 When you want to summarize your experience, maybe then others will start. Fun. So you had fun watching other people do all the hard work. What about you, Manjunath? Sorry to pick on you, there are a few people's names that I know, so. Yesterday was pretty good. A couple of sessions I thought it would have been more interactive. But I think the kind of questions that were coming up, the opportunity was there and the learning was pretty good. Actually I was more impressed that there is so many people who are looking to learn and adopt from this forum, which is very, very good. Okay, cool. Alright, that's good. So I think we have gained successfully the tag. It's 9.30, so we're going to get started. So the title of the talk is Adaptive Chain Cycle, and it's kind of inspired by guns and roses. Anyone listens to guns and roses? The band guns and roses? Alright, so you must have heard Axl Rose, November rain song, right? And then that says nothing lasts forever, not even November rain. So that's what we're going to kick on off with. But first let's do some warm-up exercises, some teasers, right? No push-ups, just simple brain warm-up exercises. Which line in those three lines do you think is the longest? All the same. All the same? Wow, you guys are really smart. That's the optical illusion, right? It's Mueller's Liars Optical Illusion, which was, you know, this guy first kind of demonstrated how the brain tricks you into believing things, which is not true, but it kind of makes you feel sure that, you know, what you're actually thinking is right. The line 2 looks longer for some reason compared to other lines, right? Let's look at another one, which orange circle in the 2 is bigger? Both are same. What is the name of this optical illusion? Relative to its neighbors, right? What makes you feel that it is bigger than the other? What does this tell you about how brain functions? It's messed up. I think it's quite elegant in the way the brain functions. And what this optical illusion tries to highlight is that we as humans are not equipped to deal with absolute measures or absolute numbers. We as humans are built to deal with the relative things, right? We can look at something relative to its neighbor and decide, but if we kind of do these optical illusions, you know, you see how the brain functions, the inner workings of the brain. Let's look at another thing here. What do you see in the picture? Two faces. A flat vase. The beauty of this optical illusion, this is one of the best optical illusions ever created, the beauty of this optical illusion is that your brain, typically in average human brains, flickers between seeing the two faces and then the vase, and vase in the face, and then it keeps flickering back and forth, right? If it is not, then I think you should get an MRI check. Why is this happening? Ever wondered, you must be wondering what nonsense is this guy talking early morning Saturday. I'm here to talk about agile, and this guy is talking about some nonsense about showing two people's faces in a vase. Like what's that going to do with software development? Is that question coming to your mind? Yes. Yes, that's good. So you'll have to hold on to that question for a while because the suspense of the movie will only be revealed towards the end. All right, now I want you to pay, now that you've done a little bit of warm-up, you're kind of attentive, now I want you to pay a little more attention to a video that I'm going to play next, all right? In this video, there are two teams. One is wearing a white T-shirt, one is wearing a black T-shirt, and they're passing a basketball between themselves. And what I want you to do is I want you to count the number of passes the T-shirt, the white T-shirt team is doing, all right? How many people have watched this video before? So you've got to shut up, right? For the rest of you, pay attention. How many passes did you count? Four people. But did you see the gorilla? Yes. So how many people saw the gorilla? Right, again, time to go get an MRI done. How many people did not see the gorilla? Right, quite a few of them. When they did this experiment, about 60% of human beings did not see the gorilla, about 40% of the people saw the gorilla. For those who saw the gorilla, they told them that they need to get some medical checkup done. Right? For those who did not see the gorilla, they had just averaged normal human beings, or they were actually paying attention to what was asked, right? And the others were not. What is this referred to as? This is referred to as selective attention, right? Where, if you can please turn off the phone. So selective attention is basically the idea that, at any given point in time, our human brain is getting a lot of different stimuli, right? It's getting input from a lot of different places. And what should the brain focus on? So you're watching the guys passing the basketball, you're watching someone walk by. There's a lot of things going on. The way the brain works is it's going to focus on the task at hand, and it's going to blur everything out for you, right? So it's only going to focus on the things that you asked, even though it's getting a lot of input. So this is, again, something interesting to ask how the brain functions, right? And what does this have to do with software development? Coming quickly, good commercial break. I think you guys have already seen my introduction. I live in Mumbai. I have a company that builds this product, I run a bunch of products, run a bunch of conferences, also organize some conference-related software called Bidakon. A few other products or toy stuff that I've done, and that's kind of a quick background about myself. So what I'm trying to highlight with my background is that I'm all over the place. I don't focus on one specific thing. I try to do all kinds of random stuff. Why? Because I'll talk about that in a minute. But let's talk about a little experiment that was done at elderly homes in a bunch of you, you know, elderly homes in the US. So what they had is they had people, researchers, visiting these elderly homes and they divided the group into two groups. There was one group which they called as the high-control group. And the other one was a low-control group. So the high-control group, basically both of these groups got pet plants. So they got plants that are their pets and they have to take care of these plants. But the high-control group basically had control over when they can water the plants, what they can do to the plants, how they will take care of the plants. That's the high-control group. And the low-control group could not decide when they will water the plants or things like that. Someone else would make those decisions on their behalf. And what they realized is when they started looking at the health conditions of these people, the high-control group was doing much better. Their health improved much better. They were off medications and, you know, a lot of improvement, positive improvements in their health. While the low-control group had a lot of their health deteriorated, it was no other change, right? Everything else remained the same, but their health deteriorated. While the other group increased, what do you think was going on over here? Having these pet plants and then some being able to control it and some not being able to control it led to quite an impact on their health conditions, right? What was going on? There was some kind of a purpose to live and that made them feel better, get more healthier. That doesn't make sense. But you're right, there's something going on here, right? Their own brain was working for them. Their own brain was working for them. Their own brain was working for, say, planting. But the other case, it was someone else triggering them what to do. Someone