Don’t Let Scrum Make You Fragile

My spouse has a personal trainer who makes him perform all kinds of healthy practices. One week it’s the lateral lift, another week it’s the biceps curl. Yesterday the focus was on the bench press, but tomorrow it could be the knuckle buckle. What’s important here is that it’s never the same thing. In order to become healthy a human body should not get too comfortable performing the same routine. The real value is in continuous adaptation to unanticipated stress.

Nassim Taleb called it antifragile, and described it like this:

Now the crux of complex systems, those with interacting parts, is that they convey information to these component parts through stressors, or thanks to these stressors: your body gets information about the environment not through your logical apparatus, your intelligence and ability to reason, compute, and calculate, but through stress. […] Thanks to variability, small variations make them [complex systems] adapt and change continuously by learning from the environment and being, sort of, continuously under pressure to be fit.

– Nassim Nicholas Taleb, Antifragile

Different words, same idea.


A complex system benefits from not following the same practices over and over again. By adding a little bit of stress, and continuous variability in the environment, the system learns to become fit and healthy.

Here’s another example.

I’ve always said that every child deserves a good dose of bacteria and viruses. This significantly boosts the child’s immune system. Instead, we raise generations of offspring with more sicknesses and allergies, because we’re protecting our children from healthy infections. Again, it’s the same thing. Short-term stress leads to long-term resilience. If you protect their bodies from a little bit of harm, you raise your kids to be quite a bit fragile. Do them a favor, and feed them some dirt. Because you love them.

It’s no different for organizations.

You should grow your teams to be antifragile.

No Frameworks, No Methods

I am in favor of unexpected changes, which is why I am against rigid and defined frameworks. Like Taiichi Ohno, the father of Lean thinking, I am against codification of methods. When you prescribe safe practices you introduce stagnation. It’s a short-term benefit, leading to a long-term danger. A short term protection from harm, culminating in long-term fragility.

Of course, people cannot learn values and principles without practices. We all have to start at the Shu level. But when you present a collection of good practices as a method or framework, you forget about the nature of complex systems. We learn from uncertainty, variability, and surprise. What strengthens your health is not that weekly relaxed lounge in the sauna. It’s the unexpected and dreadful plunge in an ice-cold bath afterwards.

Every regular practice works, until it doesn’t. Are the daily standups losing value? Try daily water cooler talks. Are people getting too comfortable sitting together? Move them around. Are the retrospectives not working? Buy them some drinks at Starbucks. Is a team too dependent on its task board? Hide it in the kitchen. Force people to do Scrum not by the book, and change things unexpectedly without notice. As I wrote before, ScrumButs are the best part of Scrum.

A complex system that gets too comfortable with certain behaviors runs the risk of becoming complacent, stagnant, and fragile.

Let people get used to surprise.

Workout Practices

For management practices it is no different. I recently asked participants in a webinar if they thought organizations need a management method or framework. Roughly 67% said “yes”.

I understand people’s need for more concrete management practices, but I am against the definition of a management method. The inevitable result would be a certification program that validates whether people understand and apply the method correctly. And learning would come to a standstill. This would be at odds with complexity science, and incompatible with systems thinking.

I prefer the workout metaphor.

Everyone understands 20 push-ups a day are healthy, but not required. It’s perfectly fine to replace this practice with something else. In fact, as your personal trainer knows, you should! Likewise, you could implement a work expo, until you are tired of it. You can play moving motivators, until people know too well how it works. You should try kudo cards, until they lose their value. And you can’t go wrong organizing exploration days, until you don’t need them anymore.

Do you want your organization to be fit and healthy?

Next week introduce a change that your well-performing team did not expect. Add a new stressor. Feed them some dirt, mixed with a bit of love.


  • CHAMPFROGS... The 10 Intrinsic Desires
  • Value Creation over Value Extraction
Related Posts
free book
“How to Change the World”
  • Mike Leber

    Totally agree Jurgen. As human beings we didn’t develop to where we are now by following the same procedures over million of years. We tried out new things, we innovated. Probably in the early stages this was risking lives.
    However, also today we have to take risk, if we want to get further – risk for the sake of future benefits, happiness, greatness, or whatever the intrinsic stuff might be, driving the one or the other anti-conformist.
    Learning is key, double-loop learning, asking about purpose, even more. And this requires experiments and variability instead of just streamlined nice behaving processes. Processes, which end up like wild animals behind fences in zoos. Same as you refer to Scrum, I do by referring to sticking notes walls, trying to optimize the detail (whether flow, WIP or cycle time). These patterns are also nice. But they are far away from substantial change.
    What really brings us further as individuals and organizations, is processes from the wild, experiments, outbreak, freedom for creativity, innovation by taking at least some risk.

