jacopo-romei-eventstorming-iterative-governance-in-heavy-industry

[00:00:00] Double checking this clicker.
[00:00:04] It's not working.
[00:00:08] Yes, okay.
[00:00:10] Let's start this one.
[00:00:12] So, nice to meet you. I'm Jacopo. So, a brief disclaimer and introduction. Merci.
[00:00:21] So, this is going to be a story, okay? So, this is going to be...
[00:00:30] I'm in charge of delivering the last session of the last day, okay? So I don't want to bury under theory, under world text, text walls and stuff like that. Okay, so I'm going to tell a story about a project that I had a few times, some time ago. And... What you have to know about me, it's going to be very quick. So I'm a nerd, okay? So I started coding in 1984. I was more or less like, no, I was, even that time I was older. And you have to think about someone like, coming from Stranger Things. I got my own BMX, I got my own Commodore, and I used to play role-playing games. Okay, that's basically the profile. I founded two companies in the early years of this century. I destroyed the first two and the third one went much better.
[00:01:24] And this is the most important slide for today. So, when I was developing software, I realized that if you develop the perfect software with a crappy process, then the process is going to be like defining a local maxima. So you start caring about the process. And when you start consulting about process, then you discover that if you have a crappy business model, then you find other local maxima. And so I ended up like 20, sorry, 10 to 12 years ago, validating business models for startups or accelerators or venture capital, so assessing their risk and helping them to reduce the risk. But that's enough about me, more or less. You only have to know that this is the company I'm working for right now. I love this company, either outside and inside. And actually, this is what I do in my spare time, okay? So like, and this was controlled with a PS2 remote control, okay? So it was like I reversed engineered the protocol, the intranet protocol to control this stuff. That was during the pandemic, so I had plenty of time. But now, I'm assuming that the audience is familiar with many of the lean thinking concepts. There is a strong intersection with the domain-driven design community also. I will talk about event storming, but this is not to introduce event storming as a technique. I leave this to many other people in the domain-driven design community, especially Alberto Brandolini, who is a friend. I'm going to basically tell an introductory story that is compatible, that is suitable for people who are beginners with the tool, but My hope is that the most expert practitioners could find some nuances, some relationship with other tools that I will tell you about. So this is, again, this is my first talk in Paris. Actually, I already had the talk in FlowCon in 2019, and this was the last. Session of mine before COVID. And it's so nice that this is my first session as a French citizen. It's too much, but resident. Okay, okay, fine. So the reason why I'm storytelling is that I want to explicitly, on purpose, avoid definitions and cortical models. I don't want to give you a map with a model of reality. I just want to give you a perspective on how different tools can be glued together in some systemic way.
[00:04:00] So it's more like a recipe, you know? You see, like we have Evan Storming, we will talk about games that I performed, we will bring in spreadsheets, and we will bring ways to negotiate contracts and ways to experiment and develop options. All topics that I've heard about during these two days, okay? So if you want, another way you can leave this session through is as a wrap-up of many other things that you've heard about during the rest of the day. And this is to restate that holistic nature of our relationship with processes, okay? That should be systemic, holistic. You know already enough about it. I'm not going to... Spend slides about this stuff, but I want this session to be somehow an object-oriented, like an instance of this class, okay?
[00:04:56] Okay, it's really okay. So do you have questions? It's a three-level check. So is it clear, which might be not true, but can you understand what I'm saying? Cool. I hope this is consistent. And about the validity, third level, let's wait for the end, okay? Okay, now. So. We will talk about who was the customer in my story, what was the request. The first session, I will describe the first session, the half-day sessions that we had, and what was the result.
[00:05:29] And hopefully I will have some time for questions and answers. I hope to be fast enough to give some time for questions or to go earlier home. We might be compatible with your status. So let's start with what was the customer. So the title is about heavy industry, and I cannot think of any heavier industry than this. So CSC Baglioni is a...
[00:05:57] He's a company.
[00:06:00] It's a company that builds tanks, not tanks like the war ones, like tanks like this, tanks for oil and gas, nuclear power plants, and hydrogen technology. But maybe if it sounds too high-end, even this one is built by them. Okay? Apart from the fact that it's Michelin, which makes particular sense here today. But still, so this steel tank is built by them. They are a very small company, and when I worked with them, I came to know that the market of steel tanks is dominated by four players worldwide. Okay? So they are one of those four players. And you see every bike shop that you go in, there is something like this. And they probably did it. Okay, so this is the customer. Actually, the least digital market I can think of. Okay, so basically everything, I was like, ah, we could try some like test-driven development, stuff like that. It was like, what are you talking about? So basically, I was not there as an IT guy. Okay, I was there as a process guy.
[00:07:08] The request came from the
[00:07:14] It's very Italian here. So he was the son of the current owner. Okay, so it's a family company.
