136: Jurgen Appelo

Listen to full episode:

Joe Krebs speaks with Jurgen Appelo about Unfix, a pattern catalog for organizational change and agility he created. Learn about the motivation of patterns over frameworks and how they are being used to create your own “buffet-style” solutions.

Transcript:

Joe Krebs 0:10

Agile FM radio for the Agile community.

www agile.fm.

Welcome to another podcast episode here of agile fm, I have Jurgen Appelo creator of unfix, which is a topic we want to talk about here today is unfix.com. That's where you can learn more about this topic. But we want to talk a little bit of what unfix is, where it came from, how old it is, how new it is, and what it can do for organizations out there. A super interesting pattern which I which is important. We want to explore what patterns are everywhere and also talk about what unfix is not. Welcome to the podcast for you. How you doing today?

Jurgen Appelo 1:05

I am great. The weather is awesome here in my in my city in Rotterdam in the Netherlands. Looking forward to the trip tomorrow to Lima, Peru, which, which is something that I have been looking forward to quite a few weeks already. So longest trip and those are nice to have every now and then. And yeah, lots of things happening.

Joe Krebs 1:29

Lima is are they interested in unfix? Or is this for pleasure

Jurgen Appelo 1:32

of course. That's what I'll be talking about. That's my keynotes agile, lean agile event. Yeah.

Joe Krebs 1:41

All right. unfix is not another scaling framework. It's not a method. It's not a framework. What is it?

Jurgen Appelo 1:49

It's a pattern library. That's that's how I call it. There are other pattern libraries such as sociocracy, triolo, and team topologies. And, and so on liberating structures, they are not frameworks, because you don't install them. That's the idea of a framework that you have something to implement. And then you can validate or verify that you did the implementation correctly. You can certify people with in the implementation roadmap, that is not what you do with a pattern library, all of the suggestions are options, there's nothing mandatory with a pattern library. So the best metaphor that I have is Lego. There is not a single block in the Lego box that is mandatory for you. All of them are optional. Some of the blocks are more obvious, then the others, so you will use them more often. Maybe nearly always. Some are more rare for special cases, but not a single part of the Lego toolbox is is mandatory. And that's that's how I see pattern languages. That's the real word that specialists use sometimes Pattern Language. Yeah. And yeah, that's that's what the unfix model is, as well.

That is interesting, because in lego, a round shaped kind of piece could be a wheel on the car, or could be a pizza on the table. Right?

Exactly.

Joe Krebs 2:49

Creative creativity here, right? It's also interesting, because you are, sometimes you build with LEGO not that I have worked on with Lego in a long time. But you could build a house, you can build a street of houses or like a road or alignment, you could build a city. You know, there are some exercises out there in the Agile community where things are being built in isolation and put together there are there is a guy called Christopher Alexander I was exposed to, in the beginning of my career with is an architect where I'm nothing really in the Agile space, but he has influenced a lot of people in that how did how do these people like Christopher Alexander or Gang of Four, and others, there's many, many people out there in the community. How did they influence you? Or unfix?

Jurgen Appelo 4:10

Yeah, I have the book actually here, one or two meters behind me the Pattern Language of Christopher Alexander where he published, I think in the 70s or something. He was the first one to recognize the benefits of micro solutions, small solutions to known problems that you have to combine to come up with larger custom made context dependent structures. And that is what cities are. So in the book Pattern Language, you find the public square as a pattern. Anyone knows what a public square is. You have public squares in New York City. I know quite a few famous ones. We have public squares here in Rotterdam. But the cities are completely different. Same with the the promenade as a pattern there are promenades in, in New York and also promenades in Rotterdam, and so on. So this book has 253 patterns that is quite a lot. But then it's up to you as an urban planner, a city designer to come up with ways of combining them that makes sense, within the context of the city, because some cities have mountains, others have lakes and rivers, and whatever, you have to work with the environment that you have. But then, within that environment, you're gonna use the familiar patterns that everyone is using that principle, you also find in while you mentioned it, design patterns are the Gang of Four and then book came out in the 90s, where they came up with familiar patterns in programming, the facade, the singleton, the model view controller, I'm sure many programmers listening to this know what I'm talking about. And it is up to you as an architect to use those patterns and combine them in any way you want, depending on what the software is supposed to do. The interesting thing is, I remember back then that some people implemented all those patterns as a framework that you could literally buy frameworks, like the dotnet implementation of all the patterns that you could then install on your computer. And, and I thought, that's, that's totally not what they meant. With the book, you should not turn those patterns into a framework that you can then install, because you're not supposed to do all of them. You only pick and choose, depending on context, what you need. And I think that is my main problem that I have with frameworks in the Agile community, where you have these rigid structures where something needs to be installed, like well, let's name the big one, the Scaled Agile Framework SAFe, they literally call the smallest version essential safe, it is in the name itself. That part is essential, it is mandatory. If you do not have an agile release, train, you do not have SAFe. So you must have an art, you must have PI planning, you must have quite a few other things that those are together the framework that needs to be installed. I do not believe in that approach. I do believe that the frameworks have lots of good patterns in them. But we have to break them down. We have to decompose them deconstruct into the smaller building blocks. And then let you in your organization, do the recombination, figure out how to combine the different patterns from different toolboxes SAFe. LESS team apologies, whatever. They all have practices that you can combine. And that's what I tried to do with the unfixed model. I just borrowed the good stuff that is already out there. Just as Christopher Alexander has done, cities existed before the book, surprisingly, good organizations that do common sense, good stuff already existed before unfix came out, I just capture the good stuff, I give it a name, I give it a visual say, well, this is what we've seen, that seems to make sense as a micro solution. We add it to the box, the little box as one of the options. And then you take it out when you think you can apply it. And the box is getting larger and larger. Because we need more options, so that you can build more stuff with the, with the pattern language.

Joe Krebs 9:02

So this is very interesting, right? Because what you just mentioned about the essential piece of in your example was SAFe, but we could probably take any, any other framework as well. Right? But when we're looking at the essential piece that does not consider the environment you're in right. So we're coming back to Christopher Alexander, he does not see that. What is what is the environment you're in? What's your view of mountains? Do you have lakes Do you have how do we build around it right, you come in with the essential piece and it might not work for that environment right to have a little bit more of a flexible approach I think that is that's a good point now is unfix like buffet style, is that what people are they have to see is like there's a collection of patterns and people go out and says I'm gonna grab this I'm gonna grab this and grab this and I get confidence in the individual pattern, but I need the skills to combine them that they make sense together

Jurgen Appelo 9:55

exactly. I like the metaphor that you're using buffet style that might make it harder to sell things to people because I'm making them do work, I have to convince people that they have to do the thinking themselves don't do just a stupid implementation or something off the shelf. That is not going to work, you have to do your own thinking, according to your context to make things work, interestingly enough, I just read a couple of weeks ago, in a very different context, the scientific results of research into body weight, or body loss or body weight loss, what is the weight loss, weight loss? Weight loss programs?

Joe Krebs 10:47

Weight loss? Yeah,

Jurgen Appelo 10:48

yeah, that was the term I was looking for. And the evidence is in none of them work. None of the standard programs work. They already know that there is scientific evidence that the only thing that works if you create your own program, out of the common sense suggestions that are captured in all those other programs, the standard programs out there, but it is so context specific, a weight loss program that you have to customize it to who you are, what kind of body you have, what kind of lifestyle you have, et cetera, et cetera. So the following any standard program is, is is going to it's going to fail. Yeah, and that is the equivalent of following a standard standard framework, it's not going to work, you have to break it apart and use the individual components good. There is a lot of good advice in there is just the whole package that is sold that you have to get rid of.

Joe Krebs 11:54

So some of the listeners, not fully familiar with with unfix might now think, throw everything out and use unfix that's not what you're saying. Right. So this is also important, because we are talking about SAFe and possibly other frameworks here right now, that does not mean that unfix is replacing these these kinds of things, right? How would they be? How do these coexist? And how to how do you envision to go into an organization say, hey, we'll, let's say there's an organization using a framework of any kind, but as unfix pair up with that approach.

Jurgen Appelo 12:33

Now, yeah, I would like to help people stop thinking in terms of frameworks. And how do we implement it well, some suggest that there could be good starting points for customization. I think the jury's out on that argument, I'm not fully convinced that they are a good starting point, I think starting from scratch might sometimes be easier than starting from a framework implementation and then adapting it. But let's give them the benefit of the doubt people who have a framework and want to customize it, they could look at what the unfix model offers in this, and then see what else is in the Lego box that we can use to start changing this implementation that we have here with continuous improvement and, and small step experiments, to turn it into something completely different. So it would be similar to beginning with a standard weight loss program, but then realizing on day one, that there's not going to help, you'll have to you'll have to change the exact diet, the exact exercises and so on to start making it work for you. That I can agree that that might my approach.

Joe Krebs 14:02

Yeah. So So that's, that's interesting. Like I for example, I teach a lot of adaptive org design courses, for example, how organizations shift, make the switch towards agility and kind of things to consider. There's a lot of talk about self management and self organization, obviously, in these courses and how to get to states like that does unfix if somebody listens now more from a leadership and managerial role, prior to this podcast is this unfix demand like a full self organized self managed like it's a radical name, right unfix? It's provocative, nice, nicely provocative, right? And it makes you think, does that also mean like we're going to the extremes with adaptive org design? How does a company steer that transition? Not necessarily top down structure, but even like any kind of structure in an organization how does unfix change that?

Um, I think What I want to achieve here is to plant a flag on the horizon and show people, this is the direction that we want to be going. And I don't expect you to be here tomorrow. But I want you to move in that direction. And that direction is being a networked organization with a fractal or design. There are some companies, not many, but a few that have evolved quite far in that in that direction, famous one is Haier, the Chinese company where it was 10,12 years ago, where they reorganized themselves into a network of 4,000 micro enterprises was a 4,000 tiny little companies that that collaborate horizontally without big fat middle management layers, no matrix structure, whatever. And they are incredibly adaptive, they are super fast in in responding to opportunities. For example, when COVID hit the Chinese economy Haier was the was the first one to start making masks or a face mask for half the country, basically, because there was an opportunity, and they could respond incredibly fast to this to this new thing emerging. While normally they make vacuum cleaners and and and fridges and whatever, but they switched to face masks short, why not? Why not? Yeah. And so this is a very inspiring company as that shows what you can do as a network company instead of a hierarchical matrix organization. And that is, as I said, the flag on the horizon that I want to offer people try and go in this direction. But I do agree that it is a step by step thing, you'll first want to move into the adjacent possible as some complexity thinkers would would say, you open up opportunities in a new direction, by making small steps, and that unlocks other doors, and then you go through that door. And if something doesn't work out, you make a step back and you move in another direction. I'm sure that is also what Haier has done that local experimentation before they did the big radical change of firing all the middle managers basically, because that was quite a revolutionary thing. That was Kaikaku not Kaizen them at that time, but I'm sure they did some Kaizen before they were sure about the big step they wanted to make. I keep telling everyone started with small experimentation. I have already 150 patterns in the in the pattern language in the model, and more are coming. And there's plenty to experiment with very small things that will harm nobody. So just start playing, get some experience. And then when something seems to be working well, you could make some more radical steps. With your org design,

assemble your city, right? Build your city like start small somewhere, right?

Jurgen Appelo 18:14

Exactly. And by the way, it's not only about organization design, about crew types, Team types and so on. There's also decision making methods. Also about goal setting patterns are coming out in the next couple of months. So more advanced version of OKRs (objecetves and key results) basically the whole OKRs and MBO (management by objectives) KPI stuff, I have deconstructed into patterns. And that's going to be awesome, I think,for people to playwith and make their own OKRs like approach with the individual patterns that we're going to offer. So yeah, organization structure, business processes and collaboration. There's lots of different angles on on the pattern library. You can start anywhere, whatever is the lowest hanging fruit the smallest pain that you can address. Start playing like with a Lego box. There are 4,000 different types of Lego pieces. Did you know that Joe 4,000 Neither did I. Yes, that's quite a lot of options that but nobody starts with with 4,000 pieces on the table now doesn't that doesn't make sense. You start with a subset of the more obvious ones and then you will dig into the rest later on with when you gained a bit of experience.

Joe Krebs 19:44

Before we go into one of the maybe we can explore one of those patterns is one thing I noticed and I just want to follow up on this because we just talked about you know, leadership etc. and organizational change. These are this is a bottom up kind of approach, right? And you just mentioned like some form of middle management and that was reduced or removed. You're not saying unfix we're not have any managers or leaders? And I think we were very clear about this. What is the role of though of leadership? If it's a bottom up movement? How can leadership support unfix? within an organization? If we're seeing on one side, there is some form of streamlining going on within an organization, which I think many organizations would benefit from, as well. Right. But on the other side, it might be the the the question of a leader that says, I don't know what my role is, in all this. How can I support unfix to make the organization a better place?

Jurgen Appelo 20:46

So well, that's where my previous work on management 3.0, comes in, I always say manage the system and not the people. And the very same thing, I suggest with the unfix model, where we have the governance crew, which is the team of chiefs, Chief Executive Officer, Chief Information Officer chief whatever. At the base level, and the base is what I'd hire would be the micro enterprise, some might call it a tribe, a self sustaining business unit, whatever small units of, of maybe up to 100, people maybe a little more, but not much, much more than that, that is the unit that we're looking at, that should be autonomous self managing, with a very specific business model proposition that they offer to either the world outside to customers or users or to other parts of the network in within the company. And that unit needs to be managed, that someone has to take responsibility for the success of that unit. All of those 4,000 micro enterprises hat Haier are managed by a chief. And actually, the fascinating thing at hire is that if the chief is not doing well, for three months, there's an automatic re-election triggered, where other people can volunteer to take up the role of the chief and see if they can do better that is interesting. But I will not go there yet with my suggestions because that makes some managers very scared of their of their job. But the fact is, the unit needs management but management of the unit of the system, how does that system work because you need to put some constraints in place on on how those 20 50,100 people in that base, collaborate with each other. And we want to make and that's what the name is about unfix we want to keep that unit as flexible as versatile as possible with its organization structure. That means that you should try and not have managers on teams. So no manager on a scrum team, no manager on a team of coaches, no manager on a platform team or anything because as soon as you have a manager to the air, you create territories, I know from personal experience, how hard it is to change those territories when you put someone there, this is this is now the place that you are going to manage and you will decide how much people get paid within this part of the organization. Then you you just put into in cement a part of your organization, you you should try and not do that. You can have a captain, on on on a crew, for example, that is something else that is like a pilot on a plane. That is a person who has the responsibility for the mission. But the pilot does not decide how much the stewards and stewardesses get paid. They don't have HR responsibility. They do report to the Chiefs how well the mission has gone and who deserves some extra credits or compliments or whatever it is they know everything that goes on, on that mission, but they do not have management responsibility. They are leaders of course captains are leaders that so we offer a captain role and what we call crews as an option, you don't need a captain maybe but it is an option that you can consider and I know companies have great success with Mission leads as they call them. For example, well I call him Captain but that's the same thing. We call that pattern the captain role. But as long as you keep that management responsibility out of it and with management I mean people management HR responsibility and desirable gets paid and career Development and so on, remove that out on the teams put that in the governance crew level, so that the rest of the bases stays flexible.

Joe Krebs 25:11

Very, that's very interesting. And I think when you just said that, and you had a smile on your face about exactly like the managing the pay and managing promotions, etc. And I think everybody out there who's listening to this right now might say, that's true in my organization that is a blocker, if we're removing that, that might change the environment. And that makes it a case for a pattern has proven micro solution for a common for a common problem. So this is, this is really cool. What I want to touch on one thing you just mentioned the word crew. What I like about the flexibility of these patterns is that you it's almost like you have name suggestions for these patterns. But you always make the link to alternatives where we say like you might have heard this word before. And this is really what it means over here. So it's like the name of the pattern, right? The same in a cookbook, where it could be a Sicilian tomato sauce, and we could be Northern Italian Italian sauce, but at the end of the day, there will be a tomato in it right? In either or, with subtle nuances in it, but you do speak about a crew. And I think that's like one example I want to take you just mentioned that. I don't know them all from the top of my head. But there are different crews, you just set the governance crew, there's I think there's a platform crew. This might be a lot of crews for somebody who looks at unfix. In the beginning. I like your approach of starting somewhere lowest hanging fruit you mentioned when but why are there different crews? What's What's the benefit of looking at the crews in different ways from different angles?

Jurgen Appelo 26:53

Well, let's let's take the two topics. separately first, naming is important. So indeed, I've used the word crew instead of team because the word team is overloaded these days people use the word team for anything. Basically, crew is a bit more specific. I use word base for what other people would call a tribe. Some people complain about cultural appropriation and things like that. So I've tried to steer away from the tribe word. Base is the home the place where people return to I like that word. And we use forum instead of Guild because by default, people assume that guilds are things that emerge bottom up that it always volunteers, that is possible. But a forum could be more formal. So it could be something bottom up. But a forum could also be installed by managers, for example, where they say we want alignment across bases on a certain technology, like we don't want five different testing platforms, we will not one, because that's cheaper. Now you go and figure out with each other which one it is going to be and we want a forum to take care of that. So language is important. But indeed, people can use their own words in the pattern language, you need words to identify things. And I think about what is the best word that comes with as little baggage as possible. But I leave it to people to use their own words in their own organization. If you like the word, pod or squad, instead of crew, go ahead, knock yourself out. The second part of the question was indeed the different crew types I borrow for from Team topologies. Actually, they simply identified 4 patterns before I did, and I credit them for that, which is the typical value stream team, Scrum Kanban. Team, whatever, we know how that works. And then the three exceptions which they call the enabling team, the complicated subsistent team and the platform team. Those are three different kinds of teams that are inward facing. And I borrowed the same ones I changed the name a little bit maybe too off topic to discuss all the reasoning behind it, but I just borrowed the same four and they added three other times that I thought were useful, which is experienced crew partnership crew and the governance crew and we just talked about so the set is seven, seven kinds of crews are teams within the base and they are like Lego blocks you use them as you want. I always tell people I hope you have as many value stream grooves as possible because that is like the most popular block. The Lego block the most beneficial ones I hope seventy percent of your teams aren't that type. But an enterprise of 100,000 people is not 20,000 Scrum teams, that doesn't work, you need some other kinds of teams to hold it all together. And that is why team topologies identify the different kinds. Because not everyone is offering value to a customer, some people are offering value internally to the other employees. And there are different kinds of behaviors that you can identify like a platform, through, as I call them, they offer a value to others on a self service basis, almost like an API, or sometimes literally through an API, in terms of technical infrastructure, or in DevOps capabilities, whatever. But I have seen kindergarten on site at a company where I was a couple of years ago. And you could bring your baby and toddler and throw them in the basement. And by the end of the day, you could pick them up. That's an API as well. That is that is also a platform crew, the kindergarten team. So that's that's that's one kind of platform crew. And then there's others the facilitation crew and capabilities. Again, alternative exceptions to the rule, you could say,

Joe Krebs 31:25

but but I do want to reiterate and get your confirmation, this is not something you would be setting up all these crews up front, right, this is you're building piece by piece, you're starting somewhere start small. So this is not completing the picture. And having a crew everywhere, this is not installing unfix, you might start with the value stream crew. And