  • John Styffe

    Jurgen we don’t need stress. We need and will continuously have stressors. It is how we respond to the stressors that is important. If I have stress then I am responding inappropriately

  • Jurgen Appelo

    I didn’t tell you to have stress. I try to follow Taleb’s definition of adding stressors. That’s different.
    As I said, I like changes (stressors). I rarely have stress.

  • J. B. Rainsberger

    I think “rigid” is a key word in this article to which I’d like to draw attention, because otherwise I fear many readers will think they never have to define nor plan anything and devolve (back?) into chaos.
    I use defined methods and practices and they help me and have helped others (though not everyone to whom I’ve introduced them). I think *intent* plays a strong role in success. As you write, when we apply defined methods and practices with the hope of magically creating universal, uniform safety, we end up with a nasty surprise. I think that when we apply defined methods and practices in order to develop a specific skill after gathering evidence that we need to improve that skill, we stand a better chance of success. So it is, I imagine, with your spouse’s workouts: is he working out because he thinks it will magically make his life better, or because he’s identified that he wants or needs to improve muscle tone or lose weight?
    Of course, that leads me to a troubling question. I generally suggest using techniques like value-stream mapping and applying the theory of constraints to help identify which issues (probably) need more urgent attention. I do this with the goal of helping groups spend more time practising techniques that are more likely to help them more significantly and sooner. Is this just another rigid framework? If it is, then is that necessarily a problem?

  • BrigittePfeifer

    Without any stress, life is not possible. Everything in our organism is related to stress on the one hand and relaxation on the other (breathing, heart beats, …).
    I agree, that stressors and stress are two more or less different things. What is important is the level in between, which is the “personal stress intensifiers”.
    With reference to these three levels you always have to keep in mind that different stressors lead do different stress reactions. What can be very positive for some of you can be dangerous for others. As written in my blog… there is no “one-fits-all-solution”. Managers must have enough empathy and experience to find the right balance.

  • Jurgen Appelo

    It is clear you have an intelligent way of applying (multiple) methods and frameworks, in a mix-and-match approach. That’s great. You’re basically shopping for good practices and techniques from multiple sources, which is what everybody should be doing, IMHO.
    I just have a problem with the words method and framework because they too easily lead to cargo cult implementations. That’s why I prefer choosing a better metaphor.

  • Rajesh Mathur

    Brilliant work Jurgen! Thanks.

  • J. B. Rainsberger

    I think I get that. If some words don’t work, then try others. Unfortunately, that has another negative effect: many people interpret that to mean that practices and methods and frameworks have little or no value. It creates a marketing nightmare: we forever chase new brand names, even if they describe the same old things. I suppose that’s just the nature of marketing, but I’d rather do better, if we can.

  • Sebastian Radics

    Hi Jurgen,
    thanks for the nice post. I fully agree with you, that there is a danger with every framework to end in stagnation and becoming fragile. The same can happen with Scrum – if not used with sense and its implied constant inspection and adaption.
    The risk I see is the tendency to use your mentioned negative ScrumBut before entering the RI (of Shu-Ha-Ri), because it’s the easy way. It’s all to easy to remove significant parts from the memeplex Scrum and therewith losing the power of the method – and most often without knowing that you destroy it.
    As I understand in my daily practice, in the beginning it’s really helpful and with sense to use Scrum as it’s described. If you really got the message and used it, you anyway do a constant inspect & adapt and sharpen the process. And honestly – the described framework has not that many elements that narrow your development space that extremely. Most of the parts are common sense and leave enough room for having adaption through your mentioned stress. For me it’s the guiding frame that’s important – like you need having constraints within the system should grow.
    All your fantastic additions and ideas – Kudos, Moving motivators, Work expo and many more are helpful even with the context of using Scrum. They help creating ideas and enable to find more ways for necessary innovation and adaption. Seen from Scrum – there is at least in my opinion not a ScrumBut in using your suggested additions (we even start using it btw).
    Maybe I’m already to narrow focussed? But can you give some concrete examples, where you think Scrum prevents you from being non fragile?
    And what about the memeplex thing you mentioned in management 3.0? When is it with sense to remove a meme?
    Thanks for your help,

  • Jurgen Appelo

    Hi Sebastian, thanks for your feedback. I agree with your comments. I’m just warning people not to get stuck at the Shu level.
    The examples you ask for I already gave in my blog post.
    And don’t confuse Scrum with Management 3.0. I did not suggest to change Scrum by mixing Management 3.0 practices (although that is perfectly fine, as you say). They are two different memeplexes.
    What I mean is that I don’t want Management 3.0 to become a framework, just like Scrum is now. People should feel comfortable adding/removing management practices in management. Just like they must feel comfortable adding/removing Scrum practices in projects.