[00:07:23] And the son of the owner attended a workshop of mine about digital transformation. So after a while, I received this call. I said, hey, Jacopo, I like what you said about digital transformation. You should help us reshape our process because, I mean, I want to bring some fresh air in the governance. And you seem the right person to do this. Okay. So.
[00:07:55] I knew he had the right perspective on my approach because he attended the workshop in first place, but I wanted to make sure a few diktats, some conditions that were non-negotiable. First, this is a homage to, so if you let Italy experiment on this side, usually something bad happens a few decades later. Okay, somewhere else in the world.
[00:08:23] They asked me explicitly, so after a meeting, we decided I could help them fix the process. Actually, I've been like bought roadblocks and the facts and real, so direct waste, okay?
[00:08:39] They had a problem dealing with acquiring new customers because there were a huge stream of standard requests. But when you build a power plant, or you build a nuclear power plant, the request is quite custom. And so you have to accommodate the customer request, which, I mean, it's like... Disclaimer, not exactly part of the Haitian law, but I mean, it's like that the small percentage of the request accounted for most of the time spent in the manufacturing line. And third, the number of customers had exploded recently. Which is, I mean, this is more common than it should make sense. How many of you have ever seen people frowning because there were new customers available? Tons of times, like, no, we have other customers. I mean, it's good news. And it's amazing how many organizations, they just feel the pain of managing new demand, as if money came from other sources. I drink. I have a bucket of sand in my...
[00:09:49] Okay, cool. So, and for this request, I decided that I should absolutely have, first, a room in which all the departments of the organization could be gathered together. Okay? So I wanted no division, no unit, no department, however they would call it, no division, no group excluded from the process. Maybe not included in total, so it's entirely, but... Definitely represented in the room. I didn't want... So basically, I wanted to avoid the risk of having blind spots in the analysis, okay? Because again, we care about systemic view. But actually, I framed this need of mine about the systemic view, the holistic analysis, in terms of a condition. Without explaining them what is systemic view, what is holism, what is the whole rather than the parts and blah blah blah blah blah. Okay, so it's just, hey, I will not work with you if not in this way. Okay? Cool. You call me. I didn't call you.
[00:10:58] Second, this would be led by me, which means, I know it sounds a little bit granted, but they call me, but actually, in my experience, many times you have a customer that calls you to do something, and then they start negotiating what you will do, which is, I mean, if you delegate, for the process of fixing the problem, I'm going to manage the risk that you feel, the lack of trust that you still might feel on waiting on you, I will manage that risk in another way, and I will show you how. But actually, you will not negotiate the way I'm going to... Drive you through the change management process by negotiating on the tools and the techniques and stuff like that. If we do event storming, we do event storming. We're not doing event storming a la CS Baglioni, okay?
[00:11:54] And one promise that I did is like, hey, I promise I will not disrupt the workflow. So we are going to fix the airplane while we're flying. So I will never ask you, like, let's go one week in a company retreat and shut the door of the company for five days, okay? So the maximum amount of time I'm going to ask you is a few hours with lots of breaks in between so that you can keep on replying emails and extinguishing fires that are starting outside of the room through the morning. Okay? So this was a promise for them. And these are the options that I offered. I'm talking about the quotation.
[00:12:38] The quotation part of the project, because I think, and I wrote a book about this, this is not the subject for today, though, I still believe that shaping their offer is the beginning of the consulting process. You cannot work the way you are meant to work or the way you want to work if you don't shape the right offer that gives you freedom enough to work the way you want. Just in a statement, there's no point in talking about agile development if you have fixed price, fixed scope contracts, okay? There is a word, look for it on Wikipedia, this is kayfabe, and that's the fiction that is created between the audience And wrestling fighters, that we all know that it's fake. We fight. They all know that it's fake, but they are supporting us. And we all think that we all fake that it's true. Well, it's all fake, and we all know that it's fake. Okay? So I don't want this. And the way I can let them free to accept all those conditions that I put is give them options.
[00:13:48] Matthew today talked about optionality. Optionality has many definitions, and one is, I love Taleb's one, which optionality, giving options, having the chance to go back from some place, to come back from some place, it's a valid substitute for intelligence. So you can be dumb if you have optionality, okay? And so I gave them many options.
[00:14:12] One was a three-hour session for a game about flow, something that could be like Matthew's game today, so a game which is showing how you should take care of the whole rather than the parts. Second part was the doc game plus an event storming-based process mapping session. So, okay, now that you have discussed the flow and the theory of constraints and lean thinking and blah, blah, blah, then why don't you map the process so that you can spot the problems in real life? C, it was everything plus a quest-based process redesign session. So basically, hey, how can you envision another process?
[00:14:05] You can be dumb if you have optionality, okay? And so I gave them many options.
