Alberto Brandolini - 50,000 Orange Stickies Later

[Music] so well welcome everybody and the reason I gave I thought about this format is because well basically because of feeling feeling kind of guilty because of things that are published about even storming not so many and I'm writing the book and I filled a lot of pressure about finishing it but still there's so many people started to practice it and that's a little bit more than it's already available so I needed to find something - yeah put things in the right place in some way so I've been doing it a lot I don't know if this number is exactly real it's probably not so precise here but well I think even 3m saw a spike in the consumption of usage of orange stick you know so we're talking about a worldwide phenomenon now it's just really hard so I'm gonna tell you a little bit about me and a little bit about my story fear in storming like everybody just in this room I guess you having a hard time explaining your job evanka parrot is my company but I still don't know what what is what is my job I think it doesn't have a name what do I do is just I try to model basically everything with sticky notes markers and a paper roll and yes I started to call this thing even storming how it started this stuff that the whole thing started almost by chance just like I was at that time I was going around with an IKEA paper roll in my backpack don't ask me about why but that was my way of working and some people from the Italian Dedede community organized the meet up exactly with this purpose I I have this program I would like to see your aggregates I would like to see how would you model that and okay there's no time to use your mouth took that paper or put by sticking out so they are remodeled at the whole stuff and I realized okay where's your model and everybody was silent and looking at MIT but what did you do I just model the process that we were talking about what did he do in the meanwhile nothing ok looks like I had something in my pocket then it became a training tool for learning aggregates for people that didn't have a PD background then it became something more I mean we started making more and more experiments and probably the tipping moment has been when I was a special guest at Vaughn Varun's iDVD tour and we were in a room full of people I mean a lot of speakers from this converse were actually in that room on that day and a lot of people started experimenting a deep variation bringing me a lot of good feedback and many other things so that was just the beginning of a long journey and now in 2017 it's growing a lot of people are starting people that I never met are already using it and they're happy but the thing that is bothering me is just like I never really clarified that the different formats in in a nice way just like I said in some cases like it's getting more like like pizza with multiple recipes he aren't gonna talk about the different recipes and the different ways to use it so let's not ruin the big picture the big picture is probably the most famous or shocking if you've never been in one of them because well the recipe is relatively easy you invite the right people business IT maybe user experience and it's already a problem some organization just like how do I invite the right people they are not coming I'm just a developer that my boss is not showing up second thing just provide unlimited modeling space well it's a little bit of hack I have a paper or it's usually long enough to contain everything that we need but it's also a limited number of markers stickies and that is the thing that is making people feeling weird like well the whole business line we are so using yeah let's focus on a portion let's only talk about this that we forget the big picture well for one day in your life let's see about the whole that's right to understand how this company is making money how this organization is delivering value doesn't have to be greedy necessarily it's a lot like discovering what is a collaboration that brings value to the users members whatever how do we do that in practice well after giving these tones of orange sticky notes and markers we try to work on a timeline and yes there's some facilitator magic because the first five minutes in well actually last two minutes in an even storming you're really what the hell is going on here and this awkward feeling can be paralyzing for some people so if somebody pretend to be dead and then you know these usual tricks and but then it works then it starts working and it's surprising like oh whoa is this what we're doing that that's actually big I wasn't thinking this was going to be possible and the the other building block is all in sticky notes domain event orange verbal past tense and relevant for domain expert we put everything on a timeline it's not as easy as it gets usually gets messy then we try to order it later but at least you start to see progressively something that looks like everything and I feel like okay now it's something I can control now something I can analyze the center can put some layers of structural or extra extra saying what is what is the problem behind it but why is this so interesting I think the big lie in software development is the feeling that we just need to understand the business and translate it into working code or there are requirements so I just gave you the requirements you just have to translate it into code that's totally crappy I mean it's never going to work in that way and it's assuming this is living in this little bubble assuming that well the business guy know the business no that's not exactly true I mean they they know very well their own portion of the business but it's still living in a place like this one I mean silos in the organization there are these are silos I'm sorry for if somebody felt something different but everybody everybody thought something different sorry about it but the real thing here is the culture nobody knows of stuff better oh we are really good at the other departments research and and it's not a word role I mean in many organization you see that there is this kind of unhealthy competition like every department just wants to be the second worst the the worst one is usually the IT department always late and then you provide in yeah the the back-up plan for the rest of the organization and and everything is fine still we have the second worst but and we don't have really do care about the whole stuff and if you think about the resulting