else was controlling what needs to be done. In this case, they were using their own brains to control what things need to be done. So they did one mother experiment, right? They can't just settle with one experiment. They want to do a few more experiments. So there's one other experiment where now, again, high control group, low control group, high control group controls when students can come and visit them, right? So the elderly people can have some time to chat around with those guys. The low control group, the students decide when they will come and visit. The patients cannot decide when the students will come or the volunteers will come and visit. And they saw the same thing again, right? So they saw as the high control groups, health conditions improved drastically, the low control group, the health deteriorated completely. Experiment was done, they got their results, they closed the experiment and they left. Six months later, they went back and visited this elderly home and they found that majority of the patients who were in the high control group, their health had completely deteriorated. About 50% of them had died unexpectedly. So what was going on? Why is this going on? So they started doing more research to find out why this phenomenon, why such a profound impact on the health conditions of people. Another experiment is basically how many people buy lotteries over here? One, two, few people. There's nothing wrong with buying lotteries, so feel free, right? The lottery industry, right, hit a stagnation point at one point where there was not much people kind of buying lotteries anymore, right? So they came up with a very interesting idea. They said you can pick the number on the lottery. Earlier you were given a lottery and then the number is hidden in it. Now they said you can pick a lottery and you can pick the number on the lottery, right? And suddenly you saw that there was a big spike in the lottery sales, right? Quite interesting. The same thing the casino guys tried. How many people here go to casinos? One, two, maybe. But you know what happens in a casino, right? I mean people go there and they play these little games and then there's money involved and this is where people think it's bad. But earlier there was a guy at the table who would throw the dice for you and then they changed it to saying that you throw the dice, right? Let not somebody else throw the dice, but if you're betting then you get to throw the dice. And suddenly same phenomena, big improvement in the number of people gambling in the casinos. Do you think it'll make any difference if someone threw the dice or you threw the dice? But then all these guys are idiots who started playing, spending a lot of money and gambling. End of self-confidence. There's a self-confidence in me throwing the dice. Somehow I can influence what number will show up on it, right? You realize what your brain is doing to you in all these examples. Your brain is tricking you, right? Your brain is tricking you into believing something which actually does not make any logical sense, right? If you put aside everything and you just look at it from a logical point of view, all these behaviors, these spikes in lottery sales, this spike in people's health improvement, this spike in, you know, this casino sales, they don't make any sense because it's completely illogical. But obviously it's a pretty well-proven phenomena across different places. I mean health conditions across gambling, across lottery, a lot of different things. What is going on? This phenomena is referred to as illusion of control, right? We are born with a strong desire to control things. We want to be in control. We want to control things. If your brain cannot, it tricks you into believing that you are in control, right? Throwing the dice makes you feel like you are in control of, you know, winning. Picking a lottery number makes you feel you are in control and you will be able to win. Being able to control when you water the plants or not makes you feel that you are in control and your health conditions improve, right? All of this generally is referred to as the illusion of control, the desire to control things that we cannot and the brain trying to trick you into it. Let's look at one last experiment. Electric shock experiment, a shock therapy experiment. Same phenomena. High shock group, low shock group. High shock group got five volts, five shocks or five volts each. Low control group got three shocks of two to four volts, slightly varying voltages, right? These guys got every 10 seconds a shock, while these guys got shock at a random time, right? So one group got five shocks of five voltage, another group got only three shocks of lesser voltage between two and four. Which group do you think sweated more, felt more fearsome, which group was more afraid in this whole experiment? No one. Second group, right? That's very obvious. But why didn't those guys figure out that they're actually getting less shock and less number of shocks, and they should have been much more easier, right? Uncertainty. It is uncertainty, right? It's the uncertainty that's actually tricking your brain to get fearsome, right? And why is this... Why do we fear uncertainty? Why don't we humans like uncertainty? Loss of control. Loss of control, right? Because of our desire to control, because we cannot control, now you feel afraid. Is it making any sense? Yes. Now I'm going to turn it and kind of start relating it to a little bit of what we do in software, right? So how do we deal with uncertainty in software? What do we do to deal with uncertainty in software? A lot of questions actually came up yesterday during the discussion, right? You've already estimated how do you deal with this, how do you do project planning, fix scope, this, that, lots of different things, right? So what do we do to deal with uncertainty? A, we forecast, right? How do we forecast? We break down the project, we do requirements analysis, we do estimations, we do all kinds of interesting things to forecast, right? And then based on the forecast, we plan what needs to be done. We distribute the work, we plan, we do a lot of interesting things. To be sure, we plan again, because last time when we had planned, it didn't quite go well, right? So let's be double sure this time, let's plan again, right? And then eventually if you're done with all the planning someday, then you get to the execution, right? One problem with this approach is what is referred to as the predictability paradox. The more I know, the more unknowns get revealed. And I want to spend more time, a little more time trying to resolve those unknowns that I got. And then when I start trying to understand those unknowns, I end up spending more time and more unknowns get revealed. So at what point do you draw the line and you say I know enough to get started, right? So it's a paradox. When do you know you've got enough information the more you learn about something, the more you realize you don't know something and you want to spend more time on it. Is it making any sense so far? Anyone's experienced the predictability paradox? No one? We all experience it, right? Whether software or not, we want to spend more time to analyze a little bit more so we'll have more details so we can plan better, right? And then when we start doing that, we realize that we need to spend a little more time to know for sure that this is actually going to work. And it's very hard to draw the line that I have got enough information to get started, right? So let's hear from Dave Snowden. Dave Snowden is one of the leading guys in the knowledge management community. He does a lot of interesting work. For example, he helps the Singapore government do the antiterrorism program. And he does a lot of interesting