Last Modified:                                                                                                



Homepage / Publications & Opinion / Silicon.com

Peter Cochrane's Uncommon Sense: Scales of grey
When to be a control freak, when to let creativity flourish...
BT's R&D operations were streamlined under silicon.com columnist Peter Cochrane's stewardship. What did he learn?

During my decades in industry I have observed people struggle with organisational structure and management methods. Broadly speaking, when a company gets into some difficulty the board will call in the consultants. The consultants trawl the company talking to people - finding out what they think - and put together a report that broadly reflects everyone's thoughts. There is nothing like telling the customer what they want to hear to ensure that you get the next contract!

The usual outcome of one of these very expensive exercises is simple. If the company happens to be centralised in its management and decision-making, the consultants will advise a break-up, a move to distributed management control and finance.

If, however, the consultants discover the company is fully distributed, they will immediately propose centralisation as the answer. And so over a period of time you will see corporations imitating a concertina as they go in and out, from centralised to distributed and back again.

It has become increasingly apparent that a primary reason for continual reorganisation has been the inadequacy of management structures and techniques in a world that is changing faster, where competition can come from any sector and where managers move from one job to another like butterflies.

Every time a new CEO arrives he is duty bound to do something because he has to be seen to be effective. As a general rule, such rapid change is not good and usually leads to significant damage to companies and their employees.

The reality is the management techniques and structures being used today are right out of the 19th century and more akin to a world of Charles Dickens than a world of 21st century technology. They are rigid, hierarchical and unsuited to a technologically driven world.

Let's examine two extremes of business. Suppose we have a huge manufacturing operation that is in the business of assembling automobiles. In fact, no one manufactures automobiles today. The plants and the companies have become the systems integrators. They harvest components from all over the planet, bring them together in one plant and with a unique set of jigs and fixtures and design capability create automobiles.

In such a plant there is a place for hierarchy, for all must be known and ordered. The design, the construction and the performance of every element are well understood and documented. The last thing required here is innovation - it would jeopardise everything.

This arena is about yield, whether it be computer hard drives, washing machines, television sets or automobiles - on the manufacture and production line yield is the key. In such organisations it is not unusual to find between four and 10 levels of management. It works and so it should. What is more, it is necessary. Everyone has to have a clearly defined role and know precisely what they're about. This is a well-ordered regime where everything runs like clockwork.

The second example is the research organisation where absolutely nothing is defined apart from a general direction of trying to create solutions. In industrial research people are generally aware they are there to find solutions to known problems or the creation of new technology to reduce costs of existing solutions or the creation of something new that we have yet to imagine. In such an operation hierarchy is a killer. Nothing is defined and little is known apart from the general direction.

Here strong leadership is required. Marshalling the brains of a multitude of PhDs is about as difficult as trying to herd cats. Investigations are initiated, paths are searched - and the vast majority are dead-ends. But search they must and explore they have to, for no one knows the correct course and there are always surprises.

Lasers, for example, spent over 30 years as a solution looking for a problem and today they are manufactured like jelly beans as vital parts in DVD/CD drives and optical fibre systems. Until the 1970s no one had a use for coherent light but we may well see lasers employed to create quantum devices that operate very close to absolute zero to give us computational power beyond our wildest imagination.

In such endeavours, research needs hierarchy like a scientist needs a hole in the head. A hierarchy of two - leader, manager, people - is generally adequate and necessary. But a hierarchy that is steeper just disables the operation because with great hierarchy comes the control freak, the person who wants to know what money is being spent on down to every pencil and what the outcome and benefit has been.

There is no doubt about it, research is a wasteful operation and it is often difficult to quantify the final outcome. But when optical fibres were first being drawn and lasers were first being developed and people were complaining of the outrageous expenditure for zero return, none of them had 0.1 per cent of the necessary imagination to see a world powered by light and dominated by IT.

Somewhere on that span of hierarchical spectrum sit people who have to make difficult decisions on the basis of often-incomplete information. It is apparent they will have to become much more flexible in their approach and that the co-existence of manufacturing and research, and all activities in between, is going to be increasingly difficult. The most likely outcome is more outsourcing on these scales of grey.

This column was dictated onto an Olympus recorder, typed on an Apple iBook by my secretary, emailed to me via a cable modem connection, revised on my laptop while sitting on the porch of a log cabin in the Rocky Mountains of Colorado - and despatched to my editor over a European GSM mobile from a location I thought the bits might not span!