[00:14:12] One was a three-hour session for a game about flow, something that could be like Matthew's game today, so a game which is showing how you should take care of the whole rather than the parts. Second part was the doc game plus an event storming-based process mapping session. So, okay, now that you have discussed the flow and the theory of constraints and mean thinking and blah, blah, blah, then why don't you map the process so that you can spot the problems in real life? C, it was everything plus a quest-based process redesign session. So basically, hey, how can you envision another process?
[00:14:51] All of this was proposed with a money-back guarantee. This is one of the key points of today's session. It's just one line. It might be somehow sneaking in as a concept. But the point is that as long as my consulting is bound to a clause like this, you know that what I'm saying to you at least is a honest message. It might be not valid, but it's honest. And so you're not allowed any level in the company, from the workshop up to the white-collar owners and blah, blah, blah. You're not allowed to think I'm a consulting selling smoke.
[00:15:27] You're not allowed anymore because anytime, what I say to them is that anytime you feel you are losing, I'm wasting your time, you quit. I don't get the money. You only give me the reimbursement for the expenses.
[00:15:40] quite bold, someone might think, but actually it was like just a few alpha days for me. The cost of our jobs, the cost of our work, is always so little compared to the outcome that we promise. Imagine, reflect about this, okay? If I die now, remember me like this, okay? So like, actually, there is a very big asymmetry between our cost, which is mainly for many of us, many, not all of us, but many of us, it's just time. Just time.
[00:16:14] But I mean, thinking is not capital intensive. That's the point. And so they decided to go for the option C. That's why I have been storing the title of the session. And you see, I mentioned Quest. What is Quest? Quest is a method, a suggestion on how to proceed to fix, to analyze problems promoted by Milo Fermin, that I had the chance to know in 20...
[00:16:46] Plot twist. What happened? Okay, fine. Then I had the chance to meet at the Cabaret Sauvage, ici in Paris. So there was this event called We Share, and she was there, I was there. So basically I got her book signed by her. Yeah, that's a nice coincidence. Still, this book,
[00:17:09] It's nice.
[00:17:11] I suggest the reading mildly. So one of the key concepts in the book is the air sandwich. And it's nice if I read an excerpt from the book and check also the time. That's perfect. Okay. So an air sandwich is a strategy that has a clear vision and future direction on the top layer. Day-to-day action on the bottom and virtually nothing in the middle. No mythic decisions that connect the two layers, no rich, chewy feeling to align the new direction with new actions within the company. By doing strategy the old way, you know, this is like white colors and blue colors, these distinctions like this.
[00:17:51] We're missing the substance of the business, the debate of options, the understanding of our capabilities, the sharing of the underlying assumptions, and so on. Do you see the point? At least, is it clear? So basically what she's stating is that we should stop thinking in terms of very manufacturing ways, and then I'll go back to the fact that this is a manufacturing company, that the digital mindset brought in the notion that actually between strategy and tactics, there is a little distance. Because if you buy the code by someone who's not skilled enough to write down a nice architecture, then you're going to end up with technical debt in two years and you have to write from scratch. And that's a strategic goal. problem, which was not in the slides of any manager and any white collar in the beginning. And so the tactical decisions are influencing the strategy, not just the other way around. You see?
[00:18:51] Questions? Because this is a key point, why I have brought all the people in the same room.
[00:18:57] Even if it was a manufacturing company. So I decided to break the division between white colors and blue colors, even if we were in the manufacturing context. So I decided, okay, we will discuss the process and the governance of your business with everyone involved, which is something like Genba.
[00:19:22] The other way around. So why don't we invite the workshop people inside the white collar world?
[00:19:30] The aim is the same. And what I want to avoid with this, again, social technical system, does this ring a bell? Social technical system for someone? Yes, someone? Okay, fine. So I want to... I want to make sure that the feeling and the notion of the system is the same for everybody, creating alignment and avoiding that. Someone could say this after a while. Really? We have decided that? Ah, wow, I didn't know. Or even worse, that someone after a few months could say, yeah, yeah, that was the decision, but you decided that. I was not agreeing.
[00:20:11] And I hope these two statements strike some chord deep down in your experience, because these are two very toxic dynamics in the governance of any organization, no matter the size, by the way. Even a two-people system can act this way.
[00:20:30] And obviously the chances that this happens grow exponentially with the amount of people, blah, blah, blah, blah, blah. I'm assuming an audience that I don't have to fill of details, okay? But please interrupt me if you need to ask some questions. Also, a good image for this is that I very much prefer having all together at a second best solution, agreeing on the second best solution, than having someone here and risking to lose the rest of the company. You see? Because the actual second order effect of alignment can bring everybody on top just after the first decisions. Okay? So I'm taking into account the social dynamics after the rational decision of how to arrange the processes. You see the point? Again, this is another key point. So I see faces that are a bit lost.
[00:21:31] Quest, yeah. So basically what I'm saying is that we have many options to go, and you have your ideas, you have your ideas, and we might agree on a third way to act, that it's suboptimal for you and suboptimal for you, but good enough for the three of us. I would opt for this second way because our alignment on what we're doing allows us to iterate incrementally in the next future rather than hoping all in that your idea or your idea is the good one. And then having someone else saying, I told you.
