Taking the Hard-Headed View of Change

By Tom Barker

Change may be inevitable but it can make us very twitchy. We all know the feeling: we watch major changes taking place in our companies, whether they be in data processing systems, budgeting procedures or personnel practices and ask ourselves whether it’s worth the business disruption and staff bewilderment.

One explanation may be that the momentum generated within an organization by its efforts to respond to market changes and technological advances allows a wave of change to roll on, bringing with it a flood of much more arbitrary changes. The question is whether the individual manager can convince the organization, where necessary, to resist such changes.

The simple answer is that the individual must make sure any proposed changes will achieve the desired end. Too often the cost-benefit analysis for a new system assumes, unchallenged, that it will correct “unsatisfactory” performance. Often the change is adopted simply because a management that has been presiding over disappointing performance must be seen to take action lest it appears to condone it. It is right that underachievement often spurs a change; so it is achievement below expectations and targets that is often the spur for change — and rightly so.

Successful organizations are learning that the best course is to analyze the facts behind historical performance and initiate changes only when these are understood.

It is the manager who is faced with the task of opening up the discussion and questioning assumptions.

Obviously it is no good raising clouds of doubt about the causes of bad system performance without supporting figures. The manager needs to be armed with a coherent set of performance histories for all key areas under his control.

Quantitative techniques are available which can help this valuable exercise. These techniques are borrowed from industrial process control but can be applied in lots of other areas to great effect.

In Japan most commercial organizations use graphical methods such as control charts to monitor performance across service departments, sales forces and personnel departments as well as manufacturing. These methods rely on basic laws of probability to put performance history in context.

The Japanese became interested in the potential of statistical methods to help understand changes in quality and productivity in the decade after the Second World War. Some leading businessmen invited the American consultant W. Edwards Deming to visit Japan and teach managers and supervisors practical techniques for monitoring performance.

Deming’s basic thesis was that there are two important types of performance variation — which he classed as common causes and special causes. Common causes are due to natural variations, present in all systems both natural and man-made, from weather to sales. Special causes are freak events or disruptions such as sudden legislation, major price promotions, extreme weather, equipment breakdowns or illness epidemics.

The result of wrongly diagnosing the reason for bad performance is twofold: on the one hand there may be a waste of resources on an unnecessary systems redesign; on the other there may be needless expenditure of valuable supervisory time chasing explanations for every unexpected result.

For example, say an accounts department has a computer system which is required to process a minimum of 1,000 invoices every month. Then it fails to do so for two months running.

The reason could be either a common or special cause. If it is a common cause there was nothing extraordinary about the performance during these two months. Such a result was bound to come about sooner or later because of the system’s performance characteristics. If this is completely unsatisfactory the system must be redesigned and rebuilt. No amount of investigation or disciplinary action against staff will reduce the likelihood of a repeat.

If the two month’s poor performance was not bound to happen according to the system’s historical performance capability, a special cause must have been responsible. It may be that there was a new release of software which contained numerous bugs, or perhaps key members of staff were absent for long periods. In this case procedures could be changed to prevent a repetition. A scheme for identifying when special causes are to blame depends on successfully gauging the extent of natural variation in the business system and noting the special causes by a process of elimination.

In Japan, such schemes are used for tracking everything from divisional sales performance to clerical staff training needs.

Before embarking on a major change to an area of business operations a manager can and should assess what the characteristics of its performance are now and what improvement is needed. Only then can he judge whether the suggested remedy will pay off.

Process Design Consultants offers workshops in both areas. For common causes, the Process Management workshop shows you how to re-engineer or re-design the processes. For special causes, the problem-solving component of the Critical Thinking Skills workshop will determine the root cause so that it can be eliminated.

Previous post:

Next post: