07:16.8 AS: Which I would say for the traditional American style, as from my perspective, it can be a bit confusing because you're starting with number one at the farthest point away instead of closest to you. Then you're going to come to number two. From a timeline perspective, it feels like you're kind of zigzagging back and forth in your thinking.

0:07:38.8 JD: Yep. You definitely are. And it takes a little bit to wrap your head around it, but we'll kind of work through this piece by piece. So let's start with the team. So you have these people on the left-hand side of this diagram. There's sort of three different groups within that team. And we've talked about this a number of times, but remember that there's this key concept when you're going to take a thinking systems or a systems view of an organization. That you have to have these three different groups of people. You have to have the people that are working on the system, the people that are working in the system, and then from Dr. Deming's perspective, you have to have somebody that has profound knowledge, has that lens. So again, someone from the outside that has profound knowledge. And then in our case, the people working in the system, generally speaking, are the students. And then you have to have the managers that have the authority to work on the system. So in our system, that would be teachers and school leaders. But this model is not specific to educational organizations. You could translate this to any other type of organization.

0:08:50.4 JD: So if we were a hospital, then perhaps the people working in the system, depending on the improvement project, could be nurses. And then the managers that have the authority to work on the system, maybe the hospital management team. And then someone from outside with profound knowledge could be either someone internally that has familiarity with the System of Profound Knowledge or someone that they bring in externally, like a consultant to help out. So the point is, is that, again, this team, whoever's working in the system is going to differ by the organizational sector that you're working in. But it translates in the system basically.

0:09:31.0 AS: It's interesting that I've seen this type of diagram or concept about work on the system, work in the system and a System of Profound Knowledge coach. But it just kind of clicked for me to think about it. It obviously, like when I work with a company, I'm working with the owners and the top management. And when I do that, we're working on the system.

0:09:58.5 JD: Yep.

0:10:00.2 AS: And I have the knowledge of the System of Profound Lnowledge. So I'm coaching them about the system. And then within the system, they have the employees who are executing on what they're trying to improve and do, but it just perfectly explains it. So I love that diagram.

0:10:17.8 JD: Yeah. And I have the same experience. And I think we've mentioned on this podcast before that in my world, we often have school or district-based improvement teams. And it's typically leaders of the organization, sometimes teachers, but almost never is it students working in the system that are a part of, or, providing significant input into the improvement. So, I think if you can combine, in our case, students working in the system, because they have things that they can identify in terms of how they experience the system that are different than the people that work on the system. And then having that third group that, or that person that has that outside profound knowledge, if you put all the three of those things together, I think you have a much better chance to improve. But I think in schools, that's probably never happening. I'm assuming that's the same in other industries as well.

0:11:08.3 AS: And this also explains why when Dr. Deming would see slogans and things like that, encouraging the workers to do better and higher quality, he was like, they don't have the authority to change the system.

0:11:22.5 JD: Right.

0:11:24.1 AS: And what you've said is the group that's working on the system has the authority or the ability to change the system.

0:11:35.4 JD: Yeah. This is one...the makeup of this team that's using this four-step process, that's one innovation that we've done to this model that would be different from the Improvement Kata. So in the Improvement Kata, there's just coach and learner. Usually sometimes there's a coach of the coach, a coach and a learner, depending on how it's represented, but this is in my view, an innovation where you have the work on the system group, the work in the system group, and then the System of Profound Knowledge coach. I haven't seen that in this model.

0:12:07.4 AS: And could that be because when Mike Rother was writing his book, he was particularly referring to Toyota.

0:12:18.7 JD: Could be. Could be.

0:12:19.5 AS: Where the workers have more authority to impact the system. Whereas in the typical American system, the worker doesn't really have the authority to stop the production line or something like that to the extent of the Japanese. So interesting point.