knowledge in a place like that I mean you cannot really trust these people they look trustworthy and that might be the problem but you can't so this this is more or less how the knowledge is distributed in an organization like this diverging options well good idea but then here we have little conflicts and then all the stuff this office that there is one guy living in this office nobody knows what this office is meant for and after you're in production you discover you should have talked with this guy and there are mysterious people that left the company knowledge that is lost and all of this possible stuff can you translate this one into requirements and say no I mean we there's something that is intrinsically in softer development it just like yeah gossip does not compile you cannot you cannot put them be within production and so you can't rely on this but I'm be with you within humans works pretty well I mean the result is not to make a fantastic process the the goal for many people is just like we're just happy to be the second worst if everything fails for somebody else's fault yeah let's put a blame on them easy stuff survival that's not how it works in many organization I can't turn this stuff into software so instead of asking thinks the people pretending that they're telling me the right thing I just try to put them all of them in the same room and try to see if they can build together a consistent story consistent narrative of what their business is and if there are ambiguities whoa they will surface they're gonna be so visible in front of my eyes how do I make them visible this way yep the typical behavior this is a little bit more specifically to corporate the expert will post a locally order sequence of events they know their job pretty well this is what I do pom pom pom pom pom this is me and the moment you put the wrong timeline in the same place with the other department timeline you realize that oh you gotta talk he just there's something you've never clarified and and so on so my tool is purple sticky notes for some reason you can't have real red with post-its that's the second best and so we put this little annotation with the okay this is where things are going wrong okay it's visible we take notes it's our learning notepad it's just huge and what happens structure will emerge II start to see some separation between different concerns Department different goals you start to see how people know some areas and with all different people's know some some other areas you might even display users roles external system and everything and every time you add something well people oh I never thought of this you add more types of conversation it had more awareness a lot more stuff the outcome the Atma starts becoming interesting the whole process is now visible just no excuse not to make it visible you have space you have markers you have time and it's not that slow we can see that and they oh yes then you've got this moment of awareness like oh this is what our company is doing and the other thing rating is just that happens in many many companies just you learn what your colleagues are doing it's not what people are going into it installing for but actually companies leer a lot learn a lot about themselves and in practice this means we start to see narratives stories user journeys from the beginning to the end and this might be the value generation path this might be just what happens in the flow the thing is visible it's a good narrative even a kid can understand it at this level it doesn't have to be hard I mean if your business so complicated that your employees don't understand it you're not in a great position on the market and yes many stories are visible they are storytelling and the humans started aggregating as a race and just around story telling stories around the fire legends means our story is not requirements and just so this works really really well other little tricks incremental notation we are not using UML BPMN we are using the simplest possible notation that allows everybody to join the conversation and in this way we can have a conversation with people coming from different tribes backgrounds so lean is fine user experience is fine agile software development organization design pure business whatever there's no choosing a strict notation is excluding people from the conversation defining the notation on the fly with some yeah professional touch just helps the conversation to happen so the net result after exploring after extending the boundaries of our exploration is just we touch those external systems and people I mean everything that could go wrong is now visible which means that the key problem the bottleneck is now in the picture the bottleneck is the way I describe it coming from a fear of constraint force for DDD people more or less the the the same ideas the core domain is the area where improvement to the system will make a difference it is visible it is there and is probably not so disputable so what do we have as a result is this moment of clarity wow this is what we have to do so this clarity in in my eyes and in participant I sigh I hope it's just now we have a clear business narrative we have a blocker somewhere and yeah you want to talk about the patterns what we do is we now let the people select what do you think is the most important thing and then people start to vote and while you might have outliers somewhere but mostly what we get in a existing organization is more or less the same so at the end of the whole day you explore the whole flow and you have arrows of people voting autonomously around one thing so this is basically what I do at the end this is my business model and just a kicky kicky in the birth of the organization that's the thing you have to do just put the people in the room understand what what is the right thing to do oh now it's obvious more than half of the people in the room said like this is the most important thing to say basically basically its end of the story and with a little bit extra awareness that is necessary we are touching the thing with the highest priority that will bring us quickly in a knob no backlog state of mind why should we have a backlog we have so many things to do before doing the most important thing or usually for a company I mean if you find the real key program for an established company I'm not expecting the problem this program to be easy be the company's full of smart people you have a problem it's probably