“Make New Mistakes. Make glorious, amazing mistakes. Make mistakes nobody’s ever made before.” Neil Gaiman
Just before Christmas – in my final catch up of the year with my manager – a pretty significant thing happened.
I was told that Bromford Lab seriously needs to up its failure rate in 2015.
Welcome to the parallel organisational universe that I exist in!
Of course this need for greater transparency of failure should be common sense to us by now.
Nielsen research suggests that “about two out of every three products are destined to fail.” However , outside of the startup community , this is rarely acknowledged and hardly ever promoted.
In the public sector , where projects take years rather than weeks, and pilots become mainstream services without any evaluation – things are worse.
Everything is a success.
If you doubt me on this pick any organisation and take a look at their annual review or report. See if you can find any mention of the things they did that failed – and what they learned as a result.
We are afraid of failing and it’s seriously constraining our creativity – and ultimately our credibility.
Here’s five ways we can get better at failure
Give people permission:
Too often targets and KPIs drive perverse incentives and lead to managers forgetting why they are even there. The Lab has no targets. Not a single one. It has an ethos though that 75% of work is permissible failure.
If we solely targeted success through the Lab – it would be catastrophic. It would drive a behaviour of pushing rubbish ideas into the organisational DNA. By giving people permission to fail you are saying “use your common sense”. That’s the only target you need.
Change the definition:
People see failure as a bad thing as it has been drummed into us through childhood and the corporate machinery. It has come to mean letting people down. Try to rebrand failure as a journey of discovery – in which you’ll learn the possibilities for you and your organisation.
Let’s remember there have been some amazing by-products of “failure”. Columbus could be said to have failed when he set out to find a new route to Asia, but his voyages led to the widespread knowledge that a new continent – America – existed. 3M certainly failed when they invented a glue that didn’t stick – but as we know it led to the creation of the Post-It note. Failure is an option. Acknowledge and prepare for it.
See it as an investment:
This seems counter intuitive but there is a strong economic argument for failing more often. As an example let’s say it takes 6 months and £100,000 to take a product from idea to launch. At best you’ll get two cycles in a year.
However, if you can do a complete cycle of learning in a week for £2000, you can get 52 cycles in a year at about half the cost. Many of those cycles of learning will fail but will have less risk and negative impact than your latest big initiative. In the words of Andrew Stanton – “be wrong as fast as you can.”
Have a scientific approach:
Embracing failure does not mean having a lax approach or work environment. Have a hypothesis and test it. Involve your data geeks and people who have no vested interests in seeing the idea succeed. If people start getting excited you have a discovery. If not – people will still feel it was worthwhile learning – but quickly move on to the next idea.
Capture the learning:
Failure is only bad if we are doomed to repeat it. Breaking our organisations out of cyclical failure is a huge challenge. Chris Bolton has asked if there’s a need for a Museum of Failed Products within public services. He’s undoubtedly right – the ‘corporate memory’ is often unreliable.
Don’t confuse this though with “We tried that before and it didn’t work.” In the Lab we are quickly filling up our Failure Shelf – but we might dust them down and give them another run out on a rainy day. Sometimes the timing just isn’t right.
And don’t be afraid of being laughed at , by colleagues , the public or competitors.
We will be laughed at as some people are willing us to fail. So embrace it.
Our job is to ceaselessly ask the question: would we do it this way if we started again?
The answer. Almost always.