S3 E5 - Donna Knapp - Probable Cause versus Root Cause
In this episode, Donna and I discuss what's wrong with the words Root Cause and Root Cause Analysis. Our original thoughts and the differences between the old and new meanings are discussed. Undoubtedly, Dr. Deming's words matter, and we try to figure out what he would have said.
Resources:
Deming's book (John Willis mentions he is writing a book about Deming)
"Beyond the Phoenix Project" - An audio book by John Willis and Gene Kim
"Beyond the Goal" by Eliyahu Goldratt - Mentioned as an influential audio book
Ishikawa diagramming (also called fishbone diagramming) - A root cause analysis technique from the Toyota Production System
Pink Elephant conference - An IT service management conference
Suresh D.P. - Mentioned as a thought leader in the DevOps and SRE community
John Allspaw - Mentioned as a mentor and influential figure in the field
Sydney Dekker's book (title not specified) - Mentioned for its analysis of post-mortems
"Learning to Lead, Leading to Learn" by Katie Anderson - A book about Toyota's leadership practices
Masaaki Imai - Mentioned as an important figure in Lean management
Steven Spear's work, particularly his dissertation published in Harvard Business Review about "Decoding the DNA of the Toyota Production System"
Donna Knapp's textbook (title not specified) - She mentions writing college textbooks
PDSA (Plan-Do-Study-Act) cycle - Deming's improvement cycle, discussed in relation to its evolution
"Thinking in Systems" by Donella Meadows - Mentioned in relation to sustainability and systems thinking
Transcript:
John Willis: [00:00:00] Hey, again, it's John Willis It's another profound podcast. Actually, I'm going to use this time a little bit news. So I can tell you Donna as well. I've got another Donna again, Donna and like, was, if you listen to the last 1, we just, we could have gone on for hours. We had such a good time, but I'll announce to Donna and everybody else.
It revolution is going to publish my Deming book and they made that official. So. And it looks, yeah, no, it's pretty exciting to know that it's fun. It's gone off the final, like, in, in general, I'm done. You know, I know you're never really done and it's going into copy editing and there might be some finishing touches, but after 26 months of it really 10 years, but 26 months of really.
Putting the pen to paper you know, just to feel, wow, it is really, really done. And then I found my right home. I, you know, I thought about other places, but in the end, the right place was it revolution. So the Kindle [00:01:00] version will be out in the summer. And there's apparently a paper shortage, supply chain, nonsense thing.
So the physical. The copy will probably be out till the end of the year, and then somewhere in between there will be the audio version. So we're going to have to just parlay it out. So anyway, that's my announcement to you, everybody else. So, Donna, what have you been up to since the last
Donna Knapp: time? And I was going to say, you know, from I'm, like, very passionate about sustainability.
And so it is a weird thing for me, like, 1 of my. So I've written 2 college textbooks when I did a new addition to 1 of my text. They told me that it was the last edition where they would produce a physical copy, that going forward, everything would be digital. And I, from this, the sustainability lover in me was like, that's awesome.
And, but the author in me was like, but wait, you're going to print at least one of them for me, right? Like, how are you going to put it on my bookshelf? Like, how, [00:02:00] like. Well, like, the idea of not having a printed copy of my book was inconceivable to me, but so congratulations. That's awesome. Well,
John Willis: you know, along those lines, me and Jean did Beyond the Phoenix Project, audio only book was very much, it was modeled, loosely modeled after Eliot Goratz's Beyond the Goal.
I had approached Jean and I'd said, this great idea, you know, it was you know, it was only about 5 years after Phoenix project, you know, cause we're both big fans of beyond the goal. It's just, it's an amazing audio book. So we did it. We did this audio book and he talked about and he talked about making and I did that.
Actually, that's where I really 1st got serious about sort of my Deming research, you know, cause I did a whole section at that. I mean, and, and and I sort of joke periodically that friends don't let friends. Yeah. Do audio only books, because you can't, you need that. There's something about going to a trade show and signing a book and showing up at a customer site [00:03:00] and sort of the, the executives ask you to sign the book.
It's, it's there's still you know, I, I think you're right about the sustainability. In fact, we probably need a 3rd podcast now. And I think because you really got me thinking about Deming and sustainability and I haven't. You know, my queue of things I want to do, it's still a little bit deep in the middle, but yeah, I want to explore that because I think that's really interesting.
Well, good.
Donna Knapp: He was passionate about it. Yeah, no, I called it that, but he was
John Willis: passionate. No, I did a couple of searches right after the podcast just to see and I'm like, you're right. And then. You know that, you know, here we go, this is me and you get into really cool tangents, but the, you know, the thing that's, you know, fascinating about Demi is his choice of words are so important and there's, there's depth to them that like, I feel sometimes it's like you.
