Safety is one of the cornerstones of a healthy work culture. Without safety, you cannot expect to see responsibility, accountability or ownership in any measure above the strict minimum.

Still, we see many of our corporate clients dismiss the idea of creating a safe culture, lest their employees would have free reign and would avoid accountability for their actions. Employers would lose control! Chaos and pandemonium would ensue!

The problem is that safety is often confused with immunity or absolution. In an unsafe environment, where failure leads to witch hunts and retributions, it can be easy to make that jump.

In this post, I will go into what safety is and how it differs from immunity and absolution. The idea of safety is not … ahem… “a hippy concept not fit for a real business”. It is, in fact, a powerful tool that can be leveraged to spur innovation, excellence, engagement, and ownership.

To create a good work culture, safety is not an option. That’s why Joshua Kerievsky lists make safety as a prerequisite as one of Modern Agile’s guiding principles. That’s also why Jurgen Appelo makes it a central argument in his books Management 3.0 and Managing for Happiness, and why writers such as Daniel Pink, Dave Grey, Frederic Laloux, Lisa Laskow Lahey and Robert Kegan spend a large portion of their arguments on it. And yet, not one of them argue for a culture of immunity to consequences or absolution.


Let’s start by defining what is a culture of safety.

In a nutshell, a culture of safety is a culture where members of the team do not have to hide for fear of retribution because of who they are, what they think or what they do.

In a safe culture, you will see people readily admitting to failures or mistakes. As a result, a remedy to problems will emerge much faster and be are much more efficient since no ones tries to save himself. Education about those issues can be shared across the company, raising awareness and reducing the risk that the same mistake will happen again.

In a safe culture, people will speak their mind and ideas will be debated, leading to better solutions. “Sparks must fly”, says James M. Whitehurst, CEO of Red Hat, in his book The Open Organization. A bit of friction while debating ideas is a good sign that people care about the subject and want the best outcome. But stay vigilant that those debates do not degenerate in spiraling arguments. Still, an occasional spiraling argument is much better than no argument at all.

In a safe organization, people feel safe to be transparent about their strengths and weaknesses, making it easier to get help working on weaknesses and raising one’s overall usefulness to the organization.

As desirable as safety might be, no organization will reach that state naturally. It requires the right conditions, time and a concerted effort. The very first step is actually the hardest one to take: get rid of the blame.


When something goes wrong, it is important to react to it. React. Note that I haven’t said either how to react nor who should react. What is important is that there is a reaction to something going badly.

In a blaming culture, the first reaction is either finding out who is to blame or making sure someone else will be blamed. Neither reaction addresses the problem, nor do they lower the chances of the failure happening again. The goal in this scenario is for individuals to protect themselves.

In a blameless culture, there is little need to react with self-preservation. The reactions tend to be swifter, as no one is trying to cover their tracks. The reaction is first focused on fixing the issue, then making sure it does not happen again.

Etsy has such blameless culture, where it is common for employees to receive a communication along the lines of “I messed up. Here’s what happened. Here’s how we reacted. Here’s what to watch out for”. They even have a good-natured annual three-armed sweater award given to the most surprising failure of the year, awarded in a ceremony and accompanied with the testimonies of the people involved on how they handled it and what has been learned. The goal here is not to humiliate them, but to spread lessons learned and raise awareness.

The only purpose of blame is to bring fear of consequences, from share to loss of employment. It is no wonder then that in a blame-based environment, many employees opt to never try anything, to do the strict minimum, and often waiting until they get direct orders from a manager so their tracks are covered and they can divert the blame toward the manager when the results end up being unsatisfactory.


There are three types of failures: Smart failures, accidental failures, and negligence failures. Each comes with their own type of reactions. These reactions are consequences of sorts, but with the aim of making things better.

Smart failures are, as far as failures go, the type you should aim for. If you are careful with your actions, including your experiments, you should plan to limit the impact of a possible failure, discuss it beforehand with people who might be impacted, act carefully and react swiftly when the worse happens. If you approach experiments in this way, then failure can be turned into a valuable learning experience for the whole organization. Someone who is able to experiment, fail and handle the recovery properly is an asset that should be celebrated. Blaming people for a smart failure would be absolutely counter-productive.

Someone who fails accidentally, through an oversight, a lack or knowledge or sheer bad luck should own it, and act swiftly to help with the recovery. Accidental failures happen all the time and should be seen as part of doing business. When an accidental failure happens, review what happened, derive some action points to help reduce the chances it will happen again and move on. Again, blaming would not improve anything.