things. So Dave Snowden talks about how there are three different models of planning things like this and how you deal with predictability paradox. And, you know, the desire to control and things like that. So I'm going to play a little video where he's going to talk about how to organize a kids party using three different approaches, all right? Let's imagine if you can that you've got to organize a party for a bunch of 11-year-old boys and you want to apply the three different types of system on your nature. But if you assume the party's chaotic, the children are acting around them, so the children can go on a personal experience of self-discovery. Your house may dare burn down in the process, but what does that matter? All property is theft and it was socially constructed in the first place. My friends in California who've tried this, I definitely recommend it. The recovery cost is high, but it's a legitimate approach. On the other hand, the one we'll be more familiar with is the order systems approach. Here it's of critical importance to construct clearly articulated learning objectives for the party itself. The learning objective should, of course, be aligned with the mission statement for education in a society to which you belong. Ideally, you should print the learning objectives off on motivational posters with pictures of eagles, sorenoma valleys and water dropping into ponds and place those around the room where you can hold the party. You then produce a project plan for the party. The project plan should have clear milestones throughout the party against which you can measure progress against ideal party outcome. Once you've done that, the senior adult can start the party with a motivational video take. After all, you don't want the children wasting time in play which isn't aligned with the learning objectives of the party itself. And then they should use PowerPoint to demonstrate their personal commitment to the objectives of the party and to show the children how pocket money is linked to the achievement of the master and targets. Of course, the third approach, the complexity approach, is even simpler. Here, we draw a line in the sand known as a boundary in complexity theory and we turn to the children and say, cross that little bastard's and you're done. And one of the things you learn pretty fast as an adult is the value of flexible, negotiable boundaries because rigid boundaries have a habit of becoming brittle and breaking catastrophically. We then use catalytic probes and I'm deliberately using the jargon of complexity theory now. We take a barbecue or computer game something which will stimulate a pattern of activity which is called an attractor. And if it's a beneficial attractor we stabilize it, we amplify it. If it's a negative attractor we doubt in it or destroy it very quickly. So what we do is we manage the emergence of beneficial coherence within attractors within boundaries. And in that simple phrase we see the promise of complexity theory for organizations and government alike. Does Dave Snowden make any sense to you? Does it strike any chord with you? No? Let's say let me give you a shot to try and explain and decode what he said. Dave Snowden came up with this model called the Canadian model. In the Canadian model he tries to break different kinds of systems into these quadrants. And he says you can break systems into these set which is predictable systems. These are systems where I can predict what's going to happen. And these are set of systems which are unordered systems which means I cannot predict that. And then in the predictable systems he talks about the ordered system he talks about there is a domain called simple domain. In simple domain for every input there is exactly one output. So if I turn on the light, if I turn on the switch I know the bulb is going to glow. For one input there is exactly one output. And that's a simple domain. In that kind of a domain I can sense I can categorize and I can respond and I can come up with best practices. Sorry Venkat, but there is this domain in which you can actually come up with best practices. Because there is a straight one is to one mapping between you doing something and something happening. So those kind of systems are very easy to deal with. Then you have the complicated systems. The complicated systems are where there is one input and there could be one of many outputs. But there is a defined set of outputs that is possible. So given an input if I sense I analyze in this case, I analyze what went on and if I have subject matter expertise I can predict what's going to happen next. Because I have subject matter expertise on it and I can predict what is going to happen. And in these cases you don't have best practices but you have what is good practices because these are practices that you can attain after a while after enough expertise in a particular subject matter. If you take of this domain like let's say your project is a very simple project then you could easily apply best practices and you can do it and maybe it requires 40 hours of training and you are often running. But when you talk about complicated domain this is where basically you need to spend more time to attain the subject matter expertise before you can apply some of these practices. Then we will come down here which is the chaotic domain. The chaotic domain is your input has no output relevance. If you do something there is no direct correlation to the output. What is an example of a chaotic system? The Indian traffic system is a classic example of a chaotic system. Is it bad? Actually chaotic systems are referred to as one of the most novel systems where innovation happens. A lot of companies fear chaotic systems but actually chaotic systems is where interesting things happen. Chaotic systems are where you act which means you have to do something and then you sense and then you respond to what has happened. This is where novel ideas come from and this is where if you have enough data points and you apply statistics on it you will be able to guess some of the patterns that are going to emerge and some of the things that is possible. If you have enough traffic system in India you will be able to figure out things but you have to do something first to figure out what needs to be done. Last we talk about what is referred to as the complex domain. In the complex domain they talk about the system which is at hand and then they talk about agents which are people doing things on the system. The agents basically the agent and the system co-evolve which means when the agent do something with the system the system will change and then the agents will respond accordingly to that system. So humans as a system if you think of human beings as a system we are inherently complex systems. For example let's say I walk up to some random person and I go slap the person do I know how the person will respond? I have no clue what the person could either hit back or the person can say I am going to take the Gandhian approach and I am going to show my other cheek to this person. I have no clue what the person is going to do but let's say the person did not respond. Can I be sure that the next time I slap this person he will behave the same way? He or she would behave the same way? I cannot. Why? Because the system and the agent will co-evolve this person might think I am going to keep quiet I don't know this guy is going to keep slapping me so I need to react. So the system and the agents can co-evolve and if I see the person reacting maybe I will behave differently. So this is kind of interesting. So if you think of if human beings were simple systems I can be pretty sure what the person is going to do. If there were complicated systems then if I had enough experience with that person 