Once you think you understand something and you read something else, and you're like, Oh, you know what? There was a lot more to that [00:04:00] set of words that he meant. And the more and more I get into it, you know, so I think the words he chose. I think you're right. I think that you know, that sustainability was really not, you know, sort of in the mainstay as as, right.
I mean, you know Danella, Danella Meadows you know, in her thinking of systems, her original work was around sustainability and green and, and Right. That was being produced right around the time that Deming passed away. So, so yes. We'll definitely have that conversation. And same
Donna Knapp: thing with customer experience, like I did some searching one time to find out, like to try to figure out, did he ever use phrases like employee experience?
And the employee experience movement was launched right after he, like same exact thing. Like he was, it was just slightly after his death that we started using those terms. Hmm. To talk about what he's been talking about [00:05:00] for. Yeah,
John Willis: yeah, no, there's there's a lot of there there. You know yeah, it was anyway.
I think there's a whole we could turn this into the Donna John show, but, all right, so last time, if anybody was teased a little bit, you know, about the way we ended the last podcast was you really wanted to go on about root, you know, wanted to understand my thoughts and your, you know, compare them to your thoughts on root cause and root cause analysis.
And, and I promised everybody that we would. Do that and here we go. So why don't you go ahead and flesh out your thoughts about this subject? And then I'll tell you what I've learned. Cause I think we both come from the same place. We've both been doing this for quite a long time. We, you know, we've been doing this before DevOps was a term.
Yeah, our experiences go back, you know, you know, I'm not going to out you on your age, but I'm, you know, I literally started my career in 1980. Basically you [00:06:00] know, working in a large corporate and you know, so we've seen a lot of transitions and a lot of sort of management principles ideas roll in and out throughout the years.
So so I, I think, like, you, when I first heard this sort of counterculture root cause, it didn't sit well with me, but I did my homework and I, and I had some great advisors. So, but I'll let you go 1st and I'll tell you what I've learned.
Donna Knapp: Okay, so I want to 1st of all, start at the present day, and then I'm going to take you back because interestingly, last the week before last week before yesterday, I went to the pink elephant conference, which is for it service management.
Professionals it's it's put on by the company Pink Elephant based in Canada. And problem management was an important topic of conversation. Companies [00:07:00] were talking about it. And interestingly they, one of the things I got asked along the way is, how do you tie SLAs to problem management? So this is some like pretty advanced thinking about problem management, which is, Which is kind of interesting in context with our conversation, which is this pushback in the DevOps and a tiny bit the SRE community when it comes to problem management.
So, let's distinguish between problem management, kind of the greater process of managing problems and recause analysis. But okay, so now let's go back in time. So, I don't know, 10 years ago or so, I went to a conference and the presenter made the comment that root cause analysis is a complete and total waste of time.
And I remember thinking in my mind, like, what the heck are you talking about? You're just crazy, [00:08:00] man. And, he, he made a couple of comments that kind of shifted some brain cells for me in terms of, you know, things like it is typically turned into a finger pointing game, a blame a game. And I thought, okay, I can understand when you're in a toxic environment and it's being misused, which is really kind of my immediate thought.
Well, you're just misusing the concept. I can understand the pushback on it, right? And I can understand the hesitation to do it. He also made the comment that there is no single root cause in a complex system. And this mantra, like I've heard now for years, right? You hear it come up in presentations over and over again.
And I always like rephrase that sentence into there is no [00:09:00] single root cause in a complex system. Right he said, there is no root cause in a complex system and I rephrase it that there's no single root cause in a complex system. Right? And and and again. In my mind, it has translated into sorry, if you're hearing some.
No worries. No worries in the background. I live in Florida. Yes, the yard people are going through, of course, right at this very moment. So again, I kind of thought to myself, well, this is just a misunderstanding of how to use root cause analysis. And 1 of the techniques that's constantly brought up is the 5 whys.
And if all, you know, about root cause analysis, it's a 5 whys. I can sort of understand how you, you think it's. It's driving to a single solution, right? Because it is. The five wise is intended for rather simple problems. It does. It does, in fact, try to [00:10:00] drive you to a single solution. But you know what?
There's a lot of recourse analysis techniques. So I always bring up things like Ishikawa diagramming, which challenges you to think about the set of causal factors that may, in fact, be causing You know, whatever this problem is, and, you know, for years, we've talked about people process technology information, right?