In either case, if you are responsible for the failure, own your mess and be part of the effort to clean it up. If you are not, avoid assigning blame as it does not bring any value except perpetuating an unwillingness to try for fear of failure and blame.

Failures caused by sheer negligence, pig-headedness or straight malevolence should be dealt with seriously. There should be consequences for these kinds of failure, but again they are not about blame. Often, having a talk with the person in a safe space will reveal some issues that can be addressed so this does not happen again. At other times, especially for a repeated offense, the best course of action might be getting rid of the bad apple.

Or not: Some deliberately developmental organizations such as Next Jump have a no fire policy and will salvage you no matter what. Still, we are not arguing for either immunity or absolution: follow-up actions, or consequences, are part of the deal.

Working with self-organized groups makes this system easier to use. Teams share success and failure and even more, they share the effort. When an organization wonders how it should react to a situation, the best course of action is to ask the team involved. Disaster recovery, lessons learned and recommendations as well as how to react to the person at the core of the problem can all be dictated by the team.


When Joshua Kerievsky developed Anzeneering in 2014 — his approach to a culture of safety — he described this work culture as something you are responsible for upholding rather than something that should be owed to you.

It should be every individual’s responsibility to ensure that coworkers are safe from a bad environment, that your organization is safe from your actions, that users are safe to use the work you produce, and so on. Nothing about that mindset is about making you safe, it’s all about making others safe. You become protected too when the mindset spreads across the organization.

That is an amazing approach: look at safety as a gift you freely give to others.


A culture of safety is not meant to exist by itself. It goes hand in hand with a culture of responsibility and ownership.

Modern Agile puts make safety as a prerequisite to remove the fear of failure and blame we so often see in organizations. As we’ve seen, fear leads to an understandable lack of desire for ownership and responsibilities. Without this fear, we are better able to reach our potential.

The other three principles of Modern Agile are the key to unlocking that potential.

Deliver value continuously forces us to revise how we approach work. To be able to deliver any kind of work in a continuous fashion, we need to divide it into smaller, independent chunks. That approach is very important. Smaller chunks mean that if we fail or miss the mark with a chunk, the damage remains small and the issue can be resolved quickly. By making these chunks independent, we make sure a failure will not spread or contaminate other chunks (ours or somebody else’s), thus limiting the impact of the failure. Small, self-contained chunks drastically reduce overall risks by making each failure insignificant. Without fear, and with insignificant risks, experiment and learn rapidly becomes a very powerful tool. It allows us to figure out solutions empirically, work out kinks to provide better quality and stimulate innovation. New discoveries are far more often the result of failed experimentation than of planned design.

With safety from reprisal and blame, with work habits that reduce the risk and the impact of failure and with constant learning from practical experimentation, we make people awesome by giving the environment they need to be stimulated by what they do, to take ownership of their work and to become the assets everyone, deep inside, wish to be. In turn, they can see it as their responsibility to spread that culture and help newcomers feel safe enough to participate and grow.

The Modern Agile logo is a wheel. I believe that to be very deliberate: each aspect builds into the next one and keeps the whole wheel turning.


In this blog, we saw what is intended with a culture of safety. Put in place properly, it can propel any organization and its members to new heights.

However, it is possible for the ideals of such a culture to be corrupted. Toxic safety is a very real danger, and one we should watch for. We will explore its forms in my next blog post: WHEN SAFETY TURN TOXIC.

See you then!

1 Comment

When safety turn toxic – The Antifragile Mind · May 16, 2017 at 2:02 pm

[…] week we looked at the advantages of a culture of safety in my post Safety is not immunity. This week we will explore the pitfalls of a badly implemented safety […]

Leave a Reply

Your email address will not be published. Required fields are marked *

Related Posts

The Antifragile Mind

What is Antifragility and how can it help me?

Let me give it to you straight: Antifragility is the ability to thrive and prosper in chaos and adversity.  The concept comes from Nassim Nicolas Taleb, author of Black Swans, Antifragile and Fooled by Randomness. Read more…

Quantum Monkeys Blog

Of the 3 leader roles, which one are you perfect for?

There are a lot of articles out there about different leadership styles. No matter what you do, there’s probably a style that defines it very clearly. I don’t care much about those. I’m a practical Read more…

Quantum Monkeys Blog

How to be an awesome change leader

Change management is not an easy thing. If you have ever lead a change endeavor, you had the pleasure to face some challenges. The fear and confusion brought by poor communication. The awakening of the Read more…