10 times after that I would have been able to guess what the next move is. But we are neither of these we are actually a complicated system and I would argue most software projects actually fit into a complex system. What that means is that let's say you show a demo to your customer of a product. It's as per what they have asked but once they look at the demo they say now that I see the demo I think this could have been done this way or we could also do this other thing or we could do this slightly differently or you have not understood what I wanted. Any one of those things so the customers changing their mind the software is evolving along the way and the people building the software are learning and evolving along the way so there are so many moving parts in that and it's not easy to predict what's going to happen next. So software fit into this context and he talks about you know how to apply the ordered system approach which is what he describes in the example where you create a project plan and you plan everything and then you kick off the party with a motivational videotape and then you align it with learning objectives or milestones of the project and you tie you know basically salary appraisals with that and that system works really well in these domains right. These are domains where things like assembly line stuff where you basically have to assemble things where you don't have to use too much cognitive power right I have to lift a bunch of bricks from here and put it over here that kind of a system really fits into these domains and that works really well right but in software which belongs to this domain if you try and apply those kind of systems to this right that kind of planning process to this is going to fail badly and that's what we see over and over again in project software projects so that's what Dave Snowden's work is all about in terms of understanding which domain your project fits into or where if you're dealing with a system which category in this it fits into and then make according decisions based on what kind of practices you can use here we talk about emergent practices which means at this point in time maybe these are set of good practices but those will emerge, those will evolve over a period of time and something that was good maybe a month ago is no longer a good practice when I go visit teams and I see them doing daily stand-up six years after they started doing scrum I feel they're like totally missed the point they totally missed the point of agile evolution the whole you know this kind of a model but I'll get back to into this in a little bit more detail but this is kind of a quick summary of what he talks in his video what he's trying to hint at that you're trying to apply the wrong planning and the wrong project management practices in a domain that it does not suit you which is why you see a lot of failures and you see a lot of problems because you're trying to apply the wrong thing clear so far? I already talked about this company yesterday but I guess you guys know which company this is I guess you know which company this is also can anyone tell me why are these guys so successful because they focus on innovation right? what else do they do? they focus on quality what else? creating a brand why is Toyota very popular for the lean manufacturing system Toyota is very popular for its Toyota production system, PPS and we know every detail about how Toyota manufactures cars how they manufacture things what does the just-in-time process that they use, what kind of mechanisms they apply, Kaisen and Kaikaku and all of those Chinese words alien words, we know everything we know how Apple works how Steve Jobs used to run things inside Apple right? but why do we only have one Toyota or one Apple? why can't we replicate what they have done because we know exactly why they were successful right? all those things, so we know exactly why these guys were successful why cannot we have another Toyota or another Apple they evolve much faster than other companies you're dealing with complex you're dealing with complex systems right? so it's not easy to take best practices because we can look at what they are doing we can copy what they are doing but guess what? you will not be as successful as them if I were to summarize in a layman's language this is how I would put it products and processes are like haircuts copying someone else rarely works why? go back to this and see where this fits into what you are trying to copy, where it fits and you know there is what we refer to as retrospective coherence which means hindsight doesn't lead to foresight knowing something happened in the past that does not mean we can predict the future do you know what happened in 9-11 when the guys flew the flights into these twin towers we know what happened during that time so they went and did a lot of study of how these guys managed the whole operation and what they realized was that there were a bunch of these guys who were taking lessons to fly flights in the US but they were doing something different everybody first thing they learn when they learn to fly is how to land the flight that's the first thing everyone learns and then they learn how to take off and then they learn how to fly the flight in air because that's an easy thing to do but these set of people were doing something very different they were not learning how to land the flight they were not learning how to take off the flight all they were learning was how to fly the flight so these guys said wow that was something that we missed if we had observed that, if we had watched for that then we should have been able to stop something like this happening because that's a that's a weird thing someone's doing that's not normal and if someone's doing something which is not normal then we should be able to catch that and stop it from happening so now we know how these guys did what they did what went on and can we stop something like this from happening in future what do you think no they are also smart guys but you would be surprised if they do the exact same thing and still get away with it the problem is they will not do the exact same thing it's a complex adaptive system they will take a slightly different take at it and they will strike back in a different way and it's very difficult because the governments or these companies tend to follow this same set of practices and they now make those as best practices and they make you take your shoes out at airports even only one guy ever tried to carry a bomb in their shoes but everyone still continues to taking their shoes off guess what the next guy is not going to take a bomb in his shoes and walk into an airplane that's not going to work so anyway what this is referred to as retrospective coherence hindsight doesn't lead to foresight just because you know how it was done in the past does not mean you will be able to predict how it's going to be done in the future but this is good but what's interesting about why Apple and Toyota are very successful and why you know other people cannot replicate it why other people are struggling replicating it but what's interesting is do you think Apple and Toyota will succeed forever well their success lasts forever because you said that they're always constantly ahead of the game they're always doing better things than the rest of them so do you think they will be able to sustain this forever nothing lasts forever right as long as they can adapt but do you already know that Apple is having a really hard time and Toyota also is having a really hard time in fact the Japanese are having a really hard time in fact there are articles talking about the downfall of the Japanese economy right because nothing lasts forever but why that's what is very important for us to