So you can actually use something like Ishikawa diagramming to like, explore all those different things. You do
John Willis: a quick, you know, sort of elevator pitch on Ishikawa. Ishikawa diagramming. I mean, I know what it is, but I think I don't. Yeah.
Donna Knapp: So it's a technique that actually came out of the Toyota production system.
So it dates. Thanks. So, hence, Deming was in the mix in this a little bit, but well, Ishikawa was a student of Dr. Deming. So, so, yes, yes, right.
So, back to that time. And what [00:11:00] Ishikawa, it's very often called fishbone dialing, right? So, kind of the simple visual to get in your head is a fishbone, right?
Where there's multiple branches, And what you do is you state the problem, and then you come up with categories of causal factors. So again, categories might be people, process, technology, information, right? And then you explore, okay, all the different, and very often you use a technique like brainstorming, right?
To brainstorm all of the various causal factors That might be contributing to this problem. And then, and then typically with with root cause analysis, what you start to do is now move into you know, scientific thinking put forth a hypothesis. What do you think is the cause? What do you think you can do to, you know, minimize the impact of or [00:12:00] prevent, you know, whatever is the problem and then you conduct some experiments, right?
And a lot of, I think a lot of times folks don't necessarily understand and appreciate that the Deming plan. Do you check act cycle? Like, it's sometimes misinterpreted as you create a plan and you execute that plan and you check that you executed the plan to the letter of the. Yeah,
John Willis: that's why he actually wanted to call.
He changed the study, right? For that example,
Donna Knapp: right? Study the results. Understand, did you, you know, put forth a hypothesis? Did you prove or disprove your hypothesis and then act on what you learned? So it should have a diagram and kind of can, in fact, combine a bunch of techniques, right? Because it can, it can, you can use brainstorming as part of that, like, throw out any and all ideas, no ideas or bad ideas.
And then and then kind of go forward from that, putting forth your [00:13:00] hypothesis. And I like the technique because it lends itself to understanding that the reality, the answer might be see all of the above. Right? We need to make some process changes. We need to upscale our folks. Yes, maybe there was some kind of technological issue, but even that, you know, we might have had a technological problem because of how that technology was implemented or configure.
Right? And so how did it get that way? So you can kind of, you know, in the end. You're hopefully I use a little idle vocabulary think and work holistically. Right? Hopefully you can understand where all the causal factors that could potentially help you right to overcome this problem. So, but ironically, hold on, I'm going to get something.
John Willis: So this is not another book where I spelt Deming
Donna Knapp: wrong. No, no, no, no. This is actually my book. Interestingly, so when I started, [00:14:00] this is in the spirit of full disclosure. So, when I started going down this journey, I looked at my book to see how, in my book, I defined root cause. Where is it at? I can't find it.
And I defined it as the singular, definitive, definitive, Causal factor, which if eliminated would prevent the problem from occurring, which is a very historical definition of root cause, right? If you actually look at if you were to Google root cause right now, that's probably what it said. And I learned, you know, I learned a lot of this stuff, like really early on 1980s.
You know, my roots are in lean, lean, lean, lean. And so that is a very historical definition. So I like kind of cringed a little bit because it's like, Oh, geez, I'm walking around preaching. There is no single root cause in a complex system, but this book is in revision. So the root cause will be different in [00:15:00] the new edition.
And, and so let's move forward. Like I thought, all right, well, we just have to revise the definition, right? We have to move off this definition that, and you know, Deming actually talked about this a lot. He Another phrase he never really that I have found used is the concept of silver bullet, but he abhorred the concept, right?
He abhorred thinking that there was like some magic solution. That you could come up with that would solve everything. You know, he, he very much was about system syncing and understanding that the organization is complex and you have to understand all its parts. So, again, I still was in the mindset that, like, okay.
We just need to understand that there is no single root cause, and we need to understand how to there are other analysis techniques and the 5 wise and and and so let's, you know, again, think a little more holistically. So then, just last year, I [00:16:00] was having a conversation with Suresh D. P. Do you know Suresh?
So, thought leader in the DevOps and SRE community. And we were talking about a concept in SRE where you move from reactive to proactive to preventative to predictive incident management. So, I asked him to explain to me, to give me a good example of how he distinguishes. preventative versus predictive.
And in the course of explaining that to me, I had this little light bulb go on. And I said to him, like, wait, can we talk about root cause analysis for a minute? And he was like, wait, where did that come from? But I think the insight I took away from that conversation was Okay, so maybe the DevOps and SRE community see it as inherently reactive, root cause analysis.