0:12:36.1 JD: Yeah, that's a really good point. My understanding of Mike Rother's work is he sort of derived this improvement model by watching, observing, working with Toyota over a very long period of time. So that very well could be the case. Cool. So we have the team, so let's go to step one, that's the challenge or direction. And I really like that because again, when we did that six part series on Goal Setting is Often an Act of Desperation, one thing that I did think was missing was like, well, still as an organization, we want to move forward. We want to improve. We want to be ambitious in how we're setting our goals, but I don't think that fully came through in the four conditions. And so I think layering this model on top of the four conditions really helps because I think it is important to be ambitious, especially when we're talking about like a mission driven organization, we need to be setting ambitious targets for student learning, coming to school, those types of things.

0:13:39.6 JD: So really what we're doing in step one of the model is we're asking the question, where do we want to be in the long run? So this is a long term goal. This is a longer range goal that would differentiate us from other schools if we achieved it. But currently when we think about this goal, it actually seems nearly impossible because it's so far from where we are currently performing. We don't know how we're going to get there. So an example in my world is, schools have been paying much closer attention to chronic absenteeism, which is when a student misses 10% or more of the school year. And those numbers basically skyrocketed towards the end of the pandemic and then for the last several years. So that's something we're focusing on as an organization. So our chronic absenteeism rate is really high, like 52%, something like that over the last several years. And we want to get that down to 5%. So there's this huge gap.

0:14:53.6 AS: That's a huge move.

0:14:54.5 JD: Huge gap, order of magnitude, right? To go from 52%, that's the voice of the process. That's what's actually happening. And the voice of the customer, what we want is 5%. And we really don't know how to get there. And that's going to be the case at the point where you're at step one, but you're doing that first. You're setting that challenge or direction. And that really is something that needs to be set, in my view, at the leadership level, at the management level. So, that's step one.

0:15:22.9 AS: And you just said something that's interesting is we really don't know how to get there.

0:15:25.6 JD: And we really don't know...

0:15:26.9 AS: I mean, if we knew how to get there, we'd probably be there.

0:15:28.6 JD: Yeah. Yep. Yep. So that's step one. That's why if you're able to view the model and you're watching the podcast and you can see the video, that's why number one happens first, even though it's on the far right hand in the upper right hand corner at the top of the mountain in the model.

0:15:45.8 AS: And is there a reason why it's a relatively vague thing, right? Challenge or direction.

0:15:54.0 JD: Yeah.

0:15:55.5 AS: Why is it vague as opposed to specific target, goal or saying something like that?

0:16:03.7 JD: Yeah. I mean, I think, I like challenge or direction. One, it fits on the page. And it sort of conveys that it's going to be a challenge. And it also, if you're going to work in this way to achieve something like that, that it's actually setting the direction of the organization, the direction that the organization is moving toward. So.

0:16:24.0 AS: In other words, is it acknowledging that we really won't, we really don't know that target. We think we know it, we see that mountain, but as we go closer to it, we want to go in that direction, but as we get closer, it'll become more clear exactly where we're going to be or want to be.

0:16:44.7 JD: Well, I think this would be something that... I think in my view, we're still learning. But when we set that challenge or direction, I guess I could see some circumstances where we would come off that, but I think we kind of want to set it in a way that really pushes us. Right. So I'd be, I mean, I think you could learn some things that would say, okay, maybe that wasn't the exact right number to set, but I'd also be careful about just adjusting it because it's hard.

0:17:13.2 AS: Okay. So you mentioned 5%.

0:17:17.9 JD: Yeah.

0:17:19.1 AS: Would that be, would you state it as achieve 5%?

0:17:25.9 JD: Yeah. 5% or less of our students are chronically absent.

0:17:30.4 AS: Okay. Keep going. I don't want to slow it down. But listeners may get it faster than I do. I'm a little bit slow and I have a lot of questions as we go along.

0:17:37.0 JD: No, no. And I think what we could do in future episodes is dig into each of the steps a little bit more too, and use this as an overview session.

0:17:46.9 AS: Yep.

0:17:48.3 JD: So that was step one. So now what's going to happen in step two, you're going to come all the way back down. Now you're at the very start of the path.