understand because then I'm going to talk about why Agile is not going to last forever and it's already declining and it's already dead in my opinion right you should move on to something else so let me explain that with what is referred to as the adaptive chain cycle the adaptive chain cycle basically has the theoretical backing of why nothing lasts forever why systems don't last forever let's imagine that there is a fresh ground where you know newly cut forest and now this forest has started growing right so you have a lot of these grass and these mushrooms kind of coming up and what they're trying to do is they're trying to take easy resources and quickly grow they kind of quickly grow and get into some kind of a state once enough of this grassland stabilizes and once enough of this habitation stabilizes then slowly you will see that that getting converted into thick forests right and thick forest starts growing in that same region so you know what they have done is the slow moving species came in took the easy resources and stabilized and gradually because of that the thick forests came up and then they started stabilizing and then once you have this thick forest do you think they will last forever these thick forests what they're trying to do is they're trying to maximize how much they can consume how much they can extract out of the environment and then how much they can preserve what they have how much they can retain what they have but because they are so vested in their specific configuration of what they're trying to do they become very vulnerable to things like forest fires or diseases or things like that and you know a typical example would be a really thick forest which is grown very tight with each other goes dry and forest fire basically burns it down right and when the forest fire burns it down it basically releases all the all the resources that it had consumed from the environment back into the environment right so they go into a release mode and then once they go into a release mode now this creates ground for new kind of species new kind of plants to kind of mutate and try something different right this time so they kind of this is how evolution works right so new kind of species try to come in bunch of those species will die out a few will stabilize and the few will basically turn back into the exploitation stage where they will quickly try to grow and this cycle keeps repeating forever right that's referred to as the adaptive chain cycle this same cycle also applies to social systems or to companies or to what we are trying to do for example let's take the automobile industry right in the automobile industry in the very early days there were bunch of people who were basically tinkerers right they were these passionate guys who were basically shrinker tinkerers and they would sit in the garage and keep trying these different experiments of how cars should be there done right they didn't even know it was cars they were just kind of tinkering and they were saying can we do something with this a few of those guys basically stabilized and they figured out a way to make something that is you know usable by people right and then they started exploiting resources they started quickly getting market share and trying to build a sustainable business model right once they started doing that and people saw you know the succeeding they turned into the big large corporations right the the fords and the general motorists of the world which basically tried to optimize everything tried to optimize conserved as much as possible in terms of cash flow and maximize what they can get out of the market right so they went into a conservation mode and rapidly grew into all kinds of interesting things what would happen when they did that they are so vested in a particular type of cars that they are making or in particular you know fuel combinations or these kind of things they would try and optimize on those and they would mass produce those things right very quickly the the consumer the consumer behavior around them would change right because in the US if you see few years ago the whole SUV trend was very big right everyone wanted an SUV now you go to the US not many people want an SUV right that the trends are changing there is the impact because of global warming because of other kinds of things and what's happening is now all the big companies are so vested in manufacturing these gigantic SUVs and nobody wants to buy that so what ends up happening they fall back right they they get into this mode where they really cannot sustain themselves they file bankruptcy because they cannot deal with it anymore and they basically release all the all the experience they have into the market for newer people to come back and do interesting things like say say or what Google is doing with personal transit or kinds of interesting things right and then that would stabilize and it would go through the same cycle over and over again so that's referred to as the adaptive chain cycle and we see this pretty much in software development every 10 years when it comes to processes if you remember back in the days people were so vested in CMMI and things like that right how many of you here CMMI being talked these days some probably but it's obviously not as much as it used to be back in the days at least in the big companies these days nobody really is talking about CMMI what people are talking about these days a lot of companies at least I visit are talking about Agile Agile is the new CMM for many companies now that's the new shiny little thing that everyone wants to go after and will that last forever we already see that a lot of companies are realizing that that's not going to work forever so let's look at a little bit of background how does this happen if you remember back in the days before even Agile existed there were all these interesting people who were working on projects on the ground and they realized that there is these heavy weight methods don't work for us so they kind of created their own little methods you know extreme programming DSDM, Scrum, Crystal lots of different things that were created by these guys and all of these guys started realizing that there's a different way of doing things and they're kind of tinkering and trying to figure out what is a better way to do software development then a bunch of these guys got together and they said you know this actually makes sense so can we come up with something which can uniform all of these guys in one package and we can sell that package and they called it Agile Agile was actually invented much after these methods existed so they got together kind of branded it they put it under the umbrella of Agile and they said well let's reorganize let's stabilize, let's try to exploit let's try and expand as quickly as possible to every single company and then what happened they were very successful at that they were very successful at that and now they're into this model which is the conservation model what they're trying to do is let's not innovate let's not try something different this works so let's just standardize this and let's expand this across the whole entire company let's make sure everyone follows the agile way of working right and what's going to happen because of that very quickly they're going to become vulnerable to you know not being able to live up to the innovation that's happening around you the whole thing with the mobile and how things are going on struggling with some of these checklist driven approaches the agile checklist here are many things that you should do to be agile kind of defeats the whole point of agile but that's where we are today as an industry and I would actually argue that we have already fallen we've already fallen down into the release mode