Maybe they see it as just inherently reactive because the [00:17:00] incidents have already happened, right? And so we're trying to figure out why they happened. Or even when we talk about proactive problem management, right, which is a concept we talk about in I told, for example, a lot of times the examples that are given are to, you know, through, you know, trend analysis, and we're talking now about I am being able to use a I.
In order to kind of better and, you know, earlier in the process, understand trends, right? And understand and be able to start to predict that an incident might happen. And so what can we do to prevent it? And I thought, well, okay, that's an interesting concept, right of of understanding that at the end of the day, can we introduce better design practices?
Can we introduce better testing practices? Right? You know, what are all the things we need to do as part of a shift left concept [00:18:00] in order to prevent incidents from occurring? So, okay, I can understand the thinking that it's inherently reactive, but, and so maybe here's where I am now, but that's an ideal, right?
That's, that's an ideal, like, in a perfect world, could we introduce practices that, you know, build quality in, right? It's really what Demi preached, right? But can we build quality in? And the answer is yes, in a perfect world we can. We don't live in a perfect world. Incidents are still happening. Major incidents in particular are increasingly visible, right?
They can significantly damage companies, reputations or cripple companies, especially as we've moved into this digital world, right? We're where we're so dependent on digital technology, so I'm still in a place where I feel like it would be irresponsible. Not [00:19:00] to use root cause analysis to try to figure out what happened there and what can we do to prevent it now.
So let's go back and just say, but use it, right? Use it in a blameless way, become skilled and using it the 3rd way, right? Mastery through practice, understand that there's lots of different techniques and the more you use those techniques, the better you can become at using them. And then and then, you know, optimally look at not just what's the short term fix that we can apply.
Yeah, that in the end might translate to a workaround. But really what behavior changes can we make? What cultural changes can we make? What process changes can we make? What skill related changes can we make that all would translate into. You know,
John Willis: so I [00:20:00] agree. I agree. But here's the thing, right? And I think this goes back.
Donna Knapp: It's like, okay,
John Willis: let's go. Yeah. So, but I think here's you go back to Deming, right? Like, words matter, you know, like, I don't know when he decided to. So, I mean, I won't give the whole diatribe of the history of P. D. S. A. But basically, it was, you know, basically, it was a short and when he, when Deming introduced in Japan, they called it the Deming wheel.
Ultimately, Ishikawa and his crew called it PDCA, like, don't quote me exact here, but and then there was some lag where Deming decided that the C was wrong because he felt words were important. And that's why he wanted it. His vision of it was PDSA. It was being
Donna Knapp: misunderstood.
John Willis: Yeah, well, because the word change just, you know, like, to him, like, the wrong word at the wrong time, you know, or the right time, the wrong time [00:21:00] is, is really important.
And I think that even if, you know, like. You know, I tried to do a little bit of a search on Deming and the word's root. Certainly there's tons of he was the, you know, 5Y creator. I have some thoughts on, on the difference between culture in Japan and what, what sort of Root means versus what root means in America.
The point is the, I think the there's 2 problems. 1 is there's the, like, everything old needs to be thrown out because we, they didn't understand and that's nonsense. Right? And like, it's not even look for some of the incredible valuable tools that exist for over 100 years at least maybe more, but certainly since Deming or Schuetz, But, but I, I, I do agree that sort of rude is the wrong word.
And so that's where it boils down to, like, I think, you know, like, [00:22:00] okay, John simple fix called probable cause. Because that's basically what it is the issue that happens and you address some of these and I, I did a little bit of a sort of a list, but and again, I think there's this great blog and I'll post it.
I serve at somebody who worked for. It was like, hospital care and, and and it was explaining about, like, 1st off the metaphor of a route. Social systems aren't a plant, right? Like, you know, they, they you know, they, they are systems. They have feedback loops. They're, they're virtualists or, or, or vicious cycles.
They are reciprocal, you know, a causes B, causes C, A can cause C, right? So you have all this, so, so when you get into the people who think about like new thinking and incident complex system systems thinking, that word root has so much baggage to it. Right and what it what it leaves, you know, 1st and foremost, I think that the, the, [00:23:00] the true people, you know, what happened to me was, I was like, you're all crazy.
I've been doing real analysis. We're on a life right now. And then the people that started arguing with you were brilliant people. And I'm like, okay, you know, right? And I need to go sort of desk check and I, you know, I spend a lot of time. In fact, so 1 of the stories was in the DevOps handbook. This is probably not known.
I don't know if they ever told this story. John Osborne had asked. Oh, no, that was situational awareness. That's another bag bag of worms. But all right, but it was very similar, but, but I, I, There was a lot of conversations I've had with John Osborne has been incredible mentor to me brilliant, brilliant person, incredibly human, you know, just so many things that are so awesome about John Osborne.