[00:22:03] Is it clear now? Thank you.
[00:22:07] Which is, I'm not saying that this is a counter argument on a good part of the narration in the Agile narrative, which is, I'm not saying that happy teams will be enough to deliver a good business model, okay?
[00:22:25] Needed and enough are two different categories. And I know that if people hate each other, success will not be there. But I'm not saying, hey, let's enjoy life together and we will be winning the air tanks war in the market. That's not going to be enough. Okay, so I already mentioned the quest. Quest is a framework that we will see. I don't need to give you the theory. It's just an acronym which stands for question, envision, select, and take. And we will see how the four sessions that we plan somehow match these steps that Mule of a Merchant offers. Okay, session one. We were in the question regime, question step, and we had a Event storming session. Who doesn't know event storming? Because I can decide, okay, you don't know it. Okay, you don't know it. Event storming is a collective technique to map processes with standard artifacts. The main one being the event. Which you can tell by the name, so event storming, it's like storming is the same storming as in brainstorming. So basically you collect ideas from everywhere. In event storming, you collect events from everywhere, which means that
[00:23:49] Everybody seeing something happening in the process in the company is entitled to create an event and stick it on the wall. And then later, we will find a way to see why it happens, who has to make it happen, who doesn't like it, who likes it, what is the blah, blah, blah, blah, blah, all the decisions. Why do we start from the event? We start from the event because events, and this is the most philosophical step of the session today, events are... Objective.
[00:24:20] Okay? So what I mean is that unless you want to bring some very refined philosophical position, which will lead very soon to a solipsistic, by the way, theories. Again, if something has happened and someone saw it happen, that is true. And we can debate on why, when, the reasons, the systems that allowed that to happen. But actually, if... All the company knows that invoices are made and only one person in the company has seen a manufacturing line broken one day last year. These facts are as true as each other.
[00:25:11] Both are true.
[00:25:13] And another artifact that we're using is the hotspot. I'm still going on for the few friends that don't know Evan Storming. Hotspot is something that deserves a better discussion about.
[00:25:25] A few events are nice, are more events than the others.
[00:25:32] These are called the pivotal events. Pivotal events are those events that somehow mark the passage from one domain or bounded context. These are not the sort of... We could have like four days workshop, okay, about event storming. So somehow these are pivotal events are those events that make the world different irreversibly.
[00:25:56] Okay.
[00:25:58] So. After having this session, so having all the people, 40 people in the room, sticking the orange post-its and the yellow ones and having the first discussion about, hey, why is this happening? There should be an hotspot here and blah, blah, blah. I decided to have this game, which is called Dot Game. I'm not the inventor of this game. I learned it in Linkin Park, Belgium in 2010. I run it like 292 times. I numbered them, yes.
[00:26:29] Two weeks ago, it was 290 the second time. Which is a game, but nothing that I've, so just to say that it's a game about flow and theory of constraint. And people can experience those concepts without having a slideshow, teaching, blah, blah, blah, blah, no matter their background. They start discussing the way cues are formed, what is the role of weak limit, what is the, how can you, you discover that the constraint is just one and there are no many more constraints, blah, blah, blah. I have this game in this point of the process, this is relevant in the room, because I want the discussion on the hotspot to be
[00:27:10] altered by this experience. So I want to inject the reflection, the first-hand reflection, the first-hand insights that they get from the game in order to alter their reflection on their own process. So before we go on with the event storming, I say, hey, break, have this game, and now let's go back to event storming so that you can... Yeah, that's what was happening during the dot game, and I see the same pattern here, and blah, blah, blah, blah, okay? So again, I'm somehow skewing the conversation towards lean thinking without talking about lean thinking.
[00:27:49] Sneaky education is a resource.
[00:27:54] Because people can sometimes, very often in my opinion, are somehow reluctant to be knowing new things and blah, blah, blah. While they are always happy to have an opinion. But if you skew their opinion, actually you end up where you want them to end up. Which again, I'm not saying that I was suggesting the outcome of the analysis. I was skewing the analysis, but the analysis was still theirs.
[00:28:19] And so we kept on adding, like, what are the systems you work on? What are the systems you work on? the policies according to which you keep on making decisions, what are the waste points, and what are the events or the systems or anything in the map that creates value. Cool.
[00:28:38] And this was for the first session. This was half a day.
[00:28:43] And after half a day, one morning, a little bit longer, like 8.30 to 1 p.m., so like four hours and a half. Now you're free to go to lunch. After some time, one week, as far as I remember, we had this session.
[00:29:03] In which we collected all the hotspots collected, so like, and clustered them, okay? And we have this, have you ever heard about affinity maps? So basically affinity maps, so like creating clusters for similarity, okay? or the systems or anything in the map that creates value. Cool.
