Black Box Thinking: Why most people never learn from our mistakes

Black Box Thinking by Matthew Syed explores how failure can be used to help individuals, companies, organisations and entire sectors operate more effectively.

Black Box Thinking by Matthew Syed.

If you have a portfolio of shares and some have done well and some badly, which are you most likely to sell?

Contrary to expectations, we are more likely to keep those that have lost money, says The Times journalist Matthew Syed in Black Box Thinking.

How so? Because a paper loss is a real loss and we don’t want to admit, even to ourselves, that we’ve made a mistake, he says.

The name of this book, which explores why most people never learn from their mistakes, comes from an industry that actively embraces failure — aviation. Today 36.4 million commercial flights carry three billion passengers, but only 210 people died in 2013.

And for every one million flights on western-built jets, airline association Iata says there were just 0.41 accidents.

In 2014, even accounting for Malaysia Airlines Flight 370’s 239 deaths, the rate per million fell to a historic low of 0.23. Compare that to the US healthcare system.

In 2013, a US study in the Journal of Patient Safety put the number of premature deaths associated with preventable harm (misdiagnosis, wrong drugs, injury during surgery, post-operative complications etc) at more than 400,000 a year — the equivalent of two jumbo jets falling out of the sky every 24 hours.

Why? In aviation, failure is regarded as a “precious learning opportunity for all pilots”; its system is data-rich and encourages incident reporting. If a near-miss is reported within 10 days, pilots enjoy immunity. In 2005, a number of reports were filed about a problem at a Kentucky airport; it turned out a nearby, lit-up mural was misleading pilots.

A warning email was sent within minutes and the mural and lights were removed within days.

Mr Syed says the issue is not just with health care. Humans have an allergic attitude to failure, he warns.

He throws out an exhausting number of examples: from the US penal system to global poverty; quantitative easing; Iraq and the fallacy of weapons of mass destruction; the technical and business worlds of Apple, Dropbox, 3M, Unilever and Dyson; and the sporting prowess of Mercedes Formula One and Team Sky cycling.

But they are powerful. Unilever biologists, he points out, used simple trial and error, patiently working their way through 449 iterations to improve a fac­tory nozzle and prevent washing powder blockages in the mach­inery. That’s 449 failures for one success.

Black Box Thinking: The Surprising Truth About Success is out in hardback and released in paperback in April 2016.

q&a denial creeps up the order

Suzanne Locke reveals more insight's from Matthew Syed's Black Box Thinking:

How do companies embrace failure in real life?

Marginal gains. Mr Syed watches the McLaren Formula One team in action at the Abu Dhabi Grand Prix of 2014 and finds the whole team, from chief strategist to Lewis Hamilton, will sit to review up to 50,000 channels of data, say from the qualifier’s pit stops, to “optimise hundreds of thousands of small items to the nth degree”.

With so much data available, how do you test?

The “gold standard” is the randomised control trial (RCT), in which a treatment group is compared with a control group. The credit card company Capital One tests everything from font types to call centre script wording. Google tested 40 different colours for its toolbar by tracking its entire Google Mail user base’s click preferences based on colour.

How do leaders handle failure?

Mr Syed says chief executives are the worst culprits. He says denial increases going up the pecking order — these are often the people with the most to lose, and use spin and PR to “bolster post hoc justifications”.

And entrepreneurs?

Generally, technology entrepreneurs are “brilliant theorists who can perform complex mathematics”, according to Mr Syed. If they can fuse those skills with a positive attitude to failure, what you get is a “minimal viable product model”, where a prototype is tested on the early adopters to see if the product is actually viable and what can be improved by what the consumers like. This “hardwires the evolutionary process into the design of the business”.

Follow The National's Business section on Twitter