0:17:56.6 AS: Back to reality.

0:18:00.6 JD: Back to reality, step two. And the first thing you have to do, okay, we've set the target, this very challenging direction we want to head into because it's the right thing to do. The next thing we're going to do is grasp the current condition. And so in step two of the model, we're going to ask, where are we now? So we know the long-term goal and now we need to study the current process and how it operates basically. So basically this study represents our current knowledge threshold about the process. And then it's going to contribute to how we define the next target condition we've set that sort of intermediate step on the way to the challenge. And so a lot of that six-part series on goal setting is often an act of desperation, a lot of that learning is right here at what we're doing at step two, because we're creating a process behavior chart in a lot of cases, and understanding how our data is performing over time in this particular area. That's what grasping the current condition means.

0:19:02.6 JD: So part of it, it's a data thing. So in this chronic absenteeism example, what I'm gonna do is I know where I want to be. Now I need to understand where are we historically. And then also as a part of grasping the current condition, I may wanna do some things like interview students and families that are chronically absent, then sort of dig into why that is. Interview teachers about why they think that is. There's a number of things that you could do at this step on the ground where the work happens to grasp the current condition. And I think there can be a sort of quantitative component to that and a qualitative component to that. Also, we sort of understand like how are things actually working on the ground that contribute to us not being where we want to in this particular area.

0:19:56.7 JD: So that's step two. That's what we're gonna do next. After we've set the challenge or direction, we wanna sort of understand the situation on the ground, grasp the current condition. And then next what we're gonna do is step three, which is establish your next target condition. So in step three of the model, we ask where do we want to be next? So we know we can't make this leap, from 52% to, 'cause we wanna decrease it down to 5%. We know we're not, that's too big a step that we're just gonna get there somehow magically. So our target condition, then it's our next goal, usually within a time bound, achieve by date. In Mike Rother's work, he suggests something on a pretty short term scale. Something like one week or one month. So something like chronic absenteeism, I think one month would be sort of where I would set the next target condition. Just having experience with something like attendance rates.

0:21:07.0 JD: And at this point we don't exactly know how we'll achieve the next target condition, but it also, it doesn't feel as impossible as the challenge. So it's a step towards the challenge. So we're gonna do that next. So we set the big challenge that may take us three years to get to. Then we understand the current conditions on the ground and we use that knowledge to set our next target condition. So that's step three. And then the fourth step is we're gonna experiment to overcome obstacles.

0:21:45.9 AS: And before you go to fourth, let me just ask a question about establish your next target condition. One of the things that's missing from that, obviously is, you know, coming from a different perspective, is that when we say, all right, here's where we want to be, and let's go back to reality, and here's where we are. Sometimes, when people work like myself and others, work with people who say, okay, let's map out all the steps to get to that vision. What are the next five things we have to do? Whereas here you're saying, let's focus on the next target condition rather than the next five.

0:22:25.4 JD: Yep. And keep in mind when I say establish the next target condition, what I literally mean is what's our next intermediate goal that we're gonna shoot for? So if we're trying to get all the way down to 5% from 52, remember decrease is good in this case, establish my next target condition, maybe over the next month, I wanna see if I can get that from 52% down to 35% or down to 40%. Part of what I would look at when I set that next target condition is what did the variation look like when I was charting in step two? So the magnitude of that variation will give me some indication of what would be a reasonable sort of next step target for step three basically.

0:23:11.9 AS: And maybe just explain for those people not familiar with Mike Rother's work and, you know, terminology that you're using, why do you say establish your next target condition?

0:23:28.0 JD: I think, I don't know. I think that, you know, really what I mean is just establish the next target, establish the next intermediate goal, basically. Now, I think using the word condition is because when you think about something like chronic absenteeism, there's conditions that probably contribute to that and part of that condition may be the things that you wanna work on. So I kind of think of like, you know, 'cause when you look at step four, you're gonna experiment. So you're creating a new set of realities, a new set of conditions in your organization. And so sort of that coincides with the metric that you're shooting for. So it's not just the metrics, it's also like what are the conditions surrounding that metric. If that makes sense.