really hard well how much time would it be needed to solve this very hard problem that we were not able to solve in the last five years you just have a couple of sprints no doesn't make any sense I'm gonna try hard till I solve it oh but I need an estimate no I won't give you and just try hard of it and it requires experiments and and this is one of the area where domain driven design and the culture of yeah we're pull model or Lean Startup Ania continuous evolution is learning it's I had a definition for the measurement design that fits very well here is the dome in different design is an approach to software where you can't do things right at first attempt where you need to iterate try again learn more till you find the solution and this means like again we are in this place but we're trying hard till we find the right solution so the best attitude for me to describe this this situation is now on dr. House's and just like I have no idea but this patient is now dying so I will try it's probably lupus by the way but I will I will try with steroids and and then whatever and see what what happens how the system reacted we found the solution and two things I don't ask estimates to dr. house and then and you don't ask dr. house - yeah treat patients with the normal flu that doesn't make any sense but just for special ops so a little bit of a difference I did it in a start-up and this is a piece of cake invitation well everybody in the same room they are already in the same room they have only one room the poor guys and so it's great but maybe they don't have enough walls so we can rent a room great attitude and this is a little bit of a problem like the moment we solve the thing we have we do something really brilliant they can't compare it with the past so it looks like it yeah this is tough dude and it's not we just save them something like six months of development and they didn't realize it big picture with a major company that's that's a different story what is the main difference between major company and and the startup oh this one a start-up is is a still innocent they didn't committed any any crime against beaut design and they have all of these bad skeletons in the closet but architectural choices or business choices I mean there's past decision that probably led them exactly to the trap the Arian and that is problem number one so inviting people it's ready and mass sometimes I mean I ran one out maybe two days worship and it might take sense like two months in order to have the people accepting the fact that they will have to be there in a meeting underlying conflicts I mean I've seen no I I can't tell how many things I've seen but I mean so much personal issues so what corporate dysfunction yes a long list legis implementation Dungeon Master's to your thoughtful llegas implementation they're still having about the influence on the company all of the stuff is there so I'm not expecting this to proceed smoothly is actually ok this is a sharp tool let's dive but yeah that is what is happening places like that special tricks might be interesting one is incremental notation like we I have a plan but I can pull my plan away if there is a better idea so we leave the people the possibility to choose what is the next thing you would like to see and that that helps shaping the notation the discussion around the interests of the participants and the other thing this is weird because at the beginning of thought like it's me being unprofessional and then I realize no I'm actually doing this for a reason so maybe it's gut feeling turning into a principal I don't provide definition I mean I provide something really really fuzzy like well we have sticky figures at the users actors role or personas are there whatever I call them people and I don't want to give a precise definition if I give a precise definition I exclude something if I give a fuzzy definition while you're putting something visible and then we have a conversation about why this is this thing or or the other one I don't care about precision I need to see things and then while we trigger good conversation it took me a little bit to arrive at this point and also a lot of time to make up this rainbow type of formatting on this this page but it started like a tool for developing aggregates now it's also a tool for discovering capacity of the company to deliver value or for an organization to be healthy and also to have a course perspective communication so we we used the same format to deal with service designer user experience designer we'd use it also for doing value stream and try to understand how to improve company's organization or software you notation is in the way the problem is the same one is a rebranded business storytelling and you can bring many many more actors personalities and skills around the same while it's not a table around the same modeling surface and and discover and discuss things notation was in the way so that's mean the main recipe for the big picture and the goal is discover is a yeah massive learning a lot of risk reduction and usually a pretty good kick in the body process modeling this is what happens when once we highlighted that some areas are really really worth exploring and modeling and here we have a little bit more of structure so the scope might be a speaker or a set of features or I don't believe in Street notation like oh this is a user story oh no this is an epic oh no this is an initiative oh I hate taxonomies so usually we challenge each one of them we just know that something is starting usually from a command and ending with a set of domain events which are good for the system but you have users and users need to see read models they need to see that one thing is happening but it's also readable for them and of course I don't trust this one so this is a good era for discovering thing and I don't trust this one written in the user story because well you didn't thought about it deeply enough so yeah let's let's do the exploration together that's the problem with user story for me like deep I mean some people define it like a placeholder for future conversation and now I see this humane dysfunction dysfunction number one is just like the conversation is what the hell does this index card means and we have to deliver it for Friday and the product owner is away and we have to do yeah throw the dices I think he meant that anyway whatever fits is three as I want to in order to scientists of the template might be the good user