How to make a great business out of failure

Just about every modern business likes to think of itself these days as a "learning organisation". This is partly because companies - as management gurus have pointed out - increasingly rely on the knowledge of their employees rather than the products of their plants. But it is also connected with the constant changes in business that make knowledge obsolete almost as soon as it has been acquired.

So how do they do this learning? Joyce Fortune and Geoff Peters of the Open University contend that they spend too much time celebrating success and not enough looking at failure. Sure, people say they learn from mistakes, but Ms Fortune and Mr Peters suggest that this tends to be done in superficial ways before the incidents are pushed under the carpet.

In their new book, Learning from Failure, the Systems Approach, they advocate a more thorough examination of what went wrong. But even when there are efforts to learn from failures, there is a tendency to concentrate on formal investigations aimed at apportioning blame and to rely on "ad hoc" partial recollections of what happened.

Their answer is a third way - "a systemic method for the analysis of failures called the Systems Failures Method". Based on 20 years of research by the pair of them, it uses several "qualitative modelling techniques" that they believe give it a significant advantage over other methods. "In addition to looking back at failures which have already occurred, it can be used to look forward to potential failures," they write.

The authors spend a fair amount of time looking back, at the Manchester Airport fire of 1985 and the previous year's Bhopal disaster in particular, but they are far more interested - and suggest that others should be too - in looking forward.

In their words, emergency planning tends to be the norm in high-risk areas, such as nuclear power generation and oil production, but many commercial organisations appear reluctant to believe that anything can go wrong. A recent US study quoted in the book found that increasing reliance on computer and IT systems was accompanied by an increase in the frequency of computers causing significant business interruptions. In addition, the same study found that less than 3 per cent of 300,000 large and medium- sized computer systems installations had disaster recovery plans in place.

But it is one thing to have such a plan, another for it to be effective. Ms Fortune and Mr Peters maintain that their method "takes the analyst from the real world (in this case the situation labelled as a failure or a potential failure) into the conceptual world, where systems thinking, qualitative modelling and comparison provide the means by which understanding can be achieved". The understanding gained is then introduced to "the real world", where it appears as a set of lessons.

They have found five recurring aspects of failures: deficiencies in the apparent organisational structure, such as lack of a "performance-measuring subsystem" or a control decision system; no clear statements of purpose; deficient performance of one or more subsystems; subsystems with ineffective means of communication; and inadequately designed subsystems.

Other researchers offer their own variants, but the point remains that a little concentration on the reasons for failure, rather than attributing the blame, might make those celebrations of success more frequent.

'Learning from Failure', John Wiley, pounds 19.99.