[00:28:38] And this was for the first session. This was half a day.
[00:28:43] And after half a day, one morning, a little bit longer, like 8.30 to 1 p.m., so like four hours and a half. Like, now you're free to go to lunch.
[00:28:55] After some time, one week, as far as I remember, we had this session. In which we collected all the hotspots collected, so like, and clustered them, okay? And we have this, have you ever heard about affinity maps? So basically affinity maps, so like creating clusters for similarity, okay? And we reach this through silent sorting. Have you ever heard about silent sorting? It's a technique. I'm skipping it. So basically, you ask people to sort stuff without speaking, which is incredibly faster than having them negotiating verbally, okay? So they can move. When things at a given point, they reach asymptotically,
[00:29:39] a steady status, you can move forward and you can ask to have every cluster, actually here you see hotspot, but all these lines are clusters of hotspot, okay? That's the title. You don't care about, I didn't translate it because you don't know, you don't have to know what it's written here in Italian. But actually, I asked the... You have to know what Pueblo is. Pueblo was everything in the room, everyone in the room that was not belonging to the board of directors. Okay, so all the people who are not directors, they were called el pueblo.
[00:30:16] Yeah. A bit of fun helped. And so I asked them to rate all the hotspots by criticality and frequency to have a ranking. Criticality and frequency are two orthogonal dimensions. There can be something very rare, which is very bad, or there can be something which is not that bad, but happens and annoys the team every single day, or twice a day, or four times a day. And so we decided that they could be like two orthogonal, independent dimensions. And also, the Pueblo was, they had the right to say, hey, this topic is going, this hotspot is going high in the list, but it's not worth discussing. And you see that I left the Pueblo having the veto power, not the board of directors. Because actually, they knew what was... Tactically annoying.
[00:31:16] Ask now and I'll see if I answer now.
[00:31:22] The veto power, not the prior. No, no, no, I'm coming. I'm not that democratic.
[00:31:29] And especially, no, no, no, jokes apart. This is something that I think it's true. Companies are not democracies. While I strongly believe in democratic values, a 40-people company is something that is owned by someone that invested their capital in risk. And you have to include everybody's, I mean, you see, I'm doing this. So I help them include everybody's opinion and everybody's experience. But it's not a non-profit, okay?
[00:32:00] I'm getting there. I'm getting there. So we had this prioritized list according to, so like a bottom-up prioritizing session. And after that, we had the Soviet, at that point, what was the natural name, the board of directors gathered in order to have a discussion on every, you see that these are the lines that you saw before, and every single person in the board of directors could evaluate the... Priority of every issue, but I had them discuss the topics in which the deviation standard of their valuation was too big.
[00:32:39] If you're not comfortable with statistics and mathematics, it doesn't Don't care. I'll tell you what it is in English. Basically, I wanted them to be aligned or at least to know why they were not agreeing. But rather than discussing what they were already agreeing about, I had them discussing what they were already disagreeing the most about. Good news takes care of themselves, okay? So if we agree on something, we don't know why, but who cares? We already agree. So let's move forward. Let's move to the next topic in which you have an idea, you have another completely different idea, and I have another completely different idea. And so that was time management, basically. We had half a day to do all this prioritizing game. And so I wanted them to discuss on the worst 10 topics rather than discussing on, hey, yeah, we agree to do this. Yes, we agree to do this. Okay, it's okay, fine. You agree, but you have to launch to talk about how much you agree.
[00:33:35] Fine. Have you noticed?
[00:33:40] Two sessions, no trace of a solution so far. We had four sessions and a half of the game was dedicated to exploring the problem.
[00:33:53] And this is the second meteor moment. If I die now, remember me for this. I really mean it. Don't rush in exploring solutions with people who don't have at least, not saying the right opinion on the problem, but at least the same
[00:34:14] toy to play with, the same representation. If we want to discuss the borders of our countries, let's at least have the same GPS system.
[00:34:25] Otherwise, ambiguity is going to strike so hard later. Okay, session three. I'm closing. It's going to be much faster. I still have 20 minutes, which might be enough. Session three. Now, envisioning options. We have a problem. We need to find solutions. The first thing that design thinking teaches me is develop options and find ways to kill those options as fast as you can. That's one of the very definitions of creativity.
[00:34:58] As an engineer, I'm an engineer. See? It's mostly the same. As engineers, we are taught to plan for the right solution from the beginning.
[00:35:12] Solution finding works much better when you develop many options and you find a cheap way to kill the ideas that don't work. Okay? And what is left? At least it was an idea better than all the rest that you killed. Okay, so envisioning options. And that's when, that's only when I introduce Quest to them. They were already doing Quest. So they had already gone through the question phase. But now I explain to them what was going to happen.
[00:35:43] And we stepped into the envisioning phase, envisioning step. And this is the moment in which we introduce cross-functional themes. So basically, we selected the five worst problems. according to the priority process, the prioritization process that I already described. And for each of the five problems, we created cross-functional teams, okay? And again, I don't need to explain why we chose cross-functional characteristics of the team. You know the cross-functional team silo and... Fine. And we used a framework, an experimental framework, which is very trivial. I mean, it's not rocket science, but it's called Popcorn Flow by Claudio Perrone. Someone might know him. Okay. And he's a very good friend of mine. So basically, you have problems to develop options to fix the problem. So like, hey, every time I go to the office, and it rains, I'm soaking wet, and I spend all my day in the office with my underpants wet, which is very annoying. That's a problem. Okay, cool. Now, what are the options? An umbrella, poncho, private driver, and building a roof from my home to the office. Okay, these are options. You can run possible experiments to evaluate the different characteristics of these solutions. You commit to the experiments, you run the experiments, then you can have other experiments to clarify more. Maybe some other options emerged, or you step to the next opportunity to experiment. Very easy. For the people in this room, I'm guessing this is very like, yeah, come on. So I'm stepping away. Is it clear for everybody though? Okay.
[00:37:26] Basically, developing options for a couple of hours. So we got plenty of ideas at the center of the third session. And so we needed to step to the selection step.
[00:37:41] And the session was over. So they had this step as a homework between the third and the first session so that they had the time to select the right options, discussing deeply. I didn't want this to be rushed into a few hours. And I mean, the core of the select step is that
[00:38:04] If you want everything, basically you want nothing. And if you develop 100 good options, you still, they are, first, it's very unlikely that all the options that you develop are good. But even if so, it's definitely true that you don't have time resources, money resources, moral resources, people resources, any resources to pursue all of them. So you have to choose. And what we did is to... To give you a brief example, it's like if you start discussing the criteria by which you choose something, the conversation, the assumption was the conversation gets much colder. So I'm with Chris and I'm with Matthew, okay? And we have to decide where to go on a holiday next weekend together. So you have your own idea. You say Cannes, Japan. Okinawa or surfing in Australia. You say, no, I love Florence, I want to go to New York and Venice Beach. And I say, Nepal,
[00:39:06] Dolomiti and South Africa, okay? Fine. If we start discussing our options in terms of just for what they are, just options, we might have a very long and some, actually not in this case, but some heated conversation.
[00:39:23] While, if we start discussing the criteria that are behind our options, that I had in mind when I proposed South Africa, Nepal, and so on, I can propose, hey, why don't we go, okay, these are our nine options, and we keep them a little bit apart for later. And then I say, we could go somewhere nature-oriented, culture-oriented, or fun-oriented. And Matthew could say, yeah, but I also want to have strong nightlife and I want to eat very well. Okay, cool. And you and Chris can have other criteria. If we start discussing and disagreeing on the criteria, what I empirically noticed is that the discussion on the criteria is much colder than the discussion on the options. And then you have agreement on the criteria, or at least more alignment. Agreement here is not a binary concept in this session. Agreement is somehow in a spectrum. So you disagree completely, and you agree completely. So if you have agreement enough, then you can go back to the options and then... Chris, Matthew, and Jakub agreed that you should eat very well and that we all need some nightlife. So basically, I'm available then. Now I'm willing to discard Nepal as an option, though the food in Nepal is very good. But I mean, I'm not expecting to party at Base Camp 3 under the Everest, okay? So that's the point. You see my point? Okay, cool. They had this long and deep conversation asynchronously when I was not there, between the third and the first session. So they had two weeks to do it. Obviously, I was there by email. We had chats. I was still supporting. They were not abandoned. Okay, 5.11.
[00:41:12] Well, I want to be on time. I'm not that Italian.
[00:41:17] So, take.
[00:41:22] The last step that Nilo Fermant shot suggests is to take action, which is one of the incredibly most overlooked steps in strategy plan. So you got slides of what describing what you want to get and not what you will do the next few weeks in the order of weeks. We talk about quarters, but we never discuss about days in slides. And actually, I want to get out of meetings knowing what I want to do, what you need to do, what you need to do until next Friday. And so we kicked off Concrete Actions. We had a demo show. So all the people... All the five teams, cross-functional teams, at the beginning of the first session presented their selected options for solution, the ones that were selected during the selection step.
[00:42:15] And we somehow arranged a solution market. So basically everybody was committing to their idea. And basically the vote was not anonymous. And by voting an idea, you were basically saying, I'm available in helping. Because otherwise you say, your idea is very good. We will win. Now do it. Okay, in Italian we have a saying, which is, armiamoci e partite, which is, let's get the weapons and you go. Okay, something like that.
[00:42:52] And I wanted to avoid this stuff, okay? I wanted to say, okay, if you express support today in this room for this idea, then I want you to commit that you will be there to support that idea in the next few weeks.