story till the demo or the user story just having the number for the specification as an attachment in JIRA so maybe four pages of boring stuff that has been written by somebody in the ancient times and then you have to finish this boring thing no they're both badly dysfunctional and so for me like well I have this place older and I have even storming as a way to structure this conversation doesn't take much time and we we just get to model the flow by using these building blocks the more I played with this the more I ended up with a repeatable structure that actually looks like as dumb as that so it's caller puzzle thinking you have read model a user is making a decision based on that and deciding to yeah click this button this goes well maybe to a system to our system to external one the result of this yeah elaboration processing is the main event and then we might have a policy like well an item was purchased and our policies whenever to with the purchase well we might go and well create the corresponding receipt for that that but the thing which is very relevant here is just like we we have a conversation like this one we need a lilac between the ORAC orange and the blue so it's it's actually just color coding a command might go into an external system triggers an event that we need polish between this and another command colos thinking even people with no exposure to softer development can can join the conversation and the main goal of this step is to investigate this this policies the Lila thing is the the little riddle that I tell to myself is whenever something happens then we react in this way so whenever event then command and the rule is just this one why is are they important what because they define your company they define exactly your your behavior and you might have the whole at the whole spectrum like they might be implicit like people doing things because that what they're used to maybe even without an explicit agreement or there are rules there is an agreement between your your peers but you might discover that while the agreement is not followed and so on the more you the more your subject of your company increase the more automation you have you might have listeners or IFTTT type of thing or something more sophisticated Sagas process manager and so on everything is in the Polish spectrum and at this level we are mostly challenging both the business narrative like well why do you do this oh well we do this well maybe it's not exactly this well I don't do exactly that well if we find inconsistencies before writing software is actually gold and discovering that the boss told the one story and the workers are telling you a different story and and discovering that this thing is not is not real it's is pretty interesting that this time I don't want to write software to discover that I was wrong no I can discover that a lot and not better another way of thinking on this level it might be this idea of challenging value so every step I mean we're still talking in storytelling way users are doing given things for a reason and we might want to see what is this reason every step in the flow can create or destroy value for a given user the idea at the beginning people start thinking about yeah here is where we get the money here's when the customer gets a item he bought for end of the story and now then then you might find out that well at the given moment somebody's happy because something is happening or because something is not happening and so you start moaning also negative events or yeah the absence of events and then but then you start considering like well happiness is also a different type of currency is not only money is not only goods then you start including well sounds like anxiety or reputation or a personal satisfaction you start to realize that people are doing things along the way and you are missing some opportunities to make people happy along the flow or to reduce some of these anxiety and you can streamline the flow not only for mechanical consistency that's what we are good at but also for yeah emotions and and yes happiness in some sometimes and it's funny because oh well this looked a lot like a customer journey type of approach yes pretty close I just didn't choose any notation in this case and just do that this emerge but even if you don't have a customer journey expert in your room just asking for this value simply put in the question you have brains in the room you have people you have humans who have sensibility even if we are nerds we have some sensibility sometimes and and yes you can ask this person then you realize you can do a better job and it's just as simple as adding a couple of extra colors to the model maybe is not perfect but it's way better than nothing and we just do it in yeah 20 minutes half an hour it's just very very very interesting once again the tool the flexibility of this this platform just enabling cross for expected conversation I'm exactly reusing the same slide but here's the area where we had pretty good feeling with people coming from user experience and service design and there is no reason to keep this discussion separated that doesn't make any sense to have developers isolated from the value or a Happiness generation type of conversation and just do stuff know just the response speed is not enough to make people happy and of course he has never said it's easy but we can we can actually do that and what is the problem for a father making this this things easy big picture is for discovery I try not to enforce any agreement I must have up to 30 35 people in the room and if I try to reach consensus between all of these people is a job in itself so I want so I will have some in math inconsistencies along the flow but when he put designers softer designers and process designers and yeah business analysts in the same room try to reach a consensus about this is exactly what we want to do well this is hard so it needs some discipline tricks and yes we it might be feeling again like really awkward what are we supposed to do the thing is we are never really trained in modeling in a collaborative fashion I mean Kay's tools UML diagrams it's you your mouse maybe a colleague that is pretending to look at your screen but but it is not really collaboration even whiteboard modelling it's way better but you you have one marker you don't write easily on somebody else's model you take terms and it might be a really really cruel sport but when everybody can write and