  • Roelof Reitsma

    I think you are completely missing the point. You state that you force people not to do Scrum by the book, but changing practices that do not work is actually doing Scrum by the book. That’s what retrospectives are for; identify things that are slowing your team down and improve your process.
    There is something very wrong with training people not to do Scrum by the book. As J.B. Rainsberger wrote: you are teaching them that the Scrum activities, artifacts and roles have little value. There are a lots of things in a software development process that can be changed, but I believe you should adhere to the simple basics of Scrum process or rethink doing Scrum at all. I would never “feel comfortable adding/removing Scrum practices”. Not having a daily, a retrospective or a sprint planning would render the process useless.

  • Jurgen Appelo

    No, it is you who is completely missing the point of my blog post (and Taleb’s book). Retrospectives and antifragility are totally different things.
    Retrospectives are about intentionally seeking improvements to the current process BY THE TEAM. Such as, maybe shortening the sprints or moving meetings to another day of the week.
    Antifragility is about intentionally causing a little bit of UNEXPECTED harm to the team BY THE ENVIRONMENT. Like, maybe hiding their task board, or injecting a virus in their code base, and see how they learn from that and improve their health.
    I suggest your read the book.
    In fact, if you do regular retrospectives in a predictable pattern that would be very much *against* antifragility.

  • Roelof Reitsma

    You are right, I did miss that point. I got thrown off by your "force people not to do Scrum by the book" remark. Why blame Scrum for becoming fragile? The problems you mention (stagnation, complacency, fragility) can occur in any development process, regardless whether you are using a predefined framework like Scrum or not. They are not caused by the frameworks or methods you are using, they are caused by a lack of change in the environment. 
    Changing the things that work (like the simple Scrum process) is a bad idea. And hiding a task board, or injecting a virus into a code base, is only a good idea if you want your team to hate you. There are plenty of other ways to introduce change. Don't forget we are professionals. We do not need management changing our coffee just to evoke a reaction. 

  • Jurgen Appelo

    I have said on many occasions that we cannot blame methods or frameworks, we can only blame people who create or use them in a bad way.
    Second, you apparently still don’t understand what antifragile means. It means improving health by intentionally inflicting a little bit of harm.
    I know a company with a team that has as their purpose to inflict harm. Like removing cables, deleting code, stealing computers, etc. Because nobody in the organization knows what will happen next, and they keep improving after every little disaster, the organization learns to become very adaptable, resilient, and secure. Thanks to the malicious team.
    If you think that people “will only hate you” for doing such things, I’m afraid you still don’t get the point of my blog post, and what it means to be antifragile.

  • Roelof Reitsma

    If you feel that frameworks are not to blame, you might consider changing the title of your blog article. 
    I think I do understand what you are trying to say about being antifragile, I just disagree. I don't think that antifragility works, the claim that organisational health will improve by inflicting harm is unfounded and has not been proven. And yes, it got me irritated that you mention Scrum as some kind of threat which can be cured by this management technique.
    The case of the company you are referring to is quite interesting, but how do they know their organisation has become more adaptable, secure or resilient? And how do their employees feel about it? 
    You should believe me about developers: if you deploy a team to steal their computer, they will hate you for it. Developers are used to dealing with unexpected issues, so they will probably find a way to quick-install new computers, but that doesn't mean they will be happy falling victim to random harm. And do you really want developers dealing with that kind of issues? I'd rather have them adding business value, than wasting their time on something else. 
    By the way: why is this a private discussion? Although I do enjoy the discussion and appreciate your response, I'd love to hear the opinion of other readers of your blog on this subject.

  • Jurgen Appelo

    Nassim Taleb refers to a huge body of research to support his arguments, so your claim that “it has not been proven” is simply not true.
    It is also a big theme in complexity science, though the scientists use different terminology. Therefore “unfounded” is also untrue.
    We will simply disagree then. It makes no sense to me to continue debating while ignoring science.
    “why is this a private discussion”
    It isn’t. Everyone can read this.

