Get Chief Innovator
Delivered To Your Inbox.

Celebrate Failure? Please Don’t!

The myth of failure and risk sounds good: celebrate failure, embrace failure, fail fast. But this is misleading at best.

We think we need to celebrate failure because we believe eventual shipwreck is a price we must sometimes pay for innovation and progress. Stuff happens and we can’t punish those who try or they won’t try at all.

Many experts are happy to expand on why failure drives innovation, even splitting people into “Type I” and “Type II” mind-sets (Silicon Valley versus the rest of us). Others say that we need to fail even faster.

Let’s be honest: Do you really want your innovation to fail—fast or slow? Who truly wants to celebrate that failure? Shouldn’t we rather learn and move on?

Interestingly, in amateur risk management, we find similar thinking. Many companies don’t want their project managers to “play it safe.” Risk taking may even be part of their appraisal. Yet, whose money are these project managers putting at risk? At worst, their career path takes a kink. What they gamble is their company’s money, however.

Are we asking the right question at all?

Tony Ulwick, CEO of Strategyn, has been battling this idea for a couple of years already. As far as I can see, research institutions like Argonne National Labs, CERN and many others don’t work that way either. They don’t ask win-lose questions. The answer to such questions has an information content of one bit anyway: success – failure, yes – no, 1 – 0. These are not good questions to ask.

Let us have a closer look at how scientists work by looking at the recent development of a novel technique to read classical texts. It’s not about whether or not we can read those old scrolls without unfolding (which would destroy them). The researchers at ESRF were asking how to bring to best use their own core capabilities. For example, parchment and ink have different absorption and refractive indices, and those two even change with the wavelength of the probing radiation. Which is the required spatial resolution? Using X-ray tomography, can a 3D-model for the distribution of the refractive index be built? Would that allow reading the scroll without unfolding?

You can indeed launch many good research projects with such questions. And of course: All that had been done long before the ESRF researchers were actually reading the scrolls. Only the sum of many such investigations culminates in such a breakthrough—and brings with it a long tail of other learnings, which certainly were not “failures” at all.

Is the mantra of “celebrate failure” coming from two worlds clashing?

Back to business…in the world of daily management, we know the relation between input and output: The number of staff ill is x, therefore production will lag behind plan by Y where Y is a function of x, Y = f(x).

In the world of innovation, it’s different. Cause and effect are often not clear. Minor changes can lead to major differences in outcomes. For example, what happens if Tempo, a German company selling facial tissues, decides to branch into toilet paper? Technically they master it all. And few innovations fail because of technical issues. Yet, they might stumble over some unsuspected cause-and-effect relations. In the case of Tempo, the company learned their brand stood for facial tissues and not their core competencies. Toilet paper sold under the same brand name severely damaged their brand perception. When companies start such a new business line “full-blown” they might indeed need to “celebrate failure”—but is that what they truly want?

When it comes to innovation, we have less knowledge about cause and effect. Therefore, we need an approach different from our take on daily management.

Managing Today’s Business vs. Creating Tomorrow’s Business


Managing today’s business

Creating tomorrow’s business


Getting stuff done

Testing hypotheses

Cause and effect

Well known

Little known

How to steer it

Projects and performance goals

Experiments and learning goals

Hold people accountable for




Perfection as only benchmark
All are one team
Scientific approach

Perfection as only benchmark
All are one team
Scientific approach

Almost by definition in the domain of innovation, the effects to causes can’t be predicted. We don’t want to fail fast with a full-scale trial run. We rather want to learn fast. Which causes are driving desired and undesired effects? How can Tempo test their hypotheses prior to launching the new product on the market? The distinction between projects and performance goals on the one hand and experimentation and learning on the other may sound subtle but it makes a tremendous difference

As a consequence, you should not hold your people accountable for the results they have achieved in their innovation endeavors. You rather need to hold them accountable for how carefully they have researched their hypotheses, set up their experiments to test them, and how much learning they generated along the way. No doubt, a professional such as Brian Joiner might say: Besides your striving for perfection and making sure all are one team and not playing against each other, also a scientific approach is key to success in both worlds, when managing today’s and when creating tomorrow’s business.

Beyond winning or losing in innovation

The art and craft of experimentation help us to grasp innovation opportunities. Not least, the important chance findings are made much more likely when your teams are used to carefully formulating their hypotheses and then designing and conducting with equal care the experiments to test them before your new offering reaches the market. You want to learn a lot. You want to open as many doors as possible. You want to raise new issues and new challenges. That may not be easy. But it is so much richer than just coming to a result such as “mission accomplished – yes or no” and the need to celebrate regardless

You should indeed celebrate:

  • How many most-cherished hypotheses you “killed.”
  • How you have set up your experiments.
  • How much you have learned with them.

But please! Don’t celebrate failure, fast or slow. Because then, most likely, you haven’t formulated carefully enough the task you want get done.