First, what's the problem?

We’ve all been there, facing a problem so large or complex it seems insoluble. Fortunately, insoluble problems are solved with great regularity. If that weren’t true, we’d still be living in caves and eating only what we could catch or pick.

What can we do to increase our chances of solving the big problems in our lives and workplaces? Here are three suggestions.

First, understand the problem. Reframe the problem in as many ways as possible. Einstein said that if he had twenty days to solve a problem, he’d start by spending 19 days understanding the problem. And Dewey stressed that properly stating the problem was half the solution. Buddhism teaches that a good answer requires a good question.

Second, don’t undermine the role of teamwork by turning a team into a committee. The purpose of teams is to bring a wide range of different skills and perspectives to bear on a problem. The purpose of a committee is to come to a consensus. Recognize that these are incompatible, at least as simultaneous processes. Make sure that each member of the team has a way to think creatively and contribute from their individual areas of interest without allowing the committee/meeting process to stifle creativity. (There is good evidence that creativity in a group process decreases after the group size reaches four or five.)

Finally, don’t make the mistake of thinking that problem solving (especially with difficult problems) is a one-step process. This is beautifully illustrated by the story of human powered flight.

In 1959 Henry Kremer, a wealthy British industrialist and the Royal Aeronautical Societyoffered 50,000 pounds ($1.5 million in today’s dollars) for the first person who could fly a human powered vehicle 10 feet above the ground in a figure of eight around two markers for a distance of 1 mile. He offered twice that for a human powered flight across the English Channel.  Over nearly 2 decades nearly fifty unsuccessful attempts were made, at considerable expense. Paul MacCready, an American aerospace engineer, had a great insight: everyone had been working on the wrong problem.  The engineering challenge was straightforward. The power available was limited to what a fit human could produce sustainably during the flight. This meant that the aircraft had to find the right balance of low weight, adequate stability,  control, lift, and durability. All the failed attempts had focused on a solution to this very challenging engineering problem, studying the problem in detail and then painstakingly building a craft based on theoretical calculations. When the craft failed, it took many months to rebuild and retest.

MacCready decided that the real problem was the slow and expensive learning cycle. He focused instead on how to build a light plane that was inexpensive and could be easily and quickly repaired, modified, and relaunched.This allowed him to test, fail, analyze, learn, rebuild, retest, re-fail, reanalyze, and relearn in a rapid and ongoing cycle. Instead of 6 to 12 months between attempts, he was able to make as many as 4 revised attempts in a single day.

On August 23, 1977,  his aluminum and nylon Gossamer Condor flew the required figure-of-eight 10 feet above the ground. On June 12, 1979, his Albatross crossed the Channel, winning both Kremer Prizes.  

The bigger and more complex the problem, the less useful it is to seek the big solution. Instead, it pays to break the problem down into smaller pieces and re-engineer the development process into rapid and affordable learning cycles. Learning and progress are the result of analyzing and learning from failure. The more you want to learn and the more progress you want to make, the more often you have to be willing - even eager - to fail. Silicon Valley’s successes are built on Fail Often, Fail Early, Fail Cheaply.  Jeff Immelt, CEO of GE, says that he evaluates candidates for projects by their answer to his favorite question: tell me about your biggest failures and what you learned.

The next time you are faced with a big or complex problem:

  • Start by reframing and reformulating the problem from as many perspectives and in as many contexts as possible.
  • Create a diverse team to maximize the breadth of knowledge and range of skills and contexts brought to bear - but make sure the advantages of team work are not destroyed by the poison of meetings and groupthink.
  • Find a way to try as many things as possible with as simple and rapid a P-D-S-Aparadigm (developed by Deming) as you can develop. Do not get sucked into thinking that a large and complex problem is best solved by developing a large and complex solution. The only large and complex solutions that ever work are those that evolve over time from multiple small and simple (but tested) solutions. (See the Laws of Systemantics.)


Footnote: I like to think that the Fail Often, Fail Early, Fail Cheaply idea comes from Samuel Beckett’s novella Worstward Ho (1983): "All of old. Nothing else ever. Ever tried. Ever failed. No matter. Try again. Fail again. Fail better."



Links to more on this topic::