[00:43:06] Two weeks. So basically, I asked them to develop. Okay, so you got your solution that you just showcased to your colleagues, and they voted it and expressed, you know, who you can rely on. But then the session was, the fourth session was all about, okay, now, if you have idea A and you have idea B, what you are going to do the next two weeks to... Progress toward that result. Two weeks. It was maybe a Thursday. The question asked was, what are you going to do until two Thursdays from now? And two weeks commitment to me was small enough for it to become impossible to reject. It's actually, whenever people say, no, I cannot, you have the options to reduce the commitment and say, okay, you cannot do anything next month. What about two weeks? No, I cannot. What about next week? No, I cannot. What about next 10 minutes? No, I cannot. Okay, so you're not here. Okay, now, if you have idea A and you have idea B, what you are going to do the next two weeks to progress toward that result? Two weeks. It was maybe a Thursday.
[00:43:38] The question asked was, what you're going to do until two Thursdays from now?
[00:43:44] And two weeks commitment to me was small enough for it to become impossible to reject. It's actually, whenever people say, no, I cannot, you have the options to reduce the commitment and say, okay, you cannot do anything next month. What about two weeks? No, I cannot. What about next week? No, I cannot. What about next 10 minutes? No, I cannot. Okay, so you're not here.
[00:44:08] But two weeks seem to me big enough for it to show things that may change. Okay, so basically like, hey, we can be here twice per month, which is 24 times per year as a steering point.
[00:44:25] You know, being an engineer, an IT engineer, I know what somebody knows, so there's something like sample rating, stuff like that, okay? If you don't sample often enough, you get aliasing, something like that, okay? Cool. And so they started having these 15 days experiments.
[00:44:47] And it's interesting because I never mentioned the word experiment so far, but I'll give you this definition of experiment, this definition of mine. Anything that you do which is cheap enough, doesn't kill you, and teach you something, it's an experiment.
[00:45:05] Anything that doesn't kill you and that leaves you at the end with something new, knowing something new, at any level, strategical, tactical, inward, outward, you can call it an experiment. And so basically, I never insisted too much with them calling, this is going to be an experiment, 15 days experiment, because otherwise I would have triggered their, no, we're not doing experiment, we are a manufacturing company. Okay, fine. So what you're going to do the next 15 days? And then you tell us if it worked or not.
[00:45:42] And we're close to the end. I took care that they nominated owners for every initiative. This is something you know from Scrum, it's very similar. The owner though was, I mean, what I made sure is that I didn't want the owners to care about the solution, but as again Matthew established this morning,
[00:46:07] I needed them to care about the problem.
[00:46:11] I made very clear that they didn't have to follow the idea that they had the beginning of the 15 days, but they had to progress towards a solution for the problem they stated at the beginning of the 15 days period.
[00:46:29] Basically, what I wanted to get in the room was that owners could become optionality fanatics. Like really, like brainwashing them about optionality. Like, we don't care. You had a good idea, right? So during the select step, you had very good ideas. You had this, wow, and everybody voted your idea because it looks so promising. I don't care a bit. If in 15 days you haven't fixed or at least mitigated or envisioned another concrete solution for the problem, we are at the starting point. Please don't love your solution, but love the fact that someone will say, my life is better now.
[00:47:15] UX. UX thinking, okay? So that's the point. Okay.
[00:47:20] And they all gather periodically every two weeks to cross-pulse check and also to report to the Soviet, which was still there, okay?
[00:47:30] The Soviets were still in the room. People from the Soviet were taking part in the teams. The hierarchy in the team was flat, or at least officially. I never knew, I didn't know whether there were some abuse of power during the action. I'll let you know what was the outcome in a while. And also, this is the third meteor point today. What I wanted the owners to do is to ask for help. I wanted them to ask for help if they are stuck. If something doesn't happen, don't wait for the 15th day to ask for help, but actually interrupt. So like raise your hand or please, the only harm that someone can do in an organization is not asking for help.
[00:48:21] You shouldn't be blamable because you don't know how to do something. You cannot fake knowledge. You cannot be blamable if you are not in the condition to express your knowledge, because actually the system is stronger than the individual. You know already. But then, if you don't ask for help, who the hell is supposed to ask for help if not you? You know you're not getting something. No one can know if you don't tell me that you know. Okay.
[00:48:51] And then I went away.
[00:48:54] After a few months,
[00:48:58] I received an unsolicited mail, okay?
[00:49:02] So it was basically out of the blue. This is the real mail, Italian, so don't worry, you don't need to know Italian. And you don't even need to read what is written in the following slides. I just want to highlight a few words, which is, there would be a lot to say about CSS, even just the fact that I see everyone really feel excited and wrapped up. Okay, fine, so feeling. Good, as I said, necessary, but not enough. What I love is that they were saying they were progressing towards solution fixing problem. This is an excerpt from their mail. And the fact that they were talking about these things is already a great sign. All of a sudden, the conversation had been triggered.