But you know, like, I started listening to his thoughts. I'm like, well, he knows what he's doing. So let me let me go start reading, you know, and I think if you read his master's thesis, you know, the things [00:24:00] that you'll probably pop up 1st. The is the, if it, it, unfortunately, in sort of the American or Western culture version of the root root cause analysis root cause doesn't enable learning the way it's it again, it's the word, like, somebody like, you could use the word root cause and root cause analysis all day long.
And I'm. Certain, and I know you're doing the right practices. Unfortunately, in mass, the, the word sort of simplifies the problem in that, you know, it implies that there's a single you've already pointed this out. There's a single cause it could be oversimplify and, and, and again, not taking account wanted them in core principles appreciation for a system, right?
Systems thinking, you know, and then you get into the whole You know, misidentifying or mistaking symptoms for causes, and then we get into the more detailed [00:25:00] special cause, common cause thinking that, you know, like, a real voice could be, you know, the sort of I sort of my quote is the misinterpreting.
Variation is the cause is the cause of all evil, right? So the right, like, if you misinterpret a special course for a common cause or a common cause for special cause, right? Right. So, I think that that's the thing is, is that word has too much baggage to a new culture. Who's trying to who's who are growing up with systems thinking non linear thinking and it just, you know, they've seen too many examples.
Of deterministic, you know, like, you know, again, I know if we could all go back, I, I dared. And I'd be scared to death for anybody to read out loud something. I wrote, you know, 10, 15, 20 years ago, right?
Donna Knapp: That's why I like,
John Willis: you know, well, anyway, I have a good, [00:26:00] I knew you would not judge me. No, no. And well, in fact, you know, there's a, there's a great story I heard about Dr Deming where somebody had asked him, somebody was in 1 of his seminars and I don't know if I said this last time.
They were Dr. Deming, Dr. Deming, I saw your seminar 3 years ago and you said, you know, X is X, X, Y, Y. And he said, I will never apologize for learning, you know, so I think that's what it comes down to. And so I think the other thing, too, is I, you know, I, I still. The reason I like probable cause, because then that fits in line with Deming, Ishikawa, you know, all the sort of follow on work, Duran.
I think they all would agree that, you know, that these are probable, probabilistic, you know, statistical process control is a system for probabilistic identification or analysis. Deming, You know, talked about the difference between analytical statistics and enumerated statistics, right? So I [00:27:00] think to me, and then I, I think you're right.
The other thing, too, is, you know, if you look up the definition, or sort of your read the quite the, so even to almost 5 wise, right? I want to believe that he didn't really mean five. I think he meant luckily, maybe by the time you start the problem by asking questions that like, that's not an any pattern, any universe or any scenario.
I think the fishbone stuff is just stuff that just doesn't get you just like SPC. Does not get used in our industry nearly enough, right? There's always this like, oh my goodness, you know Six Signor is terrible. I mean, I just got an argument with Steven Spear, Dr. Spear recently about how terrible Jack Welch was.
Well, Jack Welch was not a great leader. I worked at GE Capital. And I did to make my wife was a black belt green belt, and we work [00:28:00] like scientists. Now, it was very dramatic. And, you know, I think they went overboard on it. Also, in the early 90s, I don't many corporations that were priding and making diversity a 1st class.
Principle he was you know, they're managing modeling. So this, this idea that it's either black or white, right? It's either root cause is like that person who stood up and said. You know, this is a total waste of time. Well, let's let's be careful about our choice of words. You know, the other thing I've done a lot of research on is operational definition or operationalization.
The thing we're terrible in our history is not defining an operational deficient. So what is root cause mean? What is zero effect mean? Because I think we're going to be having this argument. About zero defect in the future i think right now opening people and this is one other point i want to make back in the day when me and you started it was the wild [00:29:00] wild wild west today it's just the wild west right and we had no tools.
Right, we and so things like idle popped up and service management and we started embracing tools and then, you know, as time goes on, it's easy to look back and say, oh, that's terrible stuff. But but there's an evolution here, so the notion when you wrote that book, and you talk about root cause analysis, that was what the world and it needed to hear right now.
It may not be right today. In a world, it's far more complex, right? So so I think that again, the, the words matter, but what, what really matters is what is our operational definition. So right now, zero defect or zero trust, sorry, zero trust.
Donna Knapp: I was going to say zero defects coming back. Cause I know, no, I,
John Willis: my bad.
Maybe I'll have to erase that. I meant to say,
Donna Knapp: who coined that term.