where people have started moving on people have started talking about lead startup people have started talking about other kinds of things which work better for them and agile does not work for them and they're trying to explore other kinds of things which works for them and then I would argue that these guys would soon stabilize, write books and become very popular and offer certifications and do all kinds of interesting things get into this mode and one day you will be standing here attending a lead startup conference where people will be saying you know what we've just fallen down into this mode and there's a new little thing called XYZ it has to go through this cycle it has to go rapidly through this cycle otherwise you cannot sustain it and in our industry we see about every 10 years we go through this cycle agile is already 10 years old so guess what we have already fallen over I would argue from that and there is too much of dogma in my opinion in the agile community today people are saying oh don't question this is how exactly it needs to be done this is standardized you ought to do this thou shalt follow this right and that kind of contradicts the whole agile manifesto in the first place but this is where we are whether the guys who wrote the manifesto like it or not this is where we are this is the current state of reality so which is why these are few quick references if you want to refer to which talks about you know adaptive change cycle and how you go through this and this is why yesterday during my talk I talked about being agile over following agile is important being adaptive being rapidly able to change and being able to respond to things is more important than following agile I follow agile I do this daily stand up I do this I do that that's not important what is important is being able to adapt being able to respond quickly and as far as you can do that you are agile enough to sustain if you are not you will be caught up in dogma and suddenly one day you will flip over and you will say this none of this stuff is working and we are so heavily invested we have changed all the titles of all our people from project managers to scrum masters we have got all of these guys certified we have invested so much money in this that now we don't know what to do right we are stuck here with this this is what happened with CMM companies were so heavily invested in this whole CMM they got all of this stuff done and they were saying well but this is actually not working it worked for a while for marketing but now even that is not working because customers are asking for agile they don't care if you are CMM level 5 anymore they want if you can do agile so that's not working now so let's go down this round and then 5 years 10 years later they say well that's not even working because customers are asking for lean startup can you do continuous deployments all those kinds of things and that's not working for us anymore so don't get caught up in following something you know focus on being agile focus on staying agile focus on staying ahead of the game alright I want to take quick questions and then I want to summarize some of these things that I have randomly thrown at you guys early morning Saturday sorry for doing that but I want to summarize the whole thing but I want to take few questions before that I am very enthusiastic it's like I come to this conference to talk about agile in this case don't do agile the ratio is said being agile so as you said yesterday we are not following agile in most cases we follow the mini waterfall in most cases so what do you think being agile and following agile would make difference for us being agile was just following agile this is the organization's in fact there is a very famous saying which talks about organizations like bad habits even though at the risk of getting themselves extinct so organizations structured themselves in the waterfall model so they created all these departments and all of these stuff so when they move to agile they don't want to throw that away so they want to make sure they keep all of that which is why you see that you have a testing team you have a development team and they end up basically taking what they were doing and squeezing it down to a mini waterfall in some time they will realize that's not working and they will reorganize themselves to be truly agile in the sense where everyone's cross functional and everyone's able to do every little thing the whole generalizing specialist model but once they get at that point they will realize that that's not enough there is something else that they need to do so the point I am trying to make is just trying to follow something you need to ask yourself what works for you, what does not work for you how can you innovate constantly how can you go through this cycle the cycle of adaptive change how can you go through this cycle rapidly I was working with one company in Mumbai it's a small company they do very interesting work and the founder said when he looked at this he said my vision is to take my company through this very rapidly what does that mean so he said we built a product the product started gaining a lot of traction we had about 10,000 users using this product we made a good amount of money and we decided to kill the product and come up with a new product because he said this is making money but if I kept pushing and try to get more and more customers at some point the company will be so vested on that product that it won't see anything else it will miss all the market opportunities at some point this product will not be able to sustain any more and it's just going to fall and collapse so what he is trying to do with this company is take something kind of stabilize it level quickly get to this stage and as soon as he realizes that there are no new problems that we are facing with our product he says it's time to slowly start cutting this product out and start building another product so that's one way to basically look at another point of view if certain practices seem to have stabilized seem to have give you benefit and you seem to have getting into this mode where it's now sustainable people are doing this on a regular basis everything looks good then that's the time when you need to stop and you need to say what in this can we throw away what practices can we throw away because we don't need them anymore and how you can kind of go into this new and try and do that and go through this cycle so if you're conscious about this cycle you'll be able to throw away things and you'll be able to experiment with new stuff in small safe pale experimentation right don't try and do it at a company wide level try and do it at smaller team level and once that stabilizes then kind of move through this cycle quickly so be cautious about this and quickly try and move through this cycle because that's going to help you be agile rather than follow the agile right does that answer your question right how many people throw away practices one two very difficult to unlearn very difficult to unlearn but it's much more difficult to suffer failure in my opinion when did and I met and Sri Lankan we were talking about how I built one of these product that I recently built I built it with zero tests right I'm kind of proud about it in the sense that I believed in something for 10 years I followed it for 10 years and it gave me a lot of benefit but I felt I'm kind of getting stuck in a dogma I think this is the only way to build product and that's very scary so what about building a product where I challenge all those practices that I followed and do something kind of very different and I realized there were some things that I had assumed that will be very difficult will be very costly or things like that which actually did not