Thursday, 12 July 2012

Tolerating failure – yeah right!

If there’s one thing that drives me nuts it's hearing senior executives exhorting their people to ‘tolerate failure’ and ‘take more risks’.  Sure we all know that innovation needs us to throw caution to the wind every now and then but ‘tolerating failure’ is such a blunt concept - it’s pretty useless.

There’s the obvious chestnut about how rewards at work are rarely lined up with risk taking.  ‘Sorry boss, I missed my targets but I did take a lot of risks’.  Hmmm!  Most people have mortgages and mouths to feed – we need a much more subtle concept of risk.

In my experience it’s far more effective to substitute the words ‘lets take more risks’ with a more positive line; ‘let's make it real now’.  Asking your people (or asking yourself) to accelerate an inkling of an idea into something that others can react to is a cornerstone of innovation.  ‘Can you draw it…make a mock up…act it out?’

The trick to effective ‘realness’ is stealth.  The cheaper and quicker you can make each mock up or simulation then the more iterative the process will be.  Over several quick fire rounds you’ll learn more about your great idea and make better development decisions.

Quick and dirty experimentation can escape under the radar - no need to waste time making a case and seeking approval.  Most innovation starts like this - a few people muttering in the corner and staying late to make something real – not writing a proposal. 

So let's replace the phrase ‘tolerate failure’ with an altogether more wholesome and useful concept of ‘making it real’.  Forcing an idea through multiple experiments is scary as hell because not all experiments work out, in fact you need failures as much as successes to learn from.  In the context of ‘making it real’ then failure becomes really valuable.

There are a couple of great stories that underscore my point about making things real.  Sir James Dyson’s account of developing the dual cyclone vacuum cleaner in ‘Against the Odds: An Autobiography" (1997, Orion Publishing) is brilliant.  He developed over 5000 prototypes and made loads of mistakes along the way. An experimental approach to solving problems has also been proven to be more effective than a single burst of work. Researchers at Stanford University asked 28 participants to work on a design to protect a raw egg in a fall.  Half the participants designed, tested and iterated their egg protection ideas after 5, 10, 15 and 25 minutes. The other participants spent all their time on one design and were not allowed to test it until the end of the session.  All had similar resources (paper, string and other materials).  The results showed that the iterators significantly outperformed their non-iterating counterparts, achieving roughly double the non-breaking drop height – in some cases at 15 feet.  Definitely one to try at home with the kids!  For more read ‘Efficacy of Prototyping Under Time Constraints’ – a paper by Steven P. Dow, Kate Heddleston, Scott R. Klemmer.  Stanford University HCI Group.  Department of Computer Science, October 26–30, 2009.