John Willis: I'm trying to, but zero defects. [00:30:00] Yeah, I don't I didn't want to even that was
Donna Knapp: a guy who coined that term later in his life actually wrote about the fact that he read the day that those
John Willis: words. Oh, really? Brilliant. Brilliant. Well,
Donna Knapp: I think he said people lost their minds.
Yeah, people lost sight
John Willis: of. Yeah. Well, let's see. This is the problem with records, right? I think it's the same thing and what I meant to say, and I'll say it like, I'm sorry for anybody to throw this off course. I meant zero trust. Zero trust. Zero trust right now. I think there'll be a time when somebody's going to look at it and say, what the heck were they talking about?
Zero trust. There's no such thing as it. Okay. You know, a 0, like, you know, like, again, these, these sort of deterministic root or deterministic 0 defect or 0 trust or right. These are not realistic. Things that exist in the physical world or social physical world. So [00:31:00] so
Donna Knapp: unless we needed slogans, and that's an exit
John Willis: or or minus clear operational definitions.
Right, then we get into this sort of like arbitrage of, is it horrible? Is it great? Is it work? Does it not work? Should we throw everything out? So so at the end of the day, I think that the word today, a root and root cause analysis is not the proper way to describe it. I don't know if I'm singly going to change the mind share of everybody.
We should probably be calling it probable cause because everything that we've learned to date. Okay. Is that the best we can do is make probable, you know, these systems are so complex, you know, and going back, you were, you were talking about the, the, the question between preventative and predictive. I think it's the same thing.
It's a probability. Preventive implies deterministic. I can stop this from happening, right? Complexity demands that [00:32:00] it's never, there's this notion of the arrow of time that it will never be exactly the same, you know, John also has this great thing that he saw. I've seen a presentation. So he loves when some, he sees a developer and it's not any push against the developer.
It could be anybody puts in a log seen this before a transaction. It runs like. A billion times an hour, maybe a minute, like, no, there's no seen this before. They're like the, the, the, the, the way these systems work and the dynamics between the humans and the complexities and the transactions in the systems, the networks, the network delay.
I mean, like, we go on and on that. That nothing, everything is going to be new. The best we can do is be probabilistic. And again, this is why I'm such a big fan of Deming. Deming's analytical statistics is sort of underutilized in our, in our [00:33:00] world. And it really is, you know, what Sheward decided was the only reasonable way to do quality analysis in manufacturing and Deming took that to all other areas.
is to use statistical, analytical statistics. You know, so I, I think that so that that's why I come to the world where I basically agree with, you know, all of the, the sort of the, the new discussion about root cause the where I disagree is to wholesale, say, everything that falls under that umbrella or anybody who has prior art knowledge of how to solve problems.
Is useless,
Donna Knapp: right? Well, and I, you know, I think what I like about something like probable cause analysis is it inspires experimentation, right? It's like, all right, but let's understand that. We now still have to prove it up, but there's another, [00:34:00] there's a quote in I don't know, 1 of the courses that says.
Root cause is the point at which we stop learning. And I thought, okay, that's interesting. It's an interesting perspective. If we're chasing the shiny bullet, right, which we tend to do. Americans in particular, we love the shiny bullet. If we're chasing the shiny bullet and we, you know, we come across the easy answer.
Let's point the finger. That's where the problem lies. But the question I would always beg is, can you prove it right? You know, even if we go back to my original definition. Then you should in fact be able to eliminate that and prove up that by eliminating whatever
John Willis: that Yeah, I think that's where we would disagree.
I, I think in complex systems you cannot eliminate. Right now you can, you can. So all we can
Donna Knapp: do is experiment.
John Willis: That's right. We can arbitrage and we can basically create better [00:35:00] practice. We, it's kaizen like basically the answer is, and I guess that's, you know, that the assumption is that we will continuously improve.
So, but we have to move away from, you know, the example that so 1st off in Sydney, Decker, Sydney, Decker has a book. I, I, I, I try not to recommend it because there's some things in there. I don't believe I agree with. Socially or philosophically, so I'll just leave it at that, but but 1 of the things he does in the book, which is basically fascinating.
It is the single I think I'm going to write about it in a way that I can point to my blog article and not attribute the book, but what he what he does is he takes the the, the postmortem that is done by Valery Legasov, which is the guy who was basically doing the postmortem on Chernobyl Scott Snook, who did the postmortem on Black Hawk Down and the Columbia Accident Investigating Board, and [00:36:00] all of them were trying to look for a root cause.