move things around its way a lot more collaborative third recipe is a softer design we are changing the recipe a little so we might have a command going into a in aggregate and and the resulting in the main event or still going into an external system well the policy has still all the spectrum and yet just notation slightly differ but not not that much again more or less the same thing this that's the scope an epic or a set of features and not so much so many difference talking about aggregates the thing to look for is yeah kind of a state machine type of logic focusing on the behavior not data actually the whole trick about learning aggregates through even storming is you get to the aggregate like the last thing you just find out everything else that happened in the system so that in terms of comments and and events of behavior so that the aggregate become obvious like there's must be something here between the blue and the orange that's more or less the event storming definition of the aggregate is the yellow thing between the blue and the orange and if you postpone naming if you don't go for all this clearly has to be a user this is clearly a customer this is an order and I know everybody tries to look smart in in the room but yes if you can postpone the naming because as everybody else you feel like you need an order because an order has yeah a customer and the total amount and all of this information no you're still thinking data we are trying to sync on behavior then you look like which are the bits of information triggering a given behavior and this is the key content of the aggregate the other information well depends on the implementation might be redundant you just propagate them through events you can hold them inside that's a matter of style but the role of the aggregate in the flow is a little local decision-making I would accept or reject the given command based on my internal state that's what we're going to write inside here and through that we can understand what is the plausible name of this aggregate so the other Holy Grail for a conference like this one is the idea of ubiquitous language and the up the large-scale big-picture exploration it turns out really obvious that we won't have one even if many companies still confused yeah an enterprise language with the idea of ubiquitous language or even worse some people want everybody to speak their own language great idea is not going to work one little recommendation well one reason why it's not going to work is well people are leaving the departments their own internal conversation is consistent you're trying to change their names because you're putting things inside the same box the cylinder thing rotating where you put names and you're pretending that you're making architecture and that but the moment you put the two things together inside the same rotating cylinder you're just finding all this name clash or start using composite names in order to solve this problem so this is not the note that this is the order history and this is a purchase order instead and you try to make this distinction and then you discover that the users are not and the stakeholders are not talking with your language they just call in them order or in customer end of the story and you know why because it's not their problem I mean it's your problem of putting things inside the same box and try to find a way to distinguish them they don't put things in the same box so it's your problem you won't have any empathy from from the rest of the company because you're solving a problem that they don't have so if you accept the fact it will be in multiple languages that will form an inconsistent whole well that would be better except that with domain events the consistent hole might be taking the shape of a good well shaped dictionary dictionary of we are specific domain event it might take a little but that thing is the is the only portion of language that has some hope of became becoming lingua franca over all the organization but at this level one of the funny thing is just like everybody thinks they know exactly the name of the thing and we are trying to introduce consistency in big picture we accepted almost everything and in softer design we we end up refining a lot of things so yeah we can't do things right at the very at the very beginning symmetries I've started to tell us something like well you see a command which hasn't had the resulting event it normally at the very beginning sorry I just thought at the very beginning yeah feels like this is too trivial if I have a command which is say by item and advancing saying item but is that really what even storming is all about I'm a really wasting my time rephrasing verbs are the path to verb at present tense no the thing is sometimes I mean not every problem is complex but we're symmetries are lacking that's where your brain should start thinking what why is it not symmetrical here maybe there is a concept which is missing or maybe there are two more concept conflicting and usually developers are looking for a winner and I'm looking for a new challenge to see because maybe both concept might have a reasonable purpose and that they need to be both in in the picture and also you might find a lot of places where you're looking for symmetries but in order to reach them you're you're peeling some part of the language and one of the key areas is usually when things are seen from the customer perspective on the outside and and instead from the inside you see things with which is speaking a different more complex language and the two things should not match oh yes we were looking for ubiquitous language no not really actually one of the hardest thing for people for Dedede practitioners just like they once they understood the problem really well they found a very complex model who's taking care of every single corner case and in a brilliant way we can master everything so what do you do while we expose this really complex model to the user showing well look at this this is fantastic it's not the user problem what you have to do usually is try to that's your brilliant solution hide it nobody should know about it I mean that's what Google does is not telling oh look at all of this option you can have for searching no put your string here make it look like like it's magic but we are so proud of our discovery that normally we feel like showing them I understood it not not yet you you will understand it when it's completely hidden under a