daheq.com logo

No Failure for the Sake of Failure

Failure Is Not an Option, but Still Okay

 /   / 6min read


Admittedly, the statement that failure is not an option is a very provocative one. So if you are wondering what's wrong with the 'fail fast' mantra of Silicon Valley's startup scene, I'd answer: nothing. At least to a certain extent. First and foremost, it depends on the way you live failure within the company. Let me explain.

Innovation Is Not about Failure

Don't Get It Wrong

Please do not get me wrong, but in fact, failure itself does not lead to innovation and innovative results. Hence, failure should never be seen as an aim. The idea that failure is part of innovation, may imply exactly this and thereby (subconsciously) focus on the wrong aspects of the process.

One thing that can lead to innovation, is how to deal with failure as an individual or in a system. Later it will be explained in more detail how this can be done collectively. First, the focus is on traits that might be relevant. The only way to learn from mistakes is to get up again after an unsuccessful attempt — resilience. This, in turn, requires a certain amount of fortitude. For this reason, it is important to have the freedom to fail, more important what follows that defeat and especially how to deal with it as an individual. Of course, culture and cooperation play a crucial role, so these traits can also be transferred to teams.

Three Stages of Mistakes

Before delving into how mistakes can lead to innovation and how to foster a culture of failure, it is important to define what mistakes are and what kind of mistakes can actually be helpful.

Well, there are different types of mistakes or even more different root causes. With the help of these, mistakes can be roughly categorised: helpful vs. non-helpful, invaluable vs. valueless, and constructive vs. destructive. Probably the worst mistakes are those that could have been avoided. Whether by reading documentations, researching, or considering specifications. It may sound harsh, but these mistakes can be traced back to recklessness or inattention.

In addition, there are inevitable mistakes, which are basically due to the complexity of the system, including both internal and external aspects. For example, these may be the market itself, influences that cannot be controlled, or technical challenges in general. In these cases, small mistakes are unavoidable, so they shouldn't be seen as bad.

Try, fail, and try something else until you find or create a solution that works. George Couros

Last but not least, there are smart mistakes that occur during experiments and research; in processes that are meant to gain insights and learn more. Processes in which, by definition, not everything is known that needs to be taken into account. These are the mistakes that can trigger innovation and have a real, constructive influence, and are therefore mainly responsible for the positive connotations of mistakes in Silicon Valley.

These smart mistakes are invaluable. Smart and unavoidable mistakes are totally ok. In either case, it is important to quickly recognise mistakes as such, learn from them and adapt the approach accordingly. This helps to avoid subsequent mistakes and to keep the process running. Try to avoid the avoidable ones.

Feedback Loop of Learning

As mentioned above, mistakes can only move you forward if they trigger or drive feedback loops. Yes, it is the good feedback loop, as known from Scrum and the like. Well, at the bottom line, innovation can be traced back to iteration and invention. But what does it actually mean to establish such a feedback loop?

Learning by mistakes is one of the most constructive but also most complex and exhausting ways to deal with them. Drawing conclusions is only a small step; the much more important one is to apply the gained knowledge. If this is taken to heart, the smart mistakes of the prototype-driven approach — as mentioned above — can maximise the learning curve. Nonetheless, it should be kept in mind that pure trial & error is not necessarily recommended, as certain managers bear responsibility for assets as well as shareholders. As always, it is a question of a well-balanced mix.

Infinite stairway

Culture of Failure

What you are afraid of is that if you actually have to try something, you might fail and that's just not a chance you wanna take. Dr. John Dorian

The quote above highlights what often discourages taking risks that are worth being taken. In most cases, fear is the biggest hurdle. The fear of making mistakes, failing, and looking incompetent in front of colleagues and managers. So how is it possible to establish a culture in which employees and colleagues are willing to take certain risks? Two complementary approaches are outlined below.

Don't Worry, You Are Safe

Self-protection is innate and a natural instinct: nobody wants to seem obtrusive and by no means incompetent, but obviously smart and helpful. This leads to a loss of opportunities that could allow us to contribute to the big picture. Whether it's expressing a risky idea, asking an interesting question, or admitting mistakes. The learning effect is lost. What Amy Edmondson defines as team psychological safety is very beneficial in tackling this problem.

Team psychological safety is defined as a shared belief that the team is safe for interpersonal risk taking. Amy Edmondson

To achieve this safety, she recommends promoting curiosity, defining work as a learning process and acknowledging one's own mistakes. It is well known that one should lead by example and since people want to impress their manager, in particular, it would be worth considering that the boss starts admitting mistakes and establishing transparency. This transparency can also be communicated externally in the form of post-mortems, which can be used as a collective memory for others to learn from books.

In my summary of Dr. Frederik Pferdt's talk at the 5th German Summit of SMEs, I've already mentioned how this can look like at Google. In regular meetings, people are encouraged to ask questions and to be curious by reviewing products that have been launched or mistakes that have occurred.

Oh Damn — Cheers Though

At the moment, I'd identify two different approaches for dealing with mistakes in public or across departments. First, not to admit mistakes in order to avoid negative effects on one's reputation. Second, to publish detailed post-mortems that not only reveal the causes, but also the lessons learned and the follow-ups. I have to admit that (until now) I've mostly worked at the former type of company. Problematically, this can result in managers getting a false sense of the likelihood of mistakes and failure.


The latter type of company even goes one step further in some cases. Among others, Google and Alphabet's X advertise to celebrate (personal or professional) failure at weekly or annual events. I can't tell how exactly this happens and what's the truth about it, but the idea sounds quite interesting because it can create psychological safety in terms of failure. Another ritual are so-called fuckup nights at which it is also possible to talk about professional failures across companies. It's actually a thing.

No matter how you end up, I think it's a great idea to have this way of talking about failures and seeing the positive after you've learned the lessons.


To summarise all the aspects is not that easy. But let's have a look back at the initial statement that failure is not an option. It should actually be:

Be willing to take risks and to try new things, because it's okay to fail if it's worth a try and there's a lesson to be learned. Do not try to focus on the mistakes themselves & avoid foolish ones. If you fail, get up, it's not the end of the world. Share and discuss your experiences. Don't take it too seriously.

Have a nice start in the Advent season.


Share  /  Twitter  /  WhatsApp  /  E-Mail