0:24:15.8 AS: Yep.

0:24:16.9 JD: Cool. And then step four then is experiment to overcome obstacles. So basically in step four of the model, we move toward the target condition with experiments. And by experiments, what I'm talking about is Plan, Do, Study, Act cycles or PDSA cycles, which uncover obstacles we'll need to work on. So the path, and that's the path in the model is windy 'cause it's this path to the target condition is not gonna be straight line, but it's gonna require this rapid learning to move in that direction basically. And so let's say we've set that next target condition to be one month from now, that's what we're shooting for. And we're gonna run a series of experiments. Maybe it's four one-week PSDA cycles, maybe it's two, two week PSDA cycles. Maybe it's one one month cycle. It depends on sort of the nature of the Plan, Do Study, Act cycle. But running these cycles where we make a plan, including a prediction, run the experiment, and then study what happens and see if it's moving us in the direction of the target condition.

0:25:40.0 JD: And so in that way, we're rapidly learning what it's gonna take to hit that next target condition. And the other important part of this, you'll see in between the grasping of the current condition at step two and running those experiments, there's this huge fault line, this huge crack in the path that you can't just jump over. And it's kinda labeled there, it says Threshold of Knowledge. And basically it's the point at which you have no facts and data to go on. That's the threshold of knowledge. There's always a threshold of knowledge. And so to see further beyond that threshold of knowledge, that's where you conduct your next experiment.

0:26:28.7 AS: Interesting.

0:26:29.8 JD: So because you, like you were saying, we wanna outline these five steps that we're gonna do. So with chronic absenteeism, I read somewhere a Harvard study where if you text parents what a kid's attendance rate is on a regular basis, they're then more likely to come to school on a frequent basis. So you could see where a school system would spend all this money to get a texting system, maybe even allocate a person or a half of a FTE of a person to run this system. And they faithfully implement this texting system, and it has no impact at their school to impact those chronic, because it had nothing to do with what the actual problem was in that context. And you've spent all this money. And that was just a hypothetical.

0:27:21.2 AS: And you could have done a pilot test of 10 parents or 20 and done it manually and sent out some messages and just tested a little bit.

0:27:31.1 JD: Yeah. You run a test with 10 chronically absent kids. Just to see if you can improve their attendance for a week. And maybe you learn something or for a month and maybe you learn something. And then if the early evidence is pointing in the right direction, then you can run that experiment with more kids or for a longer period of time or under slightly different conditions. Those types of things.

0:27:54.6 AS: So an example that I would say in relation to this for one of my clients is that we've identified that they need to get a higher gross profit margin.

0:28:04.7 JD: Okay.

0:28:05.5 JD: And their gross profit margin is about 23%. And I know that the average is about 30 in the industry. And so my work with them is how are we gonna get that profit margin to be 30 or 35%? 35 would be showing that you've really got pricing power because of something that we've done. And so, I'm pounding away that we've gotta improve this, but you know what? We don't have data to understand the current condition. And this week we've... It's taken us about a couple months to pull that data together. But now we have absolutely comprehensive data that my team has calculated on the profitability of every product, the profitability of every customer, and the profitability of every process. We know the capacity utilization of each part of the production process. So now we have the knowledge that we didn't have before that's gonna, that once get, digest this knowledge, it's gonna give us the indication of what to do next. Which is it's gonna be shut down a particular production process or increase price there. We may lose customers, but it's not worth doing it at this low price or so, but without that knowledge, we're just, it's a dream.

0:29:21.4 JD: Yeah. It sounds like you guys have done step one and step two in that process.

0:29:28.0 AS: Yeah. Which is exciting. 'Cause now Friday's meeting is gonna be about, all right, how do we take this huge amount of data and effort that we've put in and now it's time to come up with what are the steps that we're gonna take?