incredibly simple UI so and yes brute force does not work if you are going for premature optimization if you feel in like right like it's easy and there oh it's just this that we just need this I agree that class is gonna solve everything no that's not going to work so the best indicator for success in modeling session is actually how much trash are you making that's the best one couple of little thing a reed model apparently is just data they are not it's little switch like my definition of reed model is the data needed for a user to take a given decision well that will result in a command by my by item but the moment you think about this user in a given step of the process what is the data needed how can I make these decisions simple has nothing to do with the data available on the database is what is making me make the right choice typical example timesheet as every time I'm I'm late with the timesheet and as I every developer I hate filling up the timesheet how do you feel the timesheet when you are three weeks late well what is your read mode that you read model is Google Calendar where was I supposed to be on that day well maybe my github come at least I was working on that thing maybe my phone record for that day the email I've sent okay that's more or less what I've done in that day that's my read model if I can I put it in an indecent screen so that is giving me the hint to solve this problem yes but I mean read models are not exposing data they are decision-making tools and we are trying to make users go through a streamlined flow and yes don't get reused get in the way let's put a page when we show all the data that we can reuse for every possible decision in the flow no doesn't work and sorry this creates a wave of I won't do it again and okay we can't do this without a product corner we need we need the business in the room because without the business we will be doing we will be exploring corner cases which are not going to happen in reality and and we will need to make some choices that has nothing to do with the elegance of the softer the the coolness of the implementation we need product owner we need to challenge and to be challenged by the business choice and to make selection to make what we can even make shortcuts i I can accept shortcuts if they are conscious I don't accept knowledge shortcuts let's do a stupid implementation first and then we see now if we know that the good implementation is complex but it's just relevant for 1% of the use cases of the user base then we can skip it or defer it but we know that in this case we will need something like that just being dumped is not a real good strategy and if we can know things in the a cheap way that is fine I mean we're not delivering it for the next sprint but we know that this thing is gonna look like that so our implementation is a little bit more a future-proof it's all a matter of time if you if you waste three weeks exploring then it's may be wasted if you get to the same result in half an hour well this half an hour's really good design style it's funny because it might look like overkill like do we really need to do this for every feature well know I'm creating the the red carpet approach we are on the core we are trying make an experiment we need different development style from supporting subdomains so there is no over design we are in the place that is making gold plating not gold plating we are gold creating you move you move us to a different portion of the system are we going to iterate a lot make an experiment on how to implement logging or security no security is exactly the place where if you do a good job nobody knows about you if you screw it up everybody will know your name and you're gonna find messages written on your car in the parking lot all this stuff and but if you are on the right place this is where experiments pay off and how do we know that we're over it's just what I call this this Bruce Lee State well I don't have the right picture it but it's basically when you feel like some more requirements please so that that's basically it and when you try it when you challenge your your system with all the possible yeah corner cases you try that you feel like yes we can do it it is sound we just separated this thing is happening this bounded contest then it goes there here we challenge it there yeah it makes sense you look at the face of the people like yes we can put this in production we can we can really make it that's where you feel like okay guys open the IDE let's make this flow real so takeaways I'm actually closing first thing is as a set of the beginning even storming is just like a pizza has many possible recipes and so we have domain events in the timeline which is just like tomato mozzarella for a pizza then it add whatever you want with a couple of exceptions so no database tables no pineapple and second thing domain events they are working very well as lingua franca there are a better choice for describing the business and implementing the supporting software it's not because we wanted to go into even driven architecture is not because we wanted to go into even sourcing events are a better candidate for being the ubiquitous language business is speaking in domain events the policies people are reacting where something happens not not because any information is in the database that that is the clothes that it was the key and events are really really good in this in this sense so what happens you might just end up having your independent bounded context and this is good good language on the on the boundaries what happens you might you might discover along the refinement that some information is very fine-grained inside the given area and it might need more general purpose type of description many many possible things and progressively we will end up discovering them around the the possible discussion one of the thing which is interesting just like we are using the same building blocks from a very high level vision type of conversation with the with the sea or the board of the company to the implementation it just loves zooming and allows yeah backward zooming so it might allow allow we're still talking about events big picture process modern software design we still we see conflicts goals blockers boundaries and I mean this is the same story just a different level zooming we can go back and forth realize that the vision was inconsistent or that the process was not viable we can move up yeah up and down in any