end up being and some things did bite me back so there's a kind of learning lesson that you learn but for that you have to get outside your comfort zone and try and do something that you believe is sacrosanct it cannot be questioned you need to ask yourself what happens if I don't do testival development and I try to do it at a small new product and even if it blows up it's not going to stop the world from working it's a safe fail experiment if I fail, no big deal so you try and innovate or learn with such kind of practices and that's very important to do so it's important to throw away practices it's important to try new things it's important to not try for the sake of trying but try for the purpose of learning for the purpose of improving and getting better have you guys heard these two terms called kaizen and kaikaku kaizen is very popular everyone talks about kaizen it's a small gradual improvement trying to make things slightly better little by little but very few people talk about kaikaku which is disruptive change trying to do something completely disruptive how did Toyota do something very disruptive one of the examples was they tried to take basically they realized that we're running out of fuel and that is not a good way to manufacture more cars so they wanted to do something very disruptive and they realized that electric cars are not going to work so they tried doing this hybrid car they were the first guys to try this hybrid car where it runs on both of these things that was kind of against what the company was making a lot of money already in why would I want to do something when my company is already making so much money on this but they said that at some point is going to run out they're doing some disruptive changes today to be able to sustain in the future have they got it perfect probably not but they're on the right path in the sense of trying to constantly challenge their own best practices any other questions that's great so I'm going to try and quickly summarize what we have talked about in terms of 5 key takeaways from this session what are the 5 key key takeaways from this session beware of the illusion of control and predictability paradox right is that a takeaway from this session you have to be aware of this illusion of control how your brain can trick you into believing something that might not make logical sense right but while you are in that mindset you will always think this is the best thing to do so you have to step out of that mindset and see if this logically makes any sense a friend of mine always keeps saying action precedes clarity right you got to act before you get clarity you cannot be sitting there and saying let's plan everything let's try and get everything down to the finest details before we start anything right because the cost of failure will be very high if you do that guess what you will never be able to do because you all spend all your time planning and by the time you're done with planning you're already obsolete right so we always say planning is important but plans are useless and planning is not a one time thing it's an ongoing thing let's take an example of how we run this conference we want to run this conference and we want to know every detail about how this conference will be run how many people will show up I want to know the exact number right because my planning is based on that how many people will show up what our cost will be how we will divide the cost so we can break even so I need to know exactly how many people will show up at the conference can I know that three weeks before we had 26 people registered for this conference going by that we should have taken a small hall somewhere and run this conference but last two weeks we had 120 people registered completely if I had planned based on that the whole thing would have gone for a toss right so what we had to do was we had to be very adaptive we had to say well let's try and plan just enough and when we get more information we will plan further right so there its action precedes clarity you need to act you need to do some things you will get more clarity as you proceed beware of the retrospective coherence copying someone else's stuff rarely works trying to mimic what Venkat and I did yesterday will not work for you guys you need to figure out what works for you but you need to try first you can't just figure out something without trying anything right you have to first try something you have to follow maybe for a while and then question why is this important can I do away with it can I do something different so beware of the retrospective coherence don't blindly copy someone else because someone says this is the best thing to do don't blindly follow it it's never the best thing to do for you how many people believe that you can achieve perfection in the first shot how many people believe you can attain perfection in the first shot when you try to do something right away you should be able to get perfection nobody believes but all of us behave like we believe in it right actions tell something else actions tell that everyone believes that you should be able to get things right in the first shot we don't have much time on planning so much emphasis on estimation because we believe we can get it right the first time right rarely in any field you will be able to do that and last thing is nothing lasts forever so all of these you should question because none of these will last forever you need to go through this adaptive chain cycle of reorganization exploration conservation and release rapidly and you need to figure out how to do this rapidly in the companies inside your organization to be able to stay ahead of the curve so you are not caught by surprise and that's basically the the final message is be careful not to do this with agile any questions I am done before time no questions then we can break for nothing because there is no coffee yet there is a question over there Karthik are you trying to say people are saying agile is already dead and you know there is something else coming up because people are still struggling to adapt to agile at an enterprise level that's a good question what do I get by scaring people struggling to adapt agile for whatever they can have their own hand number of pieces your piece of advice my piece of advice is to A get a sense of what's going on try and identify where your real issues are forget agile forget any of these other things try and understand, try to introspect and see what real issues your company is facing where are your biggest bottlenecks today is x your biggest bottleneck or is y your biggest bottleneck try and identify what's your biggest bottleneck once you identify your biggest bottleneck apply what I call is just in time practices just in time practices it can come from anywhere it can come from lean startup it can come from agile as far as that practice can help me resolve the bottleneck can move the bottleneck somewhere else I should be good so if you constantly keep introspecting and seeing where your bottleneck is and applying a just in time practice constantly make progress having said that you also need to have some kind of an overall philosophy or a value system that you believe in so it's important for organizations to also have their own value system they can't just randomly keep following whatever comes their way there are certain things, in fact last evening we were having a big conversation over here about as individuals we need to have certain value system we need to know what is okay if your company is pushing you or if your customers are pushing you to work extra it's okay you take up once then that becomes the expectation next time they push you to take more work then that becomes the new status quo at some point someone has to put