Jurgen Appelo 31:46

I know it sometimes it's more clarifying what people are already doing, or giving it a name to something that seemed sensible. Actually, I had people reach out to me literally, when I published the unfix model for the first time where people said finally now, now we have a language for what we have already been doing for several years. And it seemed natural to us only we didn't we didn't see this visualized in other frameworks like like that. Yeah, so there are several case studies like that, on the unfix website web page already use the patterns, they they didn't have the names yet, I was just giving it a name and a color and that's it. But um, so what what can help is with a pattern language is it helps people to have a conversation, like, Okay, we have a couple of Scrum teams here. That is obvious or Kanban teams, whatever your preferred, agile approach, but a few of the other people are inward facing with the things that they do instead of outward facing what, what are they? Well, that depends on how they behave, do they literally sit with the others? Are they facilitating like agile coaches and so on then they would be facilitation crew, that's a different kind, but that gives you a name it gives you it gives those people recognition like that. So we we are this we are this pattern the facilitation group pattern. And now we we know how to explain what we do to to the others. And if you don't have it, then you might want to consider it like, Okay, you have 10 Scrum teams or something, maybe want to consider facilitation crew because it could offer these benefits. They might be interesting in your in your context. So read up on the available patterns and decide whether this building block is something for you. And then you use it or you don't.

Joe Krebs 31:46

Yeah. Awesome. Yeah. So you're talking to a pattern freak, right? I love I love the thinking behind it. I have you know, read all this stuff in the past. And as you mentioned, this started a while back. So really cool stuff. And for everybody out there interested in in unfix obviously on fixed.com is the place to go to learn a little bit about it. As you said, you're on your way to a conference, you're speaking as well as the unfix conferences. What's what's your approach on You know, sharing the wealth of unfix with the world, more conferences is the training programs behind it, etc. How do you how do you multiply Jurgen around the world in a way cloning that your model is sticking?

Jurgen Appelo 34:50

I'm glad that we cannot clone ourselves because I'm a difficult enough person as it is. So I wouldn't, I wouldn't want to bother the world where have multiple copies of me. But no kidding kidding, all kidding aside, I do want the word out. And of course, I want people to play with this. I love working on the model and pattern language itself. I'm very, very happy to be doing the research and then pattern analysis and coming up with names and visualizations. I have other people who in their own way try to bring this to an audience. So for example, the unfixcon is organized by other people in Berlin. I know them, well, they use the brand, I am involved, but it is not me doing this. I have a team that is working on partnerships, so people can sign up as a partner and then use the the courseware materials to the unfix foundation classes. And again, I have team for that is not what I do. There's someone else creating an app plotter app that you can use to design on your computer, your org design with the unfixed patterns again, that somebody else so I actually want to enable a lot of people who have an idea of how can I bring this pattern language to certain users customers in any way that that seems sensible, and then enable them to do that. I have someone who creating a webshop with mugs and T shirts, and so on. All right, thumbs up. So and and someone else might be writing books on the topics or creating courseware modules, all of that I delegate to others, I just want to focus on the Patreon side because that makes me happy.

Joe Krebs 36:46

Awesome. Well, you're getting to you also speak about it. So have fun with that meet a lot of people. The unfix model is the first thing that's going to strike you when you go to the website is colors, lots of colors, and maybe that is an expression on diversity, diversity and the patterns the approach and maybe the ways of how people approach unfix in in many many different ways.

Jurgen Appelo 37:13

true! the colors well actually what I use colors for for 15 years people know me I'm not I consider myself in eternal midlife crisis and I need colors. But at the same time, it gives people a sense of playing with a Lego or having a playful toolbox. It's it should not look boring. I find I find very important. So there is also marketing psychological. It has to be colorful because life is too. Life is too short to use only boring colors. Joe

Joe Krebs 37:53

Yeah. Thank you another black T shirt guy. Jurgen thank you so much. Thank you too. And thanks for sharing your thoughts and, and good luck with that. Thank you for listening to Agile FM, the radio for the Agile community. I'm your host Joe Krebs. If you're interested in more programming and additional podcasts, please go to www.agile.fm. Talk to you soon.