  • Glen B. Alleman

    While Taleb has many “interesting” ideas, just take care that his work is not peer reviewed. He is outside the economics community.
    This doesn’t mean his ideas are “wrong,” just that they may not be as “right” has he’d like us to think. There is no way to actually “test” any idea in economics, so peer consensus is the next best, and many times over history, that consensus has been wrong in the end.
    Referring to a “body of research” is not the same as “proving a concept.” It’s the trouble with economics and other “social science” style disciplines. There is no “test” that can be performed.

  • Glen B. Alleman

    See “The Great Persuasion: Reinventing Free Markets,” Angus Burgin for a history of the “economic theory wars,” starting in the late 20’s and continuing today, between the London School of Economics and the University of Chicago.
    Taleb is just one of many authors working in the area of “risk based decision making.” But be aware Taleb is a former derivatives trader who became a scholar and philosophical essayist in 2006.
    Although he is currently Distinguished Professor of Risk Engineering at New York University’s Polytechnic Institute, he self-funds his research and operates in the manner of independent scholars.
    So academic rigor “may” not be his strong suite. Not that has gotten us very far in the economics process!!

  • Jurgen Appelo

    Thanks for pointing that out.
    And yet, I feel encouraged that Taleb seems to be supported by some very smart people, such as Nobel prize winner Daniel Kahneman and Dan Ariely, behavioral economist.

  • Glen B. Alleman

    Yes, he has support and his ideas are “interesting.” He does tend to be retrospective “I told you this would happen, after it happened.” He’s a wonderful speaker so has a popular following. A cycling group member is a University of Chicago economist (Grocery Chain CEO by day) and would say that most economist are serious dull speakers, so Taleb “gets the press.”
    But like Google, being supported by some very smart people is a popularity contest. Only time will tell if his theories hold up.
    It’s the quirky self-funded and self-published approach that is somewhat troubling – no peer review, no editorial review, no side by side comparison with field data. The economics community is wildly split from far left to far right and in other dimensions as well. Taleb is a “populist” (I know you hate that word), but he makes his living with populist ideas.
    Possible good ones, but with no way of actually testing them. So in the end it comes down to belief and “resonance” with ones personal world view. I don’t like his approach to “self serving” narrative, that is well documented in the reviews of his books. But I acknowledge that’s a personal issue.
    Always worth a read though, just take his stuff with a “grain of salt” as we would say. The Great Persuasion: Reinventing Free Markets, is well worth the read to see how economist get into “theological” battles with no way to test their ideas, just the power of persuasion. As another colleague with an economics PhD say “it’s the dismal science,” like sociology, and psychology.

  • Aolchik

    Nice post Jurgen.
    I was recently reading and posting about an article from Sarah Green, published at HBR, called Making Process Planning Cool Again. The article, based on the approach of Toyota Production System Support Center, suggests the world is in need of more processes, suggesting cautionary measures to avoid failure.
    I believe that one of the best chapters of Steve Denning, The Leader’s Guide to Radical Management, is the one about Toyota, and how difficult is to understand the real value of the often forgotten pillar of the Toyota Production System: “respect for people”.
    Your post comes in hand with what I really believe. The article helps us to avoid forgetting the nature of complex systems, and how processes or collections of “best practices” can obfuscate the importance of the learning process and the importance of the respect for people.
    Does the world really need more processes or should we gradually fight them to become more antifragile?
    Congratulations for your post!

  • Pablo R. Facal

    Great post Jurgen.
    I agree with you. Introducing a small amount of “controlled” chaos acts like a vaccine against stagnation and force us to left our comfort zone. And all we know there are no improvement inside the comfort zone.
    As the personal trainer may say, “no pain, no gain”.

  • Dennis McDonald

    Great thoughts. I totally agree with the need to be shaken up every now and then, no matter whether you are following a methodology or framework or whatever. In the real world I’ve also found it to be the case that, even when following a particular framework, external events very often transpire to force change — the management changes, there’s an emergency of some sort, sr. management has a special project that needs attention, etc.
    Also, for a good series of quotations about “moderation” go here:

  • Olivia Jennifer

    i agree with you that scrum is very useful for a project. also offers agile
    scrum certification
    Master Certification
    ) trainings across USA. Visist
    for details.

  • Pingback: Don't Let Scrum Make You Fragile | Leadership, ...

How to Change the World - free Workout - free