0:29:40.4 JD: Yeah. And I think even just in that situation, even just acknowledging that there's the threshold of knowledge. Even just getting people to acknowledge that in a room that they actually don't know what's gonna happen. That's the power of the PDSA because it makes you predict, okay, you say this thing is gonna work and when you put in this plan in place, this is your prediction. And then when you come back next week and it doesn't work, then you have to explain that, you know, it's not a gotcha, but it very quickly makes you think in a different way.

0:30:13.0 AS: It keeps a record so someone has gone back, well, I didn't think it was gonna work, you know, for sure.

0:30:18.8 JD: Well, right. And it's usually very like, some of the things that I found in that is when people are off on their predictions, it's very mundane things that they didn't account for. We're in student recruitment season and we set a goal for the number of calls we're gonna make to prospective families. And then hypothetically a recruitment director could fall short and it's like, well what happened? It's like, well, oh, the two part-time people that we had, I forgot they are actually out two days last week right? And so it's usually things like that are actually getting in the way of us accomplishing these grand targets that we have set.

0:31:05.5 AS: By the way, where does the threshold of knowledge fit? We've got number one challenge or direction, number two, grasp the current condition. It's after the grasp the current condition that we come to the threshold of knowledge.

0:31:17.7 JD: Yeah. Because, well, we have somewhat of an understanding of the condition on the ground, but we don't know what's gonna improve it until we run the experiments. So we start running the experiments and we try to sort of narrow that knowledge gap basically. And this is sort of the final part is basically like what do you do when you get to that experiment and when you hit that target condition, when you reach that by the achieve-by date, well now there's a new condition and you repeat the four steps because you haven't reached the challenge or the direction. You just met that sort of intermediate goal. And you basically keep running this four step cycle until that learned long-term challenge is achieved.

0:32:12.5 AS: Okay. Great. So we've got the establish your next condition down where it could be one week, it could be one month, in some cases it could be longer, but it's really our next intermediate goal. Where do we wanna go next? What's the next right step?

0:32:28.5 JD: Yeah. Well, so you go back to step two 'cause you're not gonna change the challenge or direction. Now there's a new set of conditions 'cause you've moved ahead, right? And now you're gonna go back and say, okay, what are the current conditions like? And now we're gonna, okay, let's say we move from 52% to 42%. Now we go back and sort of understand the experiments from that last cycle. And we're gonna set that next target condition. So maybe now we wanna get it down to 25%. And we're gonna run another round of experiments in a certain amount of time to see if it hits that next target condition. And basically you're just gonna keep doing this over and over again. That's really the continual improvement model that we're operating under.

0:33:22.7 AS: So how would we wrap this up?

0:33:24.4 JD: So the big thing for me is, you sort of have to have a model to bridge that gap between current conditions and future aspirations. Beause there's always a gap between those two things. And what this model does is it gives us a scientific way of thinking and working to close this gap. It's a more powerful model than I've ever sort of seen anywhere. And then literally you put it on a piece of paper like this and then you have to explain it to people over and over and over. And then you have to actually do it with people. So we're actually doing this, getting people excited about running PDSAs. And the most important thing is that the challenge or direction, especially for leaders that are listening to this, you don't stand on this mountaintop and set it and then say, go do it. That's why this team aspect is so important. We're setting this challenge or direction as a team, and then we're working together on the ground. Putting that work in, running those experiments to try to bring this thing about, is a completely different way of working. It's not an accountability system, it's an improvement system.

0:34:39.4 AS: Yeah. That's a great overview of this system that you guys are applying and it's exciting to learn more. So I wanna thank you on behalf of everyone at The Deming Institute, John. And I thought the discussion was very interesting myself. And for listeners, remember to go to deming.org to continue your journey. You can find John's book win-win W. Edwards Deming, the System of Profound Knowledge and the Science of Improving Schools on amazon.com. This is your host, Andrew Stotz, and I wanna leave you with one of my favorite quotes from Dr. Deming. "People are entitled to joy in work."

9. Romain Monnoyeur | Moving to America & Running a Restaurant