worse so we can have this story going up and the story going out final thing just this is a pretty good platform for self-organization after seeing this one does this limit like people want self-organizing around the system they don't understand just I'm removing a blocker I cannot guarantee that people will finally now self-organized they want self organize if you threaten them for many other reason whatever but this is a big blocker if the system is clear understandable you can see some very many magic things happen this is what happens in in our company we have even storming for having new cameras in the company we have a policy of radical transparency all the business numbers are available to everybody and the culture of rewarding exploration what happens people are making experiment they are improving the organization even when I'm not around so I'm kind of kind of happy and it ended up being also this one a tool for a Purpose Driven organization who are we making happy that that's actually the whole thing that was my closing slide so I am okay to take to take pictures to take questions and except this one and so yeah thank you okay I think we have time for a couple questions anybody has this is my first time to hear the event storming within my company we are practice safe the skilled I do we have a concept of the future grooming and I see a lot of commonality of that the stickers colorful stickers the wolf paper and are going to event discuss the conflict potential challenge or those kind of things my question is what are the differences well I I don't know exactly how they do it they do it in in safe so there are a few place similar approaches for discovering at the future at the future level use the story mapping or well I actually don't know what is the format for for safe the thing which is relevant for us is just we are starting to build the story and that the story the business has to be consistent at the very beginning so the the it has to make sense which we find usually we end up finding inconsistencies on the day one of the exploration so established companies with the making a lot of money with a lot of employees you just make the business feel visible it doesn't make sense there are holes and bumps and an unnecessary frictions and this is the thing that we we want to discover as soon as possible also we look at the friction between the people so it's not about the model are we able to make those people cooperate what that goal this is exposed during the big picture and if it just already scoped around the software some of the key players might not even be in the room so it's it's my approach to get into a no-bullshit state more or less inside inside the company then well this is good way to shape the conversation for me in because of the last round we are progressively adding precision then it's just like what this really looks like softer to me we never injected soft from the very beginning we just still tried to model a granny generation and then I just need the ID I'm feeling so close to that so yeah dot dot but what happens if different for month and then we but we're really trying to optimize for the overall value generation for the users involve yeah dot dot saw give thank you for the target was great so I'm familiar with work of David Allah Karim who is the corner of the complex event processing term because I mean I just wanna because he there is a notion of event model of the organization and he argues like important of it and I'm wondering if you have any influence from his works not directly but I mean events are there Fox things that happened does you don't need to learn this concept from somewhere else she's just like well things are happening inside the organization people react to things happened it was I mean for me was just like what did it take us so much time to get there it's just uh so no influence but feels like obvious once you wants to discover that but now on couriers yeah okay I think why I would do one more question hello so we we are a group of about twenty five to thirty developers split into four teams five to six twelve person age so for such an organization which you imagine everybody coming together and doing even storming or would it be a team-based activity how many people are too many in a room well I managed up to 35 and nobody asked me for anything more than thirty-five so I don't know it worked pretty well with this with this company it takes a lot of space and an organization there might be good techniques in order to split and partition like we are gang storming is a pretty good source for that so I would say the time needed by note not grow linearly so it might be a lot of yeah merge and and well split and merge type of thing but I would say it still still viable is I mean doesn't to me the thing that doesn't excesses this learning is crucial for the success of the project doesn't make sense to split and to give people second hand learning then of course I'm not forcing anybody to be in the room you say like no I don't want to be there well your choice and everybody's invited appreciated nobody's mandatory to be in unless you feel like we really should ask Tom to be in the room because he's the only one that knows his stuff in this case like well please join us and that is the thing but this actually scales better than expected like the some people are telling the right story some people are guessing but then well people are actually self-organizing I mean in some situation it's okay to be a spectator even the process modeling we usually have three four active player years and but the people in the backseat are actually enjoying it I'm seeing this evolving this looks like the growing solution to my problem and am I just coming with some comments and remark everybody okay yes so it's a little bit easier than it should be real problem is logistics room size oxygen good good food and the beverages at least this type of thing and this might affect the result of the workshop a lot and but for the food the numbers are up 2/3 is manageable more you need more than one facilitator and yep okay thank you we're at time YUM visit event event storming calm and find out where to in between thanks yes it's this reference is here and all of this thank you [Music]

Channel: Explore DDD

Views: 38,485

Rating: 4.9429927 out of 5

Keywords: ddd, domain-driven design, software design, technology

Id: 1i6QYvYhlYQ

Length: 54min 28sec (3268 seconds)

Published: Thu Oct 19 2017