[00:49:49] And people are thrilled to work in a collaborative way that wasn't there before. I love this one. This is very strongly correlated with Taichi Hono, my thinking, which is, most importantly, they are thinking. Before, they were not paid to think. And you know, the... Taichi Ono quotes about the fact that a worker should be paid to do their work and to question their work.
[00:50:17] That way they deserve the salary. Actually, this is a key point. The fact that they were starting... So this is the owner. Owner telling me that very happily, the owner was happy that the workers were starting questioning the way they were working. Boom. To me, that's enough.
[00:50:36] And also good signs that are, and they are, all the information about what others are doing is publicly available. These silos were breaking down from a notion point of view. Notion, not in the tool-wise definition, like the original meaning. Final considerations then.
[00:50:55] Do you know this diagram, the Diablo Diamond? It's very famous in problem solving and design. Okay, cool. So basically, this structure for these four sessions was exactly compliant with this approach. We had a divergent problem phase, which was about, hey, how can And do you remember the quotation that I did, the options that I offered and so on? They were making room for the problem to be shaped in the way they would love. I didn't force them to have a four-day session. Sorry, it's for half-day sessions. I let them choose because actually if they want to just tip in with a dog game, three hours worth, worth three hours, that means it's fine. They pay me less, it's true, but also more time for me to pursue other business opportunities. So it's fine.
[00:51:50] Then we had a divergent session with Evan Storming and
[00:51:57] Somehow, I induced, it was not granted, but I induced more option thinking with the dot game. Then we converged on a shared problem definition. And then we developed options according to the popcorn flow.
[00:52:18] And then we converged, again, to the popcorn flow, to an experimentation rhythm. Okay, cool. Now, final considerations. What about AI nowadays for this stuff? I really mean it. I really mean it, actually. So it's like, okay, fine. The promise in the abstract of this session, you read 20 years old technologies that can be brought together to get something good. But okay, so just with a one less than one slide look in the next future, how can AI be integrated? How would I integrate AI in this? First, any divergent step can be helped by LLMs by definition of generative AI. So if you are worried about generative AI to be imprecise, don't worry when you have to storm options. That's not a problem. The flaws of current status AI are not a flaw when you are generating options. with Chris two days ago, the fact that you can, I said buy, but I said you can get options from any source when you need divergent thinking. You can ask your daughter, you can ask your cousin, you can ask your family, you can ask co-workers, even those people that don't care about and don't know what's your job. If you describe them the situation, you can ask them, how would you do? What would you do? And out of their, literally, ignorance, they can give you unexpected ideas and unexpected suggestions. Don't forget that brain is an associative machine. Two bad ideas can trigger a third very good idea. No matter if the first two ideas are bad. And so, LLM can give you tons of bad ideas. And actually, what you have to keep at bay is the cost of reading all these options, because actually, you don't want to read thousands of options. You can get some extra anger for options.
[00:54:22] So we could use AI to generate options for the experiments too. We could use AI to
[00:54:31] skew the problem analysis too. You can have the event storming picture it, like I mean, I've shot photographs of the, how would you say, the tableau, the board, yeah, thank you. And submit it to Claude, to GPT. And if you explain to it what is the formal conventions that you're using, and it's plenty of articles that you can use to teach LLM how to interpret those images, then it's perfectly fine. It's going to start to give you hotspots.
[00:55:10] which somehow can give a lateral view on the analysis. Okay, Q&A. I got one minute, so one question.
[00:55:27] Any kind of question. Also, like, how much was it?
[00:55:30] No, sorry.
[00:55:33] Yeah, Guillaume.
[00:55:42] Did I have to convince the top management about this? So this was an easy case because actually the owner, the young owner of the future owner of the company got me at the beginning. Remember, he had just had a digital transformation master, blah, blah, blah. So basically, he had somehow advocated for this kind of techniques. But actually, one thing. I stopped doing consulting because I hate convincing people to work the way I want. And that's why I, but still, that means, I'm not suggesting this, but to me, that means that I still do, I'm an entrepreneur mainly, but I still do consulting for those who are willing to work the way I want to, because actually, then I give them the money back guarantee. So basically, I'm not saying you work the way I do, you work the way I want, and you live with the consequence, okay? I'm putting my full skin in the game and I say them if you don't want to pay me the end, you're free to do so.
[00:56:42] And I also did it when I was depending on consulting. I did it not just because I have another company and this is my hobby. I respect the position of consultants, if there are any in this room.
[00:56:55] Now, this is my LinkedIn, if you want, and this is my Substack.
[00:57:01] Thanks for the question and for the flow con is over.
[00:57:05] I thank you, everybody, for resisting up to this point. I would be super happy to get your feedback in the SCAD application or in direct message like, you sucked, but please tell me why. Just tell me why. And also, I'm going to be available for anyone asking questions. If you send me direct messages on LinkedIn, I reply to all the messages I receive. So please do, okay? Thank you.

Transcript