I mean, the actually Scott Snook's quote was the fate of the army or the Air Force or the military is at stake if we can't basically pin this on somebody. There's the black, right? And all of them came to the same conclusion that it was basically post cold war policy enforcements that caused all 3 of them.
So the point being what I would right? It's the system. And so what I used to use to say, here's why. A group of people don't like root cause because it can lead to a false identification of solving problem. A patient dies in a hospital. So we, we do some form of why we find out that the patient was given the, the certain injections of I'm totally making this up of needles in the wrong order.
Okay. We solve the problem. It was a bad need, but needle. So we fire the the nurse, [00:37:00] right? And then so what happens 6 months later, a patient dies in hospital. We find out looks like the same problem. Maybe it isn't the nurse. Oh, it's a diet. It's the it's the dietitian switch to table. It flipped it around and so when the nurse came in, they took the 1 on the left, gave it right.
So we fired a dietitian then 6 months later again, patient dies. We realize, oh, my goodness, it can't be the dietitian. We look and we find out that it's that the pharmacy to cut costs basically start stop color coding the needles for cost, you know, and you might think, okay,
Donna Knapp: that supports that notion that repose is where we stopped learning.
John Willis: That's right exactly. That's right. That's right. We could make fatal fatal mistakes. Right, especially in in, in, in safety, critical. Environments but the you know, I think that the thing is that, you [00:38:00] know, that this idea that you just, you know, we can improve, but like the, and, you know, and I think the other thing when I went back and I looked and I said this earlier when I looked at like, and he does say root cause and root.
I found this really good article about how. You know, when you think about sort of a Western culture blame and, and the sort of the the Japanese culture on a holistic, what's good for the company again, I think they're the idea of, like, a 5 why they're where, like, blameless this. It's sort of built into the code.
I won't say that nobody in Japan blames anybody, but there tends to be a more intrinsic motivation of for the company for the whole. Right to serve a Western. And so so people are less likely because that's the other danger. I think the reason that you'll find the [00:39:00] safety critical people and the resilience people be very vehement against the term route is they believe people die.
Right. Because, because, like, you know, the, the, the idea of a root cause implies that you found a root actor, and in which case, in an environment that sort of behaves that way, nobody's going to tell and be honest and say, yeah, it was me. You know, I, I'm the one who flipped the table from fear of retribution, fear of getting fired.
So, so the whole notion of us copying some of the Lean, ultimately Toyota, techniques in some ways is misplaced. Because the Western culture is completely different than an Eastern culture. So, well,
Donna Knapp: Deming said something to that effect. 1 time, he said something about. Americans wanting to copy, but we don't know what we're copying or something.
I can't remember. There's
John Willis: an, I don't know. He spoke to that, but. There's [00:40:00] a quote, it's not, I have to dig as it was, but somebody asked, I don't know, why do you let all these American companies come over and sort of watch and copy our process? And he said, basically, because they can copy our process, but they'll never understand our culture.
Donna Knapp: Right? Yeah, but I think. So that would where are we 3 minutes to go. All right. So that brings us to the whole conversation about leadership. Right? And I think I think part of this, the challenge is then for leadership to understand 1st of all. How to appropriately apply these techniques and to coach people in how to use these techniques.
You know, if you think about Kata, there is a coaching Kata, right, which teaches leadership how to coach people on continual improvement. So, I think today, what we need is leaders and I'm not talking executive leadership. I'm talking leaders at any [00:41:00] level in an organization, really understanding how to appropriately use these techniques.
Maybe how to shift the vocabulary. I think that's a fair statement. If we need to shift our language a little, that's okay. Common language is. Having, having an effective common language is always a, a, a, a, an important first step in shifting culture. I believe it's why we adopt frameworks, right? It's adopt, it's to have a common vocabulary and then, and then the big challenge becomes for leadership to then empower people and trust people to take action.
Based on what they think is wrong, not them make a decision. Okay. The easy out is to fire that guy or that girl, but to actually empower people to make changes in their systems day in and day out, that could lead to better ways of working. I think in some organizations they're getting there. In other organizations, we're still not quite there yet.
Right? We still, certainly in enterprises, we still see a little bit of [00:42:00] command and control. And and that and that has to shift as well, right? Because I can understand the other reason I can understand people not wanting to use techniques like, let's call it problem will cause analysis is because they actually take the time.
To get together as a team to collaborate to come up with hypothesis, maybe even to conduct some experience to come up with what they think the solution is. And then leadership says, you know, we're not going to spend that money, or we're not willing to change the policies or the processes in a way in the way that you're recommending.
So, good job team. Let's move on to whatever is the next hot big. You know, issue that has their attention, so, you know, it, it always, I think a shift in mindset and language and how we talk about things and how we apply things. Comes [00:43:00] down to leadership. Oh,
John Willis: yeah, totally. And well, leadership definitely and I mean, that's.
You know, that's a, you know, a Devin cornerstone. Right. But yeah, words, I mean, words matter. And sometimes have to get changed just because they have to get changed again. The word that was appropriate at the time work may not be appropriate now or in the future. Yeah, and I think the other thing too, is that, you know, like, you know, the reason I like doing podcasts with you, because there's just, there's a lot of knowledge.
In in sort of your world, right? You know, service management. I told that sort of gets looked at from the new world of like, oh, that's all nonsense. Right. And so you know, and I think to me, you know, like, again, I've got Deming on the brain, but if I look at where like, Deming got his knowledge from, and it wasn't just sure it was pragmatism.
It was it was early physicist. It was, you know, it was, you know, it [00:44:00] was sort of the notion of quantum of the world being completely different probabilistic. I mean, he was all about probability, statistical, analytical statistics, and then what came after him like Ishikawa and. Yeah. Yeah. And I, you know, if you haven't read what Katie Anderson's Learning to Lead, Leading to Learn, it's an amazing book.
And it's, it's, she she befriended a gentleman named Yoshino, who worked at Toyota for like four decades and tells four stories. This is somebody who actually worked at Toyota. It's awesome. I mean, God bless all of our mentors who taught us about Toyota, like Rother and, and Spear, but like this guy actually did four.
And one of them was a colossal failure. And two or three of them, one of them was say, just okay. And two of them were amazing success stories. But he points to a guy named Masayo [00:45:00] Nomoto, who he says. Is and I wrote a blog article about this, he says, is was as important as and in some ways more important.
Guess who, when you read Masumoto's book, the number 1 person he talks about that you should read before you read his book is Ishikawa. And Ishikawa was part of Juice, who basically was baptized by Deming. So, it all goes back, I mean, I don't know how many times I have to say this. It all, if you think your your discipline is lean and you think Toyota is great.
Like, let's be really clear this stuff and not just came from Deming. It came from pragmatism. It came from physicists. It came from, you know, just a lot of thinking that was happening in the 2nd, scientific revolution. But, and we can't, you know, like, to me, you know, my sort of mission these days is to bring all that to the forefront [00:46:00] to say, this is.
What we all inherently know, like, when you, when you tell people system profound knowledge and you explain it, even the newest of the fanciest, you know, pipe hitting you know, K native function coding, lambda coders. Oh, my goodness. I get it. We have to sort of be careful about how we do our dialogue so that when you talk at a forum and have incredible amount of information.
People don't dismiss you because they hear you say, you know, I'm just using you as because you,
Donna Knapp: yeah. Could you use this? What
John Willis: they don't as a phrase, we don't, we, we, we make the first cardinal sin that everybody agrees in. We don't learn, right? So yep. There you go. No, good. I got it. Probably we analysis, we solve the problem, we solved it,
Donna Knapp: but so.
Just to make sure it doesn't go unsaid need it. It's important. Learn to use [00:47:00] it properly. Use it to learn to experiment and learn there's. There's
John Willis: an incredible amount of intellectual property built around problem solving incident problem, understanding, complexity, change management, all those things that are that that just incredible body of work that should just not be dismissed because.
Of the word route,
Donna Knapp: right? And actually, if you read about Toyota, it's often said that what Toyota was trying to accomplish was to create an organization of problem solvers. Yeah, so much of the was about teaching people how to learn and teaching this elite group of leaders how to solve problems and then dictate to everybody else what to do.
But every single individual in the organization day in and day out. How to how to solve problems life skill,
John Willis: right? If there's some variant, I don't ever get the quote exactly. Right. But in Steven Spears, you know, [00:48:00] basically his dissertation, which was published in HBR, the decoding, the Toyota, the coding, the, the.
T. P. S. D. N. A. or someone D. N. A. and the quote isn't exactly like this, but it was something to the order of Toyota was a community of scientists continually experimenting that's that we want to call that root cause. You want to call it probable cause you just want to call it analytical statistics. Right.
I don't care, but like, that's what I want my leaders to
Donna Knapp: basically. It's a great skill set. Yeah, it's, it's a great skill set. It's a great mindset. All right, Donna. We got to start with the little kids today too. Yeah, there you go.
John Willis: Yeah, maybe we'll get a breather and we'll do the sustainability next time. So.
For
Donna Knapp: sure. All right. All right. Thanks, John.
John Willis: Sure.