their foot down saying this is not acceptable at the risk of losing my job this is not acceptable to me I cannot deal with this and here's the reasoning behind it I'm just not being arrogant so when people do that then there is a value system there's a culture that comes in the organization and then the culture is important and then there's just in time practices and using that kind of an approach is useful the other thing that I see people struggle a lot is just the awareness of what's going on a lot of people are not aware of what's happening out in the industry and if you're going to work inside your company with the three people that you always work you will not know what's going on out there there's a lot of stuff that's happening outside the industry that you need to keep aware of you need to keep up to date with ideas come from everywhere things like twitter is a wonderful thing because you get exposed to so many ideas of people what they're talking what they're doing, curiosity raises so you need to be cautious so to summarize understand what your bottlenecks are and then use just in time practices culture or a value system for your organization and be aware or conscious of what's going on outside so you can take advantages of some of those things right, that's my advice in three sentences you said like when you adapt a change like for trying a new method you can't try an enterprise level try a theme level so maybe a method which works for my team may not work for the other team so how do you choose at an enterprise level how do you decide which is best practice that's your last sentence is how do you decide for an enterprise what is best practice is oxymoron in my experience right there is no such thing for best practice for an organization if all your projects are doing the same thing they're lifting 10 bricks from here and putting 10 bricks over here so I can define best practices but unfortunately all our projects are very different so what is common across all of this is what I was trying to highlight is the value system is the culture of the company that's what is common across all of these projects all of them believe in doing quality work all of them believe in investing in keeping sustainable base things like that, those values of those principles is what is common across the company practices that they use are useful only for their context right and they do it they get benefited out of it and then they throw it out and they move to something else why this fascination with standardization I don't get it right are we trying to convert human beings into machines if we want to do that then yes I want to standardize, I want to basically program it so that I can basically replace all the humans with machines but we are not yet there maybe in few years we will get there but we are not yet there but if you insist that we should keep trying then good luck with that alright, does that answer your question enterprise best practices is an oxymoron that does not exist what exist is organization wide culture values that we believe in as an organization some companies have these beautiful things I go into their office and there are all these posters with beautiful values written on that when you see that you know that those values only live in the posters not with the people you go to companies where they don't put these fancy posters and there you will see the values actually stay with people the moment people have put it on the walls you know this organization is doomed it's standardized now everyone ought to follow these integrity is one thing that I see everywhere when you look at people and you say what on earth is integrity like let me look up the dictionary again ask questions before we break yeah, there you go Nanishka I think I just wanted to just expand his question probably in organization they have these standards like ISOS, PMMI and I've seen that in our firms that we have these standards that we are agile but due to the fees that price can't flag we can't see it's completely agile but then we have to face a non-confirmance group call so that's why the how to explain to them because from the auditor's perspective from agile that's all they've read in the book and they see that okay in user stories this is how it needs to be written but your user story doesn't match that they need to have these clothes which they have read in the agile so that I want to format yes right and then your stories don't confirm to the standards which means you're not agile yes what do you do? so now it's like just for the auditing purpose we make them this week those and say that okay these are the new sexes since criteria and just we face it but then we'll do it as we see it's like it's working good for our project we adapted and just for the auditors it was we adapted to those names and tried to convince them that this is what's agile but I agree to you you want to succeed you need to just do the retrospective and think about what works and this is quite difficult to convince people who have got the standards like I also okay if you're going agile it's fine they'll just compare with two projects you're seeing that you're following agile the other project is also following agile what's important is to ask why they have these confirmants or audits in the organization what's the purpose of these audits right? the purpose of these audits the purpose of these audits I believe because I've had this argument with many people in many companies right the purpose of this audit system is to give like this is the statement they will make is to give the executive management a dashboard to look at across all the projects which project needs attention and I was I was with one organization recently and I said this is fantastic this makes perfect sense but tell me how many projects that you audited and they showed green on your dashboard they were all good and then they backfired suddenly what percentage of projects the guy hesitated I insisted like how many just tell the number it doesn't have to be accurate just give me a ring and he said about 60-70% of the projects they looked green on the dashboard but suddenly they backfired and I told how many of your projects which were actually highlighted as red they did confirm they didn't have something and they were in your red zone and how many of those projects actually succeeded with you know CSAT that's a customer rating of you know 515 and he said out of 5 4 of them and your time by bringing me and paying me all this money because the answer is very obvious what you're trying to do does not work so why try and apply it on agile projects because that's going to make it even worse it does not already work now you're saying oh it doesn't work but I want to also apply it on agile projects where are we heading with this so I think you need to again that's what I was saying you need to have a value system and you need to question you need to put your foot down if you're not willing to do that then you have to deal with it it's as simple as that if you want best of both you're not going to get it and I want to confirm also and I want to be agile also that's an oxymoron that's not going to happen you have to help people understand why this is not going to work and why it's not already working they can see it someone needs to just probe it what should be done this is not working what should be done that's something we need to figure out for our organization we can't copy what somebody else has done because that's not going to work either and if organizations are not willing to do that that's perfectly fine nature has its way of going through that cycle you don't have to worry the nature will take care of it go through that cycle that